applied platformio structure

Fix missing colon

GUITION_SETUP.md

@@ -0,0 +1,189 @@
+# Guition JC4827W543C Board Setup Guide
+
+This guide explains how to configure and build squeezelite-esp32 for the Guition JC4827W543C development board.
+
+## Board Features
+
+- **Processor**: ESP32-S3-WROOM-1 (dual-core, 240MHz)
+- **Display**: 4.3" ILI9488 LCD (480x272 pixels)
+- **Touch**: GT911 capacitive touch controller
+- **Interface**: QSPI display interface
+- **Memory**: 4MB Flash, 4MB PSRAM
+- **Connectivity**: WiFi, Bluetooth
+
+## Hardware Configuration
+
+### Display Connections (QSPI)
+- **CLK**: GPIO47
+- **DATA0**: GPIO21
+- **DATA1**: GPIO48  
+- **DATA2**: GPIO40
+- **DATA3**: GPIO39
+- **CS**: GPIO45
+- **DC**: Not used with QSPI
+- **RST**: GPIO38 (if available)
+
+### I2C Connections (Touch)
+- **SDA**: GPIO8
+- **SCL**: GPIO4
+- **Touch INT**: GPIO3
+- **Touch RST**: GPIO38
+
+### Audio I2S (External DAC Required)
+- **BCK**: GPIO15
+- **WS**: GPIO16  
+- **DO**: GPIO17
+- **DI**: (not used for output only)
+
+### Other GPIO
+- **Backlight**: GPIO1 (PWM controlled)
+
+## Software Configuration
+
+### 1. Build Configuration
+
+Use the provided build script:
+
+```bash
+# Make the script executable
+chmod +x build-guition.sh
+
+# Configure and build
+./build-guition.sh
+```
+
+Or manually:
+
+```bash
+# Set ESP32-S3 target
+export IDF_TARGET=esp32s3
+
+# Copy Guition configuration
+cp squeezelite-esp32-Guition-sdkconfig.defaults sdkconfig.defaults
+
+# Configure project
+idf.py menuconfig
+
+# Build
+idf.py build
+```
+
+### 2. Menuconfig Settings
+
+In `idf.py menuconfig`, select:
+
+**Target Configuration**:
+- `Squeezelite-ESP32` → `Guition JC4827W543C`
+
+**Audio Settings**:
+- Configure I2S GPIO pins for your external DAC
+- Default: BCK=15, WS=16, DO=17
+
+**Display Settings**:
+- Should be automatically configured as:
+  - Type: QSPI
+  - Driver: ILI9488
+  - Width: 480
+  - Height: 272
+  - CS: 45
+  - Speed: 20MHz
+
+**I2C Settings**:
+- SDA: 8
+- SCL: 4
+- Speed: 400kHz
+- Port: 0
+
+### 3. External Audio DAC
+
+The Guition board does not have a built-in audio DAC. You must connect an external I2S DAC such as:
+
+- PCM5102
+- TAS575x series
+- ES8388
+- AC101
+
+Connect the DAC to the I2S pins and configure it in the NVS settings or through the web UI.
+
+## Flashing the Firmware
+
+```bash
+# Flash to the board
+idf.py -p /dev/ttyUSB0 flash
+
+# Monitor output (optional)
+idf.py -p /dev/ttyUSB0 monitor
+```
+
+## First Boot Configuration
+
+1. Connect to the WiFi AP created by the device (SSID: SqueezeESP32-XXXXXX)
+2. Open the web configuration interface
+3. Configure your WiFi network
+4. Set up your audio DAC if needed
+5. Configure LMS server connection
+
+## Display Features
+
+The ILI9488 driver supports:
+- 16-bit RGB565 color
+- Hardware acceleration
+- Shadow buffering for performance
+- Automatic dirty region tracking
+- Rotation and flip options
+
+## Touch Support
+
+Touch functionality is planned but not yet implemented. The GT911 touch controller is connected via I2C.
+
+## Troubleshooting
+
+### Display Not Working
+- Check QSPI GPIO connections
+- Verify CS pin is correctly set to GPIO45
+- Ensure proper power supply (3.3V for display logic)
+
+### Audio Not Working  
+- Verify external DAC is properly connected
+- Check I2S GPIO assignments
+- Configure DAC model in web UI if using a supported DAC
+
+### Build Issues
+- Ensure ESP-IDF supports ESP32-S3
+- Check that all required components are included
+- Verify GPIO pin assignments don't conflict
+
+## Performance Notes
+
+- The 480x272 display requires significant memory bandwidth
+- PSRAM is used for display framebuffer (≈260KB)
+- Audio performance may be affected at high display refresh rates
+- Consider reducing display update rate if audio issues occur
+
+## GPIO Map Summary
+
+| GPIO | Function | Direction |
+|------|----------|-----------|
+| 1    | Backlight PWM | Out |
+| 3    | Touch INT | In |
+| 4    | I2C SCL | Out |
+| 8    | I2C SDA | I/O |
+| 15   | I2S BCK | Out |
+| 16   | I2S WS | Out |
+| 17   | I2S DO | Out |
+| 21   | QSPI DATA0 | Out |
+| 39   | QSPI DATA3 | Out |
+| 40   | QSPI DATA2 | Out |
+| 45   | QSPI CS | Out |
+| 47   | QSPI CLK | Out |
+| 48   | QSPI DATA1 | Out |
+| 38   | Touch RST | Out |
+
+## Support
+
+For issues specific to the Guition board implementation:
+1. Check this README first
+2. Review the squeezelite-esp32 documentation
+3. Open an issue on the GitHub repository
+
+For general squeezelite-esp32 questions, refer to the main project documentation and forums.

README-PlatformIO.md

@@ -0,0 +1,124 @@
+# SqueezeLite ESP32 - PlatformIO Configuration
+
+This project has been configured to work with PlatformIO while maintaining compatibility with the original ESP-IDF structure.
+
+## Project Structure
+
+```
+squeezelite-esp32-gronod/
+├── platformio.ini          # PlatformIO configuration file
+├── src/                    # Main source files
+│   ├── esp_app_main.c      # Application entry point
+│   ├── platform_esp32.h    # ESP32 platform definitions
+│   └── Params.txt          # Configuration parameters
+├── lib/                    # Component libraries (moved from components/)
+│   ├── wifi-manager/       # WiFi management component
+│   ├── squeezelite/        # Squeezelite audio player
+│   ├── platform_console/   # Platform console
+│   ├── display/            # Display drivers
+│   ├── telnet/             # Telnet interface
+│   ├── audio/              # Audio processing
+│   ├── codecs/             # Audio codecs
+│   ├── spotify/            # Spotify integration (CSpot)
+│   └── ...                 # Other components
+├── include/                # Global headers
+├── partitions.csv          # Partition table
+└── build_scripts/          # Custom build scripts
+```
+
+## Key Differences from ESP-IDF
+
+### 1. Directory Structure
+- `main/` → `src/` (PlatformIO standard)
+- `components/` → `lib/` (PlatformIO libraries)
+- `include/` created for global headers
+
+### 2. Build System
+- Uses PlatformIO's build system instead of CMake
+- Components are configured as libraries with `library.json` files
+- Build flags and options configured in `platformio.ini`
+
+### 3. Multiple Targets
+The original ESP-IDF project builds both `recovery` and `squeezelite` binaries. PlatformIO builds the main squeezelite application. Recovery functionality can be added as a separate build environment if needed.
+
+## Building with PlatformIO
+
+### Prerequisites
+- PlatformIO IDE or command-line tools
+- ESP32 development board
+
+### Build Commands
+```bash
+# Build the project
+pio run
+
+# Upload to device
+pio run --target upload
+
+# Monitor serial output
+pio device monitor
+
+# Clean build
+pio run --target clean
+```
+
+### Configuration
+- Edit `platformio.ini` for board-specific settings
+- Adjust `src/Params.txt` for application configuration
+- Modify build flags in `platformio.ini` as needed
+
+## Component Libraries
+
+Each major component has been converted to a PlatformIO library with a `library.json` file:
+
+- **wifi-manager**: WiFi network management
+- **squeezelite**: Core audio player functionality
+- **platform_console**: System console and commands
+- **display**: Display drivers and UI
+- **telnet**: Telnet interface for remote control
+- **audio**: Audio processing and output
+- **codecs**: Audio codec support
+- **spotify**: Spotify integration via CSpot
+
+## ESP-IDF Compatibility
+
+The project maintains ESP-IDF compatibility:
+- All ESP-IDF components are still available
+- ESP-IDF configuration files (sdkconfig.*) are preserved
+- Original ESP-IDF build system can still be used
+
+## Migration Notes
+
+### What Changed
+1. **platformio.ini**: Main configuration file
+2. **Directory structure**: Reorganized for PlatformIO
+3. **Library metadata**: Added library.json files
+4. **Build flags**: Migrated to platformio.ini format
+
+### What Remained
+- All source code files remain unchanged
+- ESP-IDF component structure preserved in lib/
+- Configuration files (sdkconfig, partitions.csv)
+- Original functionality maintained
+
+### Known Limitations
+1. **Multiple ELF files**: PlatformIO doesn't natively support building both recovery and squeezelite binaries in one build
+2. **Complex build scripts**: Some advanced ESP-IDF features may need custom PlatformIO scripts
+3. **Kconfig**: ESP-IDF's Kconfig system is not directly supported in PlatformIO
+
+## Troubleshooting
+
+### Build Issues
+- Ensure all required libraries are in lib/ directory
+- Check build flags in platformio.ini
+- Verify ESP-IDF environment variables if using local ESP-IDF
+
+### Component Issues
+- Each component needs a library.json file
+- Check component dependencies in platformio.ini
+- Verify include paths and source file locations
+
+### Flash Issues
+- Check partition table configuration
+- Verify flash size settings
+- Ensure correct board configuration in platformio.ini

README.md

@@ -157,6 +157,21 @@ XMT - 3.3V
 
 Use the `squeezelite-esp32-I2S-4MFlash-sdkconfig.defaults` configuration file.
 
+### ESP32 LyraT Mini v1.2
+
+This board is one of the [audio developpement board](https://docs.espressif.com/projects/esp-adf/en/latest/design-guide/dev-boards/get-started-esp32-lyrat-mini.html) designed by espressif for their ESP-ADF (Espressif Audio Development Framework). It uses ESP32-WROVER-E as it core and ES8311 as DAC. 
+A jack connector can be used or an HP output.
+It also contains 2 LEDS (one green, one blue) and 8 Buttons. Those buttons are not supported for now.
+An ADC is also present on the board, but not used in squeezelite use case.
+
+As for now, audio playback and LEDs are working.
+
+Use the `squeezelite-esp32-I2S-4MFlash-sdkconfig.defaults` configuration file.
+
+- i2c_config: `scl=25,sda=18,speed=400000,port=1`
+- dac_config: `model=es8311,bck=5,ws=25,do=26,sda=18,scl=23,i2c=24`
+- set_GPIO: `22=green,27=red,19=jack` 
+
 ### SqueezeAmpToo !
 
 And the super cool project https://github.com/rochuck/squeeze-amp-too
@@ -184,11 +199,11 @@ Default and only "host" is 1 as others are used already by flash and spiram. The
 ### DAC/I2S
 The NVS parameter "dac_config" set the gpio used for i2s communication with your DAC. You can define the defaults at compile time but nvs parameter takes precedence except for named configurations
 ```
-bck=<gpio>,ws=<gpio>,do=<gpio>[,mck=0|1|2][,mute=<gpio>[:0|1][,model=TAS57xx|TAS5713|AC101|WM8978|ES8388|I2S][,sda=<gpio>,scl=<gpio>[,i2c=<addr>]]
+bck=<gpio>,ws=<gpio>,do=<gpio>[,mck=0|1|2][,mute=<gpio>[:0|1][,model=TAS57xx|TAS5713|AC101|WM8978|ES8388|ES8311|I2S][,sda=<gpio>,scl=<gpio>[,i2c=<addr>]]
 ```
 if "model" is not set or is not recognized, then default "I2S" is used. The option "mck" is used for some codecs that require a master clock (although they should not). By default GPIO0 is used as MCLK and only recent builds (post mid-2023) can use 1 or 2. Also be aware that this cannot coexit with RMII Ethernet (see ethernet section below). I2C parameters are optional and only needed if your DAC requires an I2C control (See 'dac_controlset' below). Note that "i2c" parameters are decimal, hex notation is not allowed.
 
-So far, TAS57xx, TAS5713, AC101, WM8978 and ES8388 are recognized models where the proper init sequence/volume/power controls are sent. For other codecs that might require an I2C commands, please use the parameter "dac_controlset" that allows definition of simple commands to be sent over i2c for init, power, speaker and headset on and off using a JSON syntax:
+So far, TAS57xx, TAS5713, AC101, WM8978, ES8388 and ES8311 are recognized models where the proper init sequence/volume/power controls are sent. For other codecs that might require an I2C commands, please use the parameter "dac_controlset" that allows definition of simple commands to be sent over i2c for init, power, speaker and headset on and off using a JSON syntax:
 ```json
 { <command>: [ <item1>, <item2>, ... <item3> ],
   <command>: [ <item1>, <item2>, ... <item3> ],
@@ -478,11 +493,12 @@ The benefit of the "raw" mode is that you can build a player which is as close a
 There is no good or bad option, it's your choice. Use the NVS parameter "lms_ctrls_raw" to change that option
 	
 **Note that gpio 36 and 39 are input only and cannot use interrupt. When using them for a button, a 100ms polling is started which is expensive. Long press is also likely to not work very well**
+
+**Note:** Touch buttons that can be found on some board like the LyraT V4.3 are not supported currently.
+
 ### Ethernet 
 Wired ethernet is supported by esp32 with various options but squeezeESP32 is only supporting a Microchip LAN8720 with a RMII interface like [this](https://www.aliexpress.com/item/32858432526.html) or SPI-ethernet bridges like Davicom DM9051 [that](https://www.amazon.com/dp/B08JLFWX9Z) or W5500 like [this](https://www.aliexpress.com/item/32312441357.html).
 
-**Note:** Touch buttons that can be find on some board like the LyraT V4.3 are not supported currently.
-
 #### RMII (LAN8720)	
 - RMII PHY wiring is fixed and can not be changed
 

WINDOWS_BUILD_GUIDE.md

@@ -0,0 +1,216 @@
+# Windows Build Guide for Guition JC4827W543C
+
+This guide explains how to build squeezelite-esp32 for the Guition board on Windows.
+
+## Prerequisites
+
+1. **ESP-IDF for Windows**
+   - Download and install ESP-IDF from: https://dl.espressif.com/dl/esp-idf/
+   - Run the installer and follow the setup instructions
+   - Make sure to install the required tools (Git, Python, CMake)
+
+2. **Visual Studio Build Tools** (if not already installed by ESP-IDF)
+   - Install from Visual Studio Installer
+   - Select "C++ build tools"
+
+3. **Hardware**
+   - Guition JC4827W543C board
+   - USB cable for power and programming
+   - External I2S audio DAC (required)
+
+## Setup Instructions
+
+### 1. Install ESP-IDF
+```cmd
+# Run the ESP-IDF installer
+# Launch ESP-IDF Command Prompt from Start Menu
+```
+
+### 2. Initialize ESP-IDF Environment
+```cmd
+# In ESP-IDF Command Prompt
+cd C:\Espressif\esp-idf
+export.bat
+```
+
+### 3. Clone and Setup Project
+```cmd
+# Navigate to your workspace
+cd C:\Users\YourName\Projects
+
+# Clone the repository (if not already done)
+git clone <repository-url> squeezelite-esp32-gronod
+cd squeezelite-esp32-gronod
+```
+
+### 4. Set Environment for Project
+```cmd
+# Set ESP-IDF environment for this project
+C:\Espressif\esp-idf\export.bat
+```
+
+## Build Methods
+
+### Method 1: PowerShell Script (Recommended)
+```powershell
+# Open PowerShell as Administrator
+# Navigate to project directory
+cd C:\Users\YourName\Projects\squeezelite-esp32-gronod
+
+# Run the build script
+.\build-guition.ps1
+
+# For clean build
+.\build-guition.ps1 -Clean
+```
+
+**Note:** If you get execution policy errors, run:
+```powershell
+Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
+```
+
+### Method 2: Batch File (Simple)
+```cmd
+# Open Command Prompt
+# Navigate to project directory
+cd C:\Users\YourName\Projects\squeezelite-esp32-gronod
+
+# Run the build script
+build-guition.bat
+
+# For clean build
+build-guition.bat clean
+```
+
+### Method 3: Manual Commands
+```cmd
+# Set ESP-IDF environment
+C:\Espressif\esp-idf\export.bat
+
+# Set target
+set IDF_TARGET=esp32s3
+
+# Copy Guition config
+copy /Y squeezelite-esp32-Guition-sdkconfig.defaults sdkconfig.defaults
+
+# Configure (opens menu interface)
+idf.py menuconfig
+
+# Build
+idf.py build
+```
+
+## Menuconfig Configuration
+
+When `idf.py menuconfig` runs:
+
+1. **Select Target Hardware**:
+   - Navigate to: `Squeezelite-ESP32` → `Guition JC4827W543C`
+
+2. **Configure Audio** (if needed):
+   - Navigate to: `Audio settings` → `DAC settings`
+   - Set I2S GPIO pins for your external DAC
+   - Default: BCK=15, WS=16, DO=17
+
+3. **Save and Exit**:
+   - Press `S` to save
+   - Press `Q` to exit
+
+## Flashing the Firmware
+
+### Find Your COM Port
+```cmd
+# List all COM ports
+mode
+
+# Or check Device Manager under "Ports (COM & LPT)"
+```
+
+### Flash and Monitor
+```cmd
+# Flash firmware
+idf.py -p COM3 flash
+
+# Flash and monitor (recommended)
+idf.py -p COM3 flash monitor
+
+# Just monitor (if already flashed)
+idf.py -p COM3 monitor
+```
+
+### Using the Scripts
+Both PowerShell and batch scripts will ask if you want to flash after building.
+
+## Troubleshooting
+
+### Common Issues
+
+1. **"ESP-IDF environment not found"**
+   - Make sure you're running from ESP-IDF Command Prompt
+   - Or run `C:\Espressif\esp-idf\export.bat` first
+
+2. **"Python not found"**
+   - ESP-IDF installer should include Python
+   - Check that Python is in your PATH
+
+3. **"Build tools not found"**
+   - Install Visual Studio Build Tools
+   - Make sure C++ tools are selected
+
+4. **"Permission denied" (PowerShell)**
+   - Run PowerShell as Administrator
+   - Or set execution policy: `Set-ExecutionPolicy RemoteSigned`
+
+5. **"COM port not found"**
+   - Check device connections
+   - Install USB drivers if needed (CH340/CP210x)
+   - Check Device Manager
+
+### Clean Build
+If you encounter build issues:
+```cmd
+# Using batch file
+build-guition.bat clean
+
+# Or manually
+idf.py fullclean
+idf.py build
+```
+
+### Verifying Installation
+```cmd
+# Check ESP-IDF version
+idf.py --version
+
+# Check target
+echo %IDF_TARGET%
+
+# List available targets
+idf.py --help | findstr target
+```
+
+## Development Workflow
+
+1. **Initial Setup**: Run the build script once to configure
+2. **Code Changes**: Make your modifications
+3. **Build**: Run `.\build-guition.ps1` or `build-guition.bat`
+4. **Flash**: Let the script flash automatically, or use `idf.py -p COMX flash`
+5. **Monitor**: Use `idf.py -p COMX monitor` to view logs
+
+## Performance Tips
+
+- **SSD Storage**: Build on SSD for much faster compilation
+- **RAM**: 8GB+ recommended for large projects
+- **CPU**: Multi-core CPU helps with parallel compilation
+- **Antivirus**: Exclude project directory from real-time scanning
+
+## Next Steps
+
+After successful build and flash:
+
+1. Connect to WiFi AP created by device
+2. Configure your network settings
+3. Set up audio DAC configuration
+4. Connect to your Logitech Media Server
+
+For more detailed configuration, see `GUITION_SETUP.md`.

build-guition.bat

@@ -0,0 +1,80 @@
+@echo off
+REM Build script for Guition JC4827W543C board - Batch file version
+REM Usage: build-guition.bat [clean]
+
+echo Building Squeezelite-ESP32 for Guition JC4827W543C
+
+REM Check if ESP-IDF environment is set up
+if "%IDF_PATH%"=="" (
+    echo Error: ESP-IDF environment not found!
+    echo Please run esp-idf\export.bat first
+    pause
+    exit /b 1
+)
+
+REM Set target to ESP32-S3
+set IDF_TARGET=esp32s3
+echo Target set to: %IDF_TARGET%
+
+REM Copy Guition-specific configuration
+if exist "squeezelite-esp32-Guition-sdkconfig.defaults" (
+    echo Using Guition configuration...
+    copy /Y "squeezelite-esp32-Guition-sdkconfig.defaults" "sdkconfig.defaults"
+) else (
+    echo Warning: Guition configuration file not found
+)
+
+REM Clean if requested
+if /i "%1"=="clean" (
+    echo Cleaning build...
+    idf.py fullclean
+    if errorlevel 1 (
+        echo Clean failed!
+        pause
+        exit /b 1
+    )
+)
+
+REM Configure the project
+echo Configuring project...
+echo Running: idf.py menuconfig
+echo Select: Squeezelite-ESP32 -^> Guition JC4827W543C
+echo.
+echo Press any key to continue to menuconfig...
+pause > nul
+
+idf.py menuconfig
+if errorlevel 1 (
+    echo Configuration failed!
+    pause
+    exit /b 1
+)
+
+REM Build the project
+echo Building firmware...
+idf.py build
+if errorlevel 1 (
+    echo Build failed!
+    pause
+    exit /b 1
+)
+
+echo Build complete!
+echo Firmware location: build\squeezelite.bin
+echo.
+echo To flash the firmware:
+echo idf.py -p ^<PORT^> flash
+echo.
+echo To monitor output:
+echo idf.py -p ^<PORT^> monitor
+
+REM Ask if user wants to flash
+echo.
+set /p flash="Do you want to flash the firmware now? (y/n) "
+if /i "%flash%"=="y" (
+    set /p port="Enter COM port (e.g., COM3): "
+    echo Flashing to %port%...
+    idf.py -p %port% flash monitor
+)
+
+pause

build-guition.ps1

@@ -0,0 +1,76 @@
+# Build script for Guition JC4827W543C board - PowerShell version
+# Usage: .\build-guition.ps1 [-Clean]
+
+param(
+    [switch]$Clean
+)
+
+Write-Host "Building Squeezelite-ESP32 for Guition JC4827W543C" -ForegroundColor Green
+
+# Check if ESP-IDF environment is set up
+if (-not $env:IDF_PATH) {
+    Write-Host "Error: ESP-IDF environment not found!" -ForegroundColor Red
+    Write-Host "Please run esp-idf/export.bat first" -ForegroundColor Yellow
+    exit 1
+}
+
+# Set target to ESP32-S3
+$env:IDF_TARGET = "esp32s3"
+Write-Host "Target set to: $env:IDF_TARGET" -ForegroundColor Cyan
+
+# Copy Guition-specific configuration
+if (Test-Path "squeezelite-esp32-Guition-sdkconfig.defaults") {
+    Write-Host "Using Guition configuration..." -ForegroundColor Cyan
+    Copy-Item "squeezelite-esp32-Guition-sdkconfig.defaults" "sdkconfig.defaults" -Force
+} else {
+    Write-Host "Warning: Guition configuration file not found" -ForegroundColor Yellow
+}
+
+# Clean if requested
+if ($Clean) {
+    Write-Host "Cleaning build..." -ForegroundColor Cyan
+    idf.py fullclean
+    if ($LASTEXITCODE -ne 0) {
+        Write-Host "Clean failed!" -ForegroundColor Red
+        exit 1
+    }
+}
+
+# Configure the project
+Write-Host "Configuring project..." -ForegroundColor Cyan
+Write-Host "Running: idf.py menuconfig" -ForegroundColor White
+Write-Host "Select: Squeezelite-ESP32 -> Guition JC4827W543C" -ForegroundColor Yellow
+Write-Host "Press Enter to continue to menuconfig..." -ForegroundColor White
+Read-Host
+
+idf.py menuconfig
+if ($LASTEXITCODE -ne 0) {
+    Write-Host "Configuration failed!" -ForegroundColor Red
+    exit 1
+}
+
+# Build the project
+Write-Host "Building firmware..." -ForegroundColor Cyan
+idf.py build
+if ($LASTEXITCODE -ne 0) {
+    Write-Host "Build failed!" -ForegroundColor Red
+    exit 1
+}
+
+Write-Host "Build complete!" -ForegroundColor Green
+Write-Host "Firmware location: build\squeezelite.bin" -ForegroundColor Cyan
+Write-Host ""
+Write-Host "To flash the firmware:" -ForegroundColor White
+Write-Host "idf.py -p <PORT> flash" -ForegroundColor Yellow
+Write-Host ""
+Write-Host "To monitor output:" -ForegroundColor White
+Write-Host "idf.py -p <PORT> monitor" -ForegroundColor Yellow
+
+# Ask if user wants to flash
+Write-Host ""
+$flash = Read-Host "Do you want to flash the firmware now? (y/n)"
+if ($flash -eq 'y' -or $flash -eq 'Y') {
+    $port = Read-Host "Enter COM port (e.g., COM3)"
+    Write-Host "Flashing to $port..." -ForegroundColor Cyan
+    idf.py -p $port flash monitor
+}

build-guition.sh

@@ -0,0 +1,40 @@
+#!/bin/bash
+
+# Build script for Guition JC4827W543C board
+# Usage: ./build-guition.sh [clean]
+
+set -e
+
+echo "Building Squeezelite-ESP32 for Guition JC4827W543C"
+
+# Set target to ESP32-S3
+export IDF_TARGET=esp32s3
+
+# Copy Guition-specific configuration
+if [ -f "squeezelite-esp32-Guition-sdkconfig.defaults" ]; then
+    echo "Using Guition configuration..."
+    cp squeezelite-esp32-Guition-sdkconfig.defaults sdkconfig.defaults
+fi
+
+# Clean if requested
+if [ "$1" == "clean" ]; then
+    echo "Cleaning build..."
+    idf.py fullclean
+fi
+
+# Configure with Guition target
+echo "Configuring project..."
+idf.py menuconfig
+
+# Build the project
+echo "Building firmware..."
+idf.py build
+
+echo "Build complete!"
+echo "Firmware location: build/squeezelite.bin"
+echo ""
+echo "To flash the firmware:"
+echo "idf.py -p <PORT> flash"
+echo ""
+echo "To monitor output:"
+echo "idf.py -p <PORT> monitor"

components/display/ILI9488.c

@@ -0,0 +1,297 @@
+/**
+ * ILI9488 Display Driver for Guition JC4827W543C
+ * Supports QSPI interface for 480x272 resolution
+ * Based on ILI9341 driver adapted for ILI9488
+ * 
+ * (c) Guition Support 2026
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+#define PAGE_BLOCK		4096
+#define ENABLE_WRITE	0x2c
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "ILI9488";
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	struct {
+		uint16_t Height, Width;
+	} Offset;
+	uint8_t MADCtl, PageSize;
+	uint8_t Model;
+};
+
+// ILI9488 Commands
+static const uint8_t ILI9488_INIT_SEQUENCE[] = {
+	// Software reset
+	0x01, 0x80, 150,	// SWRESET, delay 150ms
+	// Power control
+	0xD0, 3, 0x07, 0x42, 0x18,	// Power Control
+	0xD1, 3, 0x00, 0x07, 0x10,	// VCOM Control
+	0xD2, 1, 0x01,				// Power Control for Normal Mode
+	0xC0, 2, 0x10, 0x3B,		// Panel Driving Setting
+	0xC1, 1, 0x10,				// Frame Rate Control
+	0xC5, 5, 0x0A, 0x3A, 0x28, 0x28, 0x02, // MCU Control
+	0xC6, 1, 0x00,				// Frame Rate Control
+	0xB1, 2, 0x00, 0x1B,		// Display Function Control
+	0xB4, 1, 0x02,				// Inversion Control
+	0xB6, 3, 0x02, 0x02, 0x3B,	// Display Function Control
+	0xB7, 1, 0xC6,				// Entry Mode Set
+	0xE0, 16, 0x00, 0x07, 0x10, 0x0E, 0x09, 0x16, 0x06, 0x0A,
+				0x0E, 0x09, 0x15, 0x0D, 0x0E, 0x11, 0x0F, 0x12, // Positive Gamma Control
+	0xE1, 16, 0x00, 0x17, 0x1A, 0x04, 0x0E, 0x06, 0x2F, 0x24,
+				0x1B, 0x1B, 0x22, 0x1F, 0x1E, 0x37, 0x3F, 0x00, // Negative Gamma Control
+	0x36, 1, 0xE8,				// Memory Access Control (MX, MY, RGB mode)
+	0x3A, 1, 0x55,				// Interface Pixel Format (16bpp)
+	0x11, 0x80, 150,			// Sleep Out, delay 150ms
+	0x29, 0x80, 50,				// Display On, delay 50ms
+	0xFF, 0x00					// End of sequence
+};
+
+static void WriteByte( struct GDS_Device* Device, uint8_t Data ) {
+	Device->WriteData( Device, &Data, 1 );
+}
+
+static void SetColumnAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2A );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void SetRowAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2B );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void Update16( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+#ifdef SHADOW_BUFFER
+	uint32_t *optr = (uint32_t*) Private->Shadowbuffer, *iptr = (uint32_t*) Device->Framebuffer;
+	int FirstCol = Device->Width / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+	
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width is always a multiple of 2)
+		for (int c = 0; c < Device->Width / 2; c++, iptr++, optr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+
+		// wait for a large enough window - careful that window size might increase by more than a line at once !
+		if (FirstRow < 0 || ((LastCol - FirstCol + 1) * (r - FirstRow + 1) * 4 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol *= 2;
+		LastCol = LastCol * 2 + 1;
+		SetRowAddress( Device, FirstRow + Private->Offset.Height, LastRow + Private->Offset.Height);
+		SetColumnAddress( Device, FirstCol + Private->Offset.Width, LastCol + Private->Offset.Width );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 2;
+			
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM <= (PAGE_BLOCK - ChunkSize) && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize );
+		}
+		
+		FirstCol = Device->Width / 2;
+		LastCol = 0;
+		FirstRow = -1;
+	}
+#endif
+}
+
+static void Clear( struct GDS_Device* Device ) {
+	memset( Device->Framebuffer, 0, Device->Width * Device->Height * 2 );
+	Device->Update( Device);
+}
+
+static void SetPixel( struct GDS_Device* Device, int X, int Y, uint32_t Color ) {
+	if( X < 0 || X >= Device->Width || Y < 0 || Y >= Device->Height) return;
+	
+	*((uint16_t*) Device->Framebuffer + (Y * Device->Width) + X) = (uint16_t) Color;
+}
+
+static uint32_t GetPixel( struct GDS_Device* Device, int X, int Y ) {
+	if( X < 0 || X >= Device->Width || Y < 0 || Y >= Device->Height) return 0;
+	
+	return *((uint16_t*) Device->Framebuffer + (Y * Device->Width) + X);
+}
+
+static void DrawPixel( struct GDS_Device* Device, int X, int Y, uint32_t Color ) {
+	SetPixel( Device, X, Y, Color );
+}
+
+static void DrawPixelFast( struct GDS_Device* Device, int X, int Y, uint32_t Color ) {
+	*((uint16_t*) Device->Framebuffer + (Y * Device->Width) + X) = (uint16_t) Color;
+}
+
+static void DrawCBR( struct GDS_Device* Device, int X, int Y, int Width, int Height, uint32_t Color ) {
+	uint16_t *fb = (uint16_t*) Device->Framebuffer + Y * Device->Width + X;
+	
+	for( int y = 0; y < Height; y++) {
+		for( int x = 0; x < Width; x++) {
+			fb[x] = (uint16_t) Color;
+		}
+		fb += Device->Width;
+	}
+}
+
+static void DrawHLine( struct GDS_Device* Device, int X0, int X1, int Y, uint32_t Color ) {
+	if( Y < 0 || Y >= Device->Height) return;
+	
+	if( X0 > X1) {
+		int Temp = X0;
+		X0 = X1;
+		X1 = Temp;
+	}
+	
+	if( X0 < 0) X0 = 0;
+	if( X1 >= Device->Width) X1 = Device->Width - 1;
+	
+	uint16_t *fb = (uint16_t*) Device->Framebuffer + Y * Device->Width + X0;
+	
+	for( int x = X0; x <= X1; x++) {
+		*fb++ = (uint16_t) Color;
+	}
+}
+
+static void DrawVLine( struct GDS_Device* Device, int X, int Y0, int Y1, uint32_t Color ) {
+	if( X < 0 || X >= Device->Width) return;
+	
+	if( Y0 > Y1) {
+		int Temp = Y0;
+		Y0 = Y1;
+		Y1 = Temp;
+	}
+	
+	if( Y0 < 0) Y0 = 0;
+	if( Y1 >= Device->Height) Y1 = Device->Height - 1;
+	
+	uint16_t *fb = (uint16_t*) Device->Framebuffer + Y0 * Device->Width + X;
+	
+	for( int y = Y0; y <= Y1; y++) {
+		*fb = (uint16_t) Color;
+		fb += Device->Width;
+	}
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	const uint8_t *p = ILI9488_INIT_SEQUENCE;
+	
+	ESP_LOGI(TAG, "Initializing ILI9488 display %dx%d", Device->Width, Device->Height);
+	
+	// Allocate IRAM buffer if available
+	Private->iRAM = (uint8_t*) heap_caps_malloc(PAGE_BLOCK, MALLOC_CAP_DMA | MALLOC_CAP_INTERNAL);
+	if (!Private->iRAM) {
+		ESP_LOGW(TAG, "Could not allocate IRAM buffer, using direct write");
+	}
+
+#ifdef SHADOW_BUFFER
+	Private->Shadowbuffer = (uint8_t*) heap_caps_malloc(Device->Width * Device->Height * 2, MALLOC_CAP_DMA);
+	if (!Private->Shadowbuffer) {
+		ESP_LOGE(TAG, "Could not allocate shadow buffer");
+		if (Private->iRAM) free(Private->iRAM);
+		return false;
+	}
+	memset(Private->Shadowbuffer, 0, Device->Width * Device->Height * 2);
+#endif
+
+	// Send initialization sequence
+	while(*p != 0xFF) {
+		uint8_t cmd = *p++;
+		uint8_t len = (*p & 0x7F);
+		bool delay = (*p++ & 0x80);
+		
+		Device->WriteCommand(Device, cmd);
+		if(len) Device->WriteData(Device, (uint8_t*)p, len);
+		p += len;
+		
+		if(delay) {
+			uint8_t ms = *p++;
+			vTaskDelay(ms / portTICK_PERIOD_MS);
+		}
+	}
+
+	// Set orientation and layout
+	Private->MADCtl = 0xE8; // MX=1, MY=1, RGB=1, BGR=0
+	Device->WriteCommand(Device, 0x36);
+	Device->WriteData(Device, &Private->MADCtl, 1);
+
+	// Set pixel format to 16-bit
+	uint8_t fmt = 0x55;
+	Device->WriteCommand(Device, 0x3A);
+	Device->WriteData(Device, &fmt, 1);
+
+	// Turn on display
+	Device->WriteCommand(Device, 0x29);
+	
+	ESP_LOGI(TAG, "ILI9488 initialization complete");
+	return true;
+}
+
+static void Deinit( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	
+	if (Private->iRAM) free(Private->iRAM);
+#ifdef SHADOW_BUFFER
+	if (Private->Shadowbuffer) free(Private->Shadowbuffer);
+#endif
+}
+
+struct GDS_Device* ILI9488_Detect( char *Driver, struct GDS_Device *Device ) {
+	if(strcasecmp(Driver, "ILI9488") != 0) return NULL;
+
+	Device->Private = calloc(1, sizeof(struct PrivateSpace));
+	if(!Device->Private) {
+		ESP_LOGE(TAG, "Cannot allocate private data");
+		return NULL;
+	}
+
+	Device->Mode = GDS_RGB565;
+	Device->Depth = 16;
+	Device->Update = Update16;
+	Device->Clear = Clear;
+	Device->SetPixel = SetPixel;
+	Device->GetPixel = GetPixel;
+	Device->DrawPixel = DrawPixel;
+	Device->DrawPixelFast = DrawPixelFast;
+	Device->DrawCBR = DrawCBR;
+	Device->DrawHLine = DrawHLine;
+	Device->DrawVLine = DrawVLine;
+	Device->Init = Init;
+	Device->Deinit = Deinit;
+
+	ESP_LOGI(TAG, "ILI9488 driver loaded");
+	return Device;
+}

components/display/core/gds_default_if.h

@@ -13,6 +13,10 @@ bool GDS_I2CAttachDevice( struct GDS_Device* Device, int Width, int Height, int
 bool GDS_SPIInit( int SPI, int DC );
 bool GDS_SPIAttachDevice( struct GDS_Device* Device, int Width, int Height, int CSPin, int RSTPin, int Speed, int BacklightPin, int Mode );
 
+bool GDS_QSPIInit( int QSPI, int DC );
+bool GDS_QSPIAttachDevice( struct GDS_Device* Device, int Width, int Height, int CSPin, int RSTPin, int Speed, int BacklightPin, int Mode );
+bool GDS_QSPIBusInit( int MOSIPin, int MISOPin, int CLKPin, int Host );
+
 #ifdef __cplusplus
 }
 #endif

components/display/core/ifaces/default_if_qspi.c

@@ -0,0 +1,161 @@
+/**
+ * QSPI Interface for Guition JC4827W543C ILI9488 Display
+ * Supports ESP32-S3 QSPI peripheral
+ * 
+ * (c) Guition Support 2026
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <string.h>
+#include <driver/spi_master.h>
+#include <driver/gpio.h>
+#include <freertos/task.h>
+#include "gds.h"
+#include "gds_err.h"
+#include "gds_private.h"
+#include "gds_default_if.h"
+
+static const int GDS_QSPI_Command_Mode = 0;
+static const int GDS_QSPI_Data_Mode = 1;
+
+static spi_host_device_t QSPIHost;
+static int DCPin;
+
+static bool QSPIDefaultWriteBytes( spi_device_handle_t QSPIHandle, int WriteMode, const uint8_t* Data, size_t DataLength );
+static bool QSPIDefaultWriteCommand( struct GDS_Device* Device, uint8_t Command );
+static bool QSPIDefaultWriteData( struct GDS_Device* Device, const uint8_t* Data, size_t DataLength );
+
+bool GDS_QSPIInit( int QSPI, int DC ) {
+	QSPIHost = QSPI;
+	DCPin = DC;
+    return true;
+}
+
+bool GDS_QSPIAttachDevice( struct GDS_Device* Device, int Width, int Height, int CSPin, int RSTPin, int BackLightPin, int Speed, int Mode ) {
+    spi_device_interface_config_t QSPIDeviceConfig = { };
+    spi_device_handle_t QSPIDevice;
+
+    NullCheck( Device, return false );
+	
+	if (CSPin >= 0) {
+		ESP_ERROR_CHECK_NONFATAL( gpio_set_direction( CSPin, GPIO_MODE_OUTPUT ), return false );
+		ESP_ERROR_CHECK_NONFATAL( gpio_set_level( CSPin, 0 ), return false );
+	}
+	
+    QSPIDeviceConfig.clock_speed_hz = Speed > 0 ? Speed : SPI_MASTER_FREQ_20M;
+    QSPIDeviceConfig.spics_io_num = CSPin;
+    QSPIDeviceConfig.queue_size = 4;
+    QSPIDeviceConfig.mode = Mode;
+    QSPIDeviceConfig.flags = SPI_DEVICE_HALFDUPLEX;
+    QSPIDeviceConfig.duty_cycle_pos = 128; // 50% duty cycle
+    
+    // QSPI specific configuration for ESP32-S3
+    QSPIDeviceConfig.command_bits = 8;
+    QSPIDeviceConfig.address_bits = 24;
+    QSPIDeviceConfig.address_len = 3;
+    
+    if( spi_bus_add_device( QSPIHost, &QSPIDeviceConfig, &QSPIDevice ) != ESP_OK ) {
+        GDS_LOG_ERROR( "Failed to add QSPI device to host" );
+        return false;
+    }
+    
+    if( DCPin >= 0 ) {
+        ESP_ERROR_CHECK_NONFATAL( gpio_set_direction( DCPin, GPIO_MODE_OUTPUT ), return false );
+        ESP_ERROR_CHECK_NONFATAL( gpio_set_level( DCPin, 1 ), return false );
+    }
+    
+    if( RSTPin >= 0 ) {
+        ESP_ERROR_CHECK_NONFATAL( gpio_set_direction( RSTPin, GPIO_MODE_OUTPUT ), return false );
+        ESP_ERROR_CHECK_NONFATAL( gpio_set_level( RSTPin, 1 ), return false );
+    }
+    
+    if( BackLightPin >= 0 ) {
+        ESP_ERROR_CHECK_NONFATAL( gpio_set_direction( BackLightPin, GPIO_MODE_OUTPUT ), return false );
+        ESP_ERROR_CHECK_NONFATAL( gpio_set_level( BackLightPin, 1 ), return false );
+    }
+
+    Device->WriteCommand = QSPIDefaultWriteCommand;
+    Device->WriteData = QSPIDefaultWriteData;
+    Device->DeviceHandle = QSPIDevice;
+    
+    // Hardware reset if pin is available
+    if( RSTPin >= 0 ) {
+        gpio_set_level( RSTPin, 0 );
+        vTaskDelay( pdMS_TO_TICKS( 10 ) );
+        gpio_set_level( RSTPin, 1 );
+        vTaskDelay( pdMS_TO_TICKS( 120 ) );
+    }
+
+    return true;
+}
+
+static bool QSPIDefaultWriteCommand( struct GDS_Device* Device, uint8_t Command ) {
+    return QSPIDefaultWriteBytes( Device->DeviceHandle, GDS_QSPI_Command_Mode, &Command, 1 );
+}
+
+static bool QSPIDefaultWriteData( struct GDS_Device* Device, const uint8_t* Data, size_t DataLength ) {
+    if( DCPin >= 0 ) {
+        gpio_set_level( DCPin, 1 );
+    }
+    
+    bool Result = QSPIDefaultWriteBytes( Device->DeviceHandle, GDS_QSPI_Data_Mode, Data, DataLength );
+    
+    if( DCPin >= 0 ) {
+        gpio_set_level( DCPin, 0 );
+    }
+    
+    return Result;
+}
+
+static bool QSPIDefaultWriteBytes( spi_device_handle_t QSPIHandle, int WriteMode, const uint8_t* Data, size_t DataLength ) {
+    spi_transaction_ext_t SPITransaction = { };
+    SPITransaction.base.flags = SPI_TRANS_VARIABLE_CMD | SPI_TRANS_VARIABLE_ADDR | SPI_TRANS_VARIABLE_DUMMY;
+    SPITransaction.base.cmd = 0;
+    SPITransaction.base.addr = 0;
+    SPITransaction.base.length = DataLength * 8;
+    SPITransaction.base.tx_buffer = Data;
+    SPITransaction.base.rx_buffer = NULL;
+    SPITransaction.command_bits = 8;
+    SPITransaction.address_bits = 0;
+    SPITransaction.dummy_bits = 0;
+    
+    if( WriteMode == GDS_QSPI_Command_Mode ) {
+        SPITransaction.command_bits = 8;
+        SPITransaction.base.cmd = Data[0];
+        SPITransaction.base.length = 0;
+        SPITransaction.base.tx_buffer = NULL;
+    }
+
+    if( spi_device_transmit( QSPIHandle, (spi_transaction_t*) &SPITransaction ) != ESP_OK ) {
+        GDS_LOG_ERROR( "QSPI transaction failed" );
+        return false;
+    }
+
+    return true;
+}
+
+bool GDS_QSPIBusInit( int MOSIPin, int MISOPin, int CLKPin, int Host ) {
+    spi_bus_config_t BusConfig = { };
+
+    NullCheck( Host >= SPI2_HOST && Host <= SPI3_HOST, return false );
+    
+    // QSPI pin configuration for ESP32-S3
+    BusConfig.mosi_io_num = MOSIPin;
+    BusConfig.miso_io_num = MISOPin;
+    BusConfig.sclk_io_num = CLKPin;
+    BusConfig.quadwp_io_num = -1; // Not used for ILI9488
+    BusConfig.quadhd_io_num = -1; // Not used for ILI9488
+    BusConfig.max_transfer_sz = 65536;
+    BusConfig.flags = SPICOMMON_BUSFLAG_MASTER | SPICOMMON_BUSFLAG_NATIVE_PINS;
+    
+    if( spi_bus_initialize( Host, &BusConfig, DMA_CH_AUTO ) != ESP_OK ) {
+        GDS_LOG_ERROR( "Failed to initialize QSPI bus" );
+        return false;
+    }
+
+    return true;
+}

components/display/display.c

@@ -73,6 +73,7 @@ static const char *known_drivers[] = {"SH1106",
 		"ST7735",
 		"ST7789",
 		"ILI9341",
+		"ILI9488",
 		NULL
 	};
     
@@ -80,8 +81,8 @@ static void displayer_task(void *args);
 static void display_sleep(void);
 
 struct GDS_Device *display;   
-extern GDS_DetectFunc SSD1306_Detect, SSD132x_Detect, SH1106_Detect, SH1122_Detect, SSD1675_Detect, SSD1322_Detect, SSD1351_Detect, ST77xx_Detect, ILI9341_Detect;
-GDS_DetectFunc *drivers[] = { SH1106_Detect, SH1122_Detect, SSD1306_Detect, SSD132x_Detect, SSD1675_Detect, SSD1322_Detect, SSD1351_Detect, ST77xx_Detect, ILI9341_Detect, NULL };
+extern GDS_DetectFunc SSD1306_Detect, SSD132x_Detect, SH1106_Detect, SH1122_Detect, SSD1675_Detect, SSD1322_Detect, SSD1351_Detect, ST77xx_Detect, ILI9341_Detect, ILI9488_Detect;
+GDS_DetectFunc *drivers[] = { SH1106_Detect, SH1122_Detect, SSD1306_Detect, SSD132x_Detect, SSD1675_Detect, SSD1322_Detect, SSD1351_Detect, ST77xx_Detect, ILI9341_Detect, ILI9488_Detect, NULL };
 
 /****************************************************************************************
  * 
@@ -134,6 +135,31 @@ void display_init(char *welcome) {
 			GDS_SPIAttachDevice( display, width, height, CS_pin, RST_pin, backlight_pin, speed, mode );
 				
 			ESP_LOGI(TAG, "Display is SPI host %u with cs:%d", spi_system_host, CS_pin);
+		} else if (strcasestr(config, "QSPI") && spi_system_host != -1) {
+			int CS_pin = -1, speed = 0, mode = 0;
+		
+			PARSE_PARAM(config, "cs", '=', CS_pin);
+			PARSE_PARAM(config, "speed", '=', speed);
+			PARSE_PARAM(config, "mode", '=', mode);
+			
+			// Parse QSPI data pins
+			int data_pins[4] = {-1, -1, -1, -1};
+			char *data_str = strstr(config, "data=");
+			if (data_str) {
+				sscanf(data_str + 5, "%d,%d,%d,%d", &data_pins[0], &data_pins[1], &data_pins[2], &data_pins[3]);
+			}
+		
+			init = true;
+			
+			// Initialize QSPI bus with data pins
+			if (data_pins[0] != -1 && data_pins[1] != -1 && data_pins[2] != -1 && data_pins[3] != -1) {
+				GDS_QSPIBusInit( data_pins[0], data_pins[1], spi_system_clk_gpio, spi_system_host );
+			}
+			
+			GDS_QSPIInit( spi_system_host, spi_system_dc_gpio );
+			GDS_QSPIAttachDevice( display, width, height, CS_pin, RST_pin, backlight_pin, speed, mode );
+				
+			ESP_LOGI(TAG, "Display is QSPI host %u with cs:%d", spi_system_host, CS_pin);
 		} else {
 			display = NULL;
 			ESP_LOGI(TAG, "Unsupported display interface or serial link not configured");

components/squeezelite/external/dac_external.c

@@ -51,6 +51,19 @@ static const struct {
 			{\"reg\":26,\"val\":0}, {\"reg\":27,\"val\":0}, {\"reg\":25,\"val\":50}, {\"reg\":38,\"val\":0},		\
 			{\"reg\":39,\"val\":184}, {\"reg\":42,\"val\":184}, {\"reg\":46,\"val\":30}, {\"reg\":47,\"val\":30},	\
 			{\"reg\":48,\"val\":30}, {\"reg\":49,\"val\":30}, {\"reg\":2,\"val\":170}]}" },
+    { "es8311", true,
+		"{\"init\":[																						\
+			{\"reg\":1,\"val\":48}, {\"reg\":2,\"val\":0}, {\"reg\":3,\"val\":16}, {\"reg\":22,\"val\":36},			\
+			{\"reg\":4,\"val\":16}, {\"reg\":5,\"val\":0}, {\"reg\":11,\"val\":0}, {\"reg\":12,\"val\":0},			\
+			{\"reg\":16,\"val\":31}, {\"reg\":17,\"val\":127}, {\"reg\":0,\"val\":128}, {\"reg\":0,\"val\":128},	\
+			{\"reg\":1,\"val\":63}, {\"reg\":1,\"val\":63}, {\"reg\":2,\"val\":0}, {\"reg\":5,\"val\":0},			\
+			{\"reg\":3,\"val\":16}, {\"reg\":4,\"val\":16}, {\"reg\":7,\"val\":0}, {\"reg\":8,\"val\":255},			\
+			{\"reg\":6,\"val\":3}, {\"reg\":1,\"val\":63}, {\"reg\":6,\"val\":3}, {\"reg\":19,\"val\":16},			\
+			{\"reg\":27,\"val\":10}, {\"reg\":28,\"val\":106}, {\"reg\":9,\"val\":12}, {\"reg\":10,\"val\":12},		\
+			{\"reg\":9,\"val\":12}, {\"reg\":10,\"val\":12}, {\"reg\":50,\"val\":178}, {\"reg\":9,\"val\":12},		\
+			{\"reg\":10,\"val\":12}, {\"reg\":23,\"val\":191}, {\"reg\":14,\"val\":2} ,{\"reg\":18,\"val\":0},		\
+			{\"reg\":20,\"val\":26}, {\"reg\":20,\"val\":26}, {\"reg\":13,\"val\":1}, {\"reg\":21,\"val\":64},		\
+			{\"reg\":55,\"val\":72}, {\"reg\":69,\"val\":0}, {\"reg\":50,\"val\":200}]}" },            
 	{ NULL, false, NULL }		
 };
 

components/targets/CMakeLists.txt

@@ -1,4 +1,4 @@
-idf_component_register( SRC_DIRS . muse
+idf_component_register( SRC_DIRS . muse guition
 			INCLUDE_DIRS .
 					    PRIV_REQUIRES services
 )

components/targets/guition/CMakeLists.txt

@@ -0,0 +1,5 @@
+idf_component_register(
+    SRCS "guition.c"
+    INCLUDE_DIRS "."
+    REQUIRES driver esp_common
+)

components/targets/guition/guition.c

@@ -0,0 +1,43 @@
+/*
+ * Guition JC4827W543C board support for squeezelite-esp32
+ *
+ * (c) Guition Board Support 2026
+ *      Based on squeezelite-esp32 architecture
+ *
+ *  This software is released under the MIT License.
+ *  https://opensource.org/licenses/MIT
+ */
+
+#include <string.h>
+#include <esp_log.h>
+#include <esp_types.h>
+#include <esp_system.h>
+#include <freertos/FreeRTOS.h>
+#include <freertos/task.h>
+#include "globdefs.h"
+#include "monitor.h"
+#include "targets.h"
+
+static const char TAG[] = "guition";
+
+static bool init(void);
+
+const struct target_s target_guition = { 
+    .model = "guition", 
+    .init = init 
+};
+
+static bool init(void) { 
+    ESP_LOGI(TAG, "Initializing Guition JC4827W543C board");
+    ESP_LOGI(TAG, "Board features:");
+    ESP_LOGI(TAG, "  - ESP32-S3-WROOM-1 processor");
+    ESP_LOGI(TAG, "  - 4.3\" ILI9488 display (480x272)");
+    ESP_LOGI(TAG, "  - GT911 capacitive touch");
+    ESP_LOGI(TAG, "  - QSPI display interface");
+    ESP_LOGI(TAG, "  - I2C touch interface (GPIO8/4)");
+    
+    // No board-specific hardware initialization required
+    // All configuration is handled through NVS parameters
+    
+    return true;
+}

components/targets/targets.c

@@ -1,7 +1,7 @@
 #include "string.h"
 #include "targets.h"
 
-const struct target_s *target_set[] = { &target_muse, NULL };
+const struct target_s *target_set[] = { &target_muse, &target_guition, NULL };
 
 void target_init(char *target) { 
 	for (int i = 0; *target && target_set[i]; i++) if (strcasestr(target_set[i]->model, target)) {

components/targets/targets.h

@@ -20,3 +20,4 @@ struct target_s {
 };
 
 extern const struct target_s target_muse;
+extern const struct target_s target_guition;

components/wifi-manager/webapp/dist/src/js/test.d.ts

@@ -80,6 +80,8 @@ declare function getStatus(): {};
 declare function getStatus(): {};
 declare function getStatus(): {};
 declare function getStatus(): {};
+declare function getStatus(): {};
+declare function getRadioButton(entry: any): string;
 declare function getRadioButton(entry: any): string;
 declare function getRadioButton(entry: any): string;
 declare function getRadioButton(entry: any): string;
@@ -244,6 +246,7 @@ declare function pushStatus(): void;
 declare function pushStatus(): void;
 declare function pushStatus(): void;
 declare function pushStatus(): void;
+declare function pushStatus(): void;
 declare let sd: {};
 declare let rf: boolean;
 declare function refreshStatus(): void;

components/wifi-manager/webapp/webapp.cmake

@@ -1,5 +1,5 @@
 target_add_binary_data( __idf_wifi-manager webapp/dist/css/index.3b0bbfde52d921a84f9b.css.gz BINARY)
 target_add_binary_data( __idf_wifi-manager webapp/dist/favicon-32x32.png BINARY)
 target_add_binary_data( __idf_wifi-manager webapp/dist/index.html.gz BINARY)
-target_add_binary_data( __idf_wifi-manager webapp/dist/js/index.3fded7.bundle.js.gz BINARY)
-target_add_binary_data( __idf_wifi-manager webapp/dist/js/node_vendors.3fded7.bundle.js.gz BINARY)
+target_add_binary_data( __idf_wifi-manager webapp/dist/js/index.d35fda.bundle.js.gz BINARY)
+target_add_binary_data( __idf_wifi-manager webapp/dist/js/node_vendors.d35fda.bundle.js.gz BINARY)

components/wifi-manager/webapp/webpack.c

@@ -6,29 +6,29 @@ extern const uint8_t _favicon_32x32_png_start[] asm("_binary_favicon_32x32_png_s
 extern const uint8_t _favicon_32x32_png_end[] asm("_binary_favicon_32x32_png_end");
 extern const uint8_t _index_html_gz_start[] asm("_binary_index_html_gz_start");
 extern const uint8_t _index_html_gz_end[] asm("_binary_index_html_gz_end");
-extern const uint8_t _index_3fded7_bundle_js_gz_start[] asm("_binary_index_3fded7_bundle_js_gz_start");
-extern const uint8_t _index_3fded7_bundle_js_gz_end[] asm("_binary_index_3fded7_bundle_js_gz_end");
-extern const uint8_t _node_vendors_3fded7_bundle_js_gz_start[] asm("_binary_node_vendors_3fded7_bundle_js_gz_start");
-extern const uint8_t _node_vendors_3fded7_bundle_js_gz_end[] asm("_binary_node_vendors_3fded7_bundle_js_gz_end");
+extern const uint8_t _index_d35fda_bundle_js_gz_start[] asm("_binary_index_d35fda_bundle_js_gz_start");
+extern const uint8_t _index_d35fda_bundle_js_gz_end[] asm("_binary_index_d35fda_bundle_js_gz_end");
+extern const uint8_t _node_vendors_d35fda_bundle_js_gz_start[] asm("_binary_node_vendors_d35fda_bundle_js_gz_start");
+extern const uint8_t _node_vendors_d35fda_bundle_js_gz_end[] asm("_binary_node_vendors_d35fda_bundle_js_gz_end");
 const char * resource_lookups[] = {
 	"/css/index.3b0bbfde52d921a84f9b.css.gz",
 	"/favicon-32x32.png",
 	"/index.html.gz",
-	"/js/index.3fded7.bundle.js.gz",
-	"/js/node_vendors.3fded7.bundle.js.gz",
+	"/js/index.d35fda.bundle.js.gz",
+	"/js/node_vendors.d35fda.bundle.js.gz",
 ""
 };
 const uint8_t * resource_map_start[] = {
 	_index_3b0bbfde52d921a84f9b_css_gz_start,
 	_favicon_32x32_png_start,
 	_index_html_gz_start,
-	_index_3fded7_bundle_js_gz_start,
-	_node_vendors_3fded7_bundle_js_gz_start
+	_index_d35fda_bundle_js_gz_start,
+	_node_vendors_d35fda_bundle_js_gz_start
 };
 const uint8_t * resource_map_end[] = {
 	_index_3b0bbfde52d921a84f9b_css_gz_end,
 	_favicon_32x32_png_end,
 	_index_html_gz_end,
-	_index_3fded7_bundle_js_gz_end,
-	_node_vendors_3fded7_bundle_js_gz_end
+	_index_d35fda_bundle_js_gz_end,
+	_node_vendors_d35fda_bundle_js_gz_end
 };

components/wifi-manager/webapp/webpack.h

@@ -1,6 +1,6 @@
 /***********************************
 webpack_headers
-dist/css/index.3b0bbfde52d921a84f9b.css.gz,dist/favicon-32x32.png,dist/index.html.gz,dist/js/index.3fded7.bundle.js.gz,dist/js/node_vendors.3fded7.bundle.js.gz
+dist/css/index.3b0bbfde52d921a84f9b.css.gz,dist/favicon-32x32.png,dist/index.html.gz,dist/js/index.d35fda.bundle.js.gz,dist/js/node_vendors.d35fda.bundle.js.gz
 ***********************************/
 #pragma once
 #include <inttypes.h>

include/platform_esp32.h

@@ -0,0 +1,28 @@
+/* 
+ *  Squeezelite for esp32
+ *
+ *  (c) Sebastien 2019
+ *      Philippe G. 2019, philippe_44@outlook.com
+ *
+ *  This software is released under the MIT License.
+ *  https://opensource.org/licenses/MIT
+ *
+ */
+ 
+#pragma once
+
+#include "esp_pthread.h"
+#ifndef CONFIG_SQUEEZELITE_ESP32_RELEASE_URL
+#define CONFIG_SQUEEZELITE_ESP32_RELEASE_URL "https://github.com/sle118/squeezelite-esp32/releases"
+#endif
+
+extern  bool wait_for_wifi();
+extern void console_start();
+extern pthread_cond_t wifi_connect_suspend_cond;
+extern pthread_t wifi_connect_suspend_mutex;
+typedef enum {
+	INFO,
+	WARNING,
+	ERROR
+} message_severity_t;
+extern void set_status_message(message_severity_t severity, const char * message);

lib/_override/CMakeLists.txt

@@ -0,0 +1,24 @@
+if(IDF_TARGET STREQUAL esp32 AND IDF_VERSION_MAJOR EQUAL 4 AND IDF_VERSION_MINOR LESS 4)
+    set(lib_dir ${build_dir}/esp-idf)
+    set(driver esp32/i2s.c)
+    string(REPLACE ".c" ".c.obj" driver_obj "${driver}")
+
+    idf_component_register( SRCS ${driver}
+                        REQUIRES driver           
+                        INCLUDE_DIRS ${IDF_PATH}/components/driver
+                        PRIV_INCLUDE_DIRS ${IDF_PATH}/components/driver/include/driver
+    )
+
+    # CMake is just a pile of crap
+    message(STATUS "!! overriding ${driver} !!")
+    message(STATUS "CAREFUL, LIBRARIES STRIPPING FROM DUPLICATED COMPONENTS DEPENDS ON THIS BEING REBUILD")
+
+    add_custom_command(
+                TARGET ${COMPONENT_LIB}
+                PRE_LINK
+                COMMAND xtensa-esp32-elf-ar -d ${lib_dir}/driver/libdriver.a ${driver_obj}
+                VERBATIM
+    )
+else()
+    message(STATUS "==> NO OVERRIDE <==")
+endif()

lib/audio/CMakeLists.txt

@@ -0,0 +1,10 @@
+idf_component_register(	SRC_DIRS .
+						INCLUDE_DIRS .  inc 
+						)
+add_prebuilt_library(esp_processing lib/libesp_processing.a)
+target_link_libraries(${COMPONENT_LIB} PRIVATE esp_processing)
+#target_link_libraries(${COMPONENT_LIB} INTERFACE "-Wl,--undefined=")
+target_link_libraries(${COMPONENT_LIB} INTERFACE "-u pow")
+target_link_libraries(${COMPONENT_LIB} INTERFACE "-u cos")
+target_link_libraries(${COMPONENT_LIB} INTERFACE "-u sin")
+target_link_libraries(${COMPONENT_LIB} INTERFACE "-u sqrt")

lib/audio/component.mk

@@ -0,0 +1,10 @@
+#
+# "main" pseudo-component makefile.
+#
+# (Uses default behaviour of compiling all source files in directory, adding 'include' to include path.)
+COMPONENT_ADD_LDFLAGS=-l$(COMPONENT_NAME) 	\
+	$(COMPONENT_PATH)/lib/libesp_processing.a
+	
+
+	
+	

lib/audio/inc/esp_equalizer.h

@@ -0,0 +1,78 @@
+// Copyright 2018 Espressif Systems (Shanghai) PTE LTD
+// All rights reserved.
+
+#ifndef _ESP_EQUALIZER_H
+#define _ESP_EQUALIZER_H
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif
+
+/**
+* @brief      Initialize the equalizer handle
+*
+* @param      nch     The audio channel number
+* @param      g_rate  The audio sample rate. Four sample rates are supported: 11025Hz, 22050Hz, 44100Hz and 48000Hz.
+* @param      n_band  The number of audio sub-bands. Fixed number of 10 sub-bands is supported and this value should be set to 10.
+* @param      use_xmms_original_freqs  Currently should be set 0
+*
+* @return     The equalizer handle.
+*/
+void *esp_equalizer_init(int nch, int g_rate, int n_band, int use_xmms_original_freqs);
+
+/**
+* @brief     Uninitialize the equalizer handle.
+*
+* @param      handle   The the equalizer handle
+*/
+void esp_equalizer_uninit(void *handle);
+
+/**
+* @brief      Process the data through the equalizer
+*
+* @param      handle   The the equalizer handle
+* @param      pcm_buf  The audio pcm input & output buffer
+* @param      length   The length of current bytes in pcm_buf
+* @param      g_rate   The audio sample rate. Four sample rates are supported: 11025Hz, 22050Hz, 44100Hz and 48000Hz.
+* @param      nch      The audio channel number
+*
+* @return     Length of pcm_buf after processing
+*/
+int esp_equalizer_process(void *handle, unsigned char *pcm_buf, int length, int g_rate, int nch);
+
+/**
+* @brief      Set the number of sub-bands for the equalizer
+*
+* @param      handle  The the equalizer handle
+* @param      value   The audio sub-bands gain. unit:db. 0 means no gain.
+* @param      index   The index of audio sub-bands. e.g. 0, 1, 2, 3, 4, 5, 6, 7, 8, 9.
+* @param      nch     The audio channel number
+*/
+void esp_equalizer_set_band_value(void *handle, float value, int index, int nch);
+
+/**
+* @brief      Get the number of the equalizer sub-bands
+*
+* @param      handle   The the equalizer handle
+*
+* @return     The number of the equalizer sub-bands
+*/
+int esp_equalizer_get_band_count(void *handle);
+
+/**
+* @brief      Get the value of the equalizer sub-bands
+*
+* @param      handle  The the equalizer handle
+* @param      index   The index of audio sub-bands. Currently only support 10 sub-bands, so it should be 0-9.
+* @param      nch     The audio channel number
+*
+* @return     The number of the equalizer sub-bands
+*/
+float esp_equalizer_get_band_value(void *handle, int index, int nch);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

lib/audio/link_helper.c

@@ -0,0 +1,3 @@
+void dummy_obj() {
+	return;
+}

lib/codecs/CMakeLists.txt

@@ -0,0 +1,26 @@
+idf_component_register(
+	   INCLUDE_DIRS . ./inc inc/alac inc/helix-aac inc/mad inc/resample16 inc/soxr inc/vorbis inc/opus
+)
+
+if (DEFINED AAC_DISABLE_SBR)
+add_prebuilt_library(libhelix-aac 	lib/libhelix-aac.a ) 
+else ()
+add_prebuilt_library(libhelix-aac 	lib/libhelix-aac-sbr.a ) 
+endif()
+
+add_prebuilt_library(libmad 		lib/libmad.a)
+add_prebuilt_library(libFLAC 		lib/libFLAC.a ) 
+add_prebuilt_library(libvorbisidec 	lib/libvorbisidec.a ) 
+add_prebuilt_library(libogg 		lib/libogg.a )
+add_prebuilt_library(libalac 		lib/libalac.a ) 
+add_prebuilt_library(libresample16 	lib/libresample16.a ) 
+add_prebuilt_library(libopus 		lib/libopus.a ) 
+
+target_link_libraries(${COMPONENT_LIB} INTERFACE libmad)
+target_link_libraries(${COMPONENT_LIB} INTERFACE libFLAC)
+target_link_libraries(${COMPONENT_LIB} INTERFACE libhelix-aac)
+target_link_libraries(${COMPONENT_LIB} INTERFACE libvorbisidec)
+target_link_libraries(${COMPONENT_LIB} INTERFACE libogg)
+target_link_libraries(${COMPONENT_LIB} INTERFACE libalac)
+target_link_libraries(${COMPONENT_LIB} INTERFACE libresample16)
+target_link_libraries(${COMPONENT_LIB} INTERFACE libopus)

lib/codecs/component.mk

@@ -0,0 +1,24 @@
+#
+# "main" pseudo-component makefile.
+#
+# (Uses default behaviour of compiling all source files in directory, adding 'include' to include path.)
+COMPONENT_ADD_LDFLAGS=-l$(COMPONENT_NAME) 	\
+	$(COMPONENT_PATH)/lib/libmad.a 			\
+	$(COMPONENT_PATH)/lib/libFLAC.a			\
+	$(COMPONENT_PATH)/lib/libhelix-aac.a 	\
+	$(COMPONENT_PATH)/lib/libvorbisidec.a	\
+	$(COMPONENT_PATH)/lib/libogg.a			\
+	$(COMPONENT_PATH)/lib/libalac.a			\
+	$(COMPONENT_PATH)/lib/libresample16.a	\
+	$(COMPONENT_PATH)/lib/libopusfile.a		\
+	$(COMPONENT_PATH)/lib/libopus.a 		
+	
+	#$(COMPONENT_PATH)/lib/libFLAC.a
+	#$(COMPONENT_PATH)/lib/libesp-flac.a
+	#$(COMPONENT_PATH)/lib/libsoxr.a		
+	#$(COMPONENT_PATH)/lib/libfaad.a 	
+	#$(COMPONENT_PATH)/lib/libvorbisidec.a
+	#$(COMPONENT_PATH)/lib/libesp-opus.a 
+	#$(COMPONENT_PATH)/lib/libogg.a
+	#$(COMPONENT_PATH)/lib/libesp-tremor.a
+	#$(COMPONENT_PATH)/lib/libesp-ogg-container.a

lib/codecs/inc/FLAC/all.h

@@ -0,0 +1,450 @@
+/* libFLAC - Free Lossless Audio Codec library
+ * Copyright (C) 2000-2009  Josh Coalson
+ * Copyright (C) 2011-2022  Xiph.Org Foundation
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * - Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * - Neither the name of the Xiph.org Foundation nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+ * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef FLAC__ALL_H
+#define FLAC__ALL_H
+
+#include "export.h"
+
+#include "assert.h"
+#include "callback.h"
+#include "format.h"
+#include "metadata.h"
+#include "ordinals.h"
+#include "stream_decoder.h"
+#include "stream_encoder.h"
+
+/** \mainpage
+ *
+ * \section intro Introduction
+ *
+ * This is the documentation for the FLAC C and C++ APIs.  It is
+ * highly interconnected; this introduction should give you a top
+ * level idea of the structure and how to find the information you
+ * need.  As a prerequisite you should have at least a basic
+ * knowledge of the FLAC format, documented
+ * <A HREF="https://xiph.org/flac/format.html">here</A>.
+ *
+ * \section c_api FLAC C API
+ *
+ * The FLAC C API is the interface to libFLAC, a set of structures
+ * describing the components of FLAC streams, and functions for
+ * encoding and decoding streams, as well as manipulating FLAC
+ * metadata in files.  The public include files will be installed
+ * in your include area (for example /usr/include/FLAC/...).
+ *
+ * By writing a little code and linking against libFLAC, it is
+ * relatively easy to add FLAC support to another program.  The
+ * library is licensed under <A HREF="https://xiph.org/flac/license.html">Xiph's BSD license</A>.
+ * Complete source code of libFLAC as well as the command-line
+ * encoder and plugins is available and is a useful source of
+ * examples.
+ *
+ * Aside from encoders and decoders, libFLAC provides a powerful
+ * metadata interface for manipulating metadata in FLAC files.  It
+ * allows the user to add, delete, and modify FLAC metadata blocks
+ * and it can automatically take advantage of PADDING blocks to avoid
+ * rewriting the entire FLAC file when changing the size of the
+ * metadata.
+ *
+ * libFLAC usually only requires the standard C library and C math
+ * library. In particular, threading is not used so there is no
+ * dependency on a thread library. However, libFLAC does not use
+ * global variables and should be thread-safe.
+ *
+ * libFLAC also supports encoding to and decoding from Ogg FLAC.
+ * However the metadata editing interfaces currently have limited
+ * read-only support for Ogg FLAC files.
+ *
+ * \section cpp_api FLAC C++ API
+ *
+ * The FLAC C++ API is a set of classes that encapsulate the
+ * structures and functions in libFLAC.  They provide slightly more
+ * functionality with respect to metadata but are otherwise
+ * equivalent.  For the most part, they share the same usage as
+ * their counterparts in libFLAC, and the FLAC C API documentation
+ * can be used as a supplement.  The public include files
+ * for the C++ API will be installed in your include area (for
+ * example /usr/include/FLAC++/...).
+ *
+ * libFLAC++ is also licensed under
+ * <A HREF="https://xiph.org/flac/license.html">Xiph's BSD license</A>.
+ *
+ * \section getting_started Getting Started
+ *
+ * A good starting point for learning the API is to browse through
+ * the <A HREF="modules.html">modules</A>.  Modules are logical
+ * groupings of related functions or classes, which correspond roughly
+ * to header files or sections of header files.  Each module includes a
+ * detailed description of the general usage of its functions or
+ * classes.
+ *
+ * From there you can go on to look at the documentation of
+ * individual functions.  You can see different views of the individual
+ * functions through the links in top bar across this page.
+ *
+ * If you prefer a more hands-on approach, you can jump right to some
+ * <A HREF="https://xiph.org/flac/documentation_example_code.html">example code</A>.
+ *
+ * \section porting_guide Porting Guide
+ *
+ * Starting with FLAC 1.1.3 a \link porting Porting Guide \endlink
+ * has been introduced which gives detailed instructions on how to
+ * port your code to newer versions of FLAC.
+ *
+ * \section embedded_developers Embedded Developers
+ *
+ * libFLAC has grown larger over time as more functionality has been
+ * included, but much of it may be unnecessary for a particular embedded
+ * implementation.  Unused parts may be pruned by some simple editing of
+ * src/libFLAC/Makefile.am.  In general, the decoders, encoders, and
+ * metadata interface are all independent from each other.
+ *
+ * It is easiest to just describe the dependencies:
+ *
+ * - All modules depend on the \link flac_format Format \endlink module.
+ * - The decoders and encoders depend on the bitbuffer.
+ * - The decoder is independent of the encoder.  The encoder uses the
+ *   decoder because of the verify feature, but this can be removed if
+ *   not needed.
+ * - Parts of the metadata interface require the stream decoder (but not
+ *   the encoder).
+ * - Ogg support is selectable through the compile time macro
+ *   \c FLAC__HAS_OGG.
+ *
+ * For example, if your application only requires the stream decoder, no
+ * encoder, and no metadata interface, you can remove the stream encoder
+ * and the metadata interface, which will greatly reduce the size of the
+ * library.
+ *
+ * Also, there are several places in the libFLAC code with comments marked
+ * with "OPT:" where a \#define can be changed to enable code that might be
+ * faster on a specific platform.  Experimenting with these can yield faster
+ * binaries.
+ */
+
+/** \defgroup porting Porting Guide for New Versions
+ *
+ * This module describes differences in the library interfaces from
+ * version to version.  It assists in the porting of code that uses
+ * the libraries to newer versions of FLAC.
+ *
+ * One simple facility for making porting easier that has been added
+ * in FLAC 1.1.3 is a set of \#defines in \c export.h of each
+ * library's includes (e.g. \c include/FLAC/export.h).  The
+ * \#defines mirror the libraries'
+ * <A HREF="http://www.gnu.org/software/libtool/manual/libtool.html#Libtool-versioning">libtool version numbers</A>,
+ * e.g. in libFLAC there are \c FLAC_API_VERSION_CURRENT,
+ * \c FLAC_API_VERSION_REVISION, and \c FLAC_API_VERSION_AGE.
+ * These can be used to support multiple versions of an API during the
+ * transition phase, e.g.
+ *
+ * \code
+ * #if !defined(FLAC_API_VERSION_CURRENT) || FLAC_API_VERSION_CURRENT <= 7
+ *   legacy code
+ * #else
+ *   new code
+ * #endif
+ * \endcode
+ *
+ * The source will work for multiple versions and the legacy code can
+ * easily be removed when the transition is complete.
+ *
+ * Another available symbol is FLAC_API_SUPPORTS_OGG_FLAC (defined in
+ * include/FLAC/export.h), which can be used to determine whether or not
+ * the library has been compiled with support for Ogg FLAC.  This is
+ * simpler than trying to call an Ogg init function and catching the
+ * error.
+ */
+
+/** \defgroup porting_1_1_2_to_1_1_3 Porting from FLAC 1.1.2 to 1.1.3
+ *  \ingroup porting
+ *
+ *  \brief
+ *  This module describes porting from FLAC 1.1.2 to FLAC 1.1.3.
+ *
+ * The main change between the APIs in 1.1.2 and 1.1.3 is that they have
+ * been simplified.  First, libOggFLAC has been merged into libFLAC and
+ * libOggFLAC++ has been merged into libFLAC++.  Second, both the three
+ * decoding layers and three encoding layers have been merged into a
+ * single stream decoder and stream encoder.  That is, the functionality
+ * of FLAC__SeekableStreamDecoder and FLAC__FileDecoder has been merged
+ * into FLAC__StreamDecoder, and FLAC__SeekableStreamEncoder and
+ * FLAC__FileEncoder into FLAC__StreamEncoder.  Only the
+ * FLAC__StreamDecoder and FLAC__StreamEncoder remain.  What this means
+ * is there is now a single API that can be used to encode or decode
+ * streams to/from native FLAC or Ogg FLAC and the single API can work
+ * on both seekable and non-seekable streams.
+ *
+ * Instead of creating an encoder or decoder of a certain layer, now the
+ * client will always create a FLAC__StreamEncoder or
+ * FLAC__StreamDecoder.  The old layers are now differentiated by the
+ * initialization function.  For example, for the decoder,
+ * FLAC__stream_decoder_init() has been replaced by
+ * FLAC__stream_decoder_init_stream().  This init function takes
+ * callbacks for the I/O, and the seeking callbacks are optional.  This
+ * allows the client to use the same object for seekable and
+ * non-seekable streams.  For decoding a FLAC file directly, the client
+ * can use FLAC__stream_decoder_init_file() and pass just a filename
+ * and fewer callbacks; most of the other callbacks are supplied
+ * internally.  For situations where fopen()ing by filename is not
+ * possible (e.g. Unicode filenames on Windows) the client can instead
+ * open the file itself and supply the FILE* to
+ * FLAC__stream_decoder_init_FILE().  The init functions now returns a
+ * FLAC__StreamDecoderInitStatus instead of FLAC__StreamDecoderState.
+ * Since the callbacks and client data are now passed to the init
+ * function, the FLAC__stream_decoder_set_*_callback() functions and
+ * FLAC__stream_decoder_set_client_data() are no longer needed.  The
+ * rest of the calls to the decoder are the same as before.
+ *
+ * There are counterpart init functions for Ogg FLAC, e.g.
+ * FLAC__stream_decoder_init_ogg_stream().  All the rest of the calls
+ * and callbacks are the same as for native FLAC.
+ *
+ * As an example, in FLAC 1.1.2 a seekable stream decoder would have
+ * been set up like so:
+ *
+ * \code
+ * FLAC__SeekableStreamDecoder *decoder = FLAC__seekable_stream_decoder_new();
+ * if(decoder == NULL) do_something;
+ * FLAC__seekable_stream_decoder_set_md5_checking(decoder, true);
+ * [... other settings ...]
+ * FLAC__seekable_stream_decoder_set_read_callback(decoder, my_read_callback);
+ * FLAC__seekable_stream_decoder_set_seek_callback(decoder, my_seek_callback);
+ * FLAC__seekable_stream_decoder_set_tell_callback(decoder, my_tell_callback);
+ * FLAC__seekable_stream_decoder_set_length_callback(decoder, my_length_callback);
+ * FLAC__seekable_stream_decoder_set_eof_callback(decoder, my_eof_callback);
+ * FLAC__seekable_stream_decoder_set_write_callback(decoder, my_write_callback);
+ * FLAC__seekable_stream_decoder_set_metadata_callback(decoder, my_metadata_callback);
+ * FLAC__seekable_stream_decoder_set_error_callback(decoder, my_error_callback);
+ * FLAC__seekable_stream_decoder_set_client_data(decoder, my_client_data);
+ * if(FLAC__seekable_stream_decoder_init(decoder) != FLAC__SEEKABLE_STREAM_DECODER_OK) do_something;
+ * \endcode
+ *
+ * In FLAC 1.1.3 it is like this:
+ *
+ * \code
+ * FLAC__StreamDecoder *decoder = FLAC__stream_decoder_new();
+ * if(decoder == NULL) do_something;
+ * FLAC__stream_decoder_set_md5_checking(decoder, true);
+ * [... other settings ...]
+ * if(FLAC__stream_decoder_init_stream(
+ *   decoder,
+ *   my_read_callback,
+ *   my_seek_callback,      // or NULL
+ *   my_tell_callback,      // or NULL
+ *   my_length_callback,    // or NULL
+ *   my_eof_callback,       // or NULL
+ *   my_write_callback,
+ *   my_metadata_callback,  // or NULL
+ *   my_error_callback,
+ *   my_client_data
+ * ) != FLAC__STREAM_DECODER_INIT_STATUS_OK) do_something;
+ * \endcode
+ *
+ * or you could do;
+ *
+ * \code
+ * [...]
+ * FILE *file = fopen("somefile.flac","rb");
+ * if(file == NULL) do_somthing;
+ * if(FLAC__stream_decoder_init_FILE(
+ *   decoder,
+ *   file,
+ *   my_write_callback,
+ *   my_metadata_callback,  // or NULL
+ *   my_error_callback,
+ *   my_client_data
+ * ) != FLAC__STREAM_DECODER_INIT_STATUS_OK) do_something;
+ * \endcode
+ *
+ * or just:
+ *
+ * \code
+ * [...]
+ * if(FLAC__stream_decoder_init_file(
+ *   decoder,
+ *   "somefile.flac",
+ *   my_write_callback,
+ *   my_metadata_callback,  // or NULL
+ *   my_error_callback,
+ *   my_client_data
+ * ) != FLAC__STREAM_DECODER_INIT_STATUS_OK) do_something;
+ * \endcode
+ *
+ * Another small change to the decoder is in how it handles unparseable
+ * streams.  Before, when the decoder found an unparseable stream
+ * (reserved for when the decoder encounters a stream from a future
+ * encoder that it can't parse), it changed the state to
+ * \c FLAC__STREAM_DECODER_UNPARSEABLE_STREAM.  Now the decoder instead
+ * drops sync and calls the error callback with a new error code
+ * \c FLAC__STREAM_DECODER_ERROR_STATUS_UNPARSEABLE_STREAM.  This is
+ * more robust.  If your error callback does not discriminate on the the
+ * error state, your code does not need to be changed.
+ *
+ * The encoder now has a new setting:
+ * FLAC__stream_encoder_set_apodization().  This is for setting the
+ * method used to window the data before LPC analysis.  You only need to
+ * add a call to this function if the default is not suitable.   There
+ * are also two new convenience functions that may be useful:
+ * FLAC__metadata_object_cuesheet_calculate_cddb_id() and
+ * FLAC__metadata_get_cuesheet().
+ *
+ * The \a bytes parameter to FLAC__StreamDecoderReadCallback,
+ * FLAC__StreamEncoderReadCallback, and FLAC__StreamEncoderWriteCallback
+ * is now \c size_t instead of \c uint32_t.
+ */
+
+/** \defgroup porting_1_1_3_to_1_1_4 Porting from FLAC 1.1.3 to 1.1.4
+ *  \ingroup porting
+ *
+ *  \brief
+ *  This module describes porting from FLAC 1.1.3 to FLAC 1.1.4.
+ *
+ * There were no changes to any of the interfaces from 1.1.3 to 1.1.4.
+ * There was a slight change in the implementation of
+ * FLAC__stream_encoder_set_metadata(); the function now makes a copy
+ * of the \a metadata array of pointers so the client no longer needs
+ * to maintain it after the call.  The objects themselves that are
+ * pointed to by the array are still not copied though and must be
+ * maintained until the call to FLAC__stream_encoder_finish().
+ */
+
+/** \defgroup porting_1_1_4_to_1_2_0 Porting from FLAC 1.1.4 to 1.2.0
+ *  \ingroup porting
+ *
+ *  \brief
+ *  This module describes porting from FLAC 1.1.4 to FLAC 1.2.0.
+ *
+ * There were only very minor changes to the interfaces from 1.1.4 to 1.2.0.
+ * In libFLAC, \c FLAC__format_sample_rate_is_subset() was added.
+ * In libFLAC++, \c FLAC::Decoder::Stream::get_decode_position() was added.
+ *
+ * Finally, value of the constant \c FLAC__FRAME_HEADER_RESERVED_LEN
+ * has changed to reflect the conversion of one of the reserved bits
+ * into active use.  It used to be \c 2 and now is \c 1.  However the
+ * FLAC frame header length has not changed, so to skip the proper
+ * number of bits, use \c FLAC__FRAME_HEADER_RESERVED_LEN +
+ * \c FLAC__FRAME_HEADER_BLOCKING_STRATEGY_LEN
+ */
+
+/** \defgroup porting_1_3_4_to_1_4_0 Porting from FLAC 1.3.4 to 1.4.0
+ *  \ingroup porting
+ *
+ *  \brief
+ *  This module describes porting from FLAC 1.3.4 to FLAC 1.4.0.
+ *
+ * \section porting_1_3_4_to_1_4_0_summary Summary
+ *
+ * Between FLAC 1.3.4 and FLAC 1.4.0, there have four breaking changes
+ * - the function get_client_data_from_decoder has been renamed to
+ *   FLAC__get_decoder_client_data
+ * - some data types in the FLAC__Frame struct have changed
+ * - all functions resizing metadata blocks now return the object
+ *   untouched if memory allocation fails, whereas previously the
+ *   handling varied and was more or less undefined
+ * - all functions accepting a filename now take UTF-8 encoded filenames
+ *   on Windows instead of filenames in the current codepage
+ *
+ * Furthermore, there have been the following additions
+ * - the functions FLAC__stream_encoder_set_limit_min_bitrate,
+ *   FLAC__stream_encoder_get_limit_min_bitrate,
+ *   FLAC::encoder::file::set_limit_min_bitrate() and
+ *   FLAC::encoder::file::get_limit_min_bitrate() have been added
+ * - Added FLAC__STREAM_DECODER_ERROR_STATUS_BAD_METADATA to the
+ *   FLAC__StreamDecoderErrorStatus enum
+ *
+ * \section porting_1_3_4_to_1_4_0_breaking Breaking changes
+ *
+ * The function \b get_client_data_from_decoder was added in FLAC 1.3.3
+ * but did not follow the API naming convention and was not properly
+ * exported. The function is now renamed and properly integrated as
+ * FLAC__stream_decoder_get_client_data
+ *
+ * To accomodate encoding and decoding 32-bit int PCM, some data types
+ * in the \b FLAC__frame struct were changed. Specifically, warmup
+ * in both the FLAC__Subframe_Fixed struc and the FLAC__Subframe_LPC
+ * struct is changed from FLAC__int32 to FLAC__int64. Also, value
+ * in the FLAC__Subframe_Constant is changed from FLAC__int32 to
+ * FLAC__int64. Finally, in FLAC__Subframe_Verbatim struct data is
+ * changes from a FLAC__int32 array to a union containing a FLAC__int32
+ * array and a FLAC__int64 array. Also, a new member is added,
+ * data_type, which clarifies whether the FLAC__int32 or FLAC__int64
+ * array is in use.
+ *
+ * Furthermore, the following functions now return the object untouched
+ * if memory allocation fails, whereas previously the handling varied
+ * and was more or less undefined
+ *
+ * - FLAC__metadata_object_seektable_resize_points
+ * - FLAC__metadata_object_vorbiscomment_resize_comments
+ * - FLAC__metadata_object_cuesheet_track_resize_indices
+ * - FLAC__metadata_object_cuesheet_resize_tracks
+ *
+ * The last breaking change is that all API functions taking a filename
+ * as an argument now, on Windows, must be supplied with that filename
+ * in the UTF-8 character encoding instead of using the current code
+ * page. libFLAC internally translates these UTF-8 encoded filenames to
+ * an appropriate representation to use with _wfopen. On all other
+ * systems, filename is passed to fopen without any translation, as it
+ * in libFLAC 1.3.4 and earlier.
+ *
+ * \section porting_1_3_4_to_1_4_0_additions Additions
+ *
+ * To aid in creating properly streamable FLAC files, a set of functions
+ * was added to make it possible to enfore a minimum bitrate to files
+ * created through libFLAC's stream_encoder.h interface. With this
+ * function enabled the resulting FLAC files have a minimum bitrate of
+ * 1bit/sample independent of the number of channels, i.e. 48kbit/s for
+ * 48kHz. This can be beneficial for streaming, as very low bitrates for
+ * silent sections compressed with 'constant' subframes can result in a
+ * bitrate of 1kbit/s, creating problems with clients that aren't aware
+ * of this possibility and buffer too much data.
+ *
+ * Finally, FLAC__STREAM_DECODER_ERROR_STATUS_BAD_METADATA was added to
+ * the FLAC__StreamDecoderErrorStatus enum to signal that the decoder
+ * encountered unreadable metadata.
+ *
+ */
+
+/** \defgroup flac FLAC C API
+ *
+ * The FLAC C API is the interface to libFLAC, a set of structures
+ * describing the components of FLAC streams, and functions for
+ * encoding and decoding streams, as well as manipulating FLAC
+ * metadata in files.
+ *
+ * You should start with the format components as all other modules
+ * are dependent on it.
+ */
+
+#endif

lib/codecs/inc/FLAC/assert.h

@@ -0,0 +1,51 @@
+/* libFLAC - Free Lossless Audio Codec library
+ * Copyright (C) 2001-2009  Josh Coalson
+ * Copyright (C) 2011-2022  Xiph.Org Foundation
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * - Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * - Neither the name of the Xiph.org Foundation nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+ * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef FLAC__ASSERT_H
+#define FLAC__ASSERT_H
+
+/* we need this since some compilers (like MSVC) leave assert()s on release code (and we don't want to use their ASSERT) */
+#ifdef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
+#define FLAC__ASSERT(x) if(!(x)) __builtin_abort();
+#define FLAC__ASSERT_DECLARATION(x) x
+#else
+#ifndef NDEBUG
+#include <assert.h>
+#define FLAC__ASSERT(x) assert(x)
+#define FLAC__ASSERT_DECLARATION(x) x
+#else
+#define FLAC__ASSERT(x)
+#define FLAC__ASSERT_DECLARATION(x)
+#endif
+#endif
+
+#endif

lib/codecs/inc/FLAC/callback.h

@@ -0,0 +1,185 @@
+/* libFLAC - Free Lossless Audio Codec library
+ * Copyright (C) 2004-2009  Josh Coalson
+ * Copyright (C) 2011-2022  Xiph.Org Foundation
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * - Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * - Neither the name of the Xiph.org Foundation nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+ * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef FLAC__CALLBACK_H
+#define FLAC__CALLBACK_H
+
+#include "ordinals.h"
+#include <stdlib.h> /* for size_t */
+
+/** \file include/FLAC/callback.h
+ *
+ *  \brief
+ *  This module defines the structures for describing I/O callbacks
+ *  to the other FLAC interfaces.
+ *
+ *  See the detailed documentation for callbacks in the
+ *  \link flac_callbacks callbacks \endlink module.
+ */
+
+/** \defgroup flac_callbacks FLAC/callback.h: I/O callback structures
+ *  \ingroup flac
+ *
+ *  \brief
+ *  This module defines the structures for describing I/O callbacks
+ *  to the other FLAC interfaces.
+ *
+ *  The purpose of the I/O callback functions is to create a common way
+ *  for the metadata interfaces to handle I/O.
+ *
+ *  Originally the metadata interfaces required filenames as the way of
+ *  specifying FLAC files to operate on.  This is problematic in some
+ *  environments so there is an additional option to specify a set of
+ *  callbacks for doing I/O on the FLAC file, instead of the filename.
+ *
+ *  In addition to the callbacks, a FLAC__IOHandle type is defined as an
+ *  opaque structure for a data source.
+ *
+ *  The callback function prototypes are similar (but not identical) to the
+ *  stdio functions fread, fwrite, fseek, ftell, feof, and fclose.  If you use
+ *  stdio streams to implement the callbacks, you can pass fread, fwrite, and
+ *  fclose anywhere a FLAC__IOCallback_Read, FLAC__IOCallback_Write, or
+ *  FLAC__IOCallback_Close is required, and a FILE* anywhere a FLAC__IOHandle
+ *  is required.  \warning You generally CANNOT directly use fseek or ftell
+ *  for FLAC__IOCallback_Seek or FLAC__IOCallback_Tell since on most systems
+ *  these use 32-bit offsets and FLAC requires 64-bit offsets to deal with
+ *  large files.  You will have to find an equivalent function (e.g. ftello),
+ *  or write a wrapper.  The same is true for feof() since this is usually
+ *  implemented as a macro, not as a function whose address can be taken.
+ *
+ * \{
+ */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/** This is the opaque handle type used by the callbacks.  Typically
+ *  this is a \c FILE* or address of a file descriptor.
+ */
+typedef void* FLAC__IOHandle;
+
+/** Signature for the read callback.
+ *  The signature and semantics match POSIX fread() implementations
+ *  and can generally be used interchangeably.
+ *
+ * \param  ptr      The address of the read buffer.
+ * \param  size     The size of the records to be read.
+ * \param  nmemb    The number of records to be read.
+ * \param  handle   The handle to the data source.
+ * \retval size_t
+ *    The number of records read.
+ */
+typedef size_t (*FLAC__IOCallback_Read) (void *ptr, size_t size, size_t nmemb, FLAC__IOHandle handle);
+
+/** Signature for the write callback.
+ *  The signature and semantics match POSIX fwrite() implementations
+ *  and can generally be used interchangeably.
+ *
+ * \param  ptr      The address of the write buffer.
+ * \param  size     The size of the records to be written.
+ * \param  nmemb    The number of records to be written.
+ * \param  handle   The handle to the data source.
+ * \retval size_t
+ *    The number of records written.
+ */
+typedef size_t (*FLAC__IOCallback_Write) (const void *ptr, size_t size, size_t nmemb, FLAC__IOHandle handle);
+
+/** Signature for the seek callback.
+ *  The signature and semantics mostly match POSIX fseek() WITH ONE IMPORTANT
+ *  EXCEPTION: the offset is a 64-bit type whereas fseek() is generally 'long'
+ *  and 32-bits wide.
+ *
+ * \param  handle   The handle to the data source.
+ * \param  offset   The new position, relative to \a whence
+ * \param  whence   \c SEEK_SET, \c SEEK_CUR, or \c SEEK_END
+ * \retval int
+ *    \c 0 on success, \c -1 on error.
+ */
+typedef int (*FLAC__IOCallback_Seek) (FLAC__IOHandle handle, FLAC__int64 offset, int whence);
+
+/** Signature for the tell callback.
+ *  The signature and semantics mostly match POSIX ftell() WITH ONE IMPORTANT
+ *  EXCEPTION: the offset is a 64-bit type whereas ftell() is generally 'long'
+ *  and 32-bits wide.
+ *
+ * \param  handle   The handle to the data source.
+ * \retval FLAC__int64
+ *    The current position on success, \c -1 on error.
+ */
+typedef FLAC__int64 (*FLAC__IOCallback_Tell) (FLAC__IOHandle handle);
+
+/** Signature for the EOF callback.
+ *  The signature and semantics mostly match POSIX feof() but WATCHOUT:
+ *  on many systems, feof() is a macro, so in this case a wrapper function
+ *  must be provided instead.
+ *
+ * \param  handle   The handle to the data source.
+ * \retval int
+ *    \c 0 if not at end of file, nonzero if at end of file.
+ */
+typedef int (*FLAC__IOCallback_Eof) (FLAC__IOHandle handle);
+
+/** Signature for the close callback.
+ *  The signature and semantics match POSIX fclose() implementations
+ *  and can generally be used interchangeably.
+ *
+ * \param  handle   The handle to the data source.
+ * \retval int
+ *    \c 0 on success, \c EOF on error.
+ */
+typedef int (*FLAC__IOCallback_Close) (FLAC__IOHandle handle);
+
+/** A structure for holding a set of callbacks.
+ *  Each FLAC interface that requires a FLAC__IOCallbacks structure will
+ *  describe which of the callbacks are required.  The ones that are not
+ *  required may be set to NULL.
+ *
+ *  If the seek requirement for an interface is optional, you can signify that
+ *  a data source is not seekable by setting the \a seek field to \c NULL.
+ */
+typedef struct {
+	FLAC__IOCallback_Read read;   /**< See FLAC__IOCallbacks */
+	FLAC__IOCallback_Write write; /**< See FLAC__IOCallbacks */
+	FLAC__IOCallback_Seek seek;   /**< See FLAC__IOCallbacks */
+	FLAC__IOCallback_Tell tell;   /**< See FLAC__IOCallbacks */
+	FLAC__IOCallback_Eof eof;     /**< See FLAC__IOCallbacks */
+	FLAC__IOCallback_Close close; /**< See FLAC__IOCallbacks */
+} FLAC__IOCallbacks;
+
+/* \} */
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

lib/codecs/inc/FLAC/export.h

@@ -0,0 +1,115 @@
+/* libFLAC - Free Lossless Audio Codec library
+ * Copyright (C) 2000-2009  Josh Coalson
+ * Copyright (C) 2011-2022  Xiph.Org Foundation
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * - Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * - Neither the name of the Xiph.org Foundation nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+ * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef FLAC__EXPORT_H
+#define FLAC__EXPORT_H
+
+/** \file include/FLAC/export.h
+ *
+ *  \brief
+ *  This module contains \#defines and symbols for exporting function
+ *  calls, and providing version information and compiled-in features.
+ *
+ *  See the \link flac_export export \endlink module.
+ */
+
+/** \defgroup flac_export FLAC/export.h: export symbols
+ *  \ingroup flac
+ *
+ *  \brief
+ *  This module contains \#defines and symbols for exporting function
+ *  calls, and providing version information and compiled-in features.
+ *
+ *  If you are compiling for Windows (with Visual Studio or MinGW for
+ *  example) and will link to the static library (libFLAC++.lib) you
+ *  should define FLAC__NO_DLL in your project to make sure the symbols
+ *  are exported properly.
+ *
+ * \{
+ */
+
+/** This \#define is used internally in libFLAC and its headers to make
+ * sure the correct symbols are exported when working with shared
+ * libraries. On Windows, this \#define is set to __declspec(dllexport)
+ * when compiling libFLAC into a library and to __declspec(dllimport)
+ * when the headers are used to link to that DLL. On non-Windows systems
+ * it is used to set symbol visibility.
+ *
+ * Because of this, the define FLAC__NO_DLL must be defined when linking
+ * to libFLAC statically or linking will fail.
+ */
+/* This has grown quite complicated. FLAC__NO_DLL is used by MSVC sln
+ * files and CMake, which build either static or shared. autotools can
+ * build static, shared or **both**. Therefore, DLL_EXPORT, which is set
+ * by libtool, must override FLAC__NO_DLL on building shared components
+ */
+#if defined(_WIN32)
+
+#if defined(FLAC__NO_DLL) && !(defined(DLL_EXPORT))
+#define FLAC_API
+#else
+#ifdef FLAC_API_EXPORTS
+#define	FLAC_API __declspec(dllexport)
+#else
+#define FLAC_API __declspec(dllimport)
+#endif
+#endif
+
+#elif defined(FLAC__USE_VISIBILITY_ATTR)
+#define FLAC_API __attribute__ ((visibility ("default")))
+
+#else
+#define FLAC_API
+
+#endif
+
+/** These \#defines will mirror the libtool-based library version number, see
+ * http://www.gnu.org/software/libtool/manual/libtool.html#Libtool-versioning
+ */
+#define FLAC_API_VERSION_CURRENT 12
+#define FLAC_API_VERSION_REVISION 0 /**< see above */
+#define FLAC_API_VERSION_AGE 0 /**< see above */
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/** \c 1 if the library has been compiled with support for Ogg FLAC, else \c 0. */
+extern FLAC_API int FLAC_API_SUPPORTS_OGG_FLAC;
+
+#ifdef __cplusplus
+}
+#endif
+
+/* \} */
+
+#endif

lib/codecs/inc/FLAC/ordinals.h

@@ -0,0 +1,55 @@
+/* libFLAC - Free Lossless Audio Codec library
+ * Copyright (C) 2000-2009  Josh Coalson
+ * Copyright (C) 2011-2022  Xiph.Org Foundation
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * - Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * - Neither the name of the Xiph.org Foundation nor the names of its
+ * contributors may be used to endorse or promote products derived from
+ * this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+ * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef FLAC__ORDINALS_H
+#define FLAC__ORDINALS_H
+
+/* This of course assumes C99 headers */
+
+#include <stdint.h>
+#include <stdbool.h>
+
+typedef int8_t FLAC__int8;
+typedef uint8_t FLAC__uint8;
+
+typedef int16_t FLAC__int16;
+typedef int32_t FLAC__int32;
+typedef int64_t FLAC__int64;
+typedef uint16_t FLAC__uint16;
+typedef uint32_t FLAC__uint32;
+typedef uint64_t FLAC__uint64;
+
+typedef int FLAC__bool;
+
+typedef FLAC__uint8 FLAC__byte;
+
+#endif

lib/codecs/inc/alac/alac_wrapper.h

@@ -0,0 +1,31 @@
+/*****************************************************************************
+ * alac_wrapper.h: ALAC coder wrapper
+ *
+ * (c) Philippe G. 2019, philippe_44@outlook.com
+ *
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ * 
+ */
+ 
+#ifndef __ALAC_WRAPPER_H_
+#define __ALAC_WRAPPER_H_
+
+struct alac_codec_s;
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+struct alac_codec_s *alac_create_decoder(int magic_cookie_size, unsigned char *magic_cookie,
+								unsigned char *sample_size, unsigned *sample_rate,
+								unsigned char *channels, unsigned int *block_size);
+void alac_delete_decoder(struct alac_codec_s *codec);
+bool alac_to_pcm(struct alac_codec_s *codec, unsigned char* input,
+				 unsigned char *output, char channels, unsigned *out_frames);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

lib/codecs/inc/faad2/faad.h

@@ -0,0 +1,35 @@
+/*
+** FAAD2 - Freeware Advanced Audio (AAC) Decoder including SBR decoding
+** Copyright (C) 2003-2005 M. Bakker, Nero AG, http://www.nero.com
+**  
+** This program is free software; you can redistribute it and/or modify
+** it under the terms of the GNU General Public License as published by
+** the Free Software Foundation; either version 2 of the License, or
+** (at your option) any later version.
+** 
+** This program is distributed in the hope that it will be useful,
+** but WITHOUT ANY WARRANTY; without even the implied warranty of
+** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+** GNU General Public License for more details.
+** 
+** You should have received a copy of the GNU General Public License
+** along with this program; if not, write to the Free Software 
+** Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+**
+** Any non-GPL usage of this software or parts of this software is strictly
+** forbidden.
+**
+** The "appropriate copyright message" mentioned in section 2c of the GPLv2
+** must read: "Code from FAAD2 is copyright (c) Nero AG, www.nero.com"
+**
+** Commercial non-GPL licensing of this software is possible.
+** For more info contact Nero AG through Mpeg4AAClicense@nero.com.
+**
+** $Id: faad.h,v 1.51 2007/11/01 12:33:29 menno Exp $
+**/
+
+/* warn people for update */
+#pragma message("please update faad2 include filename and function names!")
+
+/* Backwards compatible link */
+#include "neaacdec.h"

lib/codecs/inc/faad2/neaacdec.h

@@ -0,0 +1,258 @@
+/*
+** FAAD2 - Freeware Advanced Audio (AAC) Decoder including SBR decoding
+** Copyright (C) 2003-2005 M. Bakker, Nero AG, http://www.nero.com
+**  
+** This program is free software; you can redistribute it and/or modify
+** it under the terms of the GNU General Public License as published by
+** the Free Software Foundation; either version 2 of the License, or
+** (at your option) any later version.
+** 
+** This program is distributed in the hope that it will be useful,
+** but WITHOUT ANY WARRANTY; without even the implied warranty of
+** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+** GNU General Public License for more details.
+** 
+** You should have received a copy of the GNU General Public License
+** along with this program; if not, write to the Free Software 
+** Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+**
+** Any non-GPL usage of this software or parts of this software is strictly
+** forbidden.
+**
+** The "appropriate copyright message" mentioned in section 2c of the GPLv2
+** must read: "Code from FAAD2 is copyright (c) Nero AG, www.nero.com"
+**
+** Commercial non-GPL licensing of this software is possible.
+** For more info contact Nero AG through Mpeg4AAClicense@nero.com.
+**
+** $Id: neaacdec.h,v 1.13 2009/01/26 23:51:15 menno Exp $
+**/
+
+#ifndef __NEAACDEC_H__
+#define __NEAACDEC_H__
+
+#ifdef __cplusplus
+extern "C" {
+#endif /* __cplusplus */
+
+
+#if 1
+/* MACROS FOR BACKWARDS COMPATIBILITY */
+/* structs */
+#define faacDecHandle                  NeAACDecHandle
+#define faacDecConfiguration           NeAACDecConfiguration
+#define faacDecConfigurationPtr        NeAACDecConfigurationPtr
+#define faacDecFrameInfo               NeAACDecFrameInfo
+/* functions */
+#define faacDecGetErrorMessage         NeAACDecGetErrorMessage
+#define faacDecSetConfiguration        NeAACDecSetConfiguration
+#define faacDecGetCurrentConfiguration NeAACDecGetCurrentConfiguration
+#define faacDecInit                    NeAACDecInit
+#define faacDecInit2                   NeAACDecInit2
+#define faacDecInitDRM                 NeAACDecInitDRM
+#define faacDecPostSeekReset           NeAACDecPostSeekReset
+#define faacDecOpen                    NeAACDecOpen
+#define faacDecClose                   NeAACDecClose
+#define faacDecDecode                  NeAACDecDecode
+#define AudioSpecificConfig            NeAACDecAudioSpecificConfig
+#endif
+
+
+#ifdef _WIN32
+  #pragma pack(push, 8)
+  #ifndef NEAACDECAPI
+    #define NEAACDECAPI __cdecl
+  #endif
+#else
+  #ifndef NEAACDECAPI
+    #define NEAACDECAPI
+  #endif
+#endif
+
+#define FAAD2_VERSION "2.7"
+
+/* object types for AAC */
+#define MAIN       1
+#define LC         2
+#define SSR        3
+#define LTP        4
+#define HE_AAC     5
+#define ER_LC     17
+#define ER_LTP    19
+#define LD        23
+#define DRM_ER_LC 27 /* special object type for DRM */
+
+/* header types */
+#define RAW        0
+#define ADIF       1
+#define ADTS       2
+#define LATM       3
+
+/* SBR signalling */
+#define NO_SBR           0
+#define SBR_UPSAMPLED    1
+#define SBR_DOWNSAMPLED  2
+#define NO_SBR_UPSAMPLED 3
+
+/* library output formats */
+#define FAAD_FMT_16BIT  1
+#define FAAD_FMT_24BIT  2
+#define FAAD_FMT_32BIT  3
+#define FAAD_FMT_FLOAT  4
+#define FAAD_FMT_FIXED  FAAD_FMT_FLOAT
+#define FAAD_FMT_DOUBLE 5
+
+/* Capabilities */
+#define LC_DEC_CAP           (1<<0) /* Can decode LC */
+#define MAIN_DEC_CAP         (1<<1) /* Can decode MAIN */
+#define LTP_DEC_CAP          (1<<2) /* Can decode LTP */
+#define LD_DEC_CAP           (1<<3) /* Can decode LD */
+#define ERROR_RESILIENCE_CAP (1<<4) /* Can decode ER */
+#define FIXED_POINT_CAP      (1<<5) /* Fixed point */
+
+/* Channel definitions */
+#define FRONT_CHANNEL_CENTER (1)
+#define FRONT_CHANNEL_LEFT   (2)
+#define FRONT_CHANNEL_RIGHT  (3)
+#define SIDE_CHANNEL_LEFT    (4)
+#define SIDE_CHANNEL_RIGHT   (5)
+#define BACK_CHANNEL_LEFT    (6)
+#define BACK_CHANNEL_RIGHT   (7)
+#define BACK_CHANNEL_CENTER  (8)
+#define LFE_CHANNEL          (9)
+#define UNKNOWN_CHANNEL      (0)
+
+/* DRM channel definitions */
+#define DRMCH_MONO          1
+#define DRMCH_STEREO        2
+#define DRMCH_SBR_MONO      3
+#define DRMCH_SBR_STEREO    4
+#define DRMCH_SBR_PS_STEREO 5
+
+
+/* A decode call can eat up to FAAD_MIN_STREAMSIZE bytes per decoded channel,
+   so at least so much bytes per channel should be available in this stream */
+#define FAAD_MIN_STREAMSIZE 768 /* 6144 bits/channel */
+
+
+typedef void *NeAACDecHandle;
+
+typedef struct mp4AudioSpecificConfig
+{
+    /* Audio Specific Info */
+    unsigned char objectTypeIndex;
+    unsigned char samplingFrequencyIndex;
+    unsigned long samplingFrequency;
+    unsigned char channelsConfiguration;
+
+    /* GA Specific Info */
+    unsigned char frameLengthFlag;
+    unsigned char dependsOnCoreCoder;
+    unsigned short coreCoderDelay;
+    unsigned char extensionFlag;
+    unsigned char aacSectionDataResilienceFlag;
+    unsigned char aacScalefactorDataResilienceFlag;
+    unsigned char aacSpectralDataResilienceFlag;
+    unsigned char epConfig;
+
+    char sbr_present_flag;
+    char forceUpSampling;
+    char downSampledSBR;
+} mp4AudioSpecificConfig;
+
+typedef struct NeAACDecConfiguration
+{
+    unsigned char defObjectType;
+    unsigned long defSampleRate;
+    unsigned char outputFormat;
+    unsigned char downMatrix;
+    unsigned char useOldADTSFormat;
+    unsigned char dontUpSampleImplicitSBR;
+} NeAACDecConfiguration, *NeAACDecConfigurationPtr;
+
+typedef struct NeAACDecFrameInfo
+{
+    unsigned long bytesconsumed;
+    unsigned long samples;
+    unsigned char channels;
+    unsigned char error;
+    unsigned long samplerate;
+
+    /* SBR: 0: off, 1: on; upsample, 2: on; downsampled, 3: off; upsampled */
+    unsigned char sbr;
+
+    /* MPEG-4 ObjectType */
+    unsigned char object_type;
+
+    /* AAC header type; MP4 will be signalled as RAW also */
+    unsigned char header_type;
+
+    /* multichannel configuration */
+    unsigned char num_front_channels;
+    unsigned char num_side_channels;
+    unsigned char num_back_channels;
+    unsigned char num_lfe_channels;
+    unsigned char channel_position[64];
+
+    /* PS: 0: off, 1: on */
+    unsigned char ps;
+} NeAACDecFrameInfo;
+
+char* NEAACDECAPI NeAACDecGetErrorMessage(unsigned char errcode);
+
+unsigned long NEAACDECAPI NeAACDecGetCapabilities(void);
+
+NeAACDecHandle NEAACDECAPI NeAACDecOpen(void);
+
+NeAACDecConfigurationPtr NEAACDECAPI NeAACDecGetCurrentConfiguration(NeAACDecHandle hDecoder);
+
+unsigned char NEAACDECAPI NeAACDecSetConfiguration(NeAACDecHandle hDecoder,
+                                                   NeAACDecConfigurationPtr config);
+
+/* Init the library based on info from the AAC file (ADTS/ADIF) */
+long NEAACDECAPI NeAACDecInit(NeAACDecHandle hDecoder,
+                              unsigned char *buffer,
+                              unsigned long buffer_size,
+                              unsigned long *samplerate,
+                              unsigned char *channels);
+
+/* Init the library using a DecoderSpecificInfo */
+char NEAACDECAPI NeAACDecInit2(NeAACDecHandle hDecoder,
+                               unsigned char *pBuffer,
+                               unsigned long SizeOfDecoderSpecificInfo,
+                               unsigned long *samplerate,
+                               unsigned char *channels);
+
+/* Init the library for DRM */
+char NEAACDECAPI NeAACDecInitDRM(NeAACDecHandle *hDecoder, unsigned long samplerate,
+                                 unsigned char channels);
+
+void NEAACDECAPI NeAACDecPostSeekReset(NeAACDecHandle hDecoder, long frame);
+
+void NEAACDECAPI NeAACDecClose(NeAACDecHandle hDecoder);
+
+void* NEAACDECAPI NeAACDecDecode(NeAACDecHandle hDecoder,
+                                 NeAACDecFrameInfo *hInfo,
+                                 unsigned char *buffer,
+                                 unsigned long buffer_size);
+
+void* NEAACDECAPI NeAACDecDecode2(NeAACDecHandle hDecoder,
+                                  NeAACDecFrameInfo *hInfo,
+                                  unsigned char *buffer,
+                                  unsigned long buffer_size,
+                                  void **sample_buffer,
+                                  unsigned long sample_buffer_size);
+
+char NEAACDECAPI NeAACDecAudioSpecificConfig(unsigned char *pBuffer,
+                                             unsigned long buffer_size,
+                                             mp4AudioSpecificConfig *mp4ASC);
+
+#ifdef _WIN32
+  #pragma pack(pop)
+#endif
+
+#ifdef __cplusplus
+}
+#endif /* __cplusplus */
+
+#endif

lib/codecs/inc/helix-aac/aacdec.h

@@ -0,0 +1,173 @@
+/* ***** BEGIN LICENSE BLOCK *****  
+ * Source last modified: $Id: aacdec.h,v 1.8 2005/11/10 00:15:08 margotm Exp $ 
+ *   
+ * Portions Copyright (c) 1995-2005 RealNetworks, Inc. All Rights Reserved.  
+ *       
+ * The contents of this file, and the files included with this file, 
+ * are subject to the current version of the RealNetworks Public 
+ * Source License (the "RPSL") available at 
+ * http://www.helixcommunity.org/content/rpsl unless you have licensed 
+ * the file under the current version of the RealNetworks Community 
+ * Source License (the "RCSL") available at 
+ * http://www.helixcommunity.org/content/rcsl, in which case the RCSL 
+ * will apply. You may also obtain the license terms directly from 
+ * RealNetworks.  You may not use this file except in compliance with 
+ * the RPSL or, if you have a valid RCSL with RealNetworks applicable 
+ * to this file, the RCSL.  Please see the applicable RPSL or RCSL for 
+ * the rights, obligations and limitations governing use of the 
+ * contents of the file. 
+ *   
+ * This file is part of the Helix DNA Technology. RealNetworks is the 
+ * developer of the Original Code and owns the copyrights in the 
+ * portions it created. 
+ *   
+ * This file, and the files included with this file, is distributed 
+ * and made available on an 'AS IS' basis, WITHOUT WARRANTY OF ANY 
+ * KIND, EITHER EXPRESS OR IMPLIED, AND REALNETWORKS HEREBY DISCLAIMS 
+ * ALL SUCH WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES 
+ * OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, QUIET 
+ * ENJOYMENT OR NON-INFRINGEMENT. 
+ *  
+ * Technology Compatibility Kit Test Suite(s) Location:  
+ *    http://www.helixcommunity.org/content/tck  
+ *  
+ * Contributor(s):  
+ *   
+ * ***** END LICENSE BLOCK ***** */  
+
+/**************************************************************************************
+ * Fixed-point HE-AAC decoder
+ * Jon Recker (jrecker@real.com)
+ * February 2005
+ *
+ * aacdec.h - public C API for AAC decoder
+ **************************************************************************************/
+
+#ifndef _AACDEC_H
+#define _AACDEC_H
+
+#if defined(_WIN32) && !defined(_WIN32_WCE)
+#
+#elif defined(_WIN32) && defined(_WIN32_WCE) && defined(ARM)
+#
+#elif defined(_WIN32) && defined(WINCE_EMULATOR)
+#
+#elif defined (__arm) && defined (__ARMCC_VERSION)
+#
+#elif defined(_SYMBIAN) && defined(__WINS__)
+#
+#elif defined(__GNUC__) && defined(__arm__)
+#
+#elif defined(__GNUC__) && defined(__i386__)
+#
+#elif defined(__GNUC__) && defined(__amd64__)
+#
+#elif defined(__GNUC__) && (defined(__powerpc__) || defined(__POWERPC__))
+#
+#elif defined(_OPENWAVE_SIMULATOR) || defined(_OPENWAVE_ARMULATOR)
+#
+#elif defined(_SOLARIS) && !defined(__GNUC__)
+#
+#elif defined(__XTENSA__)
+#
+#else
+#error No platform defined. See valid options in aacdec.h
+#endif
+
+#ifndef USE_DEFAULT_STDLIB
+#define USE_DEFAULT_STDLIB
+#endif
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* according to spec (13818-7 section 8.2.2, 14496-3 section 4.5.3)
+ * max size of input buffer = 
+ *    6144 bits =  768 bytes per SCE or CCE-I
+ *   12288 bits = 1536 bytes per CPE
+ *       0 bits =    0 bytes per CCE-D (uses bits from the SCE/CPE/CCE-I it is coupled to)
+ */
+#ifndef AAC_MAX_NCHANS				/* if max channels isn't set in makefile, */
+#define AAC_MAX_NCHANS		2		/* set to default max number of channels  */
+#endif
+#define AAC_MAX_NSAMPS		1024
+#define AAC_MAINBUF_SIZE	(768 * AAC_MAX_NCHANS)
+
+#define AAC_NUM_PROFILES	3
+#define AAC_PROFILE_MP		0
+#define AAC_PROFILE_LC		1
+#define AAC_PROFILE_SSR		2
+
+/* define these to enable decoder features */
+#if defined(HELIX_FEATURE_AUDIO_CODEC_AAC_SBR)
+#define AAC_ENABLE_SBR
+#endif //  HELIX_FEATURE_AUDIO_CODEC_AAC_SBR.
+#define AAC_ENABLE_MPEG4
+
+enum {
+	ERR_AAC_NONE                          =   0,
+	ERR_AAC_INDATA_UNDERFLOW              =  -1,
+	ERR_AAC_NULL_POINTER                  =  -2,
+	ERR_AAC_INVALID_ADTS_HEADER           =  -3,
+	ERR_AAC_INVALID_ADIF_HEADER           =  -4,
+	ERR_AAC_INVALID_FRAME                 =  -5,
+	ERR_AAC_MPEG4_UNSUPPORTED             =  -6,
+	ERR_AAC_CHANNEL_MAP                   =  -7,
+	ERR_AAC_SYNTAX_ELEMENT                =  -8,
+
+	ERR_AAC_DEQUANT                       =  -9,
+	ERR_AAC_STEREO_PROCESS                = -10,
+	ERR_AAC_PNS                           = -11,
+	ERR_AAC_SHORT_BLOCK_DEINT             = -12,
+	ERR_AAC_TNS                           = -13,
+	ERR_AAC_IMDCT                         = -14,
+	ERR_AAC_NCHANS_TOO_HIGH               = -15,
+
+	ERR_AAC_SBR_INIT                      = -16,
+	ERR_AAC_SBR_BITSTREAM                 = -17,
+	ERR_AAC_SBR_DATA                      = -18,
+	ERR_AAC_SBR_PCM_FORMAT                = -19,
+	ERR_AAC_SBR_NCHANS_TOO_HIGH           = -20,
+	ERR_AAC_SBR_SINGLERATE_UNSUPPORTED    = -21,
+
+	ERR_AAC_RAWBLOCK_PARAMS               = -22,
+
+	ERR_AAC_UNKNOWN						= -9999
+};
+
+typedef struct _AACFrameInfo {
+	int bitRate;
+	int nChans;
+	int sampRateCore;
+	int sampRateOut;
+	int bitsPerSample;
+	int outputSamps;
+	int profile;
+	int tnsUsed;
+	int pnsUsed;
+} AACFrameInfo;
+
+typedef void *HAACDecoder;
+
+/* public C API */
+HAACDecoder AACInitDecoder(void);
+HAACDecoder AACInitDecoderPre(void *ptr, int sz);
+void AACFreeDecoder(HAACDecoder hAACDecoder);
+int AACDecode(HAACDecoder hAACDecoder, unsigned char **inbuf, int *bytesLeft, short *outbuf);
+
+int AACFindSyncWord(unsigned char *buf, int nBytes);
+void AACGetLastFrameInfo(HAACDecoder hAACDecoder, AACFrameInfo *aacFrameInfo);
+int AACSetRawBlockParams(HAACDecoder hAACDecoder, int copyLast, AACFrameInfo *aacFrameInfo);
+int AACFlushCodec(HAACDecoder hAACDecoder);
+
+#ifdef HELIX_CONFIG_AAC_GENERATE_TRIGTABS_FLOAT
+int AACInitTrigtabsFloat(void);
+void AACFreeTrigtabsFloat(void);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif	/* _AACDEC_H */

lib/codecs/inc/mad/mad.h

@@ -0,0 +1,964 @@
+/*
+ * libmad - MPEG audio decoder library
+ * Copyright (C) 2000-2004 Underbit Technologies, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ * If you would like to negotiate alternate licensing terms, you may do
+ * so by contacting: Underbit Technologies, Inc. <info@underbit.com>
+ */
+
+# ifdef __cplusplus
+extern "C" {
+# endif
+
+# define FPM_DEFAULT
+
+# define SIZEOF_INT 4
+# define SIZEOF_LONG 4
+# define SIZEOF_LONG_LONG 8
+
+/* Id: version.h,v 1.26 2004/01/23 09:41:33 rob Exp */
+
+# ifndef LIBMAD_VERSION_H
+# define LIBMAD_VERSION_H
+
+# define MAD_VERSION_MAJOR	0
+# define MAD_VERSION_MINOR	15
+# define MAD_VERSION_PATCH	1
+# define MAD_VERSION_EXTRA	" (beta)"
+
+# define MAD_VERSION_STRINGIZE(str)	#str
+# define MAD_VERSION_STRING(num)	MAD_VERSION_STRINGIZE(num)
+
+# define MAD_VERSION		MAD_VERSION_STRING(MAD_VERSION_MAJOR) "."  \
+				MAD_VERSION_STRING(MAD_VERSION_MINOR) "."  \
+				MAD_VERSION_STRING(MAD_VERSION_PATCH)  \
+				MAD_VERSION_EXTRA
+
+# define MAD_PUBLISHYEAR	"2000-2004"
+# define MAD_AUTHOR		"Underbit Technologies, Inc."
+# define MAD_EMAIL		"info@underbit.com"
+
+extern char const mad_version[];
+extern char const mad_copyright[];
+extern char const mad_author[];
+extern char const mad_build[];
+
+# endif
+
+/* Id: fixed.h,v 1.38 2004/02/17 02:02:03 rob Exp */
+
+# ifndef LIBMAD_FIXED_H
+# define LIBMAD_FIXED_H
+
+# if SIZEOF_INT >= 4
+typedef   signed int mad_fixed_t;
+
+typedef   signed int mad_fixed64hi_t;
+typedef unsigned int mad_fixed64lo_t;
+# else
+typedef   signed long mad_fixed_t;
+
+typedef   signed long mad_fixed64hi_t;
+typedef unsigned long mad_fixed64lo_t;
+# endif
+
+# if defined(_MSC_VER)
+#  define mad_fixed64_t  signed __int64
+# elif 1 || defined(__GNUC__)
+#  define mad_fixed64_t  signed long long
+# endif
+
+# if defined(FPM_FLOAT)
+typedef double mad_sample_t;
+# else
+typedef mad_fixed_t mad_sample_t;
+# endif
+
+/*
+ * Fixed-point format: 0xABBBBBBB
+ * A == whole part      (sign + 3 bits)
+ * B == fractional part (28 bits)
+ *
+ * Values are signed two's complement, so the effective range is:
+ * 0x80000000 to 0x7fffffff
+ *       -8.0 to +7.9999999962747097015380859375
+ *
+ * The smallest representable value is:
+ * 0x00000001 == 0.0000000037252902984619140625 (i.e. about 3.725e-9)
+ *
+ * 28 bits of fractional accuracy represent about
+ * 8.6 digits of decimal accuracy.
+ *
+ * Fixed-point numbers can be added or subtracted as normal
+ * integers, but multiplication requires shifting the 64-bit result
+ * from 56 fractional bits back to 28 (and rounding.)
+ *
+ * Changing the definition of MAD_F_FRACBITS is only partially
+ * supported, and must be done with care.
+ */
+
+# define MAD_F_FRACBITS		28
+
+# if MAD_F_FRACBITS == 28
+#  define MAD_F(x)		((mad_fixed_t) (x##L))
+# else
+#  if MAD_F_FRACBITS < 28
+#   warning "MAD_F_FRACBITS < 28"
+#   define MAD_F(x)		((mad_fixed_t)  \
+				 (((x##L) +  \
+				   (1L << (28 - MAD_F_FRACBITS - 1))) >>  \
+				  (28 - MAD_F_FRACBITS)))
+#  elif MAD_F_FRACBITS > 28
+#   error "MAD_F_FRACBITS > 28 not currently supported"
+#   define MAD_F(x)		((mad_fixed_t)  \
+				 ((x##L) << (MAD_F_FRACBITS - 28)))
+#  endif
+# endif
+
+# define MAD_F_MIN		((mad_fixed_t) -0x80000000L)
+# define MAD_F_MAX		((mad_fixed_t) +0x7fffffffL)
+
+# define MAD_F_ONE		MAD_F(0x10000000)
+
+# define mad_f_tofixed(x)	((mad_fixed_t)  \
+				 ((x) * (double) (1L << MAD_F_FRACBITS) + 0.5))
+# define mad_f_todouble(x)	((double)  \
+				 ((x) / (double) (1L << MAD_F_FRACBITS)))
+
+# define mad_f_intpart(x)	((x) >> MAD_F_FRACBITS)
+# define mad_f_fracpart(x)	((x) & ((1L << MAD_F_FRACBITS) - 1))
+				/* (x should be positive) */
+
+# define mad_f_fromint(x)	((x) << MAD_F_FRACBITS)
+
+# define mad_f_add(x, y)	((x) + (y))
+# define mad_f_sub(x, y)	((x) - (y))
+
+# if defined(FPM_FLOAT)
+#  error "FPM_FLOAT not yet supported"
+
+#  undef MAD_F
+#  define MAD_F(x)		mad_f_todouble(x)
+
+#  define mad_f_mul(x, y)	((x) * (y))
+#  define mad_f_scale64
+
+#  undef ASO_ZEROCHECK
+
+# elif defined(FPM_64BIT)
+
+/*
+ * This version should be the most accurate if 64-bit types are supported by
+ * the compiler, although it may not be the most efficient.
+ */
+#  if defined(OPT_ACCURACY)
+#   define mad_f_mul(x, y)  \
+    ((mad_fixed_t)  \
+     ((((mad_fixed64_t) (x) * (y)) +  \
+       (1L << (MAD_F_SCALEBITS - 1))) >> MAD_F_SCALEBITS))
+#  else
+#   define mad_f_mul(x, y)  \
+    ((mad_fixed_t) (((mad_fixed64_t) (x) * (y)) >> MAD_F_SCALEBITS))
+#  endif
+
+#  define MAD_F_SCALEBITS  MAD_F_FRACBITS
+
+/* --- Intel --------------------------------------------------------------- */
+
+# elif defined(FPM_INTEL)
+
+#  if defined(_MSC_VER)
+#   pragma warning(push)
+#   pragma warning(disable: 4035)  /* no return value */
+static __forceinline
+mad_fixed_t mad_f_mul_inline(mad_fixed_t x, mad_fixed_t y)
+{
+  enum {
+    fracbits = MAD_F_FRACBITS
+  };
+
+  __asm {
+    mov eax, x
+    imul y
+    shrd eax, edx, fracbits
+  }
+
+  /* implicit return of eax */
+}
+#   pragma warning(pop)
+
+#   define mad_f_mul		mad_f_mul_inline
+#   define mad_f_scale64
+#  else
+/*
+ * This Intel version is fast and accurate; the disposition of the least
+ * significant bit depends on OPT_ACCURACY via mad_f_scale64().
+ */
+#   define MAD_F_MLX(hi, lo, x, y)  \
+    asm ("imull %3"  \
+	 : "=a" (lo), "=d" (hi)  \
+	 : "%a" (x), "rm" (y)  \
+	 : "cc")
+
+#   if defined(OPT_ACCURACY)
+/*
+ * This gives best accuracy but is not very fast.
+ */
+#    define MAD_F_MLA(hi, lo, x, y)  \
+    ({ mad_fixed64hi_t __hi;  \
+       mad_fixed64lo_t __lo;  \
+       MAD_F_MLX(__hi, __lo, (x), (y));  \
+       asm ("addl %2,%0\n\t"  \
+	    "adcl %3,%1"  \
+	    : "=rm" (lo), "=rm" (hi)  \
+	    : "r" (__lo), "r" (__hi), "0" (lo), "1" (hi)  \
+	    : "cc");  \
+    })
+#   endif  /* OPT_ACCURACY */
+
+#   if defined(OPT_ACCURACY)
+/*
+ * Surprisingly, this is faster than SHRD followed by ADC.
+ */
+#    define mad_f_scale64(hi, lo)  \
+    ({ mad_fixed64hi_t __hi_;  \
+       mad_fixed64lo_t __lo_;  \
+       mad_fixed_t __result;  \
+       asm ("addl %4,%2\n\t"  \
+	    "adcl %5,%3"  \
+	    : "=rm" (__lo_), "=rm" (__hi_)  \
+	    : "0" (lo), "1" (hi),  \
+	      "ir" (1L << (MAD_F_SCALEBITS - 1)), "ir" (0)  \
+	    : "cc");  \
+       asm ("shrdl %3,%2,%1"  \
+	    : "=rm" (__result)  \
+	    : "0" (__lo_), "r" (__hi_), "I" (MAD_F_SCALEBITS)  \
+	    : "cc");  \
+       __result;  \
+    })
+#   elif defined(OPT_INTEL)
+/*
+ * Alternate Intel scaling that may or may not perform better.
+ */
+#    define mad_f_scale64(hi, lo)  \
+    ({ mad_fixed_t __result;  \
+       asm ("shrl %3,%1\n\t"  \
+	    "shll %4,%2\n\t"  \
+	    "orl %2,%1"  \
+	    : "=rm" (__result)  \
+	    : "0" (lo), "r" (hi),  \
+	      "I" (MAD_F_SCALEBITS), "I" (32 - MAD_F_SCALEBITS)  \
+	    : "cc");  \
+       __result;  \
+    })
+#   else
+#    define mad_f_scale64(hi, lo)  \
+    ({ mad_fixed_t __result;  \
+       asm ("shrdl %3,%2,%1"  \
+	    : "=rm" (__result)  \
+	    : "0" (lo), "r" (hi), "I" (MAD_F_SCALEBITS)  \
+	    : "cc");  \
+       __result;  \
+    })
+#   endif  /* OPT_ACCURACY */
+
+#   define MAD_F_SCALEBITS  MAD_F_FRACBITS
+#  endif
+
+/* --- ARM ----------------------------------------------------------------- */
+
+# elif defined(FPM_ARM)
+
+/* 
+ * This ARM V4 version is as accurate as FPM_64BIT but much faster. The
+ * least significant bit is properly rounded at no CPU cycle cost!
+ */
+# if 1
+/*
+ * This is faster than the default implementation via MAD_F_MLX() and
+ * mad_f_scale64().
+ */
+#  define mad_f_mul(x, y)  \
+    ({ mad_fixed64hi_t __hi;  \
+       mad_fixed64lo_t __lo;  \
+       mad_fixed_t __result;  \
+       asm ("smull	%0, %1, %3, %4\n\t"  \
+	    "movs	%0, %0, lsr %5\n\t"  \
+	    "adc	%2, %0, %1, lsl %6"  \
+	    : "=&r" (__lo), "=&r" (__hi), "=r" (__result)  \
+	    : "%r" (x), "r" (y),  \
+	      "M" (MAD_F_SCALEBITS), "M" (32 - MAD_F_SCALEBITS)  \
+	    : "cc");  \
+       __result;  \
+    })
+# endif
+
+#  define MAD_F_MLX(hi, lo, x, y)  \
+    asm ("smull	%0, %1, %2, %3"  \
+	 : "=&r" (lo), "=&r" (hi)  \
+	 : "%r" (x), "r" (y))
+
+#  define MAD_F_MLA(hi, lo, x, y)  \
+    asm ("smlal	%0, %1, %2, %3"  \
+	 : "+r" (lo), "+r" (hi)  \
+	 : "%r" (x), "r" (y))
+
+#  define MAD_F_MLN(hi, lo)  \
+    asm ("rsbs	%0, %2, #0\n\t"  \
+	 "rsc	%1, %3, #0"  \
+	 : "=r" (lo), "=r" (hi)  \
+	 : "0" (lo), "1" (hi)  \
+	 : "cc")
+
+#  define mad_f_scale64(hi, lo)  \
+    ({ mad_fixed_t __result;  \
+       asm ("movs	%0, %1, lsr %3\n\t"  \
+	    "adc	%0, %0, %2, lsl %4"  \
+	    : "=&r" (__result)  \
+	    : "r" (lo), "r" (hi),  \
+	      "M" (MAD_F_SCALEBITS), "M" (32 - MAD_F_SCALEBITS)  \
+	    : "cc");  \
+       __result;  \
+    })
+
+#  define MAD_F_SCALEBITS  MAD_F_FRACBITS
+
+/* --- MIPS ---------------------------------------------------------------- */
+
+# elif defined(FPM_MIPS)
+
+/*
+ * This MIPS version is fast and accurate; the disposition of the least
+ * significant bit depends on OPT_ACCURACY via mad_f_scale64().
+ */
+#  define MAD_F_MLX(hi, lo, x, y)  \
+    asm ("mult	%2,%3"  \
+	 : "=l" (lo), "=h" (hi)  \
+	 : "%r" (x), "r" (y))
+
+# if defined(HAVE_MADD_ASM)
+#  define MAD_F_MLA(hi, lo, x, y)  \
+    asm ("madd	%2,%3"  \
+	 : "+l" (lo), "+h" (hi)  \
+	 : "%r" (x), "r" (y))
+# elif defined(HAVE_MADD16_ASM)
+/*
+ * This loses significant accuracy due to the 16-bit integer limit in the
+ * multiply/accumulate instruction.
+ */
+#  define MAD_F_ML0(hi, lo, x, y)  \
+    asm ("mult	%2,%3"  \
+	 : "=l" (lo), "=h" (hi)  \
+	 : "%r" ((x) >> 12), "r" ((y) >> 16))
+#  define MAD_F_MLA(hi, lo, x, y)  \
+    asm ("madd16	%2,%3"  \
+	 : "+l" (lo), "+h" (hi)  \
+	 : "%r" ((x) >> 12), "r" ((y) >> 16))
+#  define MAD_F_MLZ(hi, lo)  ((mad_fixed_t) (lo))
+# endif
+
+# if defined(OPT_SPEED)
+#  define mad_f_scale64(hi, lo)  \
+    ((mad_fixed_t) ((hi) << (32 - MAD_F_SCALEBITS)))
+#  define MAD_F_SCALEBITS  MAD_F_FRACBITS
+# endif
+
+/* --- SPARC --------------------------------------------------------------- */
+
+# elif defined(FPM_SPARC)
+
+/*
+ * This SPARC V8 version is fast and accurate; the disposition of the least
+ * significant bit depends on OPT_ACCURACY via mad_f_scale64().
+ */
+#  define MAD_F_MLX(hi, lo, x, y)  \
+    asm ("smul %2, %3, %0\n\t"  \
+	 "rd %%y, %1"  \
+	 : "=r" (lo), "=r" (hi)  \
+	 : "%r" (x), "rI" (y))
+
+/* --- PowerPC ------------------------------------------------------------- */
+
+# elif defined(FPM_PPC)
+
+/*
+ * This PowerPC version is fast and accurate; the disposition of the least
+ * significant bit depends on OPT_ACCURACY via mad_f_scale64().
+ */
+#  define MAD_F_MLX(hi, lo, x, y)  \
+    do {  \
+      asm ("mullw %0,%1,%2"  \
+	   : "=r" (lo)  \
+	   : "%r" (x), "r" (y));  \
+      asm ("mulhw %0,%1,%2"  \
+	   : "=r" (hi)  \
+	   : "%r" (x), "r" (y));  \
+    }  \
+    while (0)
+
+#  if defined(OPT_ACCURACY)
+/*
+ * This gives best accuracy but is not very fast.
+ */
+#   define MAD_F_MLA(hi, lo, x, y)  \
+    ({ mad_fixed64hi_t __hi;  \
+       mad_fixed64lo_t __lo;  \
+       MAD_F_MLX(__hi, __lo, (x), (y));  \
+       asm ("addc %0,%2,%3\n\t"  \
+	    "adde %1,%4,%5"  \
+	    : "=r" (lo), "=r" (hi)  \
+	    : "%r" (lo), "r" (__lo),  \
+	      "%r" (hi), "r" (__hi)  \
+	    : "xer");  \
+    })
+#  endif
+
+#  if defined(OPT_ACCURACY)
+/*
+ * This is slower than the truncating version below it.
+ */
+#   define mad_f_scale64(hi, lo)  \
+    ({ mad_fixed_t __result, __round;  \
+       asm ("rotrwi %0,%1,%2"  \
+	    : "=r" (__result)  \
+	    : "r" (lo), "i" (MAD_F_SCALEBITS));  \
+       asm ("extrwi %0,%1,1,0"  \
+	    : "=r" (__round)  \
+	    : "r" (__result));  \
+       asm ("insrwi %0,%1,%2,0"  \
+	    : "+r" (__result)  \
+	    : "r" (hi), "i" (MAD_F_SCALEBITS));  \
+       asm ("add %0,%1,%2"  \
+	    : "=r" (__result)  \
+	    : "%r" (__result), "r" (__round));  \
+       __result;  \
+    })
+#  else
+#   define mad_f_scale64(hi, lo)  \
+    ({ mad_fixed_t __result;  \
+       asm ("rotrwi %0,%1,%2"  \
+	    : "=r" (__result)  \
+	    : "r" (lo), "i" (MAD_F_SCALEBITS));  \
+       asm ("insrwi %0,%1,%2,0"  \
+	    : "+r" (__result)  \
+	    : "r" (hi), "i" (MAD_F_SCALEBITS));  \
+       __result;  \
+    })
+#  endif
+
+#  define MAD_F_SCALEBITS  MAD_F_FRACBITS
+
+/* --- Default ------------------------------------------------------------- */
+
+# elif defined(FPM_DEFAULT)
+
+/*
+ * This version is the most portable but it loses significant accuracy.
+ * Furthermore, accuracy is biased against the second argument, so care
+ * should be taken when ordering operands.
+ *
+ * The scale factors are constant as this is not used with SSO.
+ *
+ * Pre-rounding is required to stay within the limits of compliance.
+ */
+#  if defined(OPT_SPEED)
+#   define mad_f_mul(x, y)	(((x) >> 12) * ((y) >> 16))
+#  else
+#   define mad_f_mul(x, y)	((((x) + (1L << 11)) >> 12) *  \
+				 (((y) + (1L << 15)) >> 16))
+#  endif
+
+/* ------------------------------------------------------------------------- */
+
+# else
+#  error "no FPM selected"
+# endif
+
+/* default implementations */
+
+# if !defined(mad_f_mul)
+#  define mad_f_mul(x, y)  \
+    ({ register mad_fixed64hi_t __hi;  \
+       register mad_fixed64lo_t __lo;  \
+       MAD_F_MLX(__hi, __lo, (x), (y));  \
+       mad_f_scale64(__hi, __lo);  \
+    })
+# endif
+
+# if !defined(MAD_F_MLA)
+#  define MAD_F_ML0(hi, lo, x, y)	((lo)  = mad_f_mul((x), (y)))
+#  define MAD_F_MLA(hi, lo, x, y)	((lo) += mad_f_mul((x), (y)))
+#  define MAD_F_MLN(hi, lo)		((lo)  = -(lo))
+#  define MAD_F_MLZ(hi, lo)		((void) (hi), (mad_fixed_t) (lo))
+# endif
+
+# if !defined(MAD_F_ML0)
+#  define MAD_F_ML0(hi, lo, x, y)	MAD_F_MLX((hi), (lo), (x), (y))
+# endif
+
+# if !defined(MAD_F_MLN)
+#  define MAD_F_MLN(hi, lo)		((hi) = ((lo) = -(lo)) ? ~(hi) : -(hi))
+# endif
+
+# if !defined(MAD_F_MLZ)
+#  define MAD_F_MLZ(hi, lo)		mad_f_scale64((hi), (lo))
+# endif
+
+# if !defined(mad_f_scale64)
+#  if defined(OPT_ACCURACY)
+#   define mad_f_scale64(hi, lo)  \
+    ((((mad_fixed_t)  \
+       (((hi) << (32 - (MAD_F_SCALEBITS - 1))) |  \
+	((lo) >> (MAD_F_SCALEBITS - 1)))) + 1) >> 1)
+#  else
+#   define mad_f_scale64(hi, lo)  \
+    ((mad_fixed_t)  \
+     (((hi) << (32 - MAD_F_SCALEBITS)) |  \
+      ((lo) >> MAD_F_SCALEBITS)))
+#  endif
+#  define MAD_F_SCALEBITS  MAD_F_FRACBITS
+# endif
+
+/* C routines */
+
+mad_fixed_t mad_f_abs(mad_fixed_t);
+mad_fixed_t mad_f_div(mad_fixed_t, mad_fixed_t);
+
+# endif
+
+/* Id: bit.h,v 1.12 2004/01/23 09:41:32 rob Exp */
+
+# ifndef LIBMAD_BIT_H
+# define LIBMAD_BIT_H
+
+struct mad_bitptr {
+  unsigned char const *byte;
+  unsigned short cache;
+  unsigned short left;
+};
+
+void mad_bit_init(struct mad_bitptr *, unsigned char const *);
+
+# define mad_bit_finish(bitptr)		/* nothing */
+
+unsigned int mad_bit_length(struct mad_bitptr const *,
+			    struct mad_bitptr const *);
+
+# define mad_bit_bitsleft(bitptr)  ((bitptr)->left)
+unsigned char const *mad_bit_nextbyte(struct mad_bitptr const *);
+
+void mad_bit_skip(struct mad_bitptr *, unsigned int);
+unsigned long mad_bit_read(struct mad_bitptr *, unsigned int);
+void mad_bit_write(struct mad_bitptr *, unsigned int, unsigned long);
+
+unsigned short mad_bit_crc(struct mad_bitptr, unsigned int, unsigned short);
+
+# endif
+
+/* Id: timer.h,v 1.16 2004/01/23 09:41:33 rob Exp */
+
+# ifndef LIBMAD_TIMER_H
+# define LIBMAD_TIMER_H
+
+typedef struct {
+  signed long seconds;		/* whole seconds */
+  unsigned long fraction;	/* 1/MAD_TIMER_RESOLUTION seconds */
+} mad_timer_t;
+
+extern mad_timer_t const mad_timer_zero;
+
+# define MAD_TIMER_RESOLUTION	352800000UL
+
+enum mad_units {
+  MAD_UNITS_HOURS	 =    -2,
+  MAD_UNITS_MINUTES	 =    -1,
+  MAD_UNITS_SECONDS	 =     0,
+
+  /* metric units */
+
+  MAD_UNITS_DECISECONDS	 =    10,
+  MAD_UNITS_CENTISECONDS =   100,
+  MAD_UNITS_MILLISECONDS =  1000,
+
+  /* audio sample units */
+
+  MAD_UNITS_8000_HZ	 =  8000,
+  MAD_UNITS_11025_HZ	 = 11025,
+  MAD_UNITS_12000_HZ	 = 12000,
+
+  MAD_UNITS_16000_HZ	 = 16000,
+  MAD_UNITS_22050_HZ	 = 22050,
+  MAD_UNITS_24000_HZ	 = 24000,
+
+  MAD_UNITS_32000_HZ	 = 32000,
+  MAD_UNITS_44100_HZ	 = 44100,
+  MAD_UNITS_48000_HZ	 = 48000,
+
+  /* video frame/field units */
+
+  MAD_UNITS_24_FPS	 =    24,
+  MAD_UNITS_25_FPS	 =    25,
+  MAD_UNITS_30_FPS	 =    30,
+  MAD_UNITS_48_FPS	 =    48,
+  MAD_UNITS_50_FPS	 =    50,
+  MAD_UNITS_60_FPS	 =    60,
+
+  /* CD audio frames */
+
+  MAD_UNITS_75_FPS	 =    75,
+
+  /* video drop-frame units */
+
+  MAD_UNITS_23_976_FPS	 =   -24,
+  MAD_UNITS_24_975_FPS	 =   -25,
+  MAD_UNITS_29_97_FPS	 =   -30,
+  MAD_UNITS_47_952_FPS	 =   -48,
+  MAD_UNITS_49_95_FPS	 =   -50,
+  MAD_UNITS_59_94_FPS	 =   -60
+};
+
+# define mad_timer_reset(timer)	((void) (*(timer) = mad_timer_zero))
+
+int mad_timer_compare(mad_timer_t, mad_timer_t);
+
+# define mad_timer_sign(timer)	mad_timer_compare((timer), mad_timer_zero)
+
+void mad_timer_negate(mad_timer_t *);
+mad_timer_t mad_timer_abs(mad_timer_t);
+
+void mad_timer_set(mad_timer_t *, unsigned long, unsigned long, unsigned long);
+void mad_timer_add(mad_timer_t *, mad_timer_t);
+void mad_timer_multiply(mad_timer_t *, signed long);
+
+signed long mad_timer_count(mad_timer_t, enum mad_units);
+unsigned long mad_timer_fraction(mad_timer_t, unsigned long);
+void mad_timer_string(mad_timer_t, char *, char const *,
+		      enum mad_units, enum mad_units, unsigned long);
+
+# endif
+
+/* Id: stream.h,v 1.20 2004/02/05 09:02:39 rob Exp */
+
+# ifndef LIBMAD_STREAM_H
+# define LIBMAD_STREAM_H
+
+
+# define MAD_BUFFER_GUARD	8
+# define MAD_BUFFER_MDLEN	(511 + 2048 + MAD_BUFFER_GUARD)
+
+enum mad_error {
+  MAD_ERROR_NONE	   = 0x0000,	/* no error */
+
+  MAD_ERROR_BUFLEN	   = 0x0001,	/* input buffer too small (or EOF) */
+  MAD_ERROR_BUFPTR	   = 0x0002,	/* invalid (null) buffer pointer */
+
+  MAD_ERROR_NOMEM	   = 0x0031,	/* not enough memory */
+
+  MAD_ERROR_LOSTSYNC	   = 0x0101,	/* lost synchronization */
+  MAD_ERROR_BADLAYER	   = 0x0102,	/* reserved header layer value */
+  MAD_ERROR_BADBITRATE	   = 0x0103,	/* forbidden bitrate value */
+  MAD_ERROR_BADSAMPLERATE  = 0x0104,	/* reserved sample frequency value */
+  MAD_ERROR_BADEMPHASIS	   = 0x0105,	/* reserved emphasis value */
+
+  MAD_ERROR_BADCRC	   = 0x0201,	/* CRC check failed */
+  MAD_ERROR_BADBITALLOC	   = 0x0211,	/* forbidden bit allocation value */
+  MAD_ERROR_BADSCALEFACTOR = 0x0221,	/* bad scalefactor index */
+  MAD_ERROR_BADMODE        = 0x0222,	/* bad bitrate/mode combination */
+  MAD_ERROR_BADFRAMELEN	   = 0x0231,	/* bad frame length */
+  MAD_ERROR_BADBIGVALUES   = 0x0232,	/* bad big_values count */
+  MAD_ERROR_BADBLOCKTYPE   = 0x0233,	/* reserved block_type */
+  MAD_ERROR_BADSCFSI	   = 0x0234,	/* bad scalefactor selection info */
+  MAD_ERROR_BADDATAPTR	   = 0x0235,	/* bad main_data_begin pointer */
+  MAD_ERROR_BADPART3LEN	   = 0x0236,	/* bad audio data length */
+  MAD_ERROR_BADHUFFTABLE   = 0x0237,	/* bad Huffman table select */
+  MAD_ERROR_BADHUFFDATA	   = 0x0238,	/* Huffman data overrun */
+  MAD_ERROR_BADSTEREO	   = 0x0239	/* incompatible block_type for JS */
+};
+
+# define MAD_RECOVERABLE(error)	((error) & 0xff00)
+
+struct mad_stream {
+  unsigned char const *buffer;		/* input bitstream buffer */
+  unsigned char const *bufend;		/* end of buffer */
+  unsigned long skiplen;		/* bytes to skip before next frame */
+
+  int sync;				/* stream sync found */
+  unsigned long freerate;		/* free bitrate (fixed) */
+
+  unsigned char const *this_frame;	/* start of current frame */
+  unsigned char const *next_frame;	/* start of next frame */
+  struct mad_bitptr ptr;		/* current processing bit pointer */
+
+  struct mad_bitptr anc_ptr;		/* ancillary bits pointer */
+  unsigned int anc_bitlen;		/* number of ancillary bits */
+
+  unsigned char main_data[MAD_BUFFER_MDLEN];
+					/* Layer III main_data() */
+  unsigned int md_len;			/* bytes in main_data */
+
+  int options;				/* decoding options (see below) */
+  enum mad_error error;			/* error code (see above) */
+};
+
+enum {
+  MAD_OPTION_IGNORECRC      = 0x0001,	/* ignore CRC errors */
+  MAD_OPTION_HALFSAMPLERATE = 0x0002	/* generate PCM at 1/2 sample rate */
+# if 0  /* not yet implemented */
+  MAD_OPTION_LEFTCHANNEL    = 0x0010,	/* decode left channel only */
+  MAD_OPTION_RIGHTCHANNEL   = 0x0020,	/* decode right channel only */
+  MAD_OPTION_SINGLECHANNEL  = 0x0030	/* combine channels */
+# endif
+};
+
+void mad_stream_init(struct mad_stream *);
+void mad_stream_finish(struct mad_stream *);
+
+# define mad_stream_options(stream, opts)  \
+    ((void) ((stream)->options = (opts)))
+
+void mad_stream_buffer(struct mad_stream *,
+		       unsigned char const *, unsigned long);
+void mad_stream_skip(struct mad_stream *, unsigned long);
+
+int mad_stream_sync(struct mad_stream *);
+
+char const *mad_stream_errorstr(struct mad_stream const *);
+
+# endif
+
+/* Id: frame.h,v 1.20 2004/01/23 09:41:32 rob Exp */
+
+# ifndef LIBMAD_FRAME_H
+# define LIBMAD_FRAME_H
+
+
+enum mad_layer {
+  MAD_LAYER_I   = 1,			/* Layer I */
+  MAD_LAYER_II  = 2,			/* Layer II */
+  MAD_LAYER_III = 3			/* Layer III */
+};
+
+enum mad_mode {
+  MAD_MODE_SINGLE_CHANNEL = 0,		/* single channel */
+  MAD_MODE_DUAL_CHANNEL	  = 1,		/* dual channel */
+  MAD_MODE_JOINT_STEREO	  = 2,		/* joint (MS/intensity) stereo */
+  MAD_MODE_STEREO	  = 3		/* normal LR stereo */
+};
+
+enum mad_emphasis {
+  MAD_EMPHASIS_NONE	  = 0,		/* no emphasis */
+  MAD_EMPHASIS_50_15_US	  = 1,		/* 50/15 microseconds emphasis */
+  MAD_EMPHASIS_CCITT_J_17 = 3,		/* CCITT J.17 emphasis */
+  MAD_EMPHASIS_RESERVED   = 2		/* unknown emphasis */
+};
+
+struct mad_header {
+  enum mad_layer layer;			/* audio layer (1, 2, or 3) */
+  enum mad_mode mode;			/* channel mode (see above) */
+  int mode_extension;			/* additional mode info */
+  enum mad_emphasis emphasis;		/* de-emphasis to use (see above) */
+
+  unsigned long bitrate;		/* stream bitrate (bps) */
+  unsigned int samplerate;		/* sampling frequency (Hz) */
+
+  unsigned short crc_check;		/* frame CRC accumulator */
+  unsigned short crc_target;		/* final target CRC checksum */
+
+  int flags;				/* flags (see below) */
+  int private_bits;			/* private bits (see below) */
+
+  mad_timer_t duration;			/* audio playing time of frame */
+};
+
+struct mad_frame {
+  struct mad_header header;		/* MPEG audio header */
+
+  int options;				/* decoding options (from stream) */
+
+  mad_fixed_t sbsample[2][36][32];	/* synthesis subband filter samples */
+  mad_fixed_t overlap[2][32][18];	/* Layer III block overlap data */
+
+  mad_fixed_t xr[2][576];
+  mad_fixed_t tmp[32][3][6];
+};
+
+# define MAD_NCHANNELS(header)		((header)->mode ? 2 : 1)
+# define MAD_NSBSAMPLES(header)  \
+  ((header)->layer == MAD_LAYER_I ? 12 :  \
+   (((header)->layer == MAD_LAYER_III &&  \
+     ((header)->flags & MAD_FLAG_LSF_EXT)) ? 18 : 36))
+
+enum {
+  MAD_FLAG_NPRIVATE_III	= 0x0007,	/* number of Layer III private bits */
+  MAD_FLAG_INCOMPLETE	= 0x0008,	/* header but not data is decoded */
+
+  MAD_FLAG_PROTECTION	= 0x0010,	/* frame has CRC protection */
+  MAD_FLAG_COPYRIGHT	= 0x0020,	/* frame is copyright */
+  MAD_FLAG_ORIGINAL	= 0x0040,	/* frame is original (else copy) */
+  MAD_FLAG_PADDING	= 0x0080,	/* frame has additional slot */
+
+  MAD_FLAG_I_STEREO	= 0x0100,	/* uses intensity joint stereo */
+  MAD_FLAG_MS_STEREO	= 0x0200,	/* uses middle/side joint stereo */
+  MAD_FLAG_FREEFORMAT	= 0x0400,	/* uses free format bitrate */
+
+  MAD_FLAG_LSF_EXT	= 0x1000,	/* lower sampling freq. extension */
+  MAD_FLAG_MC_EXT	= 0x2000,	/* multichannel audio extension */
+  MAD_FLAG_MPEG_2_5_EXT	= 0x4000	/* MPEG 2.5 (unofficial) extension */
+};
+
+enum {
+  MAD_PRIVATE_HEADER	= 0x0100,	/* header private bit */
+  MAD_PRIVATE_III	= 0x001f	/* Layer III private bits (up to 5) */
+};
+
+void mad_header_init(struct mad_header *);
+
+# define mad_header_finish(header)  /* nothing */
+
+int mad_header_decode(struct mad_header *, struct mad_stream *);
+
+void mad_frame_init(struct mad_frame *);
+void mad_frame_finish(struct mad_frame *);
+
+int mad_frame_decode(struct mad_frame *, struct mad_stream *);
+
+void mad_frame_mute(struct mad_frame *);
+
+# endif
+
+/* Id: synth.h,v 1.15 2004/01/23 09:41:33 rob Exp */
+
+# ifndef LIBMAD_SYNTH_H
+# define LIBMAD_SYNTH_H
+
+
+struct mad_pcm {
+  unsigned int samplerate;		/* sampling frequency (Hz) */
+  unsigned short channels;		/* number of channels */
+  unsigned short length;		/* number of samples per channel */
+  mad_fixed_t samples[2][1152];		/* PCM output samples [ch][sample] */
+};
+
+struct mad_synth {
+  mad_fixed_t filter[2][2][2][16][8];	/* polyphase filterbank outputs */
+  					/* [ch][eo][peo][s][v] */
+
+  unsigned int phase;			/* current processing phase */
+
+  struct mad_pcm pcm;			/* PCM output */
+};
+
+/* single channel PCM selector */
+enum {
+  MAD_PCM_CHANNEL_SINGLE = 0
+};
+
+/* dual channel PCM selector */
+enum {
+  MAD_PCM_CHANNEL_DUAL_1 = 0,
+  MAD_PCM_CHANNEL_DUAL_2 = 1
+};
+
+/* stereo PCM selector */
+enum {
+  MAD_PCM_CHANNEL_STEREO_LEFT  = 0,
+  MAD_PCM_CHANNEL_STEREO_RIGHT = 1
+};
+
+void mad_synth_init(struct mad_synth *);
+
+# define mad_synth_finish(synth)  /* nothing */
+
+void mad_synth_mute(struct mad_synth *);
+
+void mad_synth_frame(struct mad_synth *, struct mad_frame const *);
+
+# endif
+
+/* Id: decoder.h,v 1.17 2004/01/23 09:41:32 rob Exp */
+
+# ifndef LIBMAD_DECODER_H
+# define LIBMAD_DECODER_H
+
+
+enum mad_decoder_mode {
+  MAD_DECODER_MODE_SYNC  = 0,
+  MAD_DECODER_MODE_ASYNC
+};
+
+enum mad_flow {
+  MAD_FLOW_CONTINUE = 0x0000,	/* continue normally */
+  MAD_FLOW_STOP     = 0x0010,	/* stop decoding normally */
+  MAD_FLOW_BREAK    = 0x0011,	/* stop decoding and signal an error */
+  MAD_FLOW_IGNORE   = 0x0020	/* ignore the current frame */
+};
+
+struct mad_decoder {
+  enum mad_decoder_mode mode;
+
+  int options;
+
+  struct {
+    long pid;
+    int in;
+    int out;
+  } async;
+
+  struct {
+    struct mad_stream stream;
+    struct mad_frame frame;
+    struct mad_synth synth;
+  } *sync;
+
+  void *cb_data;
+
+  enum mad_flow (*input_func)(void *, struct mad_stream *);
+  enum mad_flow (*header_func)(void *, struct mad_header const *);
+  enum mad_flow (*filter_func)(void *,
+			       struct mad_stream const *, struct mad_frame *);
+  enum mad_flow (*output_func)(void *,
+			       struct mad_header const *, struct mad_pcm *);
+  enum mad_flow (*error_func)(void *, struct mad_stream *, struct mad_frame *);
+  enum mad_flow (*message_func)(void *, void *, unsigned int *);
+};
+
+void mad_decoder_init(struct mad_decoder *, void *,
+		      enum mad_flow (*)(void *, struct mad_stream *),
+		      enum mad_flow (*)(void *, struct mad_header const *),
+		      enum mad_flow (*)(void *,
+					struct mad_stream const *,
+					struct mad_frame *),
+		      enum mad_flow (*)(void *,
+					struct mad_header const *,
+					struct mad_pcm *),
+		      enum mad_flow (*)(void *,
+					struct mad_stream *,
+					struct mad_frame *),
+		      enum mad_flow (*)(void *, void *, unsigned int *));
+int mad_decoder_finish(struct mad_decoder *);
+
+# define mad_decoder_options(decoder, opts)  \
+    ((void) ((decoder)->options = (opts)))
+
+int mad_decoder_run(struct mad_decoder *, enum mad_decoder_mode);
+int mad_decoder_message(struct mad_decoder *, void *, unsigned int *);
+
+# endif
+
+# ifdef __cplusplus
+}
+# endif

lib/codecs/inc/ogg/config_types.h

@@ -0,0 +1,26 @@
+#ifndef __CONFIG_TYPES_H__
+#define __CONFIG_TYPES_H__
+
+/* these are filled in by configure or cmake*/
+#define INCLUDE_INTTYPES_H 1
+#define INCLUDE_STDINT_H 1
+#define INCLUDE_SYS_TYPES_H 1
+
+#if INCLUDE_INTTYPES_H
+#  include <inttypes.h>
+#endif
+#if INCLUDE_STDINT_H
+#  include <stdint.h>
+#endif
+#if INCLUDE_SYS_TYPES_H
+#  include <sys/types.h>
+#endif
+
+typedef int16_t ogg_int16_t;
+typedef uint16_t ogg_uint16_t;
+typedef int32_t ogg_int32_t;
+typedef uint32_t ogg_uint32_t;
+typedef int64_t ogg_int64_t;
+typedef uint64_t ogg_uint64_t;
+
+#endif

lib/codecs/inc/ogg/ogg.h

@@ -0,0 +1,209 @@
+/********************************************************************
+ *                                                                  *
+ * THIS FILE IS PART OF THE OggVorbis SOFTWARE CODEC SOURCE CODE.   *
+ * USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS     *
+ * GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
+ * IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING.       *
+ *                                                                  *
+ * THE OggVorbis SOURCE CODE IS (C) COPYRIGHT 1994-2007             *
+ * by the Xiph.Org Foundation http://www.xiph.org/                  *
+ *                                                                  *
+ ********************************************************************
+
+ function: toplevel libogg include
+
+ ********************************************************************/
+#ifndef _OGG_H
+#define _OGG_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include <stddef.h>
+#include <ogg/os_types.h>
+
+typedef struct {
+  void *iov_base;
+  size_t iov_len;
+} ogg_iovec_t;
+
+typedef struct {
+  long endbyte;
+  int  endbit;
+
+  unsigned char *buffer;
+  unsigned char *ptr;
+  long storage;
+} oggpack_buffer;
+
+/* ogg_page is used to encapsulate the data in one Ogg bitstream page *****/
+
+typedef struct {
+  unsigned char *header;
+  long header_len;
+  unsigned char *body;
+  long body_len;
+} ogg_page;
+
+/* ogg_stream_state contains the current encode/decode state of a logical
+   Ogg bitstream **********************************************************/
+
+typedef struct {
+  unsigned char   *body_data;    /* bytes from packet bodies */
+  long    body_storage;          /* storage elements allocated */
+  long    body_fill;             /* elements stored; fill mark */
+  long    body_returned;         /* elements of fill returned */
+
+
+  int     *lacing_vals;      /* The values that will go to the segment table */
+  ogg_int64_t *granule_vals; /* granulepos values for headers. Not compact
+                                this way, but it is simple coupled to the
+                                lacing fifo */
+  long    lacing_storage;
+  long    lacing_fill;
+  long    lacing_packet;
+  long    lacing_returned;
+
+  unsigned char    header[282];      /* working space for header encode */
+  int              header_fill;
+
+  int     e_o_s;          /* set when we have buffered the last packet in the
+                             logical bitstream */
+  int     b_o_s;          /* set after we've written the initial page
+                             of a logical bitstream */
+  long    serialno;
+  long    pageno;
+  ogg_int64_t  packetno;  /* sequence number for decode; the framing
+                             knows where there's a hole in the data,
+                             but we need coupling so that the codec
+                             (which is in a separate abstraction
+                             layer) also knows about the gap */
+  ogg_int64_t   granulepos;
+
+} ogg_stream_state;
+
+/* ogg_packet is used to encapsulate the data and metadata belonging
+   to a single raw Ogg/Vorbis packet *************************************/
+
+typedef struct {
+  unsigned char *packet;
+  long  bytes;
+  long  b_o_s;
+  long  e_o_s;
+
+  ogg_int64_t  granulepos;
+
+  ogg_int64_t  packetno;     /* sequence number for decode; the framing
+                                knows where there's a hole in the data,
+                                but we need coupling so that the codec
+                                (which is in a separate abstraction
+                                layer) also knows about the gap */
+} ogg_packet;
+
+typedef struct {
+  unsigned char *data;
+  int storage;
+  int fill;
+  int returned;
+
+  int unsynced;
+  int headerbytes;
+  int bodybytes;
+} ogg_sync_state;
+
+/* Ogg BITSTREAM PRIMITIVES: bitstream ************************/
+
+extern void  oggpack_writeinit(oggpack_buffer *b);
+extern int   oggpack_writecheck(oggpack_buffer *b);
+extern void  oggpack_writetrunc(oggpack_buffer *b,long bits);
+extern void  oggpack_writealign(oggpack_buffer *b);
+extern void  oggpack_writecopy(oggpack_buffer *b,void *source,long bits);
+extern void  oggpack_reset(oggpack_buffer *b);
+extern void  oggpack_writeclear(oggpack_buffer *b);
+extern void  oggpack_readinit(oggpack_buffer *b,unsigned char *buf,int bytes);
+extern void  oggpack_write(oggpack_buffer *b,unsigned long value,int bits);
+extern long  oggpack_look(oggpack_buffer *b,int bits);
+extern long  oggpack_look1(oggpack_buffer *b);
+extern void  oggpack_adv(oggpack_buffer *b,int bits);
+extern void  oggpack_adv1(oggpack_buffer *b);
+extern long  oggpack_read(oggpack_buffer *b,int bits);
+extern long  oggpack_read1(oggpack_buffer *b);
+extern long  oggpack_bytes(oggpack_buffer *b);
+extern long  oggpack_bits(oggpack_buffer *b);
+extern unsigned char *oggpack_get_buffer(oggpack_buffer *b);
+
+extern void  oggpackB_writeinit(oggpack_buffer *b);
+extern int   oggpackB_writecheck(oggpack_buffer *b);
+extern void  oggpackB_writetrunc(oggpack_buffer *b,long bits);
+extern void  oggpackB_writealign(oggpack_buffer *b);
+extern void  oggpackB_writecopy(oggpack_buffer *b,void *source,long bits);
+extern void  oggpackB_reset(oggpack_buffer *b);
+extern void  oggpackB_writeclear(oggpack_buffer *b);
+extern void  oggpackB_readinit(oggpack_buffer *b,unsigned char *buf,int bytes);
+extern void  oggpackB_write(oggpack_buffer *b,unsigned long value,int bits);
+extern long  oggpackB_look(oggpack_buffer *b,int bits);
+extern long  oggpackB_look1(oggpack_buffer *b);
+extern void  oggpackB_adv(oggpack_buffer *b,int bits);
+extern void  oggpackB_adv1(oggpack_buffer *b);
+extern long  oggpackB_read(oggpack_buffer *b,int bits);
+extern long  oggpackB_read1(oggpack_buffer *b);
+extern long  oggpackB_bytes(oggpack_buffer *b);
+extern long  oggpackB_bits(oggpack_buffer *b);
+extern unsigned char *oggpackB_get_buffer(oggpack_buffer *b);
+
+/* Ogg BITSTREAM PRIMITIVES: encoding **************************/
+
+extern int      ogg_stream_packetin(ogg_stream_state *os, ogg_packet *op);
+extern int      ogg_stream_iovecin(ogg_stream_state *os, ogg_iovec_t *iov,
+                                   int count, long e_o_s, ogg_int64_t granulepos);
+extern int      ogg_stream_pageout(ogg_stream_state *os, ogg_page *og);
+extern int      ogg_stream_pageout_fill(ogg_stream_state *os, ogg_page *og, int nfill);
+extern int      ogg_stream_flush(ogg_stream_state *os, ogg_page *og);
+extern int      ogg_stream_flush_fill(ogg_stream_state *os, ogg_page *og, int nfill);
+
+/* Ogg BITSTREAM PRIMITIVES: decoding **************************/
+
+extern int      ogg_sync_init(ogg_sync_state *oy);
+extern int      ogg_sync_clear(ogg_sync_state *oy);
+extern int      ogg_sync_reset(ogg_sync_state *oy);
+extern int      ogg_sync_destroy(ogg_sync_state *oy);
+extern int      ogg_sync_check(ogg_sync_state *oy);
+
+extern char    *ogg_sync_buffer(ogg_sync_state *oy, long size);
+extern int      ogg_sync_wrote(ogg_sync_state *oy, long bytes);
+extern long     ogg_sync_pageseek(ogg_sync_state *oy,ogg_page *og);
+extern int      ogg_sync_pageout(ogg_sync_state *oy, ogg_page *og);
+extern int      ogg_stream_pagein(ogg_stream_state *os, ogg_page *og);
+extern int      ogg_stream_packetout(ogg_stream_state *os,ogg_packet *op);
+extern int      ogg_stream_packetpeek(ogg_stream_state *os,ogg_packet *op);
+
+/* Ogg BITSTREAM PRIMITIVES: general ***************************/
+
+extern int      ogg_stream_init(ogg_stream_state *os,int serialno);
+extern int      ogg_stream_clear(ogg_stream_state *os);
+extern int      ogg_stream_reset(ogg_stream_state *os);
+extern int      ogg_stream_reset_serialno(ogg_stream_state *os,int serialno);
+extern int      ogg_stream_destroy(ogg_stream_state *os);
+extern int      ogg_stream_check(ogg_stream_state *os);
+extern int      ogg_stream_eos(ogg_stream_state *os);
+
+extern void     ogg_page_checksum_set(ogg_page *og);
+
+extern int      ogg_page_version(const ogg_page *og);
+extern int      ogg_page_continued(const ogg_page *og);
+extern int      ogg_page_bos(const ogg_page *og);
+extern int      ogg_page_eos(const ogg_page *og);
+extern ogg_int64_t  ogg_page_granulepos(const ogg_page *og);
+extern int      ogg_page_serialno(const ogg_page *og);
+extern long     ogg_page_pageno(const ogg_page *og);
+extern int      ogg_page_packets(const ogg_page *og);
+
+extern void     ogg_packet_clear(ogg_packet *op);
+
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif  /* _OGG_H */

lib/codecs/inc/ogg/os_types.h

@@ -0,0 +1,158 @@
+/********************************************************************
+ *                                                                  *
+ * THIS FILE IS PART OF THE OggVorbis SOFTWARE CODEC SOURCE CODE.   *
+ * USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS     *
+ * GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
+ * IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING.       *
+ *                                                                  *
+ * THE OggVorbis SOURCE CODE IS (C) COPYRIGHT 1994-2002             *
+ * by the Xiph.Org Foundation http://www.xiph.org/                  *
+ *                                                                  *
+ ********************************************************************
+
+ function: Define a consistent set of types on each platform.
+
+ ********************************************************************/
+#ifndef _OS_TYPES_H
+#define _OS_TYPES_H
+
+/* make it easy on the folks that want to compile the libs with a
+   different malloc than stdlib */
+#define _ogg_malloc  malloc
+#define _ogg_calloc  calloc
+#define _ogg_realloc realloc
+#define _ogg_free    free
+
+#if defined(_WIN32)
+
+#  if defined(__CYGWIN__)
+#    include <stdint.h>
+     typedef int16_t ogg_int16_t;
+     typedef uint16_t ogg_uint16_t;
+     typedef int32_t ogg_int32_t;
+     typedef uint32_t ogg_uint32_t;
+     typedef int64_t ogg_int64_t;
+     typedef uint64_t ogg_uint64_t;
+#  elif defined(__MINGW32__)
+#    include <sys/types.h>
+     typedef short ogg_int16_t;
+     typedef unsigned short ogg_uint16_t;
+     typedef int ogg_int32_t;
+     typedef unsigned int ogg_uint32_t;
+     typedef long long ogg_int64_t;
+     typedef unsigned long long ogg_uint64_t;
+#  elif defined(__MWERKS__)
+     typedef long long ogg_int64_t;
+     typedef unsigned long long ogg_uint64_t;
+     typedef int ogg_int32_t;
+     typedef unsigned int ogg_uint32_t;
+     typedef short ogg_int16_t;
+     typedef unsigned short ogg_uint16_t;
+#  else
+#    if defined(_MSC_VER) && (_MSC_VER >= 1800) /* MSVC 2013 and newer */
+#      include <stdint.h>
+       typedef int16_t ogg_int16_t;
+       typedef uint16_t ogg_uint16_t;
+       typedef int32_t ogg_int32_t;
+       typedef uint32_t ogg_uint32_t;
+       typedef int64_t ogg_int64_t;
+       typedef uint64_t ogg_uint64_t;
+#    else
+       /* MSVC/Borland */
+       typedef __int64 ogg_int64_t;
+       typedef __int32 ogg_int32_t;
+       typedef unsigned __int32 ogg_uint32_t;
+       typedef unsigned __int64 ogg_uint64_t;
+       typedef __int16 ogg_int16_t;
+       typedef unsigned __int16 ogg_uint16_t;
+#    endif
+#  endif
+
+#elif (defined(__APPLE__) && defined(__MACH__)) /* MacOS X Framework build */
+
+#  include <sys/types.h>
+   typedef int16_t ogg_int16_t;
+   typedef u_int16_t ogg_uint16_t;
+   typedef int32_t ogg_int32_t;
+   typedef u_int32_t ogg_uint32_t;
+   typedef int64_t ogg_int64_t;
+   typedef u_int64_t ogg_uint64_t;
+
+#elif defined(__HAIKU__)
+
+  /* Haiku */
+#  include <sys/types.h>
+   typedef short ogg_int16_t;
+   typedef unsigned short ogg_uint16_t;
+   typedef int ogg_int32_t;
+   typedef unsigned int ogg_uint32_t;
+   typedef long long ogg_int64_t;
+   typedef unsigned long long ogg_uint64_t;
+
+#elif defined(__BEOS__)
+
+   /* Be */
+#  include <inttypes.h>
+   typedef int16_t ogg_int16_t;
+   typedef uint16_t ogg_uint16_t;
+   typedef int32_t ogg_int32_t;
+   typedef uint32_t ogg_uint32_t;
+   typedef int64_t ogg_int64_t;
+   typedef uint64_t ogg_uint64_t;
+
+#elif defined (__EMX__)
+
+   /* OS/2 GCC */
+   typedef short ogg_int16_t;
+   typedef unsigned short ogg_uint16_t;
+   typedef int ogg_int32_t;
+   typedef unsigned int ogg_uint32_t;
+   typedef long long ogg_int64_t;
+   typedef unsigned long long ogg_uint64_t;
+
+
+#elif defined (DJGPP)
+
+   /* DJGPP */
+   typedef short ogg_int16_t;
+   typedef int ogg_int32_t;
+   typedef unsigned int ogg_uint32_t;
+   typedef long long ogg_int64_t;
+   typedef unsigned long long ogg_uint64_t;
+
+#elif defined(R5900)
+
+   /* PS2 EE */
+   typedef long ogg_int64_t;
+   typedef unsigned long ogg_uint64_t;
+   typedef int ogg_int32_t;
+   typedef unsigned ogg_uint32_t;
+   typedef short ogg_int16_t;
+
+#elif defined(__SYMBIAN32__)
+
+   /* Symbian GCC */
+   typedef signed short ogg_int16_t;
+   typedef unsigned short ogg_uint16_t;
+   typedef signed int ogg_int32_t;
+   typedef unsigned int ogg_uint32_t;
+   typedef long long int ogg_int64_t;
+   typedef unsigned long long int ogg_uint64_t;
+
+#elif defined(__TMS320C6X__)
+
+   /* TI C64x compiler */
+   typedef signed short ogg_int16_t;
+   typedef unsigned short ogg_uint16_t;
+   typedef signed int ogg_int32_t;
+   typedef unsigned int ogg_uint32_t;
+   typedef long long int ogg_int64_t;
+   typedef unsigned long long int ogg_uint64_t;
+
+#else
+
+#  include <ogg/config_types.h>
+
+#endif
+
+#endif  /* _OS_TYPES_H */

lib/codecs/inc/opus/opus.h

@@ -0,0 +1,981 @@
+/* Copyright (c) 2010-2011 Xiph.Org Foundation, Skype Limited
+   Written by Jean-Marc Valin and Koen Vos */
+/*
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
+   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+/**
+ * @file opus.h
+ * @brief Opus reference implementation API
+ */
+
+#ifndef OPUS_H
+#define OPUS_H
+
+#include "opus_types.h"
+#include "opus_defines.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/**
+ * @mainpage Opus
+ *
+ * The Opus codec is designed for interactive speech and audio transmission over the Internet.
+ * It is designed by the IETF Codec Working Group and incorporates technology from
+ * Skype's SILK codec and Xiph.Org's CELT codec.
+ *
+ * The Opus codec is designed to handle a wide range of interactive audio applications,
+ * including Voice over IP, videoconferencing, in-game chat, and even remote live music
+ * performances. It can scale from low bit-rate narrowband speech to very high quality
+ * stereo music. Its main features are:
+
+ * @li Sampling rates from 8 to 48 kHz
+ * @li Bit-rates from 6 kb/s to 510 kb/s
+ * @li Support for both constant bit-rate (CBR) and variable bit-rate (VBR)
+ * @li Audio bandwidth from narrowband to full-band
+ * @li Support for speech and music
+ * @li Support for mono and stereo
+ * @li Support for multichannel (up to 255 channels)
+ * @li Frame sizes from 2.5 ms to 60 ms
+ * @li Good loss robustness and packet loss concealment (PLC)
+ * @li Floating point and fixed-point implementation
+ *
+ * Documentation sections:
+ * @li @ref opus_encoder
+ * @li @ref opus_decoder
+ * @li @ref opus_repacketizer
+ * @li @ref opus_multistream
+ * @li @ref opus_libinfo
+ * @li @ref opus_custom
+ */
+
+/** @defgroup opus_encoder Opus Encoder
+  * @{
+  *
+  * @brief This page describes the process and functions used to encode Opus.
+  *
+  * Since Opus is a stateful codec, the encoding process starts with creating an encoder
+  * state. This can be done with:
+  *
+  * @code
+  * int          error;
+  * OpusEncoder *enc;
+  * enc = opus_encoder_create(Fs, channels, application, &error);
+  * @endcode
+  *
+  * From this point, @c enc can be used for encoding an audio stream. An encoder state
+  * @b must @b not be used for more than one stream at the same time. Similarly, the encoder
+  * state @b must @b not be re-initialized for each frame.
+  *
+  * While opus_encoder_create() allocates memory for the state, it's also possible
+  * to initialize pre-allocated memory:
+  *
+  * @code
+  * int          size;
+  * int          error;
+  * OpusEncoder *enc;
+  * size = opus_encoder_get_size(channels);
+  * enc = malloc(size);
+  * error = opus_encoder_init(enc, Fs, channels, application);
+  * @endcode
+  *
+  * where opus_encoder_get_size() returns the required size for the encoder state. Note that
+  * future versions of this code may change the size, so no assuptions should be made about it.
+  *
+  * The encoder state is always continuous in memory and only a shallow copy is sufficient
+  * to copy it (e.g. memcpy())
+  *
+  * It is possible to change some of the encoder's settings using the opus_encoder_ctl()
+  * interface. All these settings already default to the recommended value, so they should
+  * only be changed when necessary. The most common settings one may want to change are:
+  *
+  * @code
+  * opus_encoder_ctl(enc, OPUS_SET_BITRATE(bitrate));
+  * opus_encoder_ctl(enc, OPUS_SET_COMPLEXITY(complexity));
+  * opus_encoder_ctl(enc, OPUS_SET_SIGNAL(signal_type));
+  * @endcode
+  *
+  * where
+  *
+  * @arg bitrate is in bits per second (b/s)
+  * @arg complexity is a value from 1 to 10, where 1 is the lowest complexity and 10 is the highest
+  * @arg signal_type is either OPUS_AUTO (default), OPUS_SIGNAL_VOICE, or OPUS_SIGNAL_MUSIC
+  *
+  * See @ref opus_encoderctls and @ref opus_genericctls for a complete list of parameters that can be set or queried. Most parameters can be set or changed at any time during a stream.
+  *
+  * To encode a frame, opus_encode() or opus_encode_float() must be called with exactly one frame (2.5, 5, 10, 20, 40 or 60 ms) of audio data:
+  * @code
+  * len = opus_encode(enc, audio_frame, frame_size, packet, max_packet);
+  * @endcode
+  *
+  * where
+  * <ul>
+  * <li>audio_frame is the audio data in opus_int16 (or float for opus_encode_float())</li>
+  * <li>frame_size is the duration of the frame in samples (per channel)</li>
+  * <li>packet is the byte array to which the compressed data is written</li>
+  * <li>max_packet is the maximum number of bytes that can be written in the packet (4000 bytes is recommended).
+  *     Do not use max_packet to control VBR target bitrate, instead use the #OPUS_SET_BITRATE CTL.</li>
+  * </ul>
+  *
+  * opus_encode() and opus_encode_float() return the number of bytes actually written to the packet.
+  * The return value <b>can be negative</b>, which indicates that an error has occurred. If the return value
+  * is 2 bytes or less, then the packet does not need to be transmitted (DTX).
+  *
+  * Once the encoder state if no longer needed, it can be destroyed with
+  *
+  * @code
+  * opus_encoder_destroy(enc);
+  * @endcode
+  *
+  * If the encoder was created with opus_encoder_init() rather than opus_encoder_create(),
+  * then no action is required aside from potentially freeing the memory that was manually
+  * allocated for it (calling free(enc) for the example above)
+  *
+  */
+
+/** Opus encoder state.
+  * This contains the complete state of an Opus encoder.
+  * It is position independent and can be freely copied.
+  * @see opus_encoder_create,opus_encoder_init
+  */
+typedef struct OpusEncoder OpusEncoder;
+
+/** Gets the size of an <code>OpusEncoder</code> structure.
+  * @param[in] channels <tt>int</tt>: Number of channels.
+  *                                   This must be 1 or 2.
+  * @returns The size in bytes.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_encoder_get_size(int channels);
+
+/**
+ */
+
+/** Allocates and initializes an encoder state.
+ * There are three coding modes:
+ *
+ * @ref OPUS_APPLICATION_VOIP gives best quality at a given bitrate for voice
+ *    signals. It enhances the  input signal by high-pass filtering and
+ *    emphasizing formants and harmonics. Optionally  it includes in-band
+ *    forward error correction to protect against packet loss. Use this
+ *    mode for typical VoIP applications. Because of the enhancement,
+ *    even at high bitrates the output may sound different from the input.
+ *
+ * @ref OPUS_APPLICATION_AUDIO gives best quality at a given bitrate for most
+ *    non-voice signals like music. Use this mode for music and mixed
+ *    (music/voice) content, broadcast, and applications requiring less
+ *    than 15 ms of coding delay.
+ *
+ * @ref OPUS_APPLICATION_RESTRICTED_LOWDELAY configures low-delay mode that
+ *    disables the speech-optimized mode in exchange for slightly reduced delay.
+ *    This mode can only be set on an newly initialized or freshly reset encoder
+ *    because it changes the codec delay.
+ *
+ * This is useful when the caller knows that the speech-optimized modes will not be needed (use with caution).
+ * @param [in] Fs <tt>opus_int32</tt>: Sampling rate of input signal (Hz)
+ *                                     This must be one of 8000, 12000, 16000,
+ *                                     24000, or 48000.
+ * @param [in] channels <tt>int</tt>: Number of channels (1 or 2) in input signal
+ * @param [in] application <tt>int</tt>: Coding mode (one of @ref OPUS_APPLICATION_VOIP, @ref OPUS_APPLICATION_AUDIO, or @ref OPUS_APPLICATION_RESTRICTED_LOWDELAY)
+ * @param [out] error <tt>int*</tt>: @ref opus_errorcodes
+ * @note Regardless of the sampling rate and number channels selected, the Opus encoder
+ * can switch to a lower audio bandwidth or number of channels if the bitrate
+ * selected is too low. This also means that it is safe to always use 48 kHz stereo input
+ * and let the encoder optimize the encoding.
+ */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusEncoder *opus_encoder_create(
+    opus_int32 Fs,
+    int channels,
+    int application,
+    int *error
+);
+
+/** Initializes a previously allocated encoder state
+  * The memory pointed to by st must be at least the size returned by opus_encoder_get_size().
+  * This is intended for applications which use their own allocator instead of malloc.
+  * @see opus_encoder_create(),opus_encoder_get_size()
+  * To reset a previously initialized state, use the #OPUS_RESET_STATE CTL.
+  * @param [in] st <tt>OpusEncoder*</tt>: Encoder state
+  * @param [in] Fs <tt>opus_int32</tt>: Sampling rate of input signal (Hz)
+ *                                      This must be one of 8000, 12000, 16000,
+ *                                      24000, or 48000.
+  * @param [in] channels <tt>int</tt>: Number of channels (1 or 2) in input signal
+  * @param [in] application <tt>int</tt>: Coding mode (one of OPUS_APPLICATION_VOIP, OPUS_APPLICATION_AUDIO, or OPUS_APPLICATION_RESTRICTED_LOWDELAY)
+  * @retval #OPUS_OK Success or @ref opus_errorcodes
+  */
+OPUS_EXPORT int opus_encoder_init(
+    OpusEncoder *st,
+    opus_int32 Fs,
+    int channels,
+    int application
+) OPUS_ARG_NONNULL(1);
+
+/** Encodes an Opus frame.
+  * @param [in] st <tt>OpusEncoder*</tt>: Encoder state
+  * @param [in] pcm <tt>opus_int16*</tt>: Input signal (interleaved if 2 channels). length is frame_size*channels*sizeof(opus_int16)
+  * @param [in] frame_size <tt>int</tt>: Number of samples per channel in the
+  *                                      input signal.
+  *                                      This must be an Opus frame size for
+  *                                      the encoder's sampling rate.
+  *                                      For example, at 48 kHz the permitted
+  *                                      values are 120, 240, 480, 960, 1920,
+  *                                      and 2880.
+  *                                      Passing in a duration of less than
+  *                                      10 ms (480 samples at 48 kHz) will
+  *                                      prevent the encoder from using the LPC
+  *                                      or hybrid modes.
+  * @param [out] data <tt>unsigned char*</tt>: Output payload.
+  *                                            This must contain storage for at
+  *                                            least \a max_data_bytes.
+  * @param [in] max_data_bytes <tt>opus_int32</tt>: Size of the allocated
+  *                                                 memory for the output
+  *                                                 payload. This may be
+  *                                                 used to impose an upper limit on
+  *                                                 the instant bitrate, but should
+  *                                                 not be used as the only bitrate
+  *                                                 control. Use #OPUS_SET_BITRATE to
+  *                                                 control the bitrate.
+  * @returns The length of the encoded packet (in bytes) on success or a
+  *          negative error code (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_encode(
+    OpusEncoder *st,
+    const opus_int16 *pcm,
+    int frame_size,
+    unsigned char *data,
+    opus_int32 max_data_bytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+/** Encodes an Opus frame from floating point input.
+  * @param [in] st <tt>OpusEncoder*</tt>: Encoder state
+  * @param [in] pcm <tt>float*</tt>: Input in float format (interleaved if 2 channels), with a normal range of +/-1.0.
+  *          Samples with a range beyond +/-1.0 are supported but will
+  *          be clipped by decoders using the integer API and should
+  *          only be used if it is known that the far end supports
+  *          extended dynamic range.
+  *          length is frame_size*channels*sizeof(float)
+  * @param [in] frame_size <tt>int</tt>: Number of samples per channel in the
+  *                                      input signal.
+  *                                      This must be an Opus frame size for
+  *                                      the encoder's sampling rate.
+  *                                      For example, at 48 kHz the permitted
+  *                                      values are 120, 240, 480, 960, 1920,
+  *                                      and 2880.
+  *                                      Passing in a duration of less than
+  *                                      10 ms (480 samples at 48 kHz) will
+  *                                      prevent the encoder from using the LPC
+  *                                      or hybrid modes.
+  * @param [out] data <tt>unsigned char*</tt>: Output payload.
+  *                                            This must contain storage for at
+  *                                            least \a max_data_bytes.
+  * @param [in] max_data_bytes <tt>opus_int32</tt>: Size of the allocated
+  *                                                 memory for the output
+  *                                                 payload. This may be
+  *                                                 used to impose an upper limit on
+  *                                                 the instant bitrate, but should
+  *                                                 not be used as the only bitrate
+  *                                                 control. Use #OPUS_SET_BITRATE to
+  *                                                 control the bitrate.
+  * @returns The length of the encoded packet (in bytes) on success or a
+  *          negative error code (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_encode_float(
+    OpusEncoder *st,
+    const float *pcm,
+    int frame_size,
+    unsigned char *data,
+    opus_int32 max_data_bytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+/** Frees an <code>OpusEncoder</code> allocated by opus_encoder_create().
+  * @param[in] st <tt>OpusEncoder*</tt>: State to be freed.
+  */
+OPUS_EXPORT void opus_encoder_destroy(OpusEncoder *st);
+
+/** Perform a CTL function on an Opus encoder.
+  *
+  * Generally the request and subsequent arguments are generated
+  * by a convenience macro.
+  * @param st <tt>OpusEncoder*</tt>: Encoder state.
+  * @param request This and all remaining parameters should be replaced by one
+  *                of the convenience macros in @ref opus_genericctls or
+  *                @ref opus_encoderctls.
+  * @see opus_genericctls
+  * @see opus_encoderctls
+  */
+OPUS_EXPORT int opus_encoder_ctl(OpusEncoder *st, int request, ...) OPUS_ARG_NONNULL(1);
+/**@}*/
+
+/** @defgroup opus_decoder Opus Decoder
+  * @{
+  *
+  * @brief This page describes the process and functions used to decode Opus.
+  *
+  * The decoding process also starts with creating a decoder
+  * state. This can be done with:
+  * @code
+  * int          error;
+  * OpusDecoder *dec;
+  * dec = opus_decoder_create(Fs, channels, &error);
+  * @endcode
+  * where
+  * @li Fs is the sampling rate and must be 8000, 12000, 16000, 24000, or 48000
+  * @li channels is the number of channels (1 or 2)
+  * @li error will hold the error code in case of failure (or #OPUS_OK on success)
+  * @li the return value is a newly created decoder state to be used for decoding
+  *
+  * While opus_decoder_create() allocates memory for the state, it's also possible
+  * to initialize pre-allocated memory:
+  * @code
+  * int          size;
+  * int          error;
+  * OpusDecoder *dec;
+  * size = opus_decoder_get_size(channels);
+  * dec = malloc(size);
+  * error = opus_decoder_init(dec, Fs, channels);
+  * @endcode
+  * where opus_decoder_get_size() returns the required size for the decoder state. Note that
+  * future versions of this code may change the size, so no assuptions should be made about it.
+  *
+  * The decoder state is always continuous in memory and only a shallow copy is sufficient
+  * to copy it (e.g. memcpy())
+  *
+  * To decode a frame, opus_decode() or opus_decode_float() must be called with a packet of compressed audio data:
+  * @code
+  * frame_size = opus_decode(dec, packet, len, decoded, max_size, 0);
+  * @endcode
+  * where
+  *
+  * @li packet is the byte array containing the compressed data
+  * @li len is the exact number of bytes contained in the packet
+  * @li decoded is the decoded audio data in opus_int16 (or float for opus_decode_float())
+  * @li max_size is the max duration of the frame in samples (per channel) that can fit into the decoded_frame array
+  *
+  * opus_decode() and opus_decode_float() return the number of samples (per channel) decoded from the packet.
+  * If that value is negative, then an error has occurred. This can occur if the packet is corrupted or if the audio
+  * buffer is too small to hold the decoded audio.
+  *
+  * Opus is a stateful codec with overlapping blocks and as a result Opus
+  * packets are not coded independently of each other. Packets must be
+  * passed into the decoder serially and in the correct order for a correct
+  * decode. Lost packets can be replaced with loss concealment by calling
+  * the decoder with a null pointer and zero length for the missing packet.
+  *
+  * A single codec state may only be accessed from a single thread at
+  * a time and any required locking must be performed by the caller. Separate
+  * streams must be decoded with separate decoder states and can be decoded
+  * in parallel unless the library was compiled with NONTHREADSAFE_PSEUDOSTACK
+  * defined.
+  *
+  */
+
+/** Opus decoder state.
+  * This contains the complete state of an Opus decoder.
+  * It is position independent and can be freely copied.
+  * @see opus_decoder_create,opus_decoder_init
+  */
+typedef struct OpusDecoder OpusDecoder;
+
+/** Gets the size of an <code>OpusDecoder</code> structure.
+  * @param [in] channels <tt>int</tt>: Number of channels.
+  *                                    This must be 1 or 2.
+  * @returns The size in bytes.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_decoder_get_size(int channels);
+
+/** Allocates and initializes a decoder state.
+  * @param [in] Fs <tt>opus_int32</tt>: Sample rate to decode at (Hz).
+  *                                     This must be one of 8000, 12000, 16000,
+  *                                     24000, or 48000.
+  * @param [in] channels <tt>int</tt>: Number of channels (1 or 2) to decode
+  * @param [out] error <tt>int*</tt>: #OPUS_OK Success or @ref opus_errorcodes
+  *
+  * Internally Opus stores data at 48000 Hz, so that should be the default
+  * value for Fs. However, the decoder can efficiently decode to buffers
+  * at 8, 12, 16, and 24 kHz so if for some reason the caller cannot use
+  * data at the full sample rate, or knows the compressed data doesn't
+  * use the full frequency range, it can request decoding at a reduced
+  * rate. Likewise, the decoder is capable of filling in either mono or
+  * interleaved stereo pcm buffers, at the caller's request.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusDecoder *opus_decoder_create(
+    opus_int32 Fs,
+    int channels,
+    int *error
+);
+
+/** Initializes a previously allocated decoder state.
+  * The state must be at least the size returned by opus_decoder_get_size().
+  * This is intended for applications which use their own allocator instead of malloc. @see opus_decoder_create,opus_decoder_get_size
+  * To reset a previously initialized state, use the #OPUS_RESET_STATE CTL.
+  * @param [in] st <tt>OpusDecoder*</tt>: Decoder state.
+  * @param [in] Fs <tt>opus_int32</tt>: Sampling rate to decode to (Hz).
+  *                                     This must be one of 8000, 12000, 16000,
+  *                                     24000, or 48000.
+  * @param [in] channels <tt>int</tt>: Number of channels (1 or 2) to decode
+  * @retval #OPUS_OK Success or @ref opus_errorcodes
+  */
+OPUS_EXPORT int opus_decoder_init(
+    OpusDecoder *st,
+    opus_int32 Fs,
+    int channels
+) OPUS_ARG_NONNULL(1);
+
+/** Decode an Opus packet.
+  * @param [in] st <tt>OpusDecoder*</tt>: Decoder state
+  * @param [in] data <tt>char*</tt>: Input payload. Use a NULL pointer to indicate packet loss
+  * @param [in] len <tt>opus_int32</tt>: Number of bytes in payload*
+  * @param [out] pcm <tt>opus_int16*</tt>: Output signal (interleaved if 2 channels). length
+  *  is frame_size*channels*sizeof(opus_int16)
+  * @param [in] frame_size Number of samples per channel of available space in \a pcm.
+  *  If this is less than the maximum packet duration (120ms; 5760 for 48kHz), this function will
+  *  not be capable of decoding some packets. In the case of PLC (data==NULL) or FEC (decode_fec=1),
+  *  then frame_size needs to be exactly the duration of audio that is missing, otherwise the
+  *  decoder will not be in the optimal state to decode the next incoming packet. For the PLC and
+  *  FEC cases, frame_size <b>must</b> be a multiple of 2.5 ms.
+  * @param [in] decode_fec <tt>int</tt>: Flag (0 or 1) to request that any in-band forward error correction data be
+  *  decoded. If no such data is available, the frame is decoded as if it were lost.
+  * @returns Number of decoded samples or @ref opus_errorcodes
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_decode(
+    OpusDecoder *st,
+    const unsigned char *data,
+    opus_int32 len,
+    opus_int16 *pcm,
+    int frame_size,
+    int decode_fec
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+/** Decode an Opus packet with floating point output.
+  * @param [in] st <tt>OpusDecoder*</tt>: Decoder state
+  * @param [in] data <tt>char*</tt>: Input payload. Use a NULL pointer to indicate packet loss
+  * @param [in] len <tt>opus_int32</tt>: Number of bytes in payload
+  * @param [out] pcm <tt>float*</tt>: Output signal (interleaved if 2 channels). length
+  *  is frame_size*channels*sizeof(float)
+  * @param [in] frame_size Number of samples per channel of available space in \a pcm.
+  *  If this is less than the maximum packet duration (120ms; 5760 for 48kHz), this function will
+  *  not be capable of decoding some packets. In the case of PLC (data==NULL) or FEC (decode_fec=1),
+  *  then frame_size needs to be exactly the duration of audio that is missing, otherwise the
+  *  decoder will not be in the optimal state to decode the next incoming packet. For the PLC and
+  *  FEC cases, frame_size <b>must</b> be a multiple of 2.5 ms.
+  * @param [in] decode_fec <tt>int</tt>: Flag (0 or 1) to request that any in-band forward error correction data be
+  *  decoded. If no such data is available the frame is decoded as if it were lost.
+  * @returns Number of decoded samples or @ref opus_errorcodes
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_decode_float(
+    OpusDecoder *st,
+    const unsigned char *data,
+    opus_int32 len,
+    float *pcm,
+    int frame_size,
+    int decode_fec
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+/** Perform a CTL function on an Opus decoder.
+  *
+  * Generally the request and subsequent arguments are generated
+  * by a convenience macro.
+  * @param st <tt>OpusDecoder*</tt>: Decoder state.
+  * @param request This and all remaining parameters should be replaced by one
+  *                of the convenience macros in @ref opus_genericctls or
+  *                @ref opus_decoderctls.
+  * @see opus_genericctls
+  * @see opus_decoderctls
+  */
+OPUS_EXPORT int opus_decoder_ctl(OpusDecoder *st, int request, ...) OPUS_ARG_NONNULL(1);
+
+/** Frees an <code>OpusDecoder</code> allocated by opus_decoder_create().
+  * @param[in] st <tt>OpusDecoder*</tt>: State to be freed.
+  */
+OPUS_EXPORT void opus_decoder_destroy(OpusDecoder *st);
+
+/** Parse an opus packet into one or more frames.
+  * Opus_decode will perform this operation internally so most applications do
+  * not need to use this function.
+  * This function does not copy the frames, the returned pointers are pointers into
+  * the input packet.
+  * @param [in] data <tt>char*</tt>: Opus packet to be parsed
+  * @param [in] len <tt>opus_int32</tt>: size of data
+  * @param [out] out_toc <tt>char*</tt>: TOC pointer
+  * @param [out] frames <tt>char*[48]</tt> encapsulated frames
+  * @param [out] size <tt>opus_int16[48]</tt> sizes of the encapsulated frames
+  * @param [out] payload_offset <tt>int*</tt>: returns the position of the payload within the packet (in bytes)
+  * @returns number of frames
+  */
+OPUS_EXPORT int opus_packet_parse(
+   const unsigned char *data,
+   opus_int32 len,
+   unsigned char *out_toc,
+   const unsigned char *frames[48],
+   opus_int16 size[48],
+   int *payload_offset
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(5);
+
+/** Gets the bandwidth of an Opus packet.
+  * @param [in] data <tt>char*</tt>: Opus packet
+  * @retval OPUS_BANDWIDTH_NARROWBAND Narrowband (4kHz bandpass)
+  * @retval OPUS_BANDWIDTH_MEDIUMBAND Mediumband (6kHz bandpass)
+  * @retval OPUS_BANDWIDTH_WIDEBAND Wideband (8kHz bandpass)
+  * @retval OPUS_BANDWIDTH_SUPERWIDEBAND Superwideband (12kHz bandpass)
+  * @retval OPUS_BANDWIDTH_FULLBAND Fullband (20kHz bandpass)
+  * @retval OPUS_INVALID_PACKET The compressed data passed is corrupted or of an unsupported type
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_packet_get_bandwidth(const unsigned char *data) OPUS_ARG_NONNULL(1);
+
+/** Gets the number of samples per frame from an Opus packet.
+  * @param [in] data <tt>char*</tt>: Opus packet.
+  *                                  This must contain at least one byte of
+  *                                  data.
+  * @param [in] Fs <tt>opus_int32</tt>: Sampling rate in Hz.
+  *                                     This must be a multiple of 400, or
+  *                                     inaccurate results will be returned.
+  * @returns Number of samples per frame.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_packet_get_samples_per_frame(const unsigned char *data, opus_int32 Fs) OPUS_ARG_NONNULL(1);
+
+/** Gets the number of channels from an Opus packet.
+  * @param [in] data <tt>char*</tt>: Opus packet
+  * @returns Number of channels
+  * @retval OPUS_INVALID_PACKET The compressed data passed is corrupted or of an unsupported type
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_packet_get_nb_channels(const unsigned char *data) OPUS_ARG_NONNULL(1);
+
+/** Gets the number of frames in an Opus packet.
+  * @param [in] packet <tt>char*</tt>: Opus packet
+  * @param [in] len <tt>opus_int32</tt>: Length of packet
+  * @returns Number of frames
+  * @retval OPUS_BAD_ARG Insufficient data was passed to the function
+  * @retval OPUS_INVALID_PACKET The compressed data passed is corrupted or of an unsupported type
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_packet_get_nb_frames(const unsigned char packet[], opus_int32 len) OPUS_ARG_NONNULL(1);
+
+/** Gets the number of samples of an Opus packet.
+  * @param [in] packet <tt>char*</tt>: Opus packet
+  * @param [in] len <tt>opus_int32</tt>: Length of packet
+  * @param [in] Fs <tt>opus_int32</tt>: Sampling rate in Hz.
+  *                                     This must be a multiple of 400, or
+  *                                     inaccurate results will be returned.
+  * @returns Number of samples
+  * @retval OPUS_BAD_ARG Insufficient data was passed to the function
+  * @retval OPUS_INVALID_PACKET The compressed data passed is corrupted or of an unsupported type
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_packet_get_nb_samples(const unsigned char packet[], opus_int32 len, opus_int32 Fs) OPUS_ARG_NONNULL(1);
+
+/** Gets the number of samples of an Opus packet.
+  * @param [in] dec <tt>OpusDecoder*</tt>: Decoder state
+  * @param [in] packet <tt>char*</tt>: Opus packet
+  * @param [in] len <tt>opus_int32</tt>: Length of packet
+  * @returns Number of samples
+  * @retval OPUS_BAD_ARG Insufficient data was passed to the function
+  * @retval OPUS_INVALID_PACKET The compressed data passed is corrupted or of an unsupported type
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_decoder_get_nb_samples(const OpusDecoder *dec, const unsigned char packet[], opus_int32 len) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2);
+
+/** Applies soft-clipping to bring a float signal within the [-1,1] range. If
+  * the signal is already in that range, nothing is done. If there are values
+  * outside of [-1,1], then the signal is clipped as smoothly as possible to
+  * both fit in the range and avoid creating excessive distortion in the
+  * process.
+  * @param [in,out] pcm <tt>float*</tt>: Input PCM and modified PCM
+  * @param [in] frame_size <tt>int</tt> Number of samples per channel to process
+  * @param [in] channels <tt>int</tt>: Number of channels
+  * @param [in,out] softclip_mem <tt>float*</tt>: State memory for the soft clipping process (one float per channel, initialized to zero)
+  */
+OPUS_EXPORT void opus_pcm_soft_clip(float *pcm, int frame_size, int channels, float *softclip_mem);
+
+
+/**@}*/
+
+/** @defgroup opus_repacketizer Repacketizer
+  * @{
+  *
+  * The repacketizer can be used to merge multiple Opus packets into a single
+  * packet or alternatively to split Opus packets that have previously been
+  * merged. Splitting valid Opus packets is always guaranteed to succeed,
+  * whereas merging valid packets only succeeds if all frames have the same
+  * mode, bandwidth, and frame size, and when the total duration of the merged
+  * packet is no more than 120 ms. The 120 ms limit comes from the
+  * specification and limits decoder memory requirements at a point where
+  * framing overhead becomes negligible.
+  *
+  * The repacketizer currently only operates on elementary Opus
+  * streams. It will not manipualte multistream packets successfully, except in
+  * the degenerate case where they consist of data from a single stream.
+  *
+  * The repacketizing process starts with creating a repacketizer state, either
+  * by calling opus_repacketizer_create() or by allocating the memory yourself,
+  * e.g.,
+  * @code
+  * OpusRepacketizer *rp;
+  * rp = (OpusRepacketizer*)malloc(opus_repacketizer_get_size());
+  * if (rp != NULL)
+  *     opus_repacketizer_init(rp);
+  * @endcode
+  *
+  * Then the application should submit packets with opus_repacketizer_cat(),
+  * extract new packets with opus_repacketizer_out() or
+  * opus_repacketizer_out_range(), and then reset the state for the next set of
+  * input packets via opus_repacketizer_init().
+  *
+  * For example, to split a sequence of packets into individual frames:
+  * @code
+  * unsigned char *data;
+  * int len;
+  * while (get_next_packet(&data, &len))
+  * {
+  *   unsigned char out[1276];
+  *   opus_int32 out_len;
+  *   int nb_frames;
+  *   int err;
+  *   int i;
+  *   err = opus_repacketizer_cat(rp, data, len);
+  *   if (err != OPUS_OK)
+  *   {
+  *     release_packet(data);
+  *     return err;
+  *   }
+  *   nb_frames = opus_repacketizer_get_nb_frames(rp);
+  *   for (i = 0; i < nb_frames; i++)
+  *   {
+  *     out_len = opus_repacketizer_out_range(rp, i, i+1, out, sizeof(out));
+  *     if (out_len < 0)
+  *     {
+  *        release_packet(data);
+  *        return (int)out_len;
+  *     }
+  *     output_next_packet(out, out_len);
+  *   }
+  *   opus_repacketizer_init(rp);
+  *   release_packet(data);
+  * }
+  * @endcode
+  *
+  * Alternatively, to combine a sequence of frames into packets that each
+  * contain up to <code>TARGET_DURATION_MS</code> milliseconds of data:
+  * @code
+  * // The maximum number of packets with duration TARGET_DURATION_MS occurs
+  * // when the frame size is 2.5 ms, for a total of (TARGET_DURATION_MS*2/5)
+  * // packets.
+  * unsigned char *data[(TARGET_DURATION_MS*2/5)+1];
+  * opus_int32 len[(TARGET_DURATION_MS*2/5)+1];
+  * int nb_packets;
+  * unsigned char out[1277*(TARGET_DURATION_MS*2/2)];
+  * opus_int32 out_len;
+  * int prev_toc;
+  * nb_packets = 0;
+  * while (get_next_packet(data+nb_packets, len+nb_packets))
+  * {
+  *   int nb_frames;
+  *   int err;
+  *   nb_frames = opus_packet_get_nb_frames(data[nb_packets], len[nb_packets]);
+  *   if (nb_frames < 1)
+  *   {
+  *     release_packets(data, nb_packets+1);
+  *     return nb_frames;
+  *   }
+  *   nb_frames += opus_repacketizer_get_nb_frames(rp);
+  *   // If adding the next packet would exceed our target, or it has an
+  *   // incompatible TOC sequence, output the packets we already have before
+  *   // submitting it.
+  *   // N.B., The nb_packets > 0 check ensures we've submitted at least one
+  *   // packet since the last call to opus_repacketizer_init(). Otherwise a
+  *   // single packet longer than TARGET_DURATION_MS would cause us to try to
+  *   // output an (invalid) empty packet. It also ensures that prev_toc has
+  *   // been set to a valid value. Additionally, len[nb_packets] > 0 is
+  *   // guaranteed by the call to opus_packet_get_nb_frames() above, so the
+  *   // reference to data[nb_packets][0] should be valid.
+  *   if (nb_packets > 0 && (
+  *       ((prev_toc & 0xFC) != (data[nb_packets][0] & 0xFC)) ||
+  *       opus_packet_get_samples_per_frame(data[nb_packets], 48000)*nb_frames >
+  *       TARGET_DURATION_MS*48))
+  *   {
+  *     out_len = opus_repacketizer_out(rp, out, sizeof(out));
+  *     if (out_len < 0)
+  *     {
+  *        release_packets(data, nb_packets+1);
+  *        return (int)out_len;
+  *     }
+  *     output_next_packet(out, out_len);
+  *     opus_repacketizer_init(rp);
+  *     release_packets(data, nb_packets);
+  *     data[0] = data[nb_packets];
+  *     len[0] = len[nb_packets];
+  *     nb_packets = 0;
+  *   }
+  *   err = opus_repacketizer_cat(rp, data[nb_packets], len[nb_packets]);
+  *   if (err != OPUS_OK)
+  *   {
+  *     release_packets(data, nb_packets+1);
+  *     return err;
+  *   }
+  *   prev_toc = data[nb_packets][0];
+  *   nb_packets++;
+  * }
+  * // Output the final, partial packet.
+  * if (nb_packets > 0)
+  * {
+  *   out_len = opus_repacketizer_out(rp, out, sizeof(out));
+  *   release_packets(data, nb_packets);
+  *   if (out_len < 0)
+  *     return (int)out_len;
+  *   output_next_packet(out, out_len);
+  * }
+  * @endcode
+  *
+  * An alternate way of merging packets is to simply call opus_repacketizer_cat()
+  * unconditionally until it fails. At that point, the merged packet can be
+  * obtained with opus_repacketizer_out() and the input packet for which
+  * opus_repacketizer_cat() needs to be re-added to a newly reinitialized
+  * repacketizer state.
+  */
+
+typedef struct OpusRepacketizer OpusRepacketizer;
+
+/** Gets the size of an <code>OpusRepacketizer</code> structure.
+  * @returns The size in bytes.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_repacketizer_get_size(void);
+
+/** (Re)initializes a previously allocated repacketizer state.
+  * The state must be at least the size returned by opus_repacketizer_get_size().
+  * This can be used for applications which use their own allocator instead of
+  * malloc().
+  * It must also be called to reset the queue of packets waiting to be
+  * repacketized, which is necessary if the maximum packet duration of 120 ms
+  * is reached or if you wish to submit packets with a different Opus
+  * configuration (coding mode, audio bandwidth, frame size, or channel count).
+  * Failure to do so will prevent a new packet from being added with
+  * opus_repacketizer_cat().
+  * @see opus_repacketizer_create
+  * @see opus_repacketizer_get_size
+  * @see opus_repacketizer_cat
+  * @param rp <tt>OpusRepacketizer*</tt>: The repacketizer state to
+  *                                       (re)initialize.
+  * @returns A pointer to the same repacketizer state that was passed in.
+  */
+OPUS_EXPORT OpusRepacketizer *opus_repacketizer_init(OpusRepacketizer *rp) OPUS_ARG_NONNULL(1);
+
+/** Allocates memory and initializes the new repacketizer with
+ * opus_repacketizer_init().
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusRepacketizer *opus_repacketizer_create(void);
+
+/** Frees an <code>OpusRepacketizer</code> allocated by
+  * opus_repacketizer_create().
+  * @param[in] rp <tt>OpusRepacketizer*</tt>: State to be freed.
+  */
+OPUS_EXPORT void opus_repacketizer_destroy(OpusRepacketizer *rp);
+
+/** Add a packet to the current repacketizer state.
+  * This packet must match the configuration of any packets already submitted
+  * for repacketization since the last call to opus_repacketizer_init().
+  * This means that it must have the same coding mode, audio bandwidth, frame
+  * size, and channel count.
+  * This can be checked in advance by examining the top 6 bits of the first
+  * byte of the packet, and ensuring they match the top 6 bits of the first
+  * byte of any previously submitted packet.
+  * The total duration of audio in the repacketizer state also must not exceed
+  * 120 ms, the maximum duration of a single packet, after adding this packet.
+  *
+  * The contents of the current repacketizer state can be extracted into new
+  * packets using opus_repacketizer_out() or opus_repacketizer_out_range().
+  *
+  * In order to add a packet with a different configuration or to add more
+  * audio beyond 120 ms, you must clear the repacketizer state by calling
+  * opus_repacketizer_init().
+  * If a packet is too large to add to the current repacketizer state, no part
+  * of it is added, even if it contains multiple frames, some of which might
+  * fit.
+  * If you wish to be able to add parts of such packets, you should first use
+  * another repacketizer to split the packet into pieces and add them
+  * individually.
+  * @see opus_repacketizer_out_range
+  * @see opus_repacketizer_out
+  * @see opus_repacketizer_init
+  * @param rp <tt>OpusRepacketizer*</tt>: The repacketizer state to which to
+  *                                       add the packet.
+  * @param[in] data <tt>const unsigned char*</tt>: The packet data.
+  *                                                The application must ensure
+  *                                                this pointer remains valid
+  *                                                until the next call to
+  *                                                opus_repacketizer_init() or
+  *                                                opus_repacketizer_destroy().
+  * @param len <tt>opus_int32</tt>: The number of bytes in the packet data.
+  * @returns An error code indicating whether or not the operation succeeded.
+  * @retval #OPUS_OK The packet's contents have been added to the repacketizer
+  *                  state.
+  * @retval #OPUS_INVALID_PACKET The packet did not have a valid TOC sequence,
+  *                              the packet's TOC sequence was not compatible
+  *                              with previously submitted packets (because
+  *                              the coding mode, audio bandwidth, frame size,
+  *                              or channel count did not match), or adding
+  *                              this packet would increase the total amount of
+  *                              audio stored in the repacketizer state to more
+  *                              than 120 ms.
+  */
+OPUS_EXPORT int opus_repacketizer_cat(OpusRepacketizer *rp, const unsigned char *data, opus_int32 len) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2);
+
+
+/** Construct a new packet from data previously submitted to the repacketizer
+  * state via opus_repacketizer_cat().
+  * @param rp <tt>OpusRepacketizer*</tt>: The repacketizer state from which to
+  *                                       construct the new packet.
+  * @param begin <tt>int</tt>: The index of the first frame in the current
+  *                            repacketizer state to include in the output.
+  * @param end <tt>int</tt>: One past the index of the last frame in the
+  *                          current repacketizer state to include in the
+  *                          output.
+  * @param[out] data <tt>const unsigned char*</tt>: The buffer in which to
+  *                                                 store the output packet.
+  * @param maxlen <tt>opus_int32</tt>: The maximum number of bytes to store in
+  *                                    the output buffer. In order to guarantee
+  *                                    success, this should be at least
+  *                                    <code>1276</code> for a single frame,
+  *                                    or for multiple frames,
+  *                                    <code>1277*(end-begin)</code>.
+  *                                    However, <code>1*(end-begin)</code> plus
+  *                                    the size of all packet data submitted to
+  *                                    the repacketizer since the last call to
+  *                                    opus_repacketizer_init() or
+  *                                    opus_repacketizer_create() is also
+  *                                    sufficient, and possibly much smaller.
+  * @returns The total size of the output packet on success, or an error code
+  *          on failure.
+  * @retval #OPUS_BAD_ARG <code>[begin,end)</code> was an invalid range of
+  *                       frames (begin < 0, begin >= end, or end >
+  *                       opus_repacketizer_get_nb_frames()).
+  * @retval #OPUS_BUFFER_TOO_SMALL \a maxlen was insufficient to contain the
+  *                                complete output packet.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_repacketizer_out_range(OpusRepacketizer *rp, int begin, int end, unsigned char *data, opus_int32 maxlen) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+/** Return the total number of frames contained in packet data submitted to
+  * the repacketizer state so far via opus_repacketizer_cat() since the last
+  * call to opus_repacketizer_init() or opus_repacketizer_create().
+  * This defines the valid range of packets that can be extracted with
+  * opus_repacketizer_out_range() or opus_repacketizer_out().
+  * @param rp <tt>OpusRepacketizer*</tt>: The repacketizer state containing the
+  *                                       frames.
+  * @returns The total number of frames contained in the packet data submitted
+  *          to the repacketizer state.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_repacketizer_get_nb_frames(OpusRepacketizer *rp) OPUS_ARG_NONNULL(1);
+
+/** Construct a new packet from data previously submitted to the repacketizer
+  * state via opus_repacketizer_cat().
+  * This is a convenience routine that returns all the data submitted so far
+  * in a single packet.
+  * It is equivalent to calling
+  * @code
+  * opus_repacketizer_out_range(rp, 0, opus_repacketizer_get_nb_frames(rp),
+  *                             data, maxlen)
+  * @endcode
+  * @param rp <tt>OpusRepacketizer*</tt>: The repacketizer state from which to
+  *                                       construct the new packet.
+  * @param[out] data <tt>const unsigned char*</tt>: The buffer in which to
+  *                                                 store the output packet.
+  * @param maxlen <tt>opus_int32</tt>: The maximum number of bytes to store in
+  *                                    the output buffer. In order to guarantee
+  *                                    success, this should be at least
+  *                                    <code>1277*opus_repacketizer_get_nb_frames(rp)</code>.
+  *                                    However,
+  *                                    <code>1*opus_repacketizer_get_nb_frames(rp)</code>
+  *                                    plus the size of all packet data
+  *                                    submitted to the repacketizer since the
+  *                                    last call to opus_repacketizer_init() or
+  *                                    opus_repacketizer_create() is also
+  *                                    sufficient, and possibly much smaller.
+  * @returns The total size of the output packet on success, or an error code
+  *          on failure.
+  * @retval #OPUS_BUFFER_TOO_SMALL \a maxlen was insufficient to contain the
+  *                                complete output packet.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_repacketizer_out(OpusRepacketizer *rp, unsigned char *data, opus_int32 maxlen) OPUS_ARG_NONNULL(1);
+
+/** Pads a given Opus packet to a larger size (possibly changing the TOC sequence).
+  * @param[in,out] data <tt>const unsigned char*</tt>: The buffer containing the
+  *                                                   packet to pad.
+  * @param len <tt>opus_int32</tt>: The size of the packet.
+  *                                 This must be at least 1.
+  * @param new_len <tt>opus_int32</tt>: The desired size of the packet after padding.
+  *                                 This must be at least as large as len.
+  * @returns an error code
+  * @retval #OPUS_OK \a on success.
+  * @retval #OPUS_BAD_ARG \a len was less than 1 or new_len was less than len.
+  * @retval #OPUS_INVALID_PACKET \a data did not contain a valid Opus packet.
+  */
+OPUS_EXPORT int opus_packet_pad(unsigned char *data, opus_int32 len, opus_int32 new_len);
+
+/** Remove all padding from a given Opus packet and rewrite the TOC sequence to
+  * minimize space usage.
+  * @param[in,out] data <tt>const unsigned char*</tt>: The buffer containing the
+  *                                                   packet to strip.
+  * @param len <tt>opus_int32</tt>: The size of the packet.
+  *                                 This must be at least 1.
+  * @returns The new size of the output packet on success, or an error code
+  *          on failure.
+  * @retval #OPUS_BAD_ARG \a len was less than 1.
+  * @retval #OPUS_INVALID_PACKET \a data did not contain a valid Opus packet.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_packet_unpad(unsigned char *data, opus_int32 len);
+
+/** Pads a given Opus multi-stream packet to a larger size (possibly changing the TOC sequence).
+  * @param[in,out] data <tt>const unsigned char*</tt>: The buffer containing the
+  *                                                   packet to pad.
+  * @param len <tt>opus_int32</tt>: The size of the packet.
+  *                                 This must be at least 1.
+  * @param new_len <tt>opus_int32</tt>: The desired size of the packet after padding.
+  *                                 This must be at least 1.
+  * @param nb_streams <tt>opus_int32</tt>: The number of streams (not channels) in the packet.
+  *                                 This must be at least as large as len.
+  * @returns an error code
+  * @retval #OPUS_OK \a on success.
+  * @retval #OPUS_BAD_ARG \a len was less than 1.
+  * @retval #OPUS_INVALID_PACKET \a data did not contain a valid Opus packet.
+  */
+OPUS_EXPORT int opus_multistream_packet_pad(unsigned char *data, opus_int32 len, opus_int32 new_len, int nb_streams);
+
+/** Remove all padding from a given Opus multi-stream packet and rewrite the TOC sequence to
+  * minimize space usage.
+  * @param[in,out] data <tt>const unsigned char*</tt>: The buffer containing the
+  *                                                   packet to strip.
+  * @param len <tt>opus_int32</tt>: The size of the packet.
+  *                                 This must be at least 1.
+  * @param nb_streams <tt>opus_int32</tt>: The number of streams (not channels) in the packet.
+  *                                 This must be at least 1.
+  * @returns The new size of the output packet on success, or an error code
+  *          on failure.
+  * @retval #OPUS_BAD_ARG \a len was less than 1 or new_len was less than len.
+  * @retval #OPUS_INVALID_PACKET \a data did not contain a valid Opus packet.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_multistream_packet_unpad(unsigned char *data, opus_int32 len, int nb_streams);
+
+/**@}*/
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* OPUS_H */

lib/codecs/inc/opus/opus_custom.h

@@ -0,0 +1,343 @@
+/* Copyright (c) 2007-2008 CSIRO
+   Copyright (c) 2007-2009 Xiph.Org Foundation
+   Copyright (c) 2008-2012 Gregory Maxwell
+   Written by Jean-Marc Valin and Gregory Maxwell */
+/*
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
+   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+/**
+  @file opus_custom.h
+  @brief Opus-Custom reference implementation API
+ */
+
+#ifndef OPUS_CUSTOM_H
+#define OPUS_CUSTOM_H
+
+#include "opus_defines.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#ifdef CUSTOM_MODES
+# define OPUS_CUSTOM_EXPORT OPUS_EXPORT
+# define OPUS_CUSTOM_EXPORT_STATIC OPUS_EXPORT
+#else
+# define OPUS_CUSTOM_EXPORT
+# ifdef OPUS_BUILD
+#  define OPUS_CUSTOM_EXPORT_STATIC static OPUS_INLINE
+# else
+#  define OPUS_CUSTOM_EXPORT_STATIC
+# endif
+#endif
+
+/** @defgroup opus_custom Opus Custom
+  * @{
+  *  Opus Custom is an optional part of the Opus specification and
+  * reference implementation which uses a distinct API from the regular
+  * API and supports frame sizes that are not normally supported.\ Use
+  * of Opus Custom is discouraged for all but very special applications
+  * for which a frame size different from 2.5, 5, 10, or 20 ms is needed
+  * (for either complexity or latency reasons) and where interoperability
+  * is less important.
+  *
+  * In addition to the interoperability limitations the use of Opus custom
+  * disables a substantial chunk of the codec and generally lowers the
+  * quality available at a given bitrate. Normally when an application needs
+  * a different frame size from the codec it should buffer to match the
+  * sizes but this adds a small amount of delay which may be important
+  * in some very low latency applications. Some transports (especially
+  * constant rate RF transports) may also work best with frames of
+  * particular durations.
+  *
+  * Libopus only supports custom modes if they are enabled at compile time.
+  *
+  * The Opus Custom API is similar to the regular API but the
+  * @ref opus_encoder_create and @ref opus_decoder_create calls take
+  * an additional mode parameter which is a structure produced by
+  * a call to @ref opus_custom_mode_create. Both the encoder and decoder
+  * must create a mode using the same sample rate (fs) and frame size
+  * (frame size) so these parameters must either be signaled out of band
+  * or fixed in a particular implementation.
+  *
+  * Similar to regular Opus the custom modes support on the fly frame size
+  * switching, but the sizes available depend on the particular frame size in
+  * use. For some initial frame sizes on a single on the fly size is available.
+  */
+
+/** Contains the state of an encoder. One encoder state is needed
+    for each stream. It is initialized once at the beginning of the
+    stream. Do *not* re-initialize the state for every frame.
+   @brief Encoder state
+ */
+typedef struct OpusCustomEncoder OpusCustomEncoder;
+
+/** State of the decoder. One decoder state is needed for each stream.
+    It is initialized once at the beginning of the stream. Do *not*
+    re-initialize the state for every frame.
+   @brief Decoder state
+ */
+typedef struct OpusCustomDecoder OpusCustomDecoder;
+
+/** The mode contains all the information necessary to create an
+    encoder. Both the encoder and decoder need to be initialized
+    with exactly the same mode, otherwise the output will be
+    corrupted. The mode MUST NOT BE DESTROYED until the encoders and
+    decoders that use it are destroyed as well.
+   @brief Mode configuration
+ */
+typedef struct OpusCustomMode OpusCustomMode;
+
+/** Creates a new mode struct. This will be passed to an encoder or
+  * decoder. The mode MUST NOT BE DESTROYED until the encoders and
+  * decoders that use it are destroyed as well.
+  * @param [in] Fs <tt>int</tt>: Sampling rate (8000 to 96000 Hz)
+  * @param [in] frame_size <tt>int</tt>: Number of samples (per channel) to encode in each
+  *        packet (64 - 1024, prime factorization must contain zero or more 2s, 3s, or 5s and no other primes)
+  * @param [out] error <tt>int*</tt>: Returned error code (if NULL, no error will be returned)
+  * @return A newly created mode
+  */
+OPUS_CUSTOM_EXPORT OPUS_WARN_UNUSED_RESULT OpusCustomMode *opus_custom_mode_create(opus_int32 Fs, int frame_size, int *error);
+
+/** Destroys a mode struct. Only call this after all encoders and
+  * decoders using this mode are destroyed as well.
+  * @param [in] mode <tt>OpusCustomMode*</tt>: Mode to be freed.
+  */
+OPUS_CUSTOM_EXPORT void opus_custom_mode_destroy(OpusCustomMode *mode);
+
+
+#if !defined(OPUS_BUILD) || defined(CELT_ENCODER_C)
+
+/* Encoder */
+/** Gets the size of an OpusCustomEncoder structure.
+  * @param [in] mode <tt>OpusCustomMode *</tt>: Mode configuration
+  * @param [in] channels <tt>int</tt>: Number of channels
+  * @returns size
+  */
+OPUS_CUSTOM_EXPORT_STATIC OPUS_WARN_UNUSED_RESULT int opus_custom_encoder_get_size(
+    const OpusCustomMode *mode,
+    int channels
+) OPUS_ARG_NONNULL(1);
+
+# ifdef CUSTOM_MODES
+/** Initializes a previously allocated encoder state
+  * The memory pointed to by st must be the size returned by opus_custom_encoder_get_size.
+  * This is intended for applications which use their own allocator instead of malloc.
+  * @see opus_custom_encoder_create(),opus_custom_encoder_get_size()
+  * To reset a previously initialized state use the OPUS_RESET_STATE CTL.
+  * @param [in] st <tt>OpusCustomEncoder*</tt>: Encoder state
+  * @param [in] mode <tt>OpusCustomMode *</tt>: Contains all the information about the characteristics of
+  *  the stream (must be the same characteristics as used for the
+  *  decoder)
+  * @param [in] channels <tt>int</tt>: Number of channels
+  * @return OPUS_OK Success or @ref opus_errorcodes
+  */
+OPUS_CUSTOM_EXPORT int opus_custom_encoder_init(
+    OpusCustomEncoder *st,
+    const OpusCustomMode *mode,
+    int channels
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2);
+# endif
+#endif
+
+
+/** Creates a new encoder state. Each stream needs its own encoder
+  * state (can't be shared across simultaneous streams).
+  * @param [in] mode <tt>OpusCustomMode*</tt>: Contains all the information about the characteristics of
+  *  the stream (must be the same characteristics as used for the
+  *  decoder)
+  * @param [in] channels <tt>int</tt>: Number of channels
+  * @param [out] error <tt>int*</tt>: Returns an error code
+  * @return Newly created encoder state.
+*/
+OPUS_CUSTOM_EXPORT OPUS_WARN_UNUSED_RESULT OpusCustomEncoder *opus_custom_encoder_create(
+    const OpusCustomMode *mode,
+    int channels,
+    int *error
+) OPUS_ARG_NONNULL(1);
+
+
+/** Destroys an encoder state.
+  * @param[in] st <tt>OpusCustomEncoder*</tt>: State to be freed.
+  */
+OPUS_CUSTOM_EXPORT void opus_custom_encoder_destroy(OpusCustomEncoder *st);
+
+/** Encodes a frame of audio.
+  * @param [in] st <tt>OpusCustomEncoder*</tt>: Encoder state
+  * @param [in] pcm <tt>float*</tt>: PCM audio in float format, with a normal range of +/-1.0.
+  *          Samples with a range beyond +/-1.0 are supported but will
+  *          be clipped by decoders using the integer API and should
+  *          only be used if it is known that the far end supports
+  *          extended dynamic range. There must be exactly
+  *          frame_size samples per channel.
+  * @param [in] frame_size <tt>int</tt>: Number of samples per frame of input signal
+  * @param [out] compressed <tt>char *</tt>: The compressed data is written here. This may not alias pcm and must be at least maxCompressedBytes long.
+  * @param [in] maxCompressedBytes <tt>int</tt>: Maximum number of bytes to use for compressing the frame
+  *          (can change from one frame to another)
+  * @return Number of bytes written to "compressed".
+  *       If negative, an error has occurred (see error codes). It is IMPORTANT that
+  *       the length returned be somehow transmitted to the decoder. Otherwise, no
+  *       decoding is possible.
+  */
+OPUS_CUSTOM_EXPORT OPUS_WARN_UNUSED_RESULT int opus_custom_encode_float(
+    OpusCustomEncoder *st,
+    const float *pcm,
+    int frame_size,
+    unsigned char *compressed,
+    int maxCompressedBytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+/** Encodes a frame of audio.
+  * @param [in] st <tt>OpusCustomEncoder*</tt>: Encoder state
+  * @param [in] pcm <tt>opus_int16*</tt>: PCM audio in signed 16-bit format (native endian).
+  *          There must be exactly frame_size samples per channel.
+  * @param [in] frame_size <tt>int</tt>: Number of samples per frame of input signal
+  * @param [out] compressed <tt>char *</tt>: The compressed data is written here. This may not alias pcm and must be at least maxCompressedBytes long.
+  * @param [in] maxCompressedBytes <tt>int</tt>: Maximum number of bytes to use for compressing the frame
+  *          (can change from one frame to another)
+  * @return Number of bytes written to "compressed".
+  *       If negative, an error has occurred (see error codes). It is IMPORTANT that
+  *       the length returned be somehow transmitted to the decoder. Otherwise, no
+  *       decoding is possible.
+ */
+OPUS_CUSTOM_EXPORT OPUS_WARN_UNUSED_RESULT int opus_custom_encode(
+    OpusCustomEncoder *st,
+    const opus_int16 *pcm,
+    int frame_size,
+    unsigned char *compressed,
+    int maxCompressedBytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+/** Perform a CTL function on an Opus custom encoder.
+  *
+  * Generally the request and subsequent arguments are generated
+  * by a convenience macro.
+  * @see opus_encoderctls
+  */
+OPUS_CUSTOM_EXPORT int opus_custom_encoder_ctl(OpusCustomEncoder * OPUS_RESTRICT st, int request, ...) OPUS_ARG_NONNULL(1);
+
+
+#if !defined(OPUS_BUILD) || defined(CELT_DECODER_C)
+/* Decoder */
+
+/** Gets the size of an OpusCustomDecoder structure.
+  * @param [in] mode <tt>OpusCustomMode *</tt>: Mode configuration
+  * @param [in] channels <tt>int</tt>: Number of channels
+  * @returns size
+  */
+OPUS_CUSTOM_EXPORT_STATIC OPUS_WARN_UNUSED_RESULT int opus_custom_decoder_get_size(
+    const OpusCustomMode *mode,
+    int channels
+) OPUS_ARG_NONNULL(1);
+
+/** Initializes a previously allocated decoder state
+  * The memory pointed to by st must be the size returned by opus_custom_decoder_get_size.
+  * This is intended for applications which use their own allocator instead of malloc.
+  * @see opus_custom_decoder_create(),opus_custom_decoder_get_size()
+  * To reset a previously initialized state use the OPUS_RESET_STATE CTL.
+  * @param [in] st <tt>OpusCustomDecoder*</tt>: Decoder state
+  * @param [in] mode <tt>OpusCustomMode *</tt>: Contains all the information about the characteristics of
+  *  the stream (must be the same characteristics as used for the
+  *  encoder)
+  * @param [in] channels <tt>int</tt>: Number of channels
+  * @return OPUS_OK Success or @ref opus_errorcodes
+  */
+OPUS_CUSTOM_EXPORT_STATIC int opus_custom_decoder_init(
+    OpusCustomDecoder *st,
+    const OpusCustomMode *mode,
+    int channels
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2);
+
+#endif
+
+
+/** Creates a new decoder state. Each stream needs its own decoder state (can't
+  * be shared across simultaneous streams).
+  * @param [in] mode <tt>OpusCustomMode</tt>: Contains all the information about the characteristics of the
+  *          stream (must be the same characteristics as used for the encoder)
+  * @param [in] channels <tt>int</tt>: Number of channels
+  * @param [out] error <tt>int*</tt>: Returns an error code
+  * @return Newly created decoder state.
+  */
+OPUS_CUSTOM_EXPORT OPUS_WARN_UNUSED_RESULT OpusCustomDecoder *opus_custom_decoder_create(
+    const OpusCustomMode *mode,
+    int channels,
+    int *error
+) OPUS_ARG_NONNULL(1);
+
+/** Destroys a decoder state.
+  * @param[in] st <tt>OpusCustomDecoder*</tt>: State to be freed.
+  */
+OPUS_CUSTOM_EXPORT void opus_custom_decoder_destroy(OpusCustomDecoder *st);
+
+/** Decode an opus custom frame with floating point output
+  * @param [in] st <tt>OpusCustomDecoder*</tt>: Decoder state
+  * @param [in] data <tt>char*</tt>: Input payload. Use a NULL pointer to indicate packet loss
+  * @param [in] len <tt>int</tt>: Number of bytes in payload
+  * @param [out] pcm <tt>float*</tt>: Output signal (interleaved if 2 channels). length
+  *  is frame_size*channels*sizeof(float)
+  * @param [in] frame_size Number of samples per channel of available space in *pcm.
+  * @returns Number of decoded samples or @ref opus_errorcodes
+  */
+OPUS_CUSTOM_EXPORT OPUS_WARN_UNUSED_RESULT int opus_custom_decode_float(
+    OpusCustomDecoder *st,
+    const unsigned char *data,
+    int len,
+    float *pcm,
+    int frame_size
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+/** Decode an opus custom frame
+  * @param [in] st <tt>OpusCustomDecoder*</tt>: Decoder state
+  * @param [in] data <tt>char*</tt>: Input payload. Use a NULL pointer to indicate packet loss
+  * @param [in] len <tt>int</tt>: Number of bytes in payload
+  * @param [out] pcm <tt>opus_int16*</tt>: Output signal (interleaved if 2 channels). length
+  *  is frame_size*channels*sizeof(opus_int16)
+  * @param [in] frame_size Number of samples per channel of available space in *pcm.
+  * @returns Number of decoded samples or @ref opus_errorcodes
+  */
+OPUS_CUSTOM_EXPORT OPUS_WARN_UNUSED_RESULT int opus_custom_decode(
+    OpusCustomDecoder *st,
+    const unsigned char *data,
+    int len,
+    opus_int16 *pcm,
+    int frame_size
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+/** Perform a CTL function on an Opus custom decoder.
+  *
+  * Generally the request and subsequent arguments are generated
+  * by a convenience macro.
+  * @see opus_genericctls
+  */
+OPUS_CUSTOM_EXPORT int opus_custom_decoder_ctl(OpusCustomDecoder * OPUS_RESTRICT st, int request, ...) OPUS_ARG_NONNULL(1);
+
+/**@}*/
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* OPUS_CUSTOM_H */

lib/codecs/inc/opus/opus_defines.h

@@ -0,0 +1,801 @@
+/* Copyright (c) 2010-2011 Xiph.Org Foundation, Skype Limited
+   Written by Jean-Marc Valin and Koen Vos */
+/*
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
+   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+/**
+ * @file opus_defines.h
+ * @brief Opus reference implementation constants
+ */
+
+#ifndef OPUS_DEFINES_H
+#define OPUS_DEFINES_H
+
+#include "opus_types.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/** @defgroup opus_errorcodes Error codes
+ * @{
+ */
+/** No error @hideinitializer*/
+#define OPUS_OK                0
+/** One or more invalid/out of range arguments @hideinitializer*/
+#define OPUS_BAD_ARG          -1
+/** Not enough bytes allocated in the buffer @hideinitializer*/
+#define OPUS_BUFFER_TOO_SMALL -2
+/** An internal error was detected @hideinitializer*/
+#define OPUS_INTERNAL_ERROR   -3
+/** The compressed data passed is corrupted @hideinitializer*/
+#define OPUS_INVALID_PACKET   -4
+/** Invalid/unsupported request number @hideinitializer*/
+#define OPUS_UNIMPLEMENTED    -5
+/** An encoder or decoder structure is invalid or already freed @hideinitializer*/
+#define OPUS_INVALID_STATE    -6
+/** Memory allocation has failed @hideinitializer*/
+#define OPUS_ALLOC_FAIL       -7
+/**@}*/
+
+/** @cond OPUS_INTERNAL_DOC */
+/**Export control for opus functions */
+
+#ifndef OPUS_EXPORT
+# if defined(_WIN32)
+#  if defined(OPUS_BUILD) && defined(DLL_EXPORT)
+#   define OPUS_EXPORT __declspec(dllexport)
+#  else
+#   define OPUS_EXPORT
+#  endif
+# elif defined(__GNUC__) && defined(OPUS_BUILD)
+#  define OPUS_EXPORT __attribute__ ((visibility ("default")))
+# else
+#  define OPUS_EXPORT
+# endif
+#endif
+
+# if !defined(OPUS_GNUC_PREREQ)
+#  if defined(__GNUC__)&&defined(__GNUC_MINOR__)
+#   define OPUS_GNUC_PREREQ(_maj,_min) \
+ ((__GNUC__<<16)+__GNUC_MINOR__>=((_maj)<<16)+(_min))
+#  else
+#   define OPUS_GNUC_PREREQ(_maj,_min) 0
+#  endif
+# endif
+
+#if (!defined(__STDC_VERSION__) || (__STDC_VERSION__ < 199901L) )
+# if OPUS_GNUC_PREREQ(3,0)
+#  define OPUS_RESTRICT __restrict__
+# elif (defined(_MSC_VER) && _MSC_VER >= 1400)
+#  define OPUS_RESTRICT __restrict
+# else
+#  define OPUS_RESTRICT
+# endif
+#else
+# define OPUS_RESTRICT restrict
+#endif
+
+#if (!defined(__STDC_VERSION__) || (__STDC_VERSION__ < 199901L) )
+# if OPUS_GNUC_PREREQ(2,7)
+#  define OPUS_INLINE __inline__
+# elif (defined(_MSC_VER))
+#  define OPUS_INLINE __inline
+# else
+#  define OPUS_INLINE
+# endif
+#else
+# define OPUS_INLINE inline
+#endif
+
+/**Warning attributes for opus functions
+  * NONNULL is not used in OPUS_BUILD to avoid the compiler optimizing out
+  * some paranoid null checks. */
+#if defined(__GNUC__) && OPUS_GNUC_PREREQ(3, 4)
+# define OPUS_WARN_UNUSED_RESULT __attribute__ ((__warn_unused_result__))
+#else
+# define OPUS_WARN_UNUSED_RESULT
+#endif
+#if !defined(OPUS_BUILD) && defined(__GNUC__) && OPUS_GNUC_PREREQ(3, 4)
+# define OPUS_ARG_NONNULL(_x)  __attribute__ ((__nonnull__(_x)))
+#else
+# define OPUS_ARG_NONNULL(_x)
+#endif
+
+/** These are the actual Encoder CTL ID numbers.
+  * They should not be used directly by applications.
+  * In general, SETs should be even and GETs should be odd.*/
+#define OPUS_SET_APPLICATION_REQUEST         4000
+#define OPUS_GET_APPLICATION_REQUEST         4001
+#define OPUS_SET_BITRATE_REQUEST             4002
+#define OPUS_GET_BITRATE_REQUEST             4003
+#define OPUS_SET_MAX_BANDWIDTH_REQUEST       4004
+#define OPUS_GET_MAX_BANDWIDTH_REQUEST       4005
+#define OPUS_SET_VBR_REQUEST                 4006
+#define OPUS_GET_VBR_REQUEST                 4007
+#define OPUS_SET_BANDWIDTH_REQUEST           4008
+#define OPUS_GET_BANDWIDTH_REQUEST           4009
+#define OPUS_SET_COMPLEXITY_REQUEST          4010
+#define OPUS_GET_COMPLEXITY_REQUEST          4011
+#define OPUS_SET_INBAND_FEC_REQUEST          4012
+#define OPUS_GET_INBAND_FEC_REQUEST          4013
+#define OPUS_SET_PACKET_LOSS_PERC_REQUEST    4014
+#define OPUS_GET_PACKET_LOSS_PERC_REQUEST    4015
+#define OPUS_SET_DTX_REQUEST                 4016
+#define OPUS_GET_DTX_REQUEST                 4017
+#define OPUS_SET_VBR_CONSTRAINT_REQUEST      4020
+#define OPUS_GET_VBR_CONSTRAINT_REQUEST      4021
+#define OPUS_SET_FORCE_CHANNELS_REQUEST      4022
+#define OPUS_GET_FORCE_CHANNELS_REQUEST      4023
+#define OPUS_SET_SIGNAL_REQUEST              4024
+#define OPUS_GET_SIGNAL_REQUEST              4025
+#define OPUS_GET_LOOKAHEAD_REQUEST           4027
+/* #define OPUS_RESET_STATE 4028 */
+#define OPUS_GET_SAMPLE_RATE_REQUEST         4029
+#define OPUS_GET_FINAL_RANGE_REQUEST         4031
+#define OPUS_GET_PITCH_REQUEST               4033
+#define OPUS_SET_GAIN_REQUEST                4034
+#define OPUS_GET_GAIN_REQUEST                4045 /* Should have been 4035 */
+#define OPUS_SET_LSB_DEPTH_REQUEST           4036
+#define OPUS_GET_LSB_DEPTH_REQUEST           4037
+#define OPUS_GET_LAST_PACKET_DURATION_REQUEST 4039
+#define OPUS_SET_EXPERT_FRAME_DURATION_REQUEST 4040
+#define OPUS_GET_EXPERT_FRAME_DURATION_REQUEST 4041
+#define OPUS_SET_PREDICTION_DISABLED_REQUEST 4042
+#define OPUS_GET_PREDICTION_DISABLED_REQUEST 4043
+/* Don't use 4045, it's already taken by OPUS_GET_GAIN_REQUEST */
+#define OPUS_SET_PHASE_INVERSION_DISABLED_REQUEST 4046
+#define OPUS_GET_PHASE_INVERSION_DISABLED_REQUEST 4047
+#define OPUS_GET_IN_DTX_REQUEST              4049
+
+/** Defines for the presence of extended APIs. */
+#define OPUS_HAVE_OPUS_PROJECTION_H
+
+/* Macros to trigger compilation errors when the wrong types are provided to a CTL */
+#define __opus_check_int(x) (((void)((x) == (opus_int32)0)), (opus_int32)(x))
+#define __opus_check_int_ptr(ptr) ((ptr) + ((ptr) - (opus_int32*)(ptr)))
+#define __opus_check_uint_ptr(ptr) ((ptr) + ((ptr) - (opus_uint32*)(ptr)))
+#define __opus_check_val16_ptr(ptr) ((ptr) + ((ptr) - (opus_val16*)(ptr)))
+/** @endcond */
+
+/** @defgroup opus_ctlvalues Pre-defined values for CTL interface
+  * @see opus_genericctls, opus_encoderctls
+  * @{
+  */
+/* Values for the various encoder CTLs */
+#define OPUS_AUTO                           -1000 /**<Auto/default setting @hideinitializer*/
+#define OPUS_BITRATE_MAX                       -1 /**<Maximum bitrate @hideinitializer*/
+
+/** Best for most VoIP/videoconference applications where listening quality and intelligibility matter most
+ * @hideinitializer */
+#define OPUS_APPLICATION_VOIP                2048
+/** Best for broadcast/high-fidelity application where the decoded audio should be as close as possible to the input
+ * @hideinitializer */
+#define OPUS_APPLICATION_AUDIO               2049
+/** Only use when lowest-achievable latency is what matters most. Voice-optimized modes cannot be used.
+ * @hideinitializer */
+#define OPUS_APPLICATION_RESTRICTED_LOWDELAY 2051
+
+#define OPUS_SIGNAL_VOICE                    3001 /**< Signal being encoded is voice */
+#define OPUS_SIGNAL_MUSIC                    3002 /**< Signal being encoded is music */
+#define OPUS_BANDWIDTH_NARROWBAND            1101 /**< 4 kHz bandpass @hideinitializer*/
+#define OPUS_BANDWIDTH_MEDIUMBAND            1102 /**< 6 kHz bandpass @hideinitializer*/
+#define OPUS_BANDWIDTH_WIDEBAND              1103 /**< 8 kHz bandpass @hideinitializer*/
+#define OPUS_BANDWIDTH_SUPERWIDEBAND         1104 /**<12 kHz bandpass @hideinitializer*/
+#define OPUS_BANDWIDTH_FULLBAND              1105 /**<20 kHz bandpass @hideinitializer*/
+
+#define OPUS_FRAMESIZE_ARG                   5000 /**< Select frame size from the argument (default) */
+#define OPUS_FRAMESIZE_2_5_MS                5001 /**< Use 2.5 ms frames */
+#define OPUS_FRAMESIZE_5_MS                  5002 /**< Use 5 ms frames */
+#define OPUS_FRAMESIZE_10_MS                 5003 /**< Use 10 ms frames */
+#define OPUS_FRAMESIZE_20_MS                 5004 /**< Use 20 ms frames */
+#define OPUS_FRAMESIZE_40_MS                 5005 /**< Use 40 ms frames */
+#define OPUS_FRAMESIZE_60_MS                 5006 /**< Use 60 ms frames */
+#define OPUS_FRAMESIZE_80_MS                 5007 /**< Use 80 ms frames */
+#define OPUS_FRAMESIZE_100_MS                5008 /**< Use 100 ms frames */
+#define OPUS_FRAMESIZE_120_MS                5009 /**< Use 120 ms frames */
+
+/**@}*/
+
+
+/** @defgroup opus_encoderctls Encoder related CTLs
+  *
+  * These are convenience macros for use with the \c opus_encode_ctl
+  * interface. They are used to generate the appropriate series of
+  * arguments for that call, passing the correct type, size and so
+  * on as expected for each particular request.
+  *
+  * Some usage examples:
+  *
+  * @code
+  * int ret;
+  * ret = opus_encoder_ctl(enc_ctx, OPUS_SET_BANDWIDTH(OPUS_AUTO));
+  * if (ret != OPUS_OK) return ret;
+  *
+  * opus_int32 rate;
+  * opus_encoder_ctl(enc_ctx, OPUS_GET_BANDWIDTH(&rate));
+  *
+  * opus_encoder_ctl(enc_ctx, OPUS_RESET_STATE);
+  * @endcode
+  *
+  * @see opus_genericctls, opus_encoder
+  * @{
+  */
+
+/** Configures the encoder's computational complexity.
+  * The supported range is 0-10 inclusive with 10 representing the highest complexity.
+  * @see OPUS_GET_COMPLEXITY
+  * @param[in] x <tt>opus_int32</tt>: Allowed values: 0-10, inclusive.
+  *
+  * @hideinitializer */
+#define OPUS_SET_COMPLEXITY(x) OPUS_SET_COMPLEXITY_REQUEST, __opus_check_int(x)
+/** Gets the encoder's complexity configuration.
+  * @see OPUS_SET_COMPLEXITY
+  * @param[out] x <tt>opus_int32 *</tt>: Returns a value in the range 0-10,
+  *                                      inclusive.
+  * @hideinitializer */
+#define OPUS_GET_COMPLEXITY(x) OPUS_GET_COMPLEXITY_REQUEST, __opus_check_int_ptr(x)
+
+/** Configures the bitrate in the encoder.
+  * Rates from 500 to 512000 bits per second are meaningful, as well as the
+  * special values #OPUS_AUTO and #OPUS_BITRATE_MAX.
+  * The value #OPUS_BITRATE_MAX can be used to cause the codec to use as much
+  * rate as it can, which is useful for controlling the rate by adjusting the
+  * output buffer size.
+  * @see OPUS_GET_BITRATE
+  * @param[in] x <tt>opus_int32</tt>: Bitrate in bits per second. The default
+  *                                   is determined based on the number of
+  *                                   channels and the input sampling rate.
+  * @hideinitializer */
+#define OPUS_SET_BITRATE(x) OPUS_SET_BITRATE_REQUEST, __opus_check_int(x)
+/** Gets the encoder's bitrate configuration.
+  * @see OPUS_SET_BITRATE
+  * @param[out] x <tt>opus_int32 *</tt>: Returns the bitrate in bits per second.
+  *                                      The default is determined based on the
+  *                                      number of channels and the input
+  *                                      sampling rate.
+  * @hideinitializer */
+#define OPUS_GET_BITRATE(x) OPUS_GET_BITRATE_REQUEST, __opus_check_int_ptr(x)
+
+/** Enables or disables variable bitrate (VBR) in the encoder.
+  * The configured bitrate may not be met exactly because frames must
+  * be an integer number of bytes in length.
+  * @see OPUS_GET_VBR
+  * @see OPUS_SET_VBR_CONSTRAINT
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>0</dt><dd>Hard CBR. For LPC/hybrid modes at very low bit-rate, this can
+  *               cause noticeable quality degradation.</dd>
+  * <dt>1</dt><dd>VBR (default). The exact type of VBR is controlled by
+  *               #OPUS_SET_VBR_CONSTRAINT.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_VBR(x) OPUS_SET_VBR_REQUEST, __opus_check_int(x)
+/** Determine if variable bitrate (VBR) is enabled in the encoder.
+  * @see OPUS_SET_VBR
+  * @see OPUS_GET_VBR_CONSTRAINT
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>0</dt><dd>Hard CBR.</dd>
+  * <dt>1</dt><dd>VBR (default). The exact type of VBR may be retrieved via
+  *               #OPUS_GET_VBR_CONSTRAINT.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_VBR(x) OPUS_GET_VBR_REQUEST, __opus_check_int_ptr(x)
+
+/** Enables or disables constrained VBR in the encoder.
+  * This setting is ignored when the encoder is in CBR mode.
+  * @warning Only the MDCT mode of Opus currently heeds the constraint.
+  *  Speech mode ignores it completely, hybrid mode may fail to obey it
+  *  if the LPC layer uses more bitrate than the constraint would have
+  *  permitted.
+  * @see OPUS_GET_VBR_CONSTRAINT
+  * @see OPUS_SET_VBR
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>0</dt><dd>Unconstrained VBR.</dd>
+  * <dt>1</dt><dd>Constrained VBR (default). This creates a maximum of one
+  *               frame of buffering delay assuming a transport with a
+  *               serialization speed of the nominal bitrate.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_VBR_CONSTRAINT(x) OPUS_SET_VBR_CONSTRAINT_REQUEST, __opus_check_int(x)
+/** Determine if constrained VBR is enabled in the encoder.
+  * @see OPUS_SET_VBR_CONSTRAINT
+  * @see OPUS_GET_VBR
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>0</dt><dd>Unconstrained VBR.</dd>
+  * <dt>1</dt><dd>Constrained VBR (default).</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_VBR_CONSTRAINT(x) OPUS_GET_VBR_CONSTRAINT_REQUEST, __opus_check_int_ptr(x)
+
+/** Configures mono/stereo forcing in the encoder.
+  * This can force the encoder to produce packets encoded as either mono or
+  * stereo, regardless of the format of the input audio. This is useful when
+  * the caller knows that the input signal is currently a mono source embedded
+  * in a stereo stream.
+  * @see OPUS_GET_FORCE_CHANNELS
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>#OPUS_AUTO</dt><dd>Not forced (default)</dd>
+  * <dt>1</dt>         <dd>Forced mono</dd>
+  * <dt>2</dt>         <dd>Forced stereo</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_FORCE_CHANNELS(x) OPUS_SET_FORCE_CHANNELS_REQUEST, __opus_check_int(x)
+/** Gets the encoder's forced channel configuration.
+  * @see OPUS_SET_FORCE_CHANNELS
+  * @param[out] x <tt>opus_int32 *</tt>:
+  * <dl>
+  * <dt>#OPUS_AUTO</dt><dd>Not forced (default)</dd>
+  * <dt>1</dt>         <dd>Forced mono</dd>
+  * <dt>2</dt>         <dd>Forced stereo</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_FORCE_CHANNELS(x) OPUS_GET_FORCE_CHANNELS_REQUEST, __opus_check_int_ptr(x)
+
+/** Configures the maximum bandpass that the encoder will select automatically.
+  * Applications should normally use this instead of #OPUS_SET_BANDWIDTH
+  * (leaving that set to the default, #OPUS_AUTO). This allows the
+  * application to set an upper bound based on the type of input it is
+  * providing, but still gives the encoder the freedom to reduce the bandpass
+  * when the bitrate becomes too low, for better overall quality.
+  * @see OPUS_GET_MAX_BANDWIDTH
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>OPUS_BANDWIDTH_NARROWBAND</dt>    <dd>4 kHz passband</dd>
+  * <dt>OPUS_BANDWIDTH_MEDIUMBAND</dt>    <dd>6 kHz passband</dd>
+  * <dt>OPUS_BANDWIDTH_WIDEBAND</dt>      <dd>8 kHz passband</dd>
+  * <dt>OPUS_BANDWIDTH_SUPERWIDEBAND</dt><dd>12 kHz passband</dd>
+  * <dt>OPUS_BANDWIDTH_FULLBAND</dt>     <dd>20 kHz passband (default)</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_MAX_BANDWIDTH(x) OPUS_SET_MAX_BANDWIDTH_REQUEST, __opus_check_int(x)
+
+/** Gets the encoder's configured maximum allowed bandpass.
+  * @see OPUS_SET_MAX_BANDWIDTH
+  * @param[out] x <tt>opus_int32 *</tt>: Allowed values:
+  * <dl>
+  * <dt>#OPUS_BANDWIDTH_NARROWBAND</dt>    <dd>4 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_MEDIUMBAND</dt>    <dd>6 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_WIDEBAND</dt>      <dd>8 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_SUPERWIDEBAND</dt><dd>12 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_FULLBAND</dt>     <dd>20 kHz passband (default)</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_MAX_BANDWIDTH(x) OPUS_GET_MAX_BANDWIDTH_REQUEST, __opus_check_int_ptr(x)
+
+/** Sets the encoder's bandpass to a specific value.
+  * This prevents the encoder from automatically selecting the bandpass based
+  * on the available bitrate. If an application knows the bandpass of the input
+  * audio it is providing, it should normally use #OPUS_SET_MAX_BANDWIDTH
+  * instead, which still gives the encoder the freedom to reduce the bandpass
+  * when the bitrate becomes too low, for better overall quality.
+  * @see OPUS_GET_BANDWIDTH
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>#OPUS_AUTO</dt>                    <dd>(default)</dd>
+  * <dt>#OPUS_BANDWIDTH_NARROWBAND</dt>    <dd>4 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_MEDIUMBAND</dt>    <dd>6 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_WIDEBAND</dt>      <dd>8 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_SUPERWIDEBAND</dt><dd>12 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_FULLBAND</dt>     <dd>20 kHz passband</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_BANDWIDTH(x) OPUS_SET_BANDWIDTH_REQUEST, __opus_check_int(x)
+
+/** Configures the type of signal being encoded.
+  * This is a hint which helps the encoder's mode selection.
+  * @see OPUS_GET_SIGNAL
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>#OPUS_AUTO</dt>        <dd>(default)</dd>
+  * <dt>#OPUS_SIGNAL_VOICE</dt><dd>Bias thresholds towards choosing LPC or Hybrid modes.</dd>
+  * <dt>#OPUS_SIGNAL_MUSIC</dt><dd>Bias thresholds towards choosing MDCT modes.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_SIGNAL(x) OPUS_SET_SIGNAL_REQUEST, __opus_check_int(x)
+/** Gets the encoder's configured signal type.
+  * @see OPUS_SET_SIGNAL
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>#OPUS_AUTO</dt>        <dd>(default)</dd>
+  * <dt>#OPUS_SIGNAL_VOICE</dt><dd>Bias thresholds towards choosing LPC or Hybrid modes.</dd>
+  * <dt>#OPUS_SIGNAL_MUSIC</dt><dd>Bias thresholds towards choosing MDCT modes.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_SIGNAL(x) OPUS_GET_SIGNAL_REQUEST, __opus_check_int_ptr(x)
+
+
+/** Configures the encoder's intended application.
+  * The initial value is a mandatory argument to the encoder_create function.
+  * @see OPUS_GET_APPLICATION
+  * @param[in] x <tt>opus_int32</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>#OPUS_APPLICATION_VOIP</dt>
+  * <dd>Process signal for improved speech intelligibility.</dd>
+  * <dt>#OPUS_APPLICATION_AUDIO</dt>
+  * <dd>Favor faithfulness to the original input.</dd>
+  * <dt>#OPUS_APPLICATION_RESTRICTED_LOWDELAY</dt>
+  * <dd>Configure the minimum possible coding delay by disabling certain modes
+  * of operation.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_APPLICATION(x) OPUS_SET_APPLICATION_REQUEST, __opus_check_int(x)
+/** Gets the encoder's configured application.
+  * @see OPUS_SET_APPLICATION
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>#OPUS_APPLICATION_VOIP</dt>
+  * <dd>Process signal for improved speech intelligibility.</dd>
+  * <dt>#OPUS_APPLICATION_AUDIO</dt>
+  * <dd>Favor faithfulness to the original input.</dd>
+  * <dt>#OPUS_APPLICATION_RESTRICTED_LOWDELAY</dt>
+  * <dd>Configure the minimum possible coding delay by disabling certain modes
+  * of operation.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_APPLICATION(x) OPUS_GET_APPLICATION_REQUEST, __opus_check_int_ptr(x)
+
+/** Gets the total samples of delay added by the entire codec.
+  * This can be queried by the encoder and then the provided number of samples can be
+  * skipped on from the start of the decoder's output to provide time aligned input
+  * and output. From the perspective of a decoding application the real data begins this many
+  * samples late.
+  *
+  * The decoder contribution to this delay is identical for all decoders, but the
+  * encoder portion of the delay may vary from implementation to implementation,
+  * version to version, or even depend on the encoder's initial configuration.
+  * Applications needing delay compensation should call this CTL rather than
+  * hard-coding a value.
+  * @param[out] x <tt>opus_int32 *</tt>:   Number of lookahead samples
+  * @hideinitializer */
+#define OPUS_GET_LOOKAHEAD(x) OPUS_GET_LOOKAHEAD_REQUEST, __opus_check_int_ptr(x)
+
+/** Configures the encoder's use of inband forward error correction (FEC).
+  * @note This is only applicable to the LPC layer
+  * @see OPUS_GET_INBAND_FEC
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>0</dt><dd>Disable inband FEC (default).</dd>
+  * <dt>1</dt><dd>Inband FEC enabled. If the packet loss rate is sufficiently high, Opus will automatically switch to SILK even at high rates to enable use of that FEC.</dd>
+  * <dt>2</dt><dd>Inband FEC enabled, but does not necessarily switch to SILK if we have music.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_INBAND_FEC(x) OPUS_SET_INBAND_FEC_REQUEST, __opus_check_int(x)
+/** Gets encoder's configured use of inband forward error correction.
+  * @see OPUS_SET_INBAND_FEC
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>0</dt><dd>Inband FEC disabled (default).</dd>
+  * <dt>1</dt><dd>Inband FEC enabled. If the packet loss rate is sufficiently high, Opus will automatically switch to SILK even at high rates to enable use of that FEC.</dd>
+  * <dt>2</dt><dd>Inband FEC enabled, but does not necessarily switch to SILK if we have music.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_INBAND_FEC(x) OPUS_GET_INBAND_FEC_REQUEST, __opus_check_int_ptr(x)
+
+/** Configures the encoder's expected packet loss percentage.
+  * Higher values trigger progressively more loss resistant behavior in the encoder
+  * at the expense of quality at a given bitrate in the absence of packet loss, but
+  * greater quality under loss.
+  * @see OPUS_GET_PACKET_LOSS_PERC
+  * @param[in] x <tt>opus_int32</tt>:   Loss percentage in the range 0-100, inclusive (default: 0).
+  * @hideinitializer */
+#define OPUS_SET_PACKET_LOSS_PERC(x) OPUS_SET_PACKET_LOSS_PERC_REQUEST, __opus_check_int(x)
+/** Gets the encoder's configured packet loss percentage.
+  * @see OPUS_SET_PACKET_LOSS_PERC
+  * @param[out] x <tt>opus_int32 *</tt>: Returns the configured loss percentage
+  *                                      in the range 0-100, inclusive (default: 0).
+  * @hideinitializer */
+#define OPUS_GET_PACKET_LOSS_PERC(x) OPUS_GET_PACKET_LOSS_PERC_REQUEST, __opus_check_int_ptr(x)
+
+/** Configures the encoder's use of discontinuous transmission (DTX).
+  * @note This is only applicable to the LPC layer
+  * @see OPUS_GET_DTX
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>0</dt><dd>Disable DTX (default).</dd>
+  * <dt>1</dt><dd>Enabled DTX.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_DTX(x) OPUS_SET_DTX_REQUEST, __opus_check_int(x)
+/** Gets encoder's configured use of discontinuous transmission.
+  * @see OPUS_SET_DTX
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>0</dt><dd>DTX disabled (default).</dd>
+  * <dt>1</dt><dd>DTX enabled.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_DTX(x) OPUS_GET_DTX_REQUEST, __opus_check_int_ptr(x)
+/** Configures the depth of signal being encoded.
+  *
+  * This is a hint which helps the encoder identify silence and near-silence.
+  * It represents the number of significant bits of linear intensity below
+  * which the signal contains ignorable quantization or other noise.
+  *
+  * For example, OPUS_SET_LSB_DEPTH(14) would be an appropriate setting
+  * for G.711 u-law input. OPUS_SET_LSB_DEPTH(16) would be appropriate
+  * for 16-bit linear pcm input with opus_encode_float().
+  *
+  * When using opus_encode() instead of opus_encode_float(), or when libopus
+  * is compiled for fixed-point, the encoder uses the minimum of the value
+  * set here and the value 16.
+  *
+  * @see OPUS_GET_LSB_DEPTH
+  * @param[in] x <tt>opus_int32</tt>: Input precision in bits, between 8 and 24
+  *                                   (default: 24).
+  * @hideinitializer */
+#define OPUS_SET_LSB_DEPTH(x) OPUS_SET_LSB_DEPTH_REQUEST, __opus_check_int(x)
+/** Gets the encoder's configured signal depth.
+  * @see OPUS_SET_LSB_DEPTH
+  * @param[out] x <tt>opus_int32 *</tt>: Input precision in bits, between 8 and
+  *                                      24 (default: 24).
+  * @hideinitializer */
+#define OPUS_GET_LSB_DEPTH(x) OPUS_GET_LSB_DEPTH_REQUEST, __opus_check_int_ptr(x)
+
+/** Configures the encoder's use of variable duration frames.
+  * When variable duration is enabled, the encoder is free to use a shorter frame
+  * size than the one requested in the opus_encode*() call.
+  * It is then the user's responsibility
+  * to verify how much audio was encoded by checking the ToC byte of the encoded
+  * packet. The part of the audio that was not encoded needs to be resent to the
+  * encoder for the next call. Do not use this option unless you <b>really</b>
+  * know what you are doing.
+  * @see OPUS_GET_EXPERT_FRAME_DURATION
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>OPUS_FRAMESIZE_ARG</dt><dd>Select frame size from the argument (default).</dd>
+  * <dt>OPUS_FRAMESIZE_2_5_MS</dt><dd>Use 2.5 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_5_MS</dt><dd>Use 5 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_10_MS</dt><dd>Use 10 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_20_MS</dt><dd>Use 20 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_40_MS</dt><dd>Use 40 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_60_MS</dt><dd>Use 60 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_80_MS</dt><dd>Use 80 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_100_MS</dt><dd>Use 100 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_120_MS</dt><dd>Use 120 ms frames.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_EXPERT_FRAME_DURATION(x) OPUS_SET_EXPERT_FRAME_DURATION_REQUEST, __opus_check_int(x)
+/** Gets the encoder's configured use of variable duration frames.
+  * @see OPUS_SET_EXPERT_FRAME_DURATION
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>OPUS_FRAMESIZE_ARG</dt><dd>Select frame size from the argument (default).</dd>
+  * <dt>OPUS_FRAMESIZE_2_5_MS</dt><dd>Use 2.5 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_5_MS</dt><dd>Use 5 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_10_MS</dt><dd>Use 10 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_20_MS</dt><dd>Use 20 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_40_MS</dt><dd>Use 40 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_60_MS</dt><dd>Use 60 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_80_MS</dt><dd>Use 80 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_100_MS</dt><dd>Use 100 ms frames.</dd>
+  * <dt>OPUS_FRAMESIZE_120_MS</dt><dd>Use 120 ms frames.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_EXPERT_FRAME_DURATION(x) OPUS_GET_EXPERT_FRAME_DURATION_REQUEST, __opus_check_int_ptr(x)
+
+/** If set to 1, disables almost all use of prediction, making frames almost
+  * completely independent. This reduces quality.
+  * @see OPUS_GET_PREDICTION_DISABLED
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>0</dt><dd>Enable prediction (default).</dd>
+  * <dt>1</dt><dd>Disable prediction.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_PREDICTION_DISABLED(x) OPUS_SET_PREDICTION_DISABLED_REQUEST, __opus_check_int(x)
+/** Gets the encoder's configured prediction status.
+  * @see OPUS_SET_PREDICTION_DISABLED
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>0</dt><dd>Prediction enabled (default).</dd>
+  * <dt>1</dt><dd>Prediction disabled.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_PREDICTION_DISABLED(x) OPUS_GET_PREDICTION_DISABLED_REQUEST, __opus_check_int_ptr(x)
+
+/**@}*/
+
+/** @defgroup opus_genericctls Generic CTLs
+  *
+  * These macros are used with the \c opus_decoder_ctl and
+  * \c opus_encoder_ctl calls to generate a particular
+  * request.
+  *
+  * When called on an \c OpusDecoder they apply to that
+  * particular decoder instance. When called on an
+  * \c OpusEncoder they apply to the corresponding setting
+  * on that encoder instance, if present.
+  *
+  * Some usage examples:
+  *
+  * @code
+  * int ret;
+  * opus_int32 pitch;
+  * ret = opus_decoder_ctl(dec_ctx, OPUS_GET_PITCH(&pitch));
+  * if (ret == OPUS_OK) return ret;
+  *
+  * opus_encoder_ctl(enc_ctx, OPUS_RESET_STATE);
+  * opus_decoder_ctl(dec_ctx, OPUS_RESET_STATE);
+  *
+  * opus_int32 enc_bw, dec_bw;
+  * opus_encoder_ctl(enc_ctx, OPUS_GET_BANDWIDTH(&enc_bw));
+  * opus_decoder_ctl(dec_ctx, OPUS_GET_BANDWIDTH(&dec_bw));
+  * if (enc_bw != dec_bw) {
+  *   printf("packet bandwidth mismatch!\n");
+  * }
+  * @endcode
+  *
+  * @see opus_encoder, opus_decoder_ctl, opus_encoder_ctl, opus_decoderctls, opus_encoderctls
+  * @{
+  */
+
+/** Resets the codec state to be equivalent to a freshly initialized state.
+  * This should be called when switching streams in order to prevent
+  * the back to back decoding from giving different results from
+  * one at a time decoding.
+  * @hideinitializer */
+#define OPUS_RESET_STATE 4028
+
+/** Gets the final state of the codec's entropy coder.
+  * This is used for testing purposes,
+  * The encoder and decoder state should be identical after coding a payload
+  * (assuming no data corruption or software bugs)
+  *
+  * @param[out] x <tt>opus_uint32 *</tt>: Entropy coder state
+  *
+  * @hideinitializer */
+#define OPUS_GET_FINAL_RANGE(x) OPUS_GET_FINAL_RANGE_REQUEST, __opus_check_uint_ptr(x)
+
+/** Gets the encoder's configured bandpass or the decoder's last bandpass.
+  * @see OPUS_SET_BANDWIDTH
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>#OPUS_AUTO</dt>                    <dd>(default)</dd>
+  * <dt>#OPUS_BANDWIDTH_NARROWBAND</dt>    <dd>4 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_MEDIUMBAND</dt>    <dd>6 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_WIDEBAND</dt>      <dd>8 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_SUPERWIDEBAND</dt><dd>12 kHz passband</dd>
+  * <dt>#OPUS_BANDWIDTH_FULLBAND</dt>     <dd>20 kHz passband</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_BANDWIDTH(x) OPUS_GET_BANDWIDTH_REQUEST, __opus_check_int_ptr(x)
+
+/** Gets the sampling rate the encoder or decoder was initialized with.
+  * This simply returns the <code>Fs</code> value passed to opus_encoder_init()
+  * or opus_decoder_init().
+  * @param[out] x <tt>opus_int32 *</tt>: Sampling rate of encoder or decoder.
+  * @hideinitializer
+  */
+#define OPUS_GET_SAMPLE_RATE(x) OPUS_GET_SAMPLE_RATE_REQUEST, __opus_check_int_ptr(x)
+
+/** If set to 1, disables the use of phase inversion for intensity stereo,
+  * improving the quality of mono downmixes, but slightly reducing normal
+  * stereo quality. Disabling phase inversion in the decoder does not comply
+  * with RFC 6716, although it does not cause any interoperability issue and
+  * is expected to become part of the Opus standard once RFC 6716 is updated
+  * by draft-ietf-codec-opus-update.
+  * @see OPUS_GET_PHASE_INVERSION_DISABLED
+  * @param[in] x <tt>opus_int32</tt>: Allowed values:
+  * <dl>
+  * <dt>0</dt><dd>Enable phase inversion (default).</dd>
+  * <dt>1</dt><dd>Disable phase inversion.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_SET_PHASE_INVERSION_DISABLED(x) OPUS_SET_PHASE_INVERSION_DISABLED_REQUEST, __opus_check_int(x)
+/** Gets the encoder's configured phase inversion status.
+  * @see OPUS_SET_PHASE_INVERSION_DISABLED
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>0</dt><dd>Stereo phase inversion enabled (default).</dd>
+  * <dt>1</dt><dd>Stereo phase inversion disabled.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_PHASE_INVERSION_DISABLED(x) OPUS_GET_PHASE_INVERSION_DISABLED_REQUEST, __opus_check_int_ptr(x)
+/** Gets the DTX state of the encoder.
+  * Returns whether the last encoded frame was either a comfort noise update
+  * during DTX or not encoded because of DTX.
+  * @param[out] x <tt>opus_int32 *</tt>: Returns one of the following values:
+  * <dl>
+  * <dt>0</dt><dd>The encoder is not in DTX.</dd>
+  * <dt>1</dt><dd>The encoder is in DTX.</dd>
+  * </dl>
+  * @hideinitializer */
+#define OPUS_GET_IN_DTX(x) OPUS_GET_IN_DTX_REQUEST, __opus_check_int_ptr(x)
+
+/**@}*/
+
+/** @defgroup opus_decoderctls Decoder related CTLs
+  * @see opus_genericctls, opus_encoderctls, opus_decoder
+  * @{
+  */
+
+/** Configures decoder gain adjustment.
+  * Scales the decoded output by a factor specified in Q8 dB units.
+  * This has a maximum range of -32768 to 32767 inclusive, and returns
+  * OPUS_BAD_ARG otherwise. The default is zero indicating no adjustment.
+  * This setting survives decoder reset.
+  *
+  * gain = pow(10, x/(20.0*256))
+  *
+  * @param[in] x <tt>opus_int32</tt>:   Amount to scale PCM signal by in Q8 dB units.
+  * @hideinitializer */
+#define OPUS_SET_GAIN(x) OPUS_SET_GAIN_REQUEST, __opus_check_int(x)
+/** Gets the decoder's configured gain adjustment. @see OPUS_SET_GAIN
+  *
+  * @param[out] x <tt>opus_int32 *</tt>: Amount to scale PCM signal by in Q8 dB units.
+  * @hideinitializer */
+#define OPUS_GET_GAIN(x) OPUS_GET_GAIN_REQUEST, __opus_check_int_ptr(x)
+
+/** Gets the duration (in samples) of the last packet successfully decoded or concealed.
+  * @param[out] x <tt>opus_int32 *</tt>: Number of samples (at current sampling rate).
+  * @hideinitializer */
+#define OPUS_GET_LAST_PACKET_DURATION(x) OPUS_GET_LAST_PACKET_DURATION_REQUEST, __opus_check_int_ptr(x)
+
+/** Gets the pitch of the last decoded frame, if available.
+  * This can be used for any post-processing algorithm requiring the use of pitch,
+  * e.g. time stretching/shortening. If the last frame was not voiced, or if the
+  * pitch was not coded in the frame, then zero is returned.
+  *
+  * This CTL is only implemented for decoder instances.
+  *
+  * @param[out] x <tt>opus_int32 *</tt>: pitch period at 48 kHz (or 0 if not available)
+  *
+  * @hideinitializer */
+#define OPUS_GET_PITCH(x) OPUS_GET_PITCH_REQUEST, __opus_check_int_ptr(x)
+
+/**@}*/
+
+/** @defgroup opus_libinfo Opus library information functions
+  * @{
+  */
+
+/** Converts an opus error code into a human readable string.
+  *
+  * @param[in] error <tt>int</tt>: Error number
+  * @returns Error string
+  */
+OPUS_EXPORT const char *opus_strerror(int error);
+
+/** Gets the libopus version string.
+  *
+  * Applications may look for the substring "-fixed" in the version string to
+  * determine whether they have a fixed-point or floating-point build at
+  * runtime.
+  *
+  * @returns Version string
+  */
+OPUS_EXPORT const char *opus_get_version_string(void);
+/**@}*/
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* OPUS_DEFINES_H */

lib/codecs/inc/opus/opus_multistream.h

@@ -0,0 +1,660 @@
+/* Copyright (c) 2011 Xiph.Org Foundation
+   Written by Jean-Marc Valin */
+/*
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
+   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+/**
+ * @file opus_multistream.h
+ * @brief Opus reference implementation multistream API
+ */
+
+#ifndef OPUS_MULTISTREAM_H
+#define OPUS_MULTISTREAM_H
+
+#include "opus.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/** @cond OPUS_INTERNAL_DOC */
+
+/** Macros to trigger compilation errors when the wrong types are provided to a
+  * CTL. */
+/**@{*/
+#define __opus_check_encstate_ptr(ptr) ((ptr) + ((ptr) - (OpusEncoder**)(ptr)))
+#define __opus_check_decstate_ptr(ptr) ((ptr) + ((ptr) - (OpusDecoder**)(ptr)))
+/**@}*/
+
+/** These are the actual encoder and decoder CTL ID numbers.
+  * They should not be used directly by applications.
+  * In general, SETs should be even and GETs should be odd.*/
+/**@{*/
+#define OPUS_MULTISTREAM_GET_ENCODER_STATE_REQUEST 5120
+#define OPUS_MULTISTREAM_GET_DECODER_STATE_REQUEST 5122
+/**@}*/
+
+/** @endcond */
+
+/** @defgroup opus_multistream_ctls Multistream specific encoder and decoder CTLs
+  *
+  * These are convenience macros that are specific to the
+  * opus_multistream_encoder_ctl() and opus_multistream_decoder_ctl()
+  * interface.
+  * The CTLs from @ref opus_genericctls, @ref opus_encoderctls, and
+  * @ref opus_decoderctls may be applied to a multistream encoder or decoder as
+  * well.
+  * In addition, you may retrieve the encoder or decoder state for an specific
+  * stream via #OPUS_MULTISTREAM_GET_ENCODER_STATE or
+  * #OPUS_MULTISTREAM_GET_DECODER_STATE and apply CTLs to it individually.
+  */
+/**@{*/
+
+/** Gets the encoder state for an individual stream of a multistream encoder.
+  * @param[in] x <tt>opus_int32</tt>: The index of the stream whose encoder you
+  *                                   wish to retrieve.
+  *                                   This must be non-negative and less than
+  *                                   the <code>streams</code> parameter used
+  *                                   to initialize the encoder.
+  * @param[out] y <tt>OpusEncoder**</tt>: Returns a pointer to the given
+  *                                       encoder state.
+  * @retval OPUS_BAD_ARG The index of the requested stream was out of range.
+  * @hideinitializer
+  */
+#define OPUS_MULTISTREAM_GET_ENCODER_STATE(x,y) OPUS_MULTISTREAM_GET_ENCODER_STATE_REQUEST, __opus_check_int(x), __opus_check_encstate_ptr(y)
+
+/** Gets the decoder state for an individual stream of a multistream decoder.
+  * @param[in] x <tt>opus_int32</tt>: The index of the stream whose decoder you
+  *                                   wish to retrieve.
+  *                                   This must be non-negative and less than
+  *                                   the <code>streams</code> parameter used
+  *                                   to initialize the decoder.
+  * @param[out] y <tt>OpusDecoder**</tt>: Returns a pointer to the given
+  *                                       decoder state.
+  * @retval OPUS_BAD_ARG The index of the requested stream was out of range.
+  * @hideinitializer
+  */
+#define OPUS_MULTISTREAM_GET_DECODER_STATE(x,y) OPUS_MULTISTREAM_GET_DECODER_STATE_REQUEST, __opus_check_int(x), __opus_check_decstate_ptr(y)
+
+/**@}*/
+
+/** @defgroup opus_multistream Opus Multistream API
+  * @{
+  *
+  * The multistream API allows individual Opus streams to be combined into a
+  * single packet, enabling support for up to 255 channels. Unlike an
+  * elementary Opus stream, the encoder and decoder must negotiate the channel
+  * configuration before the decoder can successfully interpret the data in the
+  * packets produced by the encoder. Some basic information, such as packet
+  * duration, can be computed without any special negotiation.
+  *
+  * The format for multistream Opus packets is defined in
+  * <a href="https://tools.ietf.org/html/rfc7845">RFC 7845</a>
+  * and is based on the self-delimited Opus framing described in Appendix B of
+  * <a href="https://tools.ietf.org/html/rfc6716">RFC 6716</a>.
+  * Normal Opus packets are just a degenerate case of multistream Opus packets,
+  * and can be encoded or decoded with the multistream API by setting
+  * <code>streams</code> to <code>1</code> when initializing the encoder or
+  * decoder.
+  *
+  * Multistream Opus streams can contain up to 255 elementary Opus streams.
+  * These may be either "uncoupled" or "coupled", indicating that the decoder
+  * is configured to decode them to either 1 or 2 channels, respectively.
+  * The streams are ordered so that all coupled streams appear at the
+  * beginning.
+  *
+  * A <code>mapping</code> table defines which decoded channel <code>i</code>
+  * should be used for each input/output (I/O) channel <code>j</code>. This table is
+  * typically provided as an unsigned char array.
+  * Let <code>i = mapping[j]</code> be the index for I/O channel <code>j</code>.
+  * If <code>i < 2*coupled_streams</code>, then I/O channel <code>j</code> is
+  * encoded as the left channel of stream <code>(i/2)</code> if <code>i</code>
+  * is even, or  as the right channel of stream <code>(i/2)</code> if
+  * <code>i</code> is odd. Otherwise, I/O channel <code>j</code> is encoded as
+  * mono in stream <code>(i - coupled_streams)</code>, unless it has the special
+  * value 255, in which case it is omitted from the encoding entirely (the
+  * decoder will reproduce it as silence). Each value <code>i</code> must either
+  * be the special value 255 or be less than <code>streams + coupled_streams</code>.
+  *
+  * The output channels specified by the encoder
+  * should use the
+  * <a href="https://www.xiph.org/vorbis/doc/Vorbis_I_spec.html#x1-810004.3.9">Vorbis
+  * channel ordering</a>. A decoder may wish to apply an additional permutation
+  * to the mapping the encoder used to achieve a different output channel
+  * order (e.g. for outputing in WAV order).
+  *
+  * Each multistream packet contains an Opus packet for each stream, and all of
+  * the Opus packets in a single multistream packet must have the same
+  * duration. Therefore the duration of a multistream packet can be extracted
+  * from the TOC sequence of the first stream, which is located at the
+  * beginning of the packet, just like an elementary Opus stream:
+  *
+  * @code
+  * int nb_samples;
+  * int nb_frames;
+  * nb_frames = opus_packet_get_nb_frames(data, len);
+  * if (nb_frames < 1)
+  *   return nb_frames;
+  * nb_samples = opus_packet_get_samples_per_frame(data, 48000) * nb_frames;
+  * @endcode
+  *
+  * The general encoding and decoding process proceeds exactly the same as in
+  * the normal @ref opus_encoder and @ref opus_decoder APIs.
+  * See their documentation for an overview of how to use the corresponding
+  * multistream functions.
+  */
+
+/** Opus multistream encoder state.
+  * This contains the complete state of a multistream Opus encoder.
+  * It is position independent and can be freely copied.
+  * @see opus_multistream_encoder_create
+  * @see opus_multistream_encoder_init
+  */
+typedef struct OpusMSEncoder OpusMSEncoder;
+
+/** Opus multistream decoder state.
+  * This contains the complete state of a multistream Opus decoder.
+  * It is position independent and can be freely copied.
+  * @see opus_multistream_decoder_create
+  * @see opus_multistream_decoder_init
+  */
+typedef struct OpusMSDecoder OpusMSDecoder;
+
+/**\name Multistream encoder functions */
+/**@{*/
+
+/** Gets the size of an OpusMSEncoder structure.
+  * @param streams <tt>int</tt>: The total number of streams to encode from the
+  *                              input.
+  *                              This must be no more than 255.
+  * @param coupled_streams <tt>int</tt>: Number of coupled (2 channel) streams
+  *                                      to encode.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      encoded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than 255.
+  * @returns The size in bytes on success, or a negative error code
+  *          (see @ref opus_errorcodes) on error.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_multistream_encoder_get_size(
+      int streams,
+      int coupled_streams
+);
+
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_multistream_surround_encoder_get_size(
+      int channels,
+      int mapping_family
+);
+
+
+/** Allocates and initializes a multistream encoder state.
+  * Call opus_multistream_encoder_destroy() to release
+  * this object when finished.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate of the input signal (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels in the input signal.
+  *                               This must be at most 255.
+  *                               It may be greater than the number of
+  *                               coded channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param streams <tt>int</tt>: The total number of streams to encode from the
+  *                              input.
+  *                              This must be no more than the number of channels.
+  * @param coupled_streams <tt>int</tt>: Number of coupled (2 channel) streams
+  *                                      to encode.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      encoded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than the number of input channels.
+  * @param[in] mapping <code>const unsigned char[channels]</code>: Mapping from
+  *                    encoded channels to input channels, as described in
+  *                    @ref opus_multistream. As an extra constraint, the
+  *                    multistream encoder does not allow encoding coupled
+  *                    streams for which one channel is unused since this
+  *                    is never a good idea.
+  * @param application <tt>int</tt>: The target encoder application.
+  *                                  This must be one of the following:
+  * <dl>
+  * <dt>#OPUS_APPLICATION_VOIP</dt>
+  * <dd>Process signal for improved speech intelligibility.</dd>
+  * <dt>#OPUS_APPLICATION_AUDIO</dt>
+  * <dd>Favor faithfulness to the original input.</dd>
+  * <dt>#OPUS_APPLICATION_RESTRICTED_LOWDELAY</dt>
+  * <dd>Configure the minimum possible coding delay by disabling certain modes
+  * of operation.</dd>
+  * </dl>
+  * @param[out] error <tt>int *</tt>: Returns #OPUS_OK on success, or an error
+  *                                   code (see @ref opus_errorcodes) on
+  *                                   failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusMSEncoder *opus_multistream_encoder_create(
+      opus_int32 Fs,
+      int channels,
+      int streams,
+      int coupled_streams,
+      const unsigned char *mapping,
+      int application,
+      int *error
+) OPUS_ARG_NONNULL(5);
+
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusMSEncoder *opus_multistream_surround_encoder_create(
+      opus_int32 Fs,
+      int channels,
+      int mapping_family,
+      int *streams,
+      int *coupled_streams,
+      unsigned char *mapping,
+      int application,
+      int *error
+) OPUS_ARG_NONNULL(4) OPUS_ARG_NONNULL(5) OPUS_ARG_NONNULL(6);
+
+/** Initialize a previously allocated multistream encoder state.
+  * The memory pointed to by \a st must be at least the size returned by
+  * opus_multistream_encoder_get_size().
+  * This is intended for applications which use their own allocator instead of
+  * malloc.
+  * To reset a previously initialized state, use the #OPUS_RESET_STATE CTL.
+  * @see opus_multistream_encoder_create
+  * @see opus_multistream_encoder_get_size
+  * @param st <tt>OpusMSEncoder*</tt>: Multistream encoder state to initialize.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate of the input signal (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels in the input signal.
+  *                               This must be at most 255.
+  *                               It may be greater than the number of
+  *                               coded channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param streams <tt>int</tt>: The total number of streams to encode from the
+  *                              input.
+  *                              This must be no more than the number of channels.
+  * @param coupled_streams <tt>int</tt>: Number of coupled (2 channel) streams
+  *                                      to encode.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      encoded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than the number of input channels.
+  * @param[in] mapping <code>const unsigned char[channels]</code>: Mapping from
+  *                    encoded channels to input channels, as described in
+  *                    @ref opus_multistream. As an extra constraint, the
+  *                    multistream encoder does not allow encoding coupled
+  *                    streams for which one channel is unused since this
+  *                    is never a good idea.
+  * @param application <tt>int</tt>: The target encoder application.
+  *                                  This must be one of the following:
+  * <dl>
+  * <dt>#OPUS_APPLICATION_VOIP</dt>
+  * <dd>Process signal for improved speech intelligibility.</dd>
+  * <dt>#OPUS_APPLICATION_AUDIO</dt>
+  * <dd>Favor faithfulness to the original input.</dd>
+  * <dt>#OPUS_APPLICATION_RESTRICTED_LOWDELAY</dt>
+  * <dd>Configure the minimum possible coding delay by disabling certain modes
+  * of operation.</dd>
+  * </dl>
+  * @returns #OPUS_OK on success, or an error code (see @ref opus_errorcodes)
+  *          on failure.
+  */
+OPUS_EXPORT int opus_multistream_encoder_init(
+      OpusMSEncoder *st,
+      opus_int32 Fs,
+      int channels,
+      int streams,
+      int coupled_streams,
+      const unsigned char *mapping,
+      int application
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(6);
+
+OPUS_EXPORT int opus_multistream_surround_encoder_init(
+      OpusMSEncoder *st,
+      opus_int32 Fs,
+      int channels,
+      int mapping_family,
+      int *streams,
+      int *coupled_streams,
+      unsigned char *mapping,
+      int application
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(5) OPUS_ARG_NONNULL(6) OPUS_ARG_NONNULL(7);
+
+/** Encodes a multistream Opus frame.
+  * @param st <tt>OpusMSEncoder*</tt>: Multistream encoder state.
+  * @param[in] pcm <tt>const opus_int16*</tt>: The input signal as interleaved
+  *                                            samples.
+  *                                            This must contain
+  *                                            <code>frame_size*channels</code>
+  *                                            samples.
+  * @param frame_size <tt>int</tt>: Number of samples per channel in the input
+  *                                 signal.
+  *                                 This must be an Opus frame size for the
+  *                                 encoder's sampling rate.
+  *                                 For example, at 48 kHz the permitted values
+  *                                 are 120, 240, 480, 960, 1920, and 2880.
+  *                                 Passing in a duration of less than 10 ms
+  *                                 (480 samples at 48 kHz) will prevent the
+  *                                 encoder from using the LPC or hybrid modes.
+  * @param[out] data <tt>unsigned char*</tt>: Output payload.
+  *                                           This must contain storage for at
+  *                                           least \a max_data_bytes.
+  * @param [in] max_data_bytes <tt>opus_int32</tt>: Size of the allocated
+  *                                                 memory for the output
+  *                                                 payload. This may be
+  *                                                 used to impose an upper limit on
+  *                                                 the instant bitrate, but should
+  *                                                 not be used as the only bitrate
+  *                                                 control. Use #OPUS_SET_BITRATE to
+  *                                                 control the bitrate.
+  * @returns The length of the encoded packet (in bytes) on success or a
+  *          negative error code (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_multistream_encode(
+    OpusMSEncoder *st,
+    const opus_int16 *pcm,
+    int frame_size,
+    unsigned char *data,
+    opus_int32 max_data_bytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+/** Encodes a multistream Opus frame from floating point input.
+  * @param st <tt>OpusMSEncoder*</tt>: Multistream encoder state.
+  * @param[in] pcm <tt>const float*</tt>: The input signal as interleaved
+  *                                       samples with a normal range of
+  *                                       +/-1.0.
+  *                                       Samples with a range beyond +/-1.0
+  *                                       are supported but will be clipped by
+  *                                       decoders using the integer API and
+  *                                       should only be used if it is known
+  *                                       that the far end supports extended
+  *                                       dynamic range.
+  *                                       This must contain
+  *                                       <code>frame_size*channels</code>
+  *                                       samples.
+  * @param frame_size <tt>int</tt>: Number of samples per channel in the input
+  *                                 signal.
+  *                                 This must be an Opus frame size for the
+  *                                 encoder's sampling rate.
+  *                                 For example, at 48 kHz the permitted values
+  *                                 are 120, 240, 480, 960, 1920, and 2880.
+  *                                 Passing in a duration of less than 10 ms
+  *                                 (480 samples at 48 kHz) will prevent the
+  *                                 encoder from using the LPC or hybrid modes.
+  * @param[out] data <tt>unsigned char*</tt>: Output payload.
+  *                                           This must contain storage for at
+  *                                           least \a max_data_bytes.
+  * @param [in] max_data_bytes <tt>opus_int32</tt>: Size of the allocated
+  *                                                 memory for the output
+  *                                                 payload. This may be
+  *                                                 used to impose an upper limit on
+  *                                                 the instant bitrate, but should
+  *                                                 not be used as the only bitrate
+  *                                                 control. Use #OPUS_SET_BITRATE to
+  *                                                 control the bitrate.
+  * @returns The length of the encoded packet (in bytes) on success or a
+  *          negative error code (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_multistream_encode_float(
+      OpusMSEncoder *st,
+      const float *pcm,
+      int frame_size,
+      unsigned char *data,
+      opus_int32 max_data_bytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+/** Frees an <code>OpusMSEncoder</code> allocated by
+  * opus_multistream_encoder_create().
+  * @param st <tt>OpusMSEncoder*</tt>: Multistream encoder state to be freed.
+  */
+OPUS_EXPORT void opus_multistream_encoder_destroy(OpusMSEncoder *st);
+
+/** Perform a CTL function on a multistream Opus encoder.
+  *
+  * Generally the request and subsequent arguments are generated by a
+  * convenience macro.
+  * @param st <tt>OpusMSEncoder*</tt>: Multistream encoder state.
+  * @param request This and all remaining parameters should be replaced by one
+  *                of the convenience macros in @ref opus_genericctls,
+  *                @ref opus_encoderctls, or @ref opus_multistream_ctls.
+  * @see opus_genericctls
+  * @see opus_encoderctls
+  * @see opus_multistream_ctls
+  */
+OPUS_EXPORT int opus_multistream_encoder_ctl(OpusMSEncoder *st, int request, ...) OPUS_ARG_NONNULL(1);
+
+/**@}*/
+
+/**\name Multistream decoder functions */
+/**@{*/
+
+/** Gets the size of an <code>OpusMSDecoder</code> structure.
+  * @param streams <tt>int</tt>: The total number of streams coded in the
+  *                              input.
+  *                              This must be no more than 255.
+  * @param coupled_streams <tt>int</tt>: Number streams to decode as coupled
+  *                                      (2 channel) streams.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      coded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than 255.
+  * @returns The size in bytes on success, or a negative error code
+  *          (see @ref opus_errorcodes) on error.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_multistream_decoder_get_size(
+      int streams,
+      int coupled_streams
+);
+
+/** Allocates and initializes a multistream decoder state.
+  * Call opus_multistream_decoder_destroy() to release
+  * this object when finished.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate to decode at (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels to output.
+  *                               This must be at most 255.
+  *                               It may be different from the number of coded
+  *                               channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param streams <tt>int</tt>: The total number of streams coded in the
+  *                              input.
+  *                              This must be no more than 255.
+  * @param coupled_streams <tt>int</tt>: Number of streams to decode as coupled
+  *                                      (2 channel) streams.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      coded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than 255.
+  * @param[in] mapping <code>const unsigned char[channels]</code>: Mapping from
+  *                    coded channels to output channels, as described in
+  *                    @ref opus_multistream.
+  * @param[out] error <tt>int *</tt>: Returns #OPUS_OK on success, or an error
+  *                                   code (see @ref opus_errorcodes) on
+  *                                   failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusMSDecoder *opus_multistream_decoder_create(
+      opus_int32 Fs,
+      int channels,
+      int streams,
+      int coupled_streams,
+      const unsigned char *mapping,
+      int *error
+) OPUS_ARG_NONNULL(5);
+
+/** Intialize a previously allocated decoder state object.
+  * The memory pointed to by \a st must be at least the size returned by
+  * opus_multistream_encoder_get_size().
+  * This is intended for applications which use their own allocator instead of
+  * malloc.
+  * To reset a previously initialized state, use the #OPUS_RESET_STATE CTL.
+  * @see opus_multistream_decoder_create
+  * @see opus_multistream_deocder_get_size
+  * @param st <tt>OpusMSEncoder*</tt>: Multistream encoder state to initialize.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate to decode at (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels to output.
+  *                               This must be at most 255.
+  *                               It may be different from the number of coded
+  *                               channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param streams <tt>int</tt>: The total number of streams coded in the
+  *                              input.
+  *                              This must be no more than 255.
+  * @param coupled_streams <tt>int</tt>: Number of streams to decode as coupled
+  *                                      (2 channel) streams.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      coded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than 255.
+  * @param[in] mapping <code>const unsigned char[channels]</code>: Mapping from
+  *                    coded channels to output channels, as described in
+  *                    @ref opus_multistream.
+  * @returns #OPUS_OK on success, or an error code (see @ref opus_errorcodes)
+  *          on failure.
+  */
+OPUS_EXPORT int opus_multistream_decoder_init(
+      OpusMSDecoder *st,
+      opus_int32 Fs,
+      int channels,
+      int streams,
+      int coupled_streams,
+      const unsigned char *mapping
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(6);
+
+/** Decode a multistream Opus packet.
+  * @param st <tt>OpusMSDecoder*</tt>: Multistream decoder state.
+  * @param[in] data <tt>const unsigned char*</tt>: Input payload.
+  *                                                Use a <code>NULL</code>
+  *                                                pointer to indicate packet
+  *                                                loss.
+  * @param len <tt>opus_int32</tt>: Number of bytes in payload.
+  * @param[out] pcm <tt>opus_int16*</tt>: Output signal, with interleaved
+  *                                       samples.
+  *                                       This must contain room for
+  *                                       <code>frame_size*channels</code>
+  *                                       samples.
+  * @param frame_size <tt>int</tt>: The number of samples per channel of
+  *                                 available space in \a pcm.
+  *                                 If this is less than the maximum packet duration
+  *                                 (120 ms; 5760 for 48kHz), this function will not be capable
+  *                                 of decoding some packets. In the case of PLC (data==NULL)
+  *                                 or FEC (decode_fec=1), then frame_size needs to be exactly
+  *                                 the duration of audio that is missing, otherwise the
+  *                                 decoder will not be in the optimal state to decode the
+  *                                 next incoming packet. For the PLC and FEC cases, frame_size
+  *                                 <b>must</b> be a multiple of 2.5 ms.
+  * @param decode_fec <tt>int</tt>: Flag (0 or 1) to request that any in-band
+  *                                 forward error correction data be decoded.
+  *                                 If no such data is available, the frame is
+  *                                 decoded as if it were lost.
+  * @returns Number of samples decoded on success or a negative error code
+  *          (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_multistream_decode(
+    OpusMSDecoder *st,
+    const unsigned char *data,
+    opus_int32 len,
+    opus_int16 *pcm,
+    int frame_size,
+    int decode_fec
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+/** Decode a multistream Opus packet with floating point output.
+  * @param st <tt>OpusMSDecoder*</tt>: Multistream decoder state.
+  * @param[in] data <tt>const unsigned char*</tt>: Input payload.
+  *                                                Use a <code>NULL</code>
+  *                                                pointer to indicate packet
+  *                                                loss.
+  * @param len <tt>opus_int32</tt>: Number of bytes in payload.
+  * @param[out] pcm <tt>opus_int16*</tt>: Output signal, with interleaved
+  *                                       samples.
+  *                                       This must contain room for
+  *                                       <code>frame_size*channels</code>
+  *                                       samples.
+  * @param frame_size <tt>int</tt>: The number of samples per channel of
+  *                                 available space in \a pcm.
+  *                                 If this is less than the maximum packet duration
+  *                                 (120 ms; 5760 for 48kHz), this function will not be capable
+  *                                 of decoding some packets. In the case of PLC (data==NULL)
+  *                                 or FEC (decode_fec=1), then frame_size needs to be exactly
+  *                                 the duration of audio that is missing, otherwise the
+  *                                 decoder will not be in the optimal state to decode the
+  *                                 next incoming packet. For the PLC and FEC cases, frame_size
+  *                                 <b>must</b> be a multiple of 2.5 ms.
+  * @param decode_fec <tt>int</tt>: Flag (0 or 1) to request that any in-band
+  *                                 forward error correction data be decoded.
+  *                                 If no such data is available, the frame is
+  *                                 decoded as if it were lost.
+  * @returns Number of samples decoded on success or a negative error code
+  *          (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_multistream_decode_float(
+    OpusMSDecoder *st,
+    const unsigned char *data,
+    opus_int32 len,
+    float *pcm,
+    int frame_size,
+    int decode_fec
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+/** Perform a CTL function on a multistream Opus decoder.
+  *
+  * Generally the request and subsequent arguments are generated by a
+  * convenience macro.
+  * @param st <tt>OpusMSDecoder*</tt>: Multistream decoder state.
+  * @param request This and all remaining parameters should be replaced by one
+  *                of the convenience macros in @ref opus_genericctls,
+  *                @ref opus_decoderctls, or @ref opus_multistream_ctls.
+  * @see opus_genericctls
+  * @see opus_decoderctls
+  * @see opus_multistream_ctls
+  */
+OPUS_EXPORT int opus_multistream_decoder_ctl(OpusMSDecoder *st, int request, ...) OPUS_ARG_NONNULL(1);
+
+/** Frees an <code>OpusMSDecoder</code> allocated by
+  * opus_multistream_decoder_create().
+  * @param st <tt>OpusMSDecoder</tt>: Multistream decoder state to be freed.
+  */
+OPUS_EXPORT void opus_multistream_decoder_destroy(OpusMSDecoder *st);
+
+/**@}*/
+
+/**@}*/
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* OPUS_MULTISTREAM_H */

lib/codecs/inc/opus/opus_projection.h

@@ -0,0 +1,568 @@
+/* Copyright (c) 2017 Google Inc.
+   Written by Andrew Allen */
+/*
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
+   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+/**
+ * @file opus_projection.h
+ * @brief Opus projection reference API
+ */
+
+#ifndef OPUS_PROJECTION_H
+#define OPUS_PROJECTION_H
+
+#include "opus_multistream.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/** @cond OPUS_INTERNAL_DOC */
+
+/** These are the actual encoder and decoder CTL ID numbers.
+  * They should not be used directly by applications.c
+  * In general, SETs should be even and GETs should be odd.*/
+/**@{*/
+#define OPUS_PROJECTION_GET_DEMIXING_MATRIX_GAIN_REQUEST    6001
+#define OPUS_PROJECTION_GET_DEMIXING_MATRIX_SIZE_REQUEST    6003
+#define OPUS_PROJECTION_GET_DEMIXING_MATRIX_REQUEST         6005
+/**@}*/
+
+
+/** @endcond */
+
+/** @defgroup opus_projection_ctls Projection specific encoder and decoder CTLs
+  *
+  * These are convenience macros that are specific to the
+  * opus_projection_encoder_ctl() and opus_projection_decoder_ctl()
+  * interface.
+  * The CTLs from @ref opus_genericctls, @ref opus_encoderctls,
+  * @ref opus_decoderctls, and @ref opus_multistream_ctls may be applied to a
+  * projection encoder or decoder as well.
+  */
+/**@{*/
+
+/** Gets the gain (in dB. S7.8-format) of the demixing matrix from the encoder.
+  * @param[out] x <tt>opus_int32 *</tt>: Returns the gain (in dB. S7.8-format)
+  *                                      of the demixing matrix.
+  * @hideinitializer
+  */
+#define OPUS_PROJECTION_GET_DEMIXING_MATRIX_GAIN(x) OPUS_PROJECTION_GET_DEMIXING_MATRIX_GAIN_REQUEST, __opus_check_int_ptr(x)
+
+
+/** Gets the size in bytes of the demixing matrix from the encoder.
+  * @param[out] x <tt>opus_int32 *</tt>: Returns the size in bytes of the
+  *                                      demixing matrix.
+  * @hideinitializer
+  */
+#define OPUS_PROJECTION_GET_DEMIXING_MATRIX_SIZE(x) OPUS_PROJECTION_GET_DEMIXING_MATRIX_SIZE_REQUEST, __opus_check_int_ptr(x)
+
+
+/** Copies the demixing matrix to the supplied pointer location.
+  * @param[out] x <tt>unsigned char *</tt>: Returns the demixing matrix to the
+  *                                         supplied pointer location.
+  * @param y <tt>opus_int32</tt>: The size in bytes of the reserved memory at the
+  *                              pointer location.
+  * @hideinitializer
+  */
+#define OPUS_PROJECTION_GET_DEMIXING_MATRIX(x,y) OPUS_PROJECTION_GET_DEMIXING_MATRIX_REQUEST, x, __opus_check_int(y)
+
+
+/**@}*/
+
+/** Opus projection encoder state.
+ * This contains the complete state of a projection Opus encoder.
+ * It is position independent and can be freely copied.
+ * @see opus_projection_ambisonics_encoder_create
+ */
+typedef struct OpusProjectionEncoder OpusProjectionEncoder;
+
+
+/** Opus projection decoder state.
+  * This contains the complete state of a projection Opus decoder.
+  * It is position independent and can be freely copied.
+  * @see opus_projection_decoder_create
+  * @see opus_projection_decoder_init
+  */
+typedef struct OpusProjectionDecoder OpusProjectionDecoder;
+
+
+/**\name Projection encoder functions */
+/**@{*/
+
+/** Gets the size of an OpusProjectionEncoder structure.
+  * @param channels <tt>int</tt>: The total number of input channels to encode.
+  *                               This must be no more than 255.
+  * @param mapping_family <tt>int</tt>: The mapping family to use for selecting
+  *                                     the appropriate projection.
+  * @returns The size in bytes on success, or a negative error code
+  *          (see @ref opus_errorcodes) on error.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_projection_ambisonics_encoder_get_size(
+    int channels,
+    int mapping_family
+);
+
+
+/** Allocates and initializes a projection encoder state.
+  * Call opus_projection_encoder_destroy() to release
+  * this object when finished.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate of the input signal (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels in the input signal.
+  *                               This must be at most 255.
+  *                               It may be greater than the number of
+  *                               coded channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param mapping_family <tt>int</tt>: The mapping family to use for selecting
+  *                                     the appropriate projection.
+  * @param[out] streams <tt>int *</tt>: The total number of streams that will
+  *                                     be encoded from the input.
+  * @param[out] coupled_streams <tt>int *</tt>: Number of coupled (2 channel)
+  *                                 streams that will be encoded from the input.
+  * @param application <tt>int</tt>: The target encoder application.
+  *                                  This must be one of the following:
+  * <dl>
+  * <dt>#OPUS_APPLICATION_VOIP</dt>
+  * <dd>Process signal for improved speech intelligibility.</dd>
+  * <dt>#OPUS_APPLICATION_AUDIO</dt>
+  * <dd>Favor faithfulness to the original input.</dd>
+  * <dt>#OPUS_APPLICATION_RESTRICTED_LOWDELAY</dt>
+  * <dd>Configure the minimum possible coding delay by disabling certain modes
+  * of operation.</dd>
+  * </dl>
+  * @param[out] error <tt>int *</tt>: Returns #OPUS_OK on success, or an error
+  *                                   code (see @ref opus_errorcodes) on
+  *                                   failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusProjectionEncoder *opus_projection_ambisonics_encoder_create(
+    opus_int32 Fs,
+    int channels,
+    int mapping_family,
+    int *streams,
+    int *coupled_streams,
+    int application,
+    int *error
+) OPUS_ARG_NONNULL(4) OPUS_ARG_NONNULL(5);
+
+
+/** Initialize a previously allocated projection encoder state.
+  * The memory pointed to by \a st must be at least the size returned by
+  * opus_projection_ambisonics_encoder_get_size().
+  * This is intended for applications which use their own allocator instead of
+  * malloc.
+  * To reset a previously initialized state, use the #OPUS_RESET_STATE CTL.
+  * @see opus_projection_ambisonics_encoder_create
+  * @see opus_projection_ambisonics_encoder_get_size
+  * @param st <tt>OpusProjectionEncoder*</tt>: Projection encoder state to initialize.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate of the input signal (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels in the input signal.
+  *                               This must be at most 255.
+  *                               It may be greater than the number of
+  *                               coded channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param streams <tt>int</tt>: The total number of streams to encode from the
+  *                              input.
+  *                              This must be no more than the number of channels.
+  * @param coupled_streams <tt>int</tt>: Number of coupled (2 channel) streams
+  *                                      to encode.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      encoded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than the number of input channels.
+  * @param application <tt>int</tt>: The target encoder application.
+  *                                  This must be one of the following:
+  * <dl>
+  * <dt>#OPUS_APPLICATION_VOIP</dt>
+  * <dd>Process signal for improved speech intelligibility.</dd>
+  * <dt>#OPUS_APPLICATION_AUDIO</dt>
+  * <dd>Favor faithfulness to the original input.</dd>
+  * <dt>#OPUS_APPLICATION_RESTRICTED_LOWDELAY</dt>
+  * <dd>Configure the minimum possible coding delay by disabling certain modes
+  * of operation.</dd>
+  * </dl>
+  * @returns #OPUS_OK on success, or an error code (see @ref opus_errorcodes)
+  *          on failure.
+  */
+OPUS_EXPORT int opus_projection_ambisonics_encoder_init(
+    OpusProjectionEncoder *st,
+    opus_int32 Fs,
+    int channels,
+    int mapping_family,
+    int *streams,
+    int *coupled_streams,
+    int application
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(5) OPUS_ARG_NONNULL(6);
+
+
+/** Encodes a projection Opus frame.
+  * @param st <tt>OpusProjectionEncoder*</tt>: Projection encoder state.
+  * @param[in] pcm <tt>const opus_int16*</tt>: The input signal as interleaved
+  *                                            samples.
+  *                                            This must contain
+  *                                            <code>frame_size*channels</code>
+  *                                            samples.
+  * @param frame_size <tt>int</tt>: Number of samples per channel in the input
+  *                                 signal.
+  *                                 This must be an Opus frame size for the
+  *                                 encoder's sampling rate.
+  *                                 For example, at 48 kHz the permitted values
+  *                                 are 120, 240, 480, 960, 1920, and 2880.
+  *                                 Passing in a duration of less than 10 ms
+  *                                 (480 samples at 48 kHz) will prevent the
+  *                                 encoder from using the LPC or hybrid modes.
+  * @param[out] data <tt>unsigned char*</tt>: Output payload.
+  *                                           This must contain storage for at
+  *                                           least \a max_data_bytes.
+  * @param [in] max_data_bytes <tt>opus_int32</tt>: Size of the allocated
+  *                                                 memory for the output
+  *                                                 payload. This may be
+  *                                                 used to impose an upper limit on
+  *                                                 the instant bitrate, but should
+  *                                                 not be used as the only bitrate
+  *                                                 control. Use #OPUS_SET_BITRATE to
+  *                                                 control the bitrate.
+  * @returns The length of the encoded packet (in bytes) on success or a
+  *          negative error code (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_projection_encode(
+    OpusProjectionEncoder *st,
+    const opus_int16 *pcm,
+    int frame_size,
+    unsigned char *data,
+    opus_int32 max_data_bytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+
+/** Encodes a projection Opus frame from floating point input.
+  * @param st <tt>OpusProjectionEncoder*</tt>: Projection encoder state.
+  * @param[in] pcm <tt>const float*</tt>: The input signal as interleaved
+  *                                       samples with a normal range of
+  *                                       +/-1.0.
+  *                                       Samples with a range beyond +/-1.0
+  *                                       are supported but will be clipped by
+  *                                       decoders using the integer API and
+  *                                       should only be used if it is known
+  *                                       that the far end supports extended
+  *                                       dynamic range.
+  *                                       This must contain
+  *                                       <code>frame_size*channels</code>
+  *                                       samples.
+  * @param frame_size <tt>int</tt>: Number of samples per channel in the input
+  *                                 signal.
+  *                                 This must be an Opus frame size for the
+  *                                 encoder's sampling rate.
+  *                                 For example, at 48 kHz the permitted values
+  *                                 are 120, 240, 480, 960, 1920, and 2880.
+  *                                 Passing in a duration of less than 10 ms
+  *                                 (480 samples at 48 kHz) will prevent the
+  *                                 encoder from using the LPC or hybrid modes.
+  * @param[out] data <tt>unsigned char*</tt>: Output payload.
+  *                                           This must contain storage for at
+  *                                           least \a max_data_bytes.
+  * @param [in] max_data_bytes <tt>opus_int32</tt>: Size of the allocated
+  *                                                 memory for the output
+  *                                                 payload. This may be
+  *                                                 used to impose an upper limit on
+  *                                                 the instant bitrate, but should
+  *                                                 not be used as the only bitrate
+  *                                                 control. Use #OPUS_SET_BITRATE to
+  *                                                 control the bitrate.
+  * @returns The length of the encoded packet (in bytes) on success or a
+  *          negative error code (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_projection_encode_float(
+    OpusProjectionEncoder *st,
+    const float *pcm,
+    int frame_size,
+    unsigned char *data,
+    opus_int32 max_data_bytes
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(2) OPUS_ARG_NONNULL(4);
+
+
+/** Frees an <code>OpusProjectionEncoder</code> allocated by
+  * opus_projection_ambisonics_encoder_create().
+  * @param st <tt>OpusProjectionEncoder*</tt>: Projection encoder state to be freed.
+  */
+OPUS_EXPORT void opus_projection_encoder_destroy(OpusProjectionEncoder *st);
+
+
+/** Perform a CTL function on a projection Opus encoder.
+  *
+  * Generally the request and subsequent arguments are generated by a
+  * convenience macro.
+  * @param st <tt>OpusProjectionEncoder*</tt>: Projection encoder state.
+  * @param request This and all remaining parameters should be replaced by one
+  *                of the convenience macros in @ref opus_genericctls,
+  *                @ref opus_encoderctls, @ref opus_multistream_ctls, or
+  *                @ref opus_projection_ctls
+  * @see opus_genericctls
+  * @see opus_encoderctls
+  * @see opus_multistream_ctls
+  * @see opus_projection_ctls
+  */
+OPUS_EXPORT int opus_projection_encoder_ctl(OpusProjectionEncoder *st, int request, ...) OPUS_ARG_NONNULL(1);
+
+
+/**@}*/
+
+/**\name Projection decoder functions */
+/**@{*/
+
+/** Gets the size of an <code>OpusProjectionDecoder</code> structure.
+  * @param channels <tt>int</tt>: The total number of output channels.
+  *                               This must be no more than 255.
+  * @param streams <tt>int</tt>: The total number of streams coded in the
+  *                              input.
+  *                              This must be no more than 255.
+  * @param coupled_streams <tt>int</tt>: Number streams to decode as coupled
+  *                                      (2 channel) streams.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      coded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than 255.
+  * @returns The size in bytes on success, or a negative error code
+  *          (see @ref opus_errorcodes) on error.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT opus_int32 opus_projection_decoder_get_size(
+    int channels,
+    int streams,
+    int coupled_streams
+);
+
+
+/** Allocates and initializes a projection decoder state.
+  * Call opus_projection_decoder_destroy() to release
+  * this object when finished.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate to decode at (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels to output.
+  *                               This must be at most 255.
+  *                               It may be different from the number of coded
+  *                               channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param streams <tt>int</tt>: The total number of streams coded in the
+  *                              input.
+  *                              This must be no more than 255.
+  * @param coupled_streams <tt>int</tt>: Number of streams to decode as coupled
+  *                                      (2 channel) streams.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      coded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than 255.
+  * @param[in] demixing_matrix <tt>const unsigned char[demixing_matrix_size]</tt>: Demixing matrix
+  *                         that mapping from coded channels to output channels,
+  *                         as described in @ref opus_projection and
+  *                         @ref opus_projection_ctls.
+  * @param demixing_matrix_size <tt>opus_int32</tt>: The size in bytes of the
+  *                                                  demixing matrix, as
+  *                                                  described in @ref
+  *                                                  opus_projection_ctls.
+  * @param[out] error <tt>int *</tt>: Returns #OPUS_OK on success, or an error
+  *                                   code (see @ref opus_errorcodes) on
+  *                                   failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT OpusProjectionDecoder *opus_projection_decoder_create(
+    opus_int32 Fs,
+    int channels,
+    int streams,
+    int coupled_streams,
+    unsigned char *demixing_matrix,
+    opus_int32 demixing_matrix_size,
+    int *error
+) OPUS_ARG_NONNULL(5);
+
+
+/** Intialize a previously allocated projection decoder state object.
+  * The memory pointed to by \a st must be at least the size returned by
+  * opus_projection_decoder_get_size().
+  * This is intended for applications which use their own allocator instead of
+  * malloc.
+  * To reset a previously initialized state, use the #OPUS_RESET_STATE CTL.
+  * @see opus_projection_decoder_create
+  * @see opus_projection_deocder_get_size
+  * @param st <tt>OpusProjectionDecoder*</tt>: Projection encoder state to initialize.
+  * @param Fs <tt>opus_int32</tt>: Sampling rate to decode at (in Hz).
+  *                                This must be one of 8000, 12000, 16000,
+  *                                24000, or 48000.
+  * @param channels <tt>int</tt>: Number of channels to output.
+  *                               This must be at most 255.
+  *                               It may be different from the number of coded
+  *                               channels (<code>streams +
+  *                               coupled_streams</code>).
+  * @param streams <tt>int</tt>: The total number of streams coded in the
+  *                              input.
+  *                              This must be no more than 255.
+  * @param coupled_streams <tt>int</tt>: Number of streams to decode as coupled
+  *                                      (2 channel) streams.
+  *                                      This must be no larger than the total
+  *                                      number of streams.
+  *                                      Additionally, The total number of
+  *                                      coded channels (<code>streams +
+  *                                      coupled_streams</code>) must be no
+  *                                      more than 255.
+  * @param[in] demixing_matrix <tt>const unsigned char[demixing_matrix_size]</tt>: Demixing matrix
+  *                         that mapping from coded channels to output channels,
+  *                         as described in @ref opus_projection and
+  *                         @ref opus_projection_ctls.
+  * @param demixing_matrix_size <tt>opus_int32</tt>: The size in bytes of the
+  *                                                  demixing matrix, as
+  *                                                  described in @ref
+  *                                                  opus_projection_ctls.
+  * @returns #OPUS_OK on success, or an error code (see @ref opus_errorcodes)
+  *          on failure.
+  */
+OPUS_EXPORT int opus_projection_decoder_init(
+    OpusProjectionDecoder *st,
+    opus_int32 Fs,
+    int channels,
+    int streams,
+    int coupled_streams,
+    unsigned char *demixing_matrix,
+    opus_int32 demixing_matrix_size
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(6);
+
+
+/** Decode a projection Opus packet.
+  * @param st <tt>OpusProjectionDecoder*</tt>: Projection decoder state.
+  * @param[in] data <tt>const unsigned char*</tt>: Input payload.
+  *                                                Use a <code>NULL</code>
+  *                                                pointer to indicate packet
+  *                                                loss.
+  * @param len <tt>opus_int32</tt>: Number of bytes in payload.
+  * @param[out] pcm <tt>opus_int16*</tt>: Output signal, with interleaved
+  *                                       samples.
+  *                                       This must contain room for
+  *                                       <code>frame_size*channels</code>
+  *                                       samples.
+  * @param frame_size <tt>int</tt>: The number of samples per channel of
+  *                                 available space in \a pcm.
+  *                                 If this is less than the maximum packet duration
+  *                                 (120 ms; 5760 for 48kHz), this function will not be capable
+  *                                 of decoding some packets. In the case of PLC (data==NULL)
+  *                                 or FEC (decode_fec=1), then frame_size needs to be exactly
+  *                                 the duration of audio that is missing, otherwise the
+  *                                 decoder will not be in the optimal state to decode the
+  *                                 next incoming packet. For the PLC and FEC cases, frame_size
+  *                                 <b>must</b> be a multiple of 2.5 ms.
+  * @param decode_fec <tt>int</tt>: Flag (0 or 1) to request that any in-band
+  *                                 forward error correction data be decoded.
+  *                                 If no such data is available, the frame is
+  *                                 decoded as if it were lost.
+  * @returns Number of samples decoded on success or a negative error code
+  *          (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_projection_decode(
+    OpusProjectionDecoder *st,
+    const unsigned char *data,
+    opus_int32 len,
+    opus_int16 *pcm,
+    int frame_size,
+    int decode_fec
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+
+/** Decode a projection Opus packet with floating point output.
+  * @param st <tt>OpusProjectionDecoder*</tt>: Projection decoder state.
+  * @param[in] data <tt>const unsigned char*</tt>: Input payload.
+  *                                                Use a <code>NULL</code>
+  *                                                pointer to indicate packet
+  *                                                loss.
+  * @param len <tt>opus_int32</tt>: Number of bytes in payload.
+  * @param[out] pcm <tt>opus_int16*</tt>: Output signal, with interleaved
+  *                                       samples.
+  *                                       This must contain room for
+  *                                       <code>frame_size*channels</code>
+  *                                       samples.
+  * @param frame_size <tt>int</tt>: The number of samples per channel of
+  *                                 available space in \a pcm.
+  *                                 If this is less than the maximum packet duration
+  *                                 (120 ms; 5760 for 48kHz), this function will not be capable
+  *                                 of decoding some packets. In the case of PLC (data==NULL)
+  *                                 or FEC (decode_fec=1), then frame_size needs to be exactly
+  *                                 the duration of audio that is missing, otherwise the
+  *                                 decoder will not be in the optimal state to decode the
+  *                                 next incoming packet. For the PLC and FEC cases, frame_size
+  *                                 <b>must</b> be a multiple of 2.5 ms.
+  * @param decode_fec <tt>int</tt>: Flag (0 or 1) to request that any in-band
+  *                                 forward error correction data be decoded.
+  *                                 If no such data is available, the frame is
+  *                                 decoded as if it were lost.
+  * @returns Number of samples decoded on success or a negative error code
+  *          (see @ref opus_errorcodes) on failure.
+  */
+OPUS_EXPORT OPUS_WARN_UNUSED_RESULT int opus_projection_decode_float(
+    OpusProjectionDecoder *st,
+    const unsigned char *data,
+    opus_int32 len,
+    float *pcm,
+    int frame_size,
+    int decode_fec
+) OPUS_ARG_NONNULL(1) OPUS_ARG_NONNULL(4);
+
+
+/** Perform a CTL function on a projection Opus decoder.
+  *
+  * Generally the request and subsequent arguments are generated by a
+  * convenience macro.
+  * @param st <tt>OpusProjectionDecoder*</tt>: Projection decoder state.
+  * @param request This and all remaining parameters should be replaced by one
+  *                of the convenience macros in @ref opus_genericctls,
+  *                @ref opus_decoderctls, @ref opus_multistream_ctls, or
+  *                @ref opus_projection_ctls.
+  * @see opus_genericctls
+  * @see opus_decoderctls
+  * @see opus_multistream_ctls
+  * @see opus_projection_ctls
+  */
+OPUS_EXPORT int opus_projection_decoder_ctl(OpusProjectionDecoder *st, int request, ...) OPUS_ARG_NONNULL(1);
+
+
+/** Frees an <code>OpusProjectionDecoder</code> allocated by
+  * opus_projection_decoder_create().
+  * @param st <tt>OpusProjectionDecoder</tt>: Projection decoder state to be freed.
+  */
+OPUS_EXPORT void opus_projection_decoder_destroy(OpusProjectionDecoder *st);
+
+
+/**@}*/
+
+/**@}*/
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* OPUS_PROJECTION_H */

lib/codecs/inc/opus/opus_types.h

@@ -0,0 +1,166 @@
+/* (C) COPYRIGHT 1994-2002 Xiph.Org Foundation */
+/* Modified by Jean-Marc Valin */
+/*
+   Redistribution and use in source and binary forms, with or without
+   modification, are permitted provided that the following conditions
+   are met:
+
+   - Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+
+   - Redistributions in binary form must reproduce the above copyright
+   notice, this list of conditions and the following disclaimer in the
+   documentation and/or other materials provided with the distribution.
+
+   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+   ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
+   OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+   EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+   PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+   PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+   LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+/* opus_types.h based on ogg_types.h from libogg */
+
+/**
+   @file opus_types.h
+   @brief Opus reference implementation types
+*/
+#ifndef OPUS_TYPES_H
+#define OPUS_TYPES_H
+
+#define opus_int         int                     /* used for counters etc; at least 16 bits */
+#define opus_int64       long long
+#define opus_int8        signed char
+
+#define opus_uint        unsigned int            /* used for counters etc; at least 16 bits */
+#define opus_uint64      unsigned long long
+#define opus_uint8       unsigned char
+
+/* Use the real stdint.h if it's there (taken from Paul Hsieh's pstdint.h) */
+#if (defined(__STDC__) && __STDC__ && defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L) || (defined(__GNUC__) && (defined(_STDINT_H) || defined(_STDINT_H_)) || defined (HAVE_STDINT_H))
+#include <stdint.h>
+#  undef opus_int64
+#  undef opus_int8
+#  undef opus_uint64
+#  undef opus_uint8
+   typedef int8_t opus_int8;
+   typedef uint8_t opus_uint8;
+   typedef int16_t opus_int16;
+   typedef uint16_t opus_uint16;
+   typedef int32_t opus_int32;
+   typedef uint32_t opus_uint32;
+   typedef int64_t opus_int64;
+   typedef uint64_t opus_uint64;
+#elif defined(_WIN32)
+
+#  if defined(__CYGWIN__)
+#    include <_G_config.h>
+     typedef _G_int32_t opus_int32;
+     typedef _G_uint32_t opus_uint32;
+     typedef _G_int16 opus_int16;
+     typedef _G_uint16 opus_uint16;
+#  elif defined(__MINGW32__)
+     typedef short opus_int16;
+     typedef unsigned short opus_uint16;
+     typedef int opus_int32;
+     typedef unsigned int opus_uint32;
+#  elif defined(__MWERKS__)
+     typedef int opus_int32;
+     typedef unsigned int opus_uint32;
+     typedef short opus_int16;
+     typedef unsigned short opus_uint16;
+#  else
+     /* MSVC/Borland */
+     typedef __int32 opus_int32;
+     typedef unsigned __int32 opus_uint32;
+     typedef __int16 opus_int16;
+     typedef unsigned __int16 opus_uint16;
+#  endif
+
+#elif defined(__MACOS__)
+
+#  include <sys/types.h>
+   typedef SInt16 opus_int16;
+   typedef UInt16 opus_uint16;
+   typedef SInt32 opus_int32;
+   typedef UInt32 opus_uint32;
+
+#elif (defined(__APPLE__) && defined(__MACH__)) /* MacOS X Framework build */
+
+#  include <sys/types.h>
+   typedef int16_t opus_int16;
+   typedef u_int16_t opus_uint16;
+   typedef int32_t opus_int32;
+   typedef u_int32_t opus_uint32;
+
+#elif defined(__BEOS__)
+
+   /* Be */
+#  include <inttypes.h>
+   typedef int16 opus_int16;
+   typedef u_int16 opus_uint16;
+   typedef int32_t opus_int32;
+   typedef u_int32_t opus_uint32;
+
+#elif defined (__EMX__)
+
+   /* OS/2 GCC */
+   typedef short opus_int16;
+   typedef unsigned short opus_uint16;
+   typedef int opus_int32;
+   typedef unsigned int opus_uint32;
+
+#elif defined (DJGPP)
+
+   /* DJGPP */
+   typedef short opus_int16;
+   typedef unsigned short opus_uint16;
+   typedef int opus_int32;
+   typedef unsigned int opus_uint32;
+
+#elif defined(R5900)
+
+   /* PS2 EE */
+   typedef int opus_int32;
+   typedef unsigned opus_uint32;
+   typedef short opus_int16;
+   typedef unsigned short opus_uint16;
+
+#elif defined(__SYMBIAN32__)
+
+   /* Symbian GCC */
+   typedef signed short opus_int16;
+   typedef unsigned short opus_uint16;
+   typedef signed int opus_int32;
+   typedef unsigned int opus_uint32;
+
+#elif defined(CONFIG_TI_C54X) || defined (CONFIG_TI_C55X)
+
+   typedef short opus_int16;
+   typedef unsigned short opus_uint16;
+   typedef long opus_int32;
+   typedef unsigned long opus_uint32;
+
+#elif defined(CONFIG_TI_C6X)
+
+   typedef short opus_int16;
+   typedef unsigned short opus_uint16;
+   typedef int opus_int32;
+   typedef unsigned int opus_uint32;
+
+#else
+
+   /* Give up, take a reasonable guess */
+   typedef short opus_int16;
+   typedef unsigned short opus_uint16;
+   typedef int opus_int32;
+   typedef unsigned int opus_uint32;
+
+#endif
+
+#endif  /* OPUS_TYPES_H */

lib/codecs/inc/resample16/resample16.h

@@ -0,0 +1,92 @@
+/*
+ * FILE: resample16.h
+ *
+ * The configuration constants below govern
+ * the number of bits in the input sample and filter coefficients, the 
+ * number of bits to the right of the binary-point for fixed-point math, etc.
+ *
+ */
+
+/* Conversion constants */
+#define Nhc       8
+#define Na        7
+#define Np       (Nhc+Na)
+#define Npc      (1<<Nhc)
+#define Amask    ((1<<Na)-1)
+#define Pmask    ((1<<Np)-1)
+#define Nh       16
+#define Nb       16
+#define Nhxn     14
+#define Nhg      (Nh-Nhxn)
+#define NLpScl   13
+
+/* Description of constants:
+ *
+ * Npc - is the number of look-up values available for the lowpass filter
+ *    between the beginning of its impulse response and the "cutoff time"
+ *    of the filter.  The cutoff time is defined as the reciprocal of the
+ *    lowpass-filter cut off frequence in Hz.  For example, if the
+ *    lowpass filter were a sinc function, Npc would be the index of the
+ *    impulse-response lookup-table corresponding to the first zero-
+ *    crossing of the sinc function.  (The inverse first zero-crossing
+ *    time of a sinc function equals its nominal cutoff frequency in Hz.)
+ *    Npc must be a power of 2 due to the details of the current
+ *    implementation. The default value of 512 is sufficiently high that
+ *    using linear interpolation to fill in between the table entries
+ *    gives approximately 16-bit accuracy in filter coefficients.
+ *
+ * Nhc - is log base 2 of Npc.
+ *
+ * Na - is the number of bits devoted to linear interpolation of the
+ *    filter coefficients.
+ *
+ * Np - is Na + Nhc, the number of bits to the right of the binary point
+ *    in the integer "time" variable. To the left of the point, it indexes
+ *    the input array (X), and to the right, it is interpreted as a number
+ *    between 0 and 1 sample of the input X.  Np must be less than 16 in
+ *    this implementation.
+ *
+ * Nh - is the number of bits in the filter coefficients. The sum of Nh and
+ *    the number of bits in the input data (typically 16) cannot exceed 32.
+ *    Thus Nh should be 16.  The largest filter coefficient should nearly
+ *    fill 16 bits (32767).
+ *
+ * Nb - is the number of bits in the input data. The sum of Nb and Nh cannot
+ *    exceed 32.
+ *
+ * Nhxn - is the number of bits to right shift after multiplying each input
+ *    sample times a filter coefficient. It can be as great as Nh and as
+ *    small as 0. Nhxn = Nh-2 gives 2 guard bits in the multiply-add
+ *    accumulation.  If Nhxn=0, the accumulation will soon overflow 32 bits.
+ *
+ * Nhg - is the number of guard bits in mpy-add accumulation (equal to Nh-Nhxn)
+ *
+ * NLpScl - is the number of bits allocated to the unity-gain normalization
+ *    factor.  The output of the lowpass filter is multiplied by LpScl and
+ *    then right-shifted NLpScl bits. To avoid overflow, we must have 
+ *    Nb+Nhg+NLpScl < 32.
+ */
+
+typedef char           BOOL;
+typedef short          HWORD;
+typedef unsigned short UHWORD;
+typedef int            WORD;
+typedef unsigned int   UWORD;
+
+struct resample16_s;
+typedef struct {
+	UHWORD LpScl;               /* Unity-gain scale factor */
+	UHWORD Nwing;               /* Filter table size */
+	UHWORD Nmult;               /* Filter length for up-conversions */
+	const HWORD *Imp;           /* Filter coefficients */
+	const HWORD *ImpD;          /* ImpD[n] = Imp[n+1]-Imp[n] */
+} resample16_filter_t;
+
+typedef enum { RESAMPLE16_BASIC, RESAMPLE16_LOW, RESAMPLE16_MED } resample16_filter_e;
+
+WORD 				resample16(struct resample16_s *r, HWORD X[],	int inCount, HWORD Y[]);
+struct resample16_s*  resample16_create(float factor, resample16_filter_e filter, resample16_filter_t *custom, BOOL interp);
+void 				resample16_delete(struct resample16_s *r);
+void 				resample16_flush(struct resample16_s *r);
+
+

lib/codecs/inc/soxr/soxr.h

@@ -0,0 +1,348 @@
+/* SoX Resampler Library       Copyright (c) 2007-13 robs@users.sourceforge.net
+ *
+ * This library is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU Lesser General Public License as published by
+ * the Free Software Foundation; either version 2.1 of the License, or (at
+ * your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this library; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+
+
+
+/* -------------------------------- Gubbins --------------------------------- */
+
+#if !defined soxr_included
+#define soxr_included
+
+
+#if defined __cplusplus
+  #include <cstddef>
+  extern "C" {
+#else
+  #include <stddef.h>
+#endif
+
+#if defined SOXR_DLL
+  #if defined soxr_EXPORTS
+    #define SOXR __declspec(dllexport)
+  #else
+    #define SOXR __declspec(dllimport)
+  #endif
+#elif defined SOXR_VISIBILITY && defined __GNUC__ && (__GNUC__ > 4 || __GNUC__ == 4 && __GNUC_MINOR__ >= 1)
+  #define SOXR __attribute__ ((visibility("default")))
+#else
+  #define SOXR
+#endif
+
+typedef struct soxr_io_spec soxr_io_spec_t;
+typedef struct soxr_quality_spec soxr_quality_spec_t;
+typedef struct soxr_runtime_spec soxr_runtime_spec_t;
+
+
+
+/* ---------------------------- API conventions --------------------------------
+
+Buffer lengths (and occupancies) are expressed as the number of contained
+samples per channel.
+
+Parameter names for buffer lengths have the suffix `len'.
+
+A single-character `i' or 'o' is often used in names to give context as
+input or output (e.g. ilen, olen).                                            */
+
+
+
+/* --------------------------- Version management --------------------------- */
+
+/* E.g. #if SOXR_THIS_VERSION >= SOXR_VERSION(0,1,1) ...                      */
+
+#define SOXR_VERSION(x,y,z)     (((x)<<16)|((y)<<8)|(z))
+#define SOXR_THIS_VERSION       SOXR_VERSION(0,1,2)
+#define SOXR_THIS_VERSION_STR               "0.1.2"
+
+
+
+/* --------------------------- Type declarations ---------------------------- */
+
+typedef struct soxr * soxr_t;          /* A resampler for 1 or more channels. */
+typedef char const * soxr_error_t;                /* 0:no-error; non-0:error. */
+
+typedef void       * soxr_buf_t;  /* 1 buffer of channel-interleaved samples. */
+typedef void const * soxr_cbuf_t;                        /* Ditto; read-only. */
+
+typedef soxr_buf_t const  * soxr_bufs_t;/* Or, a separate buffer for each ch. */
+typedef soxr_cbuf_t const * soxr_cbufs_t;                /* Ditto; read-only. */
+
+typedef void const * soxr_in_t;      /* Either a soxr_cbuf_t or soxr_cbufs_t,
+                                        depending on itype in soxr_io_spec_t. */
+typedef void       * soxr_out_t;     /* Either a soxr_buf_t or soxr_bufs_t,
+                                        depending on otype in soxr_io_spec_t. */
+
+
+
+/* --------------------------- API main functions --------------------------- */
+
+SOXR char const * soxr_version(void);  /* Query library version: "libsoxr-x.y.z" */
+
+#define soxr_strerror(e)               /* Soxr counterpart to strerror. */     \
+    ((e)?(e):"no error")
+
+
+/* Create a stream resampler: */
+
+SOXR soxr_t soxr_create(
+    double      input_rate,      /* Input sample-rate. */
+    double      output_rate,     /* Output sample-rate. */
+    unsigned    num_channels,    /* Number of channels to be used. */
+        /* All following arguments are optional (may be set to NULL). */
+    soxr_error_t *,              /* To report any error during creation. */
+    soxr_io_spec_t const *,      /* To specify non-default I/O formats. */
+    soxr_quality_spec_t const *, /* To specify non-default resampling quality.*/
+    soxr_runtime_spec_t const *);/* To specify non-default runtime resources.
+
+    Default io_spec      is per soxr_io_spec(SOXR_FLOAT32_I, SOXR_FLOAT32_I)
+    Default quality_spec is per soxr_quality_spec(SOXR_HQ, 0)
+    Default runtime_spec is per soxr_runtime_spec(1)                          */
+
+
+
+/* If not using an app-supplied input function, after creating a stream
+ * resampler, repeatedly call: */
+
+SOXR soxr_error_t soxr_process(
+    soxr_t      resampler,      /* As returned by soxr_create. */
+                            /* Input (to be resampled): */
+    soxr_in_t   in,             /* Input buffer(s); may be NULL (see below). */
+    size_t      ilen,           /* Input buf. length (samples per channel). */
+    size_t      * idone,        /* To return actual # samples used (<= ilen). */
+                            /* Output (resampled): */
+    soxr_out_t  out,            /* Output buffer(s).*/
+    size_t      olen,           /* Output buf. length (samples per channel). */
+    size_t      * odone);       /* To return actual # samples out (<= olen).
+
+    Note that no special meaning is associated with ilen or olen equal to
+    zero.  End-of-input (i.e. no data is available nor shall be available)
+    may be indicated by seting `in' to NULL.                                  */
+
+
+
+/* If using an app-supplied input function, it must look and behave like this:*/
+
+typedef size_t /* data_len */
+  (* soxr_input_fn_t)(         /* Supply data to be resampled. */
+    void * input_fn_state,     /* As given to soxr_set_input_fn (below). */
+    soxr_in_t * data,          /* Returned data; see below. N.B. ptr to ptr(s)*/
+    size_t requested_len);     /* Samples per channel, >= returned data_len.
+
+  data_len  *data     Indicates    Meaning
+   ------- -------   ------------  -------------------------
+     !=0     !=0       Success     *data contains data to be
+                                   input to the resampler.
+      0    !=0 (or   End-of-input  No data is available nor
+           not set)                shall be available.
+      0       0        Failure     An error occurred whilst trying to
+                                   source data to be input to the resampler.  */
+
+/* and be registered with a previously created stream resampler using: */
+
+SOXR soxr_error_t soxr_set_input_fn(/* Set (or reset) an input function.*/
+    soxr_t resampler,            /* As returned by soxr_create. */
+    soxr_input_fn_t,             /* Function to supply data to be resampled.*/
+    void * input_fn_state,       /* If needed by the input function. */
+    size_t max_ilen);            /* Maximum value for input fn. requested_len.*/
+
+/* then repeatedly call: */
+
+SOXR size_t /*odone*/ soxr_output(/* Resample and output a block of data.*/
+    soxr_t resampler,            /* As returned by soxr_create. */
+    soxr_out_t data,             /* App-supplied buffer(s) for resampled data.*/
+    size_t olen);                /* Amount of data to output; >= odone. */
+
+
+
+/* Common stream resampler operations: */
+
+SOXR soxr_error_t soxr_error(soxr_t);   /* Query error status. */
+SOXR size_t   * soxr_num_clips(soxr_t); /* Query int. clip counter (for R/W). */
+SOXR double     soxr_delay(soxr_t);  /* Query current delay in output samples.*/
+SOXR char const * soxr_engine(soxr_t p); /* Query resampling engine name. */
+
+SOXR soxr_error_t soxr_clear(soxr_t); /* Ready for fresh signal, same config. */
+SOXR void         soxr_delete(soxr_t);  /* Free resources. */
+
+
+
+/* `Short-cut', single call to resample a (probably short) signal held entirely
+ * in memory.  See soxr_create and soxr_process above for parameter details.
+ * Note that unlike soxr_create however, the default quality spec. for
+ * soxr_oneshot is per soxr_quality_spec(SOXR_LQ, 0). */
+
+SOXR soxr_error_t soxr_oneshot(
+    double         input_rate,
+    double         output_rate,
+    unsigned       num_channels,
+    soxr_in_t    in , size_t ilen, size_t * idone,
+    soxr_out_t   out, size_t olen, size_t * odone,
+    soxr_io_spec_t const *,
+    soxr_quality_spec_t const *,
+    soxr_runtime_spec_t const *);
+
+
+
+/* For variable-rate resampling. See example # 5 for how to create a
+ * variable-rate resampler and how to use this function. */
+
+SOXR soxr_error_t soxr_set_io_ratio(soxr_t, double io_ratio, size_t slew_len);
+
+
+
+/* -------------------------- API type definitions -------------------------- */
+
+typedef enum {          /* Datatypes supported for I/O to/from the resampler: */
+  /* Internal; do not use: */
+  SOXR_FLOAT32, SOXR_FLOAT64, SOXR_INT32, SOXR_INT16, SOXR_SPLIT = 4,
+
+  /* Use for interleaved channels: */
+  SOXR_FLOAT32_I = SOXR_FLOAT32, SOXR_FLOAT64_I, SOXR_INT32_I, SOXR_INT16_I,
+
+  /* Use for split channels: */
+  SOXR_FLOAT32_S = SOXR_SPLIT  , SOXR_FLOAT64_S, SOXR_INT32_S, SOXR_INT16_S
+
+} soxr_datatype_t;
+
+#define soxr_datatype_size(x)  /* Returns `sizeof' a soxr_datatype_t sample. */\
+  ((unsigned char *)"\4\10\4\2")[(x)&3]
+
+
+
+struct soxr_io_spec {                                            /* Typically */
+  soxr_datatype_t itype;     /* Input datatype.                SOXR_FLOAT32_I */
+  soxr_datatype_t otype;     /* Output datatype.               SOXR_FLOAT32_I */
+  double scale;              /* Linear gain to apply during resampling.  1    */
+  void * e;                  /* Reserved for internal use                0    */
+  unsigned long flags;       /* Per the following #defines.              0    */
+};
+
+#define SOXR_TPDF              0     /* Applicable only if otype is INT16. */
+#define SOXR_NO_DITHER         8u    /* Disable the above. */
+
+
+
+struct soxr_quality_spec {                                       /* Typically */
+  double precision;         /* Conversion precision (in bits).           20   */
+  double phase_response;    /* 0=minimum, ... 50=linear, ... 100=maximum 50   */
+  double passband_end;      /* 0dB pt. bandwidth to preserve; nyquist=1  0.913*/
+  double stopband_begin;    /* Aliasing/imaging control; > passband_end   1   */
+  void * e;                 /* Reserved for internal use.                 0   */
+  unsigned long flags;      /* Per the following #defines.                0   */
+};
+
+#define SOXR_ROLLOFF_SMALL     0u    /* <= 0.01 dB */
+#define SOXR_ROLLOFF_MEDIUM    1u    /* <= 0.35 dB */
+#define SOXR_ROLLOFF_NONE      2u    /* For Chebyshev bandwidth. */
+
+#define SOXR_MAINTAIN_3DB_PT   4u  /* Reserved for internal use. */
+#define SOXR_HI_PREC_CLOCK     8u  /* Increase `irrational' ratio accuracy. */
+#define SOXR_DOUBLE_PRECISION 16u  /* Use D.P. calcs even if precision <= 20. */
+#define SOXR_VR               32u  /* Variable-rate resampling. */
+
+
+
+struct soxr_runtime_spec {                                       /* Typically */
+  unsigned log2_min_dft_size;/* For DFT efficiency. [8,15]              10    */
+  unsigned log2_large_dft_size;/* For DFT efficiency. [16,20]           17    */
+  unsigned coef_size_kbytes; /* For SOXR_COEF_INTERP_AUTO (below).      400   */
+  unsigned num_threads;      /* If built so. 0 means `automatic'.        1    */
+  void * e;                  /* Reserved for internal use.               0    */
+  unsigned long flags;       /* Per the following #defines.              0    */
+};
+                                   /* For `irrational' ratios only: */
+#define SOXR_COEF_INTERP_AUTO  0u    /* Auto select coef. interpolation. */
+#define SOXR_COEF_INTERP_LOW   2u    /* Man. select: less CPU, more memory. */
+#define SOXR_COEF_INTERP_HIGH  3u    /* Man. select: more CPU, less memory. */
+
+#define SOXR_STRICT_BUFFERING  4u  /* Reserved for future use. */
+#define SOXR_NOSMALLINTOPT     8u  /* For test purposes only. */
+
+
+
+/* -------------------------- API type constructors ------------------------- */
+
+/* These functions allow setting of the most commonly-used structure
+ * parameters, with other parameters being given default values.  The default
+ * values may then be overridden, directly in the structure, if needed.  */
+
+SOXR soxr_quality_spec_t soxr_quality_spec(
+    unsigned long recipe,       /* Per the #defines immediately below. */
+    unsigned long flags);       /* As soxr_quality_spec_t.flags. */
+
+                                  /* The 5 standard qualities found in SoX: */
+#define SOXR_QQ                 0   /* 'Quick' cubic interpolation. */
+#define SOXR_LQ                 1   /* 'Low' 16-bit with larger rolloff. */
+#define SOXR_MQ                 2   /* 'Medium' 16-bit with medium rolloff. */
+#define SOXR_HQ                 SOXR_20_BITQ /* 'High quality'. */
+#define SOXR_VHQ                SOXR_28_BITQ /* 'Very high quality'. */
+
+#define SOXR_16_BITQ            3
+#define SOXR_20_BITQ            4
+#define SOXR_24_BITQ            5
+#define SOXR_28_BITQ            6
+#define SOXR_32_BITQ            7
+                                    /* Libsamplerate equivalent qualities: */
+#define SOXR_LSR0Q              8     /* 'Best sinc'. */
+#define SOXR_LSR1Q              9     /* 'Medium sinc'. */
+#define SOXR_LSR2Q              10    /* 'Fast sinc'. */
+
+#define SOXR_LINEAR_PHASE       0x00
+#define SOXR_INTERMEDIATE_PHASE 0x10
+#define SOXR_MINIMUM_PHASE      0x30
+#define SOXR_STEEP_FILTER       0x40
+#define SOXR_ALLOW_ALIASING     0x80  /* Reserved for future use. */
+
+
+
+SOXR soxr_runtime_spec_t soxr_runtime_spec(
+    unsigned num_threads);
+
+
+
+SOXR soxr_io_spec_t soxr_io_spec(
+    soxr_datatype_t itype,
+    soxr_datatype_t otype);
+
+
+
+/* --------------------------- Advanced use only ---------------------------- */
+
+/* For new designs, the following functions/usage will probably not be needed.
+ * They might be useful when adding soxr into an existing design where values
+ * for the resampling-rate and/or number-of-channels parameters to soxr_create
+ * are not available when that function will be called.  In such cases, the
+ * relevant soxr_create parameter(s) can be given as 0, then one or both of the
+ * following (as appropriate) later invoked (but prior to calling soxr_process
+ * or soxr_output):
+ *
+ * soxr_set_error(soxr, soxr_set_io_ratio(soxr, io_ratio, 0));
+ * soxr_set_error(soxr, soxr_set_num_channels(soxr, num_channels));
+ */
+
+SOXR soxr_error_t soxr_set_error(soxr_t, soxr_error_t);
+SOXR soxr_error_t soxr_set_num_channels(soxr_t, unsigned);
+
+
+
+#undef SOXR
+
+#if defined __cplusplus
+}
+#endif
+
+#endif

lib/codecs/inc/vorbis/ivorbiscodec.h

@@ -0,0 +1,203 @@
+/********************************************************************
+ *                                                                  *
+ * THIS FILE IS PART OF THE OggVorbis 'TREMOR' CODEC SOURCE CODE.   *
+ *                                                                  *
+ * USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS     *
+ * GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
+ * IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING.       *
+ *                                                                  *
+ * THE OggVorbis 'TREMOR' SOURCE CODE IS (C) COPYRIGHT 1994-2002    *
+ * BY THE Xiph.Org FOUNDATION http://www.xiph.org/                  *
+ *                                                                  *
+ ********************************************************************
+
+ function: libvorbis codec headers
+
+ ********************************************************************/
+
+#ifndef _vorbis_codec_h_
+#define _vorbis_codec_h_
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif /* __cplusplus */
+
+#include <ogg/ogg.h>
+
+typedef struct vorbis_info{
+  int version;
+  int channels;
+  long rate;
+
+  /* The below bitrate declarations are *hints*.
+     Combinations of the three values carry the following implications:
+     
+     all three set to the same value: 
+       implies a fixed rate bitstream
+     only nominal set: 
+       implies a VBR stream that averages the nominal bitrate.  No hard 
+       upper/lower limit
+     upper and or lower set: 
+       implies a VBR bitstream that obeys the bitrate limits. nominal 
+       may also be set to give a nominal rate.
+     none set:
+       the coder does not care to speculate.
+  */
+
+  long bitrate_upper;
+  long bitrate_nominal;
+  long bitrate_lower;
+  long bitrate_window;
+
+  void *codec_setup;
+} vorbis_info;
+
+/* vorbis_dsp_state buffers the current vorbis audio
+   analysis/synthesis state.  The DSP state belongs to a specific
+   logical bitstream ****************************************************/
+typedef struct vorbis_dsp_state{
+  int analysisp;
+  vorbis_info *vi;
+
+  ogg_int32_t **pcm;
+  ogg_int32_t **pcmret;
+  int      pcm_storage;
+  int      pcm_current;
+  int      pcm_returned;
+
+  int  preextrapolate;
+  int  eofflag;
+
+  long lW;
+  long W;
+  long nW;
+  long centerW;
+
+  ogg_int64_t granulepos;
+  ogg_int64_t sequence;
+
+  void       *backend_state;
+} vorbis_dsp_state;
+
+typedef struct vorbis_block{
+  /* necessary stream state for linking to the framing abstraction */
+  ogg_int32_t  **pcm;       /* this is a pointer into local storage */ 
+  oggpack_buffer opb;
+  
+  long  lW;
+  long  W;
+  long  nW;
+  int   pcmend;
+  int   mode;
+
+  int         eofflag;
+  ogg_int64_t granulepos;
+  ogg_int64_t sequence;
+  vorbis_dsp_state *vd; /* For read-only access of configuration */
+
+  /* local storage to avoid remallocing; it's up to the mapping to
+     structure it */
+  void               *localstore;
+  long                localtop;
+  long                localalloc;
+  long                totaluse;
+  struct alloc_chain *reap;
+
+} vorbis_block;
+
+/* vorbis_block is a single block of data to be processed as part of
+the analysis/synthesis stream; it belongs to a specific logical
+bitstream, but is independant from other vorbis_blocks belonging to
+that logical bitstream. *************************************************/
+
+struct alloc_chain{
+  void *ptr;
+  struct alloc_chain *next;
+};
+
+/* vorbis_info contains all the setup information specific to the
+   specific compression/decompression mode in progress (eg,
+   psychoacoustic settings, channel setup, options, codebook
+   etc). vorbis_info and substructures are in backends.h.
+*********************************************************************/
+
+/* the comments are not part of vorbis_info so that vorbis_info can be
+   static storage */
+typedef struct vorbis_comment{
+  /* unlimited user comment fields.  libvorbis writes 'libvorbis'
+     whatever vendor is set to in encode */
+  char **user_comments;
+  int   *comment_lengths;
+  int    comments;
+  char  *vendor;
+
+} vorbis_comment;
+
+
+/* libvorbis encodes in two abstraction layers; first we perform DSP
+   and produce a packet (see docs/analysis.txt).  The packet is then
+   coded into a framed OggSquish bitstream by the second layer (see
+   docs/framing.txt).  Decode is the reverse process; we sync/frame
+   the bitstream and extract individual packets, then decode the
+   packet back into PCM audio.
+
+   The extra framing/packetizing is used in streaming formats, such as
+   files.  Over the net (such as with UDP), the framing and
+   packetization aren't necessary as they're provided by the transport
+   and the streaming layer is not used */
+
+/* Vorbis PRIMITIVES: general ***************************************/
+
+extern void     vorbis_info_init(vorbis_info *vi);
+extern void     vorbis_info_clear(vorbis_info *vi);
+extern int      vorbis_info_blocksize(vorbis_info *vi,int zo);
+extern void     vorbis_comment_init(vorbis_comment *vc);
+extern void     vorbis_comment_add(vorbis_comment *vc, char *comment); 
+extern void     vorbis_comment_add_tag(vorbis_comment *vc, 
+				       char *tag, char *contents);
+extern char    *vorbis_comment_query(vorbis_comment *vc, char *tag, int count);
+extern int      vorbis_comment_query_count(vorbis_comment *vc, char *tag);
+extern void     vorbis_comment_clear(vorbis_comment *vc);
+
+extern int      vorbis_block_init(vorbis_dsp_state *v, vorbis_block *vb);
+extern int      vorbis_block_clear(vorbis_block *vb);
+extern void     vorbis_dsp_clear(vorbis_dsp_state *v);
+
+/* Vorbis PRIMITIVES: synthesis layer *******************************/
+extern int      vorbis_synthesis_idheader(ogg_packet *op);
+extern int      vorbis_synthesis_headerin(vorbis_info *vi,vorbis_comment *vc,
+					  ogg_packet *op);
+
+extern int      vorbis_synthesis_init(vorbis_dsp_state *v,vorbis_info *vi);
+extern int      vorbis_synthesis_restart(vorbis_dsp_state *v);
+extern int      vorbis_synthesis(vorbis_block *vb,ogg_packet *op);
+extern int      vorbis_synthesis_trackonly(vorbis_block *vb,ogg_packet *op);
+extern int      vorbis_synthesis_blockin(vorbis_dsp_state *v,vorbis_block *vb);
+extern int      vorbis_synthesis_pcmout(vorbis_dsp_state *v,ogg_int32_t ***pcm);
+extern int      vorbis_synthesis_read(vorbis_dsp_state *v,int samples);
+extern long     vorbis_packet_blocksize(vorbis_info *vi,ogg_packet *op);
+
+/* Vorbis ERRORS and return codes ***********************************/
+
+#define OV_FALSE      -1  
+#define OV_EOF        -2
+#define OV_HOLE       -3
+
+#define OV_EREAD      -128
+#define OV_EFAULT     -129
+#define OV_EIMPL      -130
+#define OV_EINVAL     -131
+#define OV_ENOTVORBIS -132
+#define OV_EBADHEADER -133
+#define OV_EVERSION   -134
+#define OV_ENOTAUDIO  -135
+#define OV_EBADPACKET -136
+#define OV_EBADLINK   -137
+#define OV_ENOSEEK    -138
+
+#ifdef __cplusplus
+}
+#endif /* __cplusplus */
+
+#endif

lib/codecs/inc/vorbis/ivorbisfile.h

@@ -0,0 +1,130 @@
+/********************************************************************
+ *                                                                  *
+ * THIS FILE IS PART OF THE OggVorbis 'TREMOR' CODEC SOURCE CODE.   *
+ *                                                                  *
+ * USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS     *
+ * GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
+ * IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING.       *
+ *                                                                  *
+ * THE OggVorbis 'TREMOR' SOURCE CODE IS (C) COPYRIGHT 1994-2002    *
+ * BY THE Xiph.Org FOUNDATION http://www.xiph.org/                  *
+ *                                                                  *
+ ********************************************************************
+
+ function: stdio-based convenience library for opening/seeking/decoding
+
+ ********************************************************************/
+
+#ifndef _OV_FILE_H_
+#define _OV_FILE_H_
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif /* __cplusplus */
+
+#include <stdio.h>
+#include "ivorbiscodec.h"
+
+#define CHUNKSIZE 65535
+#define READSIZE  1024
+/* The function prototypes for the callbacks are basically the same as for
+ * the stdio functions fread, fseek, fclose, ftell. 
+ * The one difference is that the FILE * arguments have been replaced with
+ * a void * - this is to be used as a pointer to whatever internal data these
+ * functions might need. In the stdio case, it's just a FILE * cast to a void *
+ * 
+ * If you use other functions, check the docs for these functions and return
+ * the right values. For seek_func(), you *MUST* return -1 if the stream is
+ * unseekable
+ */
+typedef struct {
+  size_t (*read_func)  (void *ptr, size_t size, size_t nmemb, void *datasource);
+  int    (*seek_func)  (void *datasource, ogg_int64_t offset, int whence);
+  int    (*close_func) (void *datasource);
+  long   (*tell_func)  (void *datasource);
+} ov_callbacks;
+
+#define  NOTOPEN   0
+#define  PARTOPEN  1
+#define  OPENED    2
+#define  STREAMSET 3
+#define  INITSET   4
+
+typedef struct OggVorbis_File {
+  void            *datasource; /* Pointer to a FILE *, etc. */
+  int              seekable;
+  ogg_int64_t      offset;
+  ogg_int64_t      end;
+  ogg_sync_state   oy;
+
+  /* If the FILE handle isn't seekable (eg, a pipe), only the current
+     stream appears */
+  int              links;
+  ogg_int64_t     *offsets;
+  ogg_int64_t     *dataoffsets;
+  ogg_uint32_t    *serialnos;
+  ogg_int64_t     *pcmlengths;
+  vorbis_info     *vi;
+  vorbis_comment  *vc;
+
+  /* Decoding working state local storage */
+  ogg_int64_t      pcm_offset;
+  int              ready_state;
+  ogg_uint32_t     current_serialno;
+  int              current_link;
+
+  ogg_int64_t      bittrack;
+  ogg_int64_t      samptrack;
+
+  ogg_stream_state os; /* take physical pages, weld into a logical
+                          stream of packets */
+  vorbis_dsp_state vd; /* central working state for the packet->PCM decoder */
+  vorbis_block     vb; /* local working space for packet->PCM decode */
+
+  ov_callbacks callbacks;
+
+} OggVorbis_File;
+
+extern int ov_clear(OggVorbis_File *vf);
+extern int ov_open(FILE *f,OggVorbis_File *vf,const char *initial,long ibytes);
+extern int ov_open_callbacks(void *datasource, OggVorbis_File *vf,
+		const char *initial, long ibytes, ov_callbacks callbacks);
+
+extern int ov_test(FILE *f,OggVorbis_File *vf,const char *initial,long ibytes);
+extern int ov_test_callbacks(void *datasource, OggVorbis_File *vf,
+		const char *initial, long ibytes, ov_callbacks callbacks);
+extern int ov_test_open(OggVorbis_File *vf);
+
+extern long ov_bitrate(OggVorbis_File *vf,int i);
+extern long ov_bitrate_instant(OggVorbis_File *vf);
+extern long ov_streams(OggVorbis_File *vf);
+extern long ov_seekable(OggVorbis_File *vf);
+extern long ov_serialnumber(OggVorbis_File *vf,int i);
+
+extern ogg_int64_t ov_raw_total(OggVorbis_File *vf,int i);
+extern ogg_int64_t ov_pcm_total(OggVorbis_File *vf,int i);
+extern ogg_int64_t ov_time_total(OggVorbis_File *vf,int i);
+
+extern int ov_raw_seek(OggVorbis_File *vf,ogg_int64_t pos);
+extern int ov_pcm_seek(OggVorbis_File *vf,ogg_int64_t pos);
+extern int ov_pcm_seek_page(OggVorbis_File *vf,ogg_int64_t pos);
+extern int ov_time_seek(OggVorbis_File *vf,ogg_int64_t pos);
+extern int ov_time_seek_page(OggVorbis_File *vf,ogg_int64_t pos);
+
+extern ogg_int64_t ov_raw_tell(OggVorbis_File *vf);
+extern ogg_int64_t ov_pcm_tell(OggVorbis_File *vf);
+extern ogg_int64_t ov_time_tell(OggVorbis_File *vf);
+
+extern vorbis_info *ov_info(OggVorbis_File *vf,int link);
+extern vorbis_comment *ov_comment(OggVorbis_File *vf,int link);
+
+extern long ov_read(OggVorbis_File *vf,char *buffer,int length,
+		    int *bitstream);
+
+#ifdef __cplusplus
+}
+#endif /* __cplusplus */
+
+#endif
+

lib/display/CMakeLists.txt

@@ -0,0 +1,16 @@
+# the JPEG library is in ROM but seems to fail randomly (PSRAM issue?)
+set(TJPGD tjpgd)
+
+idf_component_register(SRC_DIRS . core core/ifaces  fonts
+						INCLUDE_DIRS . fonts core
+						REQUIRES platform_config tools esp_common
+						PRIV_REQUIRES services freertos driver ${TJPGD}
+						EMBED_FILES note.jpg )		
+
+if (NOT TJPGD)
+add_compile_definitions(TJPGD_ROM)												
+endif()
+
+set_source_files_properties(display.c
+    PROPERTIES COMPILE_FLAGS
+   -Wno-format-overflow )

lib/display/ILI9341.c

@@ -0,0 +1,347 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ *				 2021 Mumpf and Harry1999
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+#define PAGE_BLOCK		2048
+#define ENABLE_WRITE	0x2c
+#define MADCTL_MX  0x40
+#define TFT_RGB_BGR  0x08
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "ILI9341";
+
+enum { ILI9341, ILI9341_24 };	//ILI9341_24 for future use...
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	struct {
+		uint16_t Height, Width;
+	} Offset;
+	uint8_t MADCtl, PageSize;
+	uint8_t Model;
+};
+
+// Functions are not declared to minimize # of lines
+
+static void WriteByte( struct GDS_Device* Device, uint8_t Data ) {
+	Device->WriteData( Device, &Data, 1 );
+}
+
+static void SetColumnAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2A );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void SetRowAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2B );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void Update16( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+#ifdef SHADOW_BUFFER
+	uint32_t *optr = (uint32_t*) Private->Shadowbuffer, *iptr = (uint32_t*) Device->Framebuffer;
+	int FirstCol = Device->Width / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+	
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width is always a multiple of 2)
+		for (int c = 0; c < Device->Width / 2; c++, iptr++, optr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+
+		// wait for a large enough window - careful that window size might increase by more than a line at once !
+		if (FirstRow < 0 || ((LastCol - FirstCol + 1) * (r - FirstRow + 1) * 4 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol *= 2;
+		LastCol = LastCol * 2 + 1;
+		SetRowAddress( Device, FirstRow + Private->Offset.Height, LastRow + Private->Offset.Height);
+		SetColumnAddress( Device, FirstCol + Private->Offset.Width, LastCol + Private->Offset.Width );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 2;
+			
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM <= (PAGE_BLOCK - ChunkSize) && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize );
+		}	
+
+		FirstCol = Device->Width / 2; LastCol = 0;
+		FirstRow = -1;
+	}	
+#else
+	// always update by full lines
+	SetColumnAddress( Device, Private->Offset.Width, Device->Width - 1);
+	
+	for (int r = 0; r < Device->Height; r += min(Private->PageSize, Device->Height - r)) {
+		int Height = min(Private->PageSize, Device->Height - r);
+		
+		SetRowAddress( Device, Private->Offset.Height + r, Private->Offset.Height + r + Height - 1 );
+		Device->WriteCommand(Device, ENABLE_WRITE);
+		
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width * 2, Height * Device->Width * 2 );
+			Device->WriteData( Device, Private->iRAM, Height * Device->Width * 2 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width * 2, Height * Device->Width * 2 );
+		}	
+	}	
+#endif	
+}
+
+static void Update24( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+#ifdef SHADOW_BUFFER
+	uint16_t *optr = (uint16_t*) Private->Shadowbuffer, *iptr = (uint16_t*) Device->Framebuffer;
+	int FirstCol = (Device->Width * 3) / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width always / by 2)
+		for (int c = 0; c < (Device->Width * 3) / 2; c++, optr++, iptr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+
+		// do we have enough to send (cols are divided by 3/2)
+		if (FirstRow < 0 || ((((LastCol - FirstCol + 1) * 2 ) / 3) * (r - FirstRow + 1) * 4 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol = (FirstCol * 2) / 3;
+		LastCol = (LastCol * 2 + 1 ) / 3; 
+		SetRowAddress( Device, FirstRow + Private->Offset.Height, LastRow + Private->Offset.Height);
+		SetColumnAddress( Device, FirstCol + Private->Offset.Width, LastCol + Private->Offset.Width );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 3;
+					
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 3, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM <= (PAGE_BLOCK - ChunkSize) && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}	
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 3, ChunkSize );
+		}	
+
+		FirstCol = (Device->Width * 3) / 2; LastCol = 0;
+		FirstRow = -1;
+	}	
+#else
+	// always update by full lines
+	SetColumnAddress( Device, Private->Offset.Width, Device->Width - 1);
+	
+	for (int r = 0; r < Device->Height; r += min(Private->PageSize, Device->Height - r)) {
+		int Height = min(Private->PageSize, Device->Height - r);
+		
+		SetRowAddress( Device, Private->Offset.Height + r, Private->Offset.Height + r + Height - 1 );
+		Device->WriteCommand(Device, ENABLE_WRITE);
+		
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width * 3, Height * Device->Width * 3 );
+			Device->WriteData( Device, Private->iRAM, Height * Device->Width * 3 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width * 3, Height * Device->Width * 3 );
+		}	
+	}	
+#endif	
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	ESP_LOGI(TAG, "SetLayout 197 HFlip=%d VFlip=%d Rotate=%d (1=true)", Layout->HFlip, Layout->VFlip, Layout->Rotate);
+	//        D/CX RDX WRX D17-8 D7 D6 D5 D4 D3  D2 D1 D0 HEX
+	//Command   0   1   ↑    XX  0  0  1  1  0   1  1  0  36h
+	//Parameter 1   1   ↑    XX  MY MX MV ML BGR MH 0  0  00
+	//Orientation 0: MADCtl = 0x80  =   1000 0000 (MY=1)
+	if ((Device->Height)>(Device->Width)){		//Resolution = 320x240
+		Private->MADCtl = (1 << 7);				// 0x80 = default (no Rotation an no Flip)
+		if (Layout->HFlip) {							//Flip Horizontal
+			int a = Private->MADCtl;
+			Private->MADCtl = (a ^ (1 << 7));
+		}
+		if (Layout->Rotate) {							//Rotate 180 degr.
+			int a = Private->MADCtl;
+			a = (a ^ (1 << 7));
+			Private->MADCtl = (a ^ (1 << 6));
+		}
+		if (Layout->VFlip) {							//Flip Vertical
+			int a = Private->MADCtl;
+			Private->MADCtl = (a ^ (1 << 6));
+		}
+	} else {									//Resolution = 240x320
+		Private->MADCtl = (1 << 5);				// 0x20 = default (no Rotation an no Flip)
+		if (Layout->HFlip) {							//Flip Horizontal
+			int a = Private->MADCtl;
+			Private->MADCtl = (a ^ (1 << 6));
+		}
+		if (Layout->Rotate) {							//Rotate 180 degr.
+			int a = Private->MADCtl;
+			a = (a ^ (1 << 7));
+			Private->MADCtl = (a ^ (1 << 6));
+		}
+		if (Layout->VFlip) {							//Flip Vertical
+			int a = Private->MADCtl;
+			Private->MADCtl = (a ^ (1 << 7));
+		}
+	}
+	
+	Private->MADCtl = Layout->ColorSwap ? (Private->MADCtl | (1 << 3)) : (Private->MADCtl & ~(1 << 3));
+
+	ESP_LOGI(TAG, "SetLayout 255 Private->MADCtl=%hhu", Private->MADCtl);
+
+	Device->WriteCommand( Device, 0x36 );
+	WriteByte( Device, Private->MADCtl );
+	
+	Device->WriteCommand( Device, Layout->Invert ? 0x21 : 0x20 );
+	
+
+#ifdef SHADOW_BUFFER
+	// force a full refresh (almost ...)
+	memset(Private->Shadowbuffer, 0xAA, Device->FramebufferSize);
+#endif	
+}	
+
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0x29 ); }	//DISPON =0x29
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0x28 ); }	//DISPOFF=0x28
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+	Device->WriteCommand( Device, 0x51 );
+	WriteByte( Device, Contrast );
+	
+	Device->SetContrast = NULL;
+	GDS_SetContrast( Device, Contrast );
+	Device->SetContrast = SetContrast;	// 0x00 value means the lowest brightness and 0xFF value means the highest brightness.
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	int Depth = (Device->Depth + 8 - 1) / 8;
+	
+	Private->PageSize = min(8, PAGE_BLOCK / (Device->Width * Depth));
+
+#ifdef SHADOW_BUFFER	
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#endif
+#ifdef USE_IRAM
+	Private->iRAM = heap_caps_malloc( (Private->PageSize + 1) * Device->Width * Depth, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+#endif
+
+	ESP_LOGI(TAG, "ILI9341 with bit default-depth %u, page %u, iRAM %p", Device->Depth, Private->PageSize, Private->iRAM);
+	
+	// Sleepout + Booster
+	Device->WriteCommand( Device, 0x11 );
+			
+	// set flip modes & contrast
+	GDS_SetContrast( Device, 0x7f );
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+	
+	// set screen depth (16/18) *** INTERFACE PIXEL FORMAT: 0x66=18 bit; 0x55=16 bit
+	Device->WriteCommand( Device, 0x3A );
+	if (Private->Model == ILI9341_24) WriteByte( Device, Device->Depth == 24 ? 0x66 : 0x55 );
+	else WriteByte( Device, Device->Depth == 24 ? 0x66 : 0x55 );
+
+	ESP_LOGI(TAG, "ILI9341_Init 312 device-depth %u, 0x66/0x55=0x%X", Device->Depth, Device->Depth == 24 ? 0x66 : 0x55);
+
+	// no Display Inversion (INVOFF=0x20 INVON=0x21)
+	Device->WriteCommand( Device, 0x20 );
+
+	//Gamma Correction: Enable next two line and enabel one of the Test0x Section... or build you own 15 Parameter...
+	Device->WriteCommand( Device, 0xF2 ); WriteByte( Device, 0x03 );	// 3Gamma Function: Disable = default (0x02), Enable (0x03)
+	Device->WriteCommand( Device, 0x26 ); WriteByte( Device, 0x01 );	// Gamma curve selected (0x01, 0x02, 0x04, 0x08) - A maximum of 4 fixed gamma curves can be selected
+	//Gamma Correction Test01
+	Device->WriteCommand( Device, 0xE0 );								// Positive Gamma Correction (15 Parameter)
+	WriteByte( Device, 0x0F ); WriteByte( Device, 0x31 ); WriteByte( Device, 0x2B ); WriteByte( Device, 0x0C ); WriteByte( Device, 0x0E );
+	WriteByte( Device, 0x08 ); WriteByte( Device, 0x4E ); WriteByte( Device, 0xF1 ); WriteByte( Device, 0x37 ); WriteByte( Device, 0x07 );
+	WriteByte( Device, 0x10 ); WriteByte( Device, 0x03 ); WriteByte( Device, 0x0E ); WriteByte( Device, 0x09 ); WriteByte( Device, 0x00 );
+	Device->WriteCommand( Device, 0xE1 ); 								// Negative Gamma Correction (15 Parameter)
+	WriteByte( Device, 0x00 ); WriteByte( Device, 0x0E ); WriteByte( Device, 0x14 ); WriteByte( Device, 0x03 ); WriteByte( Device, 0x11 );
+	WriteByte( Device, 0x07 ); WriteByte( Device, 0x31 ); WriteByte( Device, 0xC1 ); WriteByte( Device, 0x48 ); WriteByte( Device, 0x08 );
+	WriteByte( Device, 0x0F ); WriteByte( Device, 0x0C ); WriteByte( Device, 0x31 ); WriteByte( Device, 0x36 ); WriteByte( Device, 0x0F );
+	 
+	// gone with the wind
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+
+	return true;
+}	
+
+static const struct GDS_Device ILI9341_X = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff,
+	.SetLayout = SetLayout,
+	.Update = Update16, .Init = Init,
+	.Mode = GDS_RGB565, .Depth = 16,
+};		
+
+struct GDS_Device* ILI9341_Detect(char *Driver, struct GDS_Device* Device) {
+	uint8_t Model;
+	int Depth=16;		// 16bit colordepth
+	
+	if (strcasestr(Driver, "ILI9341")) Model = ILI9341;
+	else if (strcasestr(Driver, "ILI9341_24")) Model = ILI9341_24;	//for future use...
+	else return NULL;
+		
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+		
+	*Device = ILI9341_X;	
+	sscanf(Driver, "%*[^:]:%u", &Depth);		// NVS-Parameter driver=ILI9341[:16|18]
+	struct PrivateSpace* Private = (struct PrivateSpace*) Device->Private;
+	Private->Model = Model;
+		ESP_LOGI(TAG, "ILI9341_Detect 391 Driver= %s   Depth=%d", Driver, Depth);
+
+	if (Depth == 18) {
+		Device->Mode = GDS_RGB888;
+		Device->Depth = 24;
+		Device->Update = Update24;
+	} 	
+	
+	if (Model == ILI9341_24) Device->SetContrast = SetContrast;
+
+	return Device;
+}

lib/display/ILI9488.c

@@ -0,0 +1,297 @@
+/**
+ * ILI9488 Display Driver for Guition JC4827W543C
+ * Supports QSPI interface for 480x272 resolution
+ * Based on ILI9341 driver adapted for ILI9488
+ * 
+ * (c) Guition Support 2026
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+#define PAGE_BLOCK		4096
+#define ENABLE_WRITE	0x2c
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "ILI9488";
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	struct {
+		uint16_t Height, Width;
+	} Offset;
+	uint8_t MADCtl, PageSize;
+	uint8_t Model;
+};
+
+// ILI9488 Commands
+static const uint8_t ILI9488_INIT_SEQUENCE[] = {
+	// Software reset
+	0x01, 0x80, 150,	// SWRESET, delay 150ms
+	// Power control
+	0xD0, 3, 0x07, 0x42, 0x18,	// Power Control
+	0xD1, 3, 0x00, 0x07, 0x10,	// VCOM Control
+	0xD2, 1, 0x01,				// Power Control for Normal Mode
+	0xC0, 2, 0x10, 0x3B,		// Panel Driving Setting
+	0xC1, 1, 0x10,				// Frame Rate Control
+	0xC5, 5, 0x0A, 0x3A, 0x28, 0x28, 0x02, // MCU Control
+	0xC6, 1, 0x00,				// Frame Rate Control
+	0xB1, 2, 0x00, 0x1B,		// Display Function Control
+	0xB4, 1, 0x02,				// Inversion Control
+	0xB6, 3, 0x02, 0x02, 0x3B,	// Display Function Control
+	0xB7, 1, 0xC6,				// Entry Mode Set
+	0xE0, 16, 0x00, 0x07, 0x10, 0x0E, 0x09, 0x16, 0x06, 0x0A,
+				0x0E, 0x09, 0x15, 0x0D, 0x0E, 0x11, 0x0F, 0x12, // Positive Gamma Control
+	0xE1, 16, 0x00, 0x17, 0x1A, 0x04, 0x0E, 0x06, 0x2F, 0x24,
+				0x1B, 0x1B, 0x22, 0x1F, 0x1E, 0x37, 0x3F, 0x00, // Negative Gamma Control
+	0x36, 1, 0xE8,				// Memory Access Control (MX, MY, RGB mode)
+	0x3A, 1, 0x55,				// Interface Pixel Format (16bpp)
+	0x11, 0x80, 150,			// Sleep Out, delay 150ms
+	0x29, 0x80, 50,				// Display On, delay 50ms
+	0xFF, 0x00					// End of sequence
+};
+
+static void WriteByte( struct GDS_Device* Device, uint8_t Data ) {
+	Device->WriteData( Device, &Data, 1 );
+}
+
+static void SetColumnAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2A );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void SetRowAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2B );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void Update16( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+#ifdef SHADOW_BUFFER
+	uint32_t *optr = (uint32_t*) Private->Shadowbuffer, *iptr = (uint32_t*) Device->Framebuffer;
+	int FirstCol = Device->Width / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+	
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width is always a multiple of 2)
+		for (int c = 0; c < Device->Width / 2; c++, iptr++, optr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+
+		// wait for a large enough window - careful that window size might increase by more than a line at once !
+		if (FirstRow < 0 || ((LastCol - FirstCol + 1) * (r - FirstRow + 1) * 4 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol *= 2;
+		LastCol = LastCol * 2 + 1;
+		SetRowAddress( Device, FirstRow + Private->Offset.Height, LastRow + Private->Offset.Height);
+		SetColumnAddress( Device, FirstCol + Private->Offset.Width, LastCol + Private->Offset.Width );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 2;
+			
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM <= (PAGE_BLOCK - ChunkSize) && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize );
+		}
+		
+		FirstCol = Device->Width / 2;
+		LastCol = 0;
+		FirstRow = -1;
+	}
+#endif
+}
+
+static void Clear( struct GDS_Device* Device ) {
+	memset( Device->Framebuffer, 0, Device->Width * Device->Height * 2 );
+	Device->Update( Device);
+}
+
+static void SetPixel( struct GDS_Device* Device, int X, int Y, uint32_t Color ) {
+	if( X < 0 || X >= Device->Width || Y < 0 || Y >= Device->Height) return;
+	
+	*((uint16_t*) Device->Framebuffer + (Y * Device->Width) + X) = (uint16_t) Color;
+}
+
+static uint32_t GetPixel( struct GDS_Device* Device, int X, int Y ) {
+	if( X < 0 || X >= Device->Width || Y < 0 || Y >= Device->Height) return 0;
+	
+	return *((uint16_t*) Device->Framebuffer + (Y * Device->Width) + X);
+}
+
+static void DrawPixel( struct GDS_Device* Device, int X, int Y, uint32_t Color ) {
+	SetPixel( Device, X, Y, Color );
+}
+
+static void DrawPixelFast( struct GDS_Device* Device, int X, int Y, uint32_t Color ) {
+	*((uint16_t*) Device->Framebuffer + (Y * Device->Width) + X) = (uint16_t) Color;
+}
+
+static void DrawCBR( struct GDS_Device* Device, int X, int Y, int Width, int Height, uint32_t Color ) {
+	uint16_t *fb = (uint16_t*) Device->Framebuffer + Y * Device->Width + X;
+	
+	for( int y = 0; y < Height; y++) {
+		for( int x = 0; x < Width; x++) {
+			fb[x] = (uint16_t) Color;
+		}
+		fb += Device->Width;
+	}
+}
+
+static void DrawHLine( struct GDS_Device* Device, int X0, int X1, int Y, uint32_t Color ) {
+	if( Y < 0 || Y >= Device->Height) return;
+	
+	if( X0 > X1) {
+		int Temp = X0;
+		X0 = X1;
+		X1 = Temp;
+	}
+	
+	if( X0 < 0) X0 = 0;
+	if( X1 >= Device->Width) X1 = Device->Width - 1;
+	
+	uint16_t *fb = (uint16_t*) Device->Framebuffer + Y * Device->Width + X0;
+	
+	for( int x = X0; x <= X1; x++) {
+		*fb++ = (uint16_t) Color;
+	}
+}
+
+static void DrawVLine( struct GDS_Device* Device, int X, int Y0, int Y1, uint32_t Color ) {
+	if( X < 0 || X >= Device->Width) return;
+	
+	if( Y0 > Y1) {
+		int Temp = Y0;
+		Y0 = Y1;
+		Y1 = Temp;
+	}
+	
+	if( Y0 < 0) Y0 = 0;
+	if( Y1 >= Device->Height) Y1 = Device->Height - 1;
+	
+	uint16_t *fb = (uint16_t*) Device->Framebuffer + Y0 * Device->Width + X;
+	
+	for( int y = Y0; y <= Y1; y++) {
+		*fb = (uint16_t) Color;
+		fb += Device->Width;
+	}
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	const uint8_t *p = ILI9488_INIT_SEQUENCE;
+	
+	ESP_LOGI(TAG, "Initializing ILI9488 display %dx%d", Device->Width, Device->Height);
+	
+	// Allocate IRAM buffer if available
+	Private->iRAM = (uint8_t*) heap_caps_malloc(PAGE_BLOCK, MALLOC_CAP_DMA | MALLOC_CAP_INTERNAL);
+	if (!Private->iRAM) {
+		ESP_LOGW(TAG, "Could not allocate IRAM buffer, using direct write");
+	}
+
+#ifdef SHADOW_BUFFER
+	Private->Shadowbuffer = (uint8_t*) heap_caps_malloc(Device->Width * Device->Height * 2, MALLOC_CAP_DMA);
+	if (!Private->Shadowbuffer) {
+		ESP_LOGE(TAG, "Could not allocate shadow buffer");
+		if (Private->iRAM) free(Private->iRAM);
+		return false;
+	}
+	memset(Private->Shadowbuffer, 0, Device->Width * Device->Height * 2);
+#endif
+
+	// Send initialization sequence
+	while(*p != 0xFF) {
+		uint8_t cmd = *p++;
+		uint8_t len = (*p & 0x7F);
+		bool delay = (*p++ & 0x80);
+		
+		Device->WriteCommand(Device, cmd);
+		if(len) Device->WriteData(Device, (uint8_t*)p, len);
+		p += len;
+		
+		if(delay) {
+			uint8_t ms = *p++;
+			vTaskDelay(ms / portTICK_PERIOD_MS);
+		}
+	}
+
+	// Set orientation and layout
+	Private->MADCtl = 0xE8; // MX=1, MY=1, RGB=1, BGR=0
+	Device->WriteCommand(Device, 0x36);
+	Device->WriteData(Device, &Private->MADCtl, 1);
+
+	// Set pixel format to 16-bit
+	uint8_t fmt = 0x55;
+	Device->WriteCommand(Device, 0x3A);
+	Device->WriteData(Device, &fmt, 1);
+
+	// Turn on display
+	Device->WriteCommand(Device, 0x29);
+	
+	ESP_LOGI(TAG, "ILI9488 initialization complete");
+	return true;
+}
+
+static void Deinit( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	
+	if (Private->iRAM) free(Private->iRAM);
+#ifdef SHADOW_BUFFER
+	if (Private->Shadowbuffer) free(Private->Shadowbuffer);
+#endif
+}
+
+struct GDS_Device* ILI9488_Detect( char *Driver, struct GDS_Device *Device ) {
+	if(strcasecmp(Driver, "ILI9488") != 0) return NULL;
+
+	Device->Private = calloc(1, sizeof(struct PrivateSpace));
+	if(!Device->Private) {
+		ESP_LOGE(TAG, "Cannot allocate private data");
+		return NULL;
+	}
+
+	Device->Mode = GDS_RGB565;
+	Device->Depth = 16;
+	Device->Update = Update16;
+	Device->Clear = Clear;
+	Device->SetPixel = SetPixel;
+	Device->GetPixel = GetPixel;
+	Device->DrawPixel = DrawPixel;
+	Device->DrawPixelFast = DrawPixelFast;
+	Device->DrawCBR = DrawCBR;
+	Device->DrawHLine = DrawHLine;
+	Device->DrawVLine = DrawVLine;
+	Device->Init = Init;
+	Device->Deinit = Deinit;
+
+	ESP_LOGI(TAG, "ILI9488 driver loaded");
+	return Device;
+}

lib/display/SH1106.c

@@ -0,0 +1,164 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G. 
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+
+static char TAG[] = "SH1106";
+
+struct PrivateSpace {
+	uint8_t *Shadowbuffer;
+};
+
+// Functions are not declared to minimize # of lines
+
+static void SetColumnAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	// well, unfortunately this driver is 132 colums but most displays are 128...
+	if (Device->Width != 132) Start += 2;
+	Device->WriteCommand( Device, 0x10 | (Start >> 4) );
+	Device->WriteCommand( Device, 0x00 | (Start & 0x0f) );
+}
+
+static void SetPageAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0xB0 | Start );	
+}	
+
+static void Update( struct GDS_Device* Device ) {
+#ifdef SHADOW_BUFFER
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	// not sure the compiler does not have to redo all calculation in for loops, so local it is
+	int width = Device->Width, rows = Device->Height / 8;
+	uint8_t *optr = Private->Shadowbuffer, *iptr = Device->Framebuffer;
+	
+	// by row, find first and last columns that have been updated
+	for (int r = 0; r < rows; r++) {
+		uint8_t first = 0, last;	
+		for (int c = 0; c < width; c++) {
+			if (*iptr != *optr) {
+				if (!first) first = c + 1;
+				last = c ;
+			}	
+			*optr++ = *iptr++;
+		}
+		
+		// now update the display by "byte rows"
+		if (first--) {
+			SetColumnAddress( Device, first, last );
+			SetPageAddress( Device, r, r);
+			Device->WriteData( Device, Private->Shadowbuffer + r*width + first, last - first + 1);
+		}
+	}	
+#else	
+	// SH1106 requires a page-by-page update and has no end Page/Column
+	for (int i = 0; i < Device->Height / 8 ; i++) {
+		SH1106_SetPageAddress( Device, i, 0);
+		SH1106_SetColumnAddress( Device, 0, 0);			
+		SH1106_WriteData( Device, Device->Framebuffer + i*Device->Width, Device->Width );
+	}	
+#endif	
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+	Device->WriteCommand( Device, Layout->HFlip ? 0xA1 : 0xA0 );
+	Device->WriteCommand( Device, Layout->VFlip ? 0xC8 : 0xC0 );
+    Device->WriteCommand( Device, Layout->Invert ? 0xA7 : 0xA6 );
+}	
+
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAF ); }
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAE ); }
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+    Device->WriteCommand( Device, 0x81 );
+    Device->WriteCommand( Device, Contrast );
+}
+
+static void SPIParams(int Speed, uint8_t *mode, uint16_t *CS_pre, uint8_t *CS_post) {
+	*CS_post = Speed / (8*1000*1000);
+}
+
+static bool Init( struct GDS_Device* Device ) {
+#ifdef SHADOW_BUFFER	
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+#ifdef USE_IRAM
+	if (Device->IF == GDS_IF_SPI) Private->Shadowbuffer = heap_caps_malloc( Device->FramebufferSize, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+	else 
+#endif
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	NullCheck( Private->Shadowbuffer, return false );
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#endif	
+		
+	// need to be off and disable display RAM
+	Device->DisplayOff( Device );
+    Device->WriteCommand( Device, 0xA5 );
+	
+	// charge pump regulator, do direct init
+	Device->WriteCommand( Device, 0xAD );
+	Device->WriteCommand( Device, 0x8B ); 
+			
+	// COM pins HW config (alternative:EN) - some display might need something difference
+	Device->WriteCommand( Device, 0xDA );
+	Device->WriteCommand( Device, 1 << 4);
+		
+	// MUX Ratio
+    Device->WriteCommand( Device, 0xA8 );
+    Device->WriteCommand( Device, Device->Height - 1);
+	// Display Offset
+    Device->WriteCommand( Device, 0xD3 );
+    Device->WriteCommand( Device, 0 );
+	// Display Start Line
+    Device->WriteCommand( Device, 0x40 + 0x00 );
+	Device->SetContrast( Device, 0x7F );
+	// set flip modes
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+	// no Display Inversion
+    Device->WriteCommand( Device, 0xA6 );
+	// set Clocks
+    Device->WriteCommand( Device, 0xD5 );
+    Device->WriteCommand( Device, ( 0x08 << 4 ) | 0x00 );
+		
+	// gone with the wind
+	Device->WriteCommand( Device, 0xA4 );
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+	
+	return true;
+}	
+
+static const struct GDS_Device SH1106 = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff, .SetContrast = SetContrast,
+	.SetLayout = SetLayout,
+	.Update = Update, .Init = Init,
+	.Depth = 1,
+	.SPIParams = SPIParams,
+#if !defined SHADOW_BUFFER && defined USE_IRAM	
+	.Alloc = GDS_ALLOC_IRAM_SPI;
+#endif		
+};	
+
+struct GDS_Device* SH1106_Detect(char *Driver, struct GDS_Device* Device) {
+	if (!strcasestr(Driver, "SH1106")) return NULL;
+	
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+	*Device = SH1106;
+
+	ESP_LOGI(TAG, "SH1106 driver");
+	
+	return Device;
+}

lib/display/SH1122.c

@@ -0,0 +1,176 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define PAGE_BLOCK	1024
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "SH1122";
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	uint8_t PageSize;
+};
+
+// Functions are not declared to minimize # of lines
+
+static void SetColumnAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+    Device->WriteCommand( Device, 0x10 | (Start >> 4) );
+	Device->WriteCommand( Device, 0x00 | (Start & 0x0f) );
+}
+
+static void SetRowAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0xB0 );
+	Device->WriteCommand( Device, Start );
+}
+
+static void Update( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+	// RAM is by columns of 4 pixels ...
+	SetColumnAddress( Device, 0, Device->Width / 4 - 1);
+	
+#ifdef SHADOW_BUFFER
+	uint16_t *optr = (uint16_t*) Private->Shadowbuffer, *iptr = (uint16_t*) Device->Framebuffer;
+	bool dirty = false;
+	
+	for (int r = 0, page = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width always / by 2)
+		for (int c = Device->Width / 2 / 2; --c >= 0;) {
+			if (*optr != *iptr) {
+				dirty = true;
+				*optr = *iptr;
+			}
+			iptr++; optr++;
+		}
+		
+		// one line done, check for page boundary
+		if (++page == Private->PageSize) {
+			if (dirty) {
+				SetRowAddress( Device, r - page + 1, r );
+                if (Private->iRAM) {
+                    memcpy(Private->iRAM, Private->Shadowbuffer + (r - page + 1) * Device->Width / 2, page * Device->Width / 2 );
+                    Device->WriteData( Device, Private->iRAM, Device->Width * page / 2 );
+                } else {
+                    Device->WriteData( Device, Private->Shadowbuffer + (r - page + 1) * Device->Width / 2, page * Device->Width / 2);
+                }    
+				dirty = false;
+			}	
+			page = 0;
+		}	
+	}	
+#else
+    SetRowAddress( Device, 0, Device->Height - 1 );
+	for (int r = 0; r < Device->Height; r += Private->PageSize) {
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width / 2, Private->PageSize * Device->Width / 2 );
+			Device->WriteData( Device, Private->iRAM, Private->PageSize * Device->Width / 2 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width / 2, Private->PageSize * Device->Width / 2 );
+		}	
+	}	
+#endif	
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+    if (Layout->HFlip) {
+        Device->WriteCommand( Device, 0x40 + 0x20 );
+        Device->WriteCommand( Device, 0xA1 );
+    } else {
+        Device->WriteCommand( Device, 0x40 + 0x00 );
+        Device->WriteCommand( Device, 0xA0 );
+    }
+	Device->WriteCommand( Device, Layout->VFlip ? 0xC8 : 0xC0 );
+    Device->WriteCommand( Device, Layout->Invert ? 0xA7 : 0xA6 );
+}	
+
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAF ); }
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAE ); }
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+    Device->WriteCommand( Device, 0x81 );
+    Device->WriteCommand( Device, Contrast );
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	
+	// find a page size that is not too small is an integer of height
+	Private->PageSize = min(8, PAGE_BLOCK / (Device->Width / 2));
+	while (Private->PageSize && Device->Height != (Device->Height / Private->PageSize) * Private->PageSize) Private->PageSize--;
+	
+#ifdef SHADOW_BUFFER	
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#endif
+
+    // only use iRAM for SPI
+    if (Device->IF == GDS_IF_SPI) {
+        Private->iRAM = heap_caps_malloc( Private->PageSize * Device->Width / 2, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+    }
+
+	ESP_LOGI(TAG, "SH1122 page %u, iRAM %p", Private->PageSize, Private->iRAM);
+			
+	// need to be off and disable display RAM
+	Device->DisplayOff( Device );
+    Device->WriteCommand( Device, 0xA5 );
+	
+	// Display Offset
+    Device->WriteCommand( Device, 0xD3 );
+    Device->WriteCommand( Device, 0 );
+    
+ 	// set flip modes
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+    	
+	// set Clocks => check value
+    Device->WriteCommand( Device, 0xD5 );
+    Device->WriteCommand( Device, ( 0x04 << 4 ) | 0x00 );
+	
+	// MUX Ratio => fixed
+    Device->WriteCommand( Device, 0xA8 );
+    Device->WriteCommand( Device, Device->Height - 1);
+			
+	// no Display Inversion
+    Device->WriteCommand( Device, 0xA6 );
+	
+	// gone with the wind
+	Device->WriteCommand( Device, 0xA4 );
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+	
+	return true;
+}	
+
+static const struct GDS_Device SH1122 = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff, .SetContrast = SetContrast,
+	.SetLayout = SetLayout,
+	.Update = Update, .Init = Init,
+	.Mode = GDS_GRAYSCALE, .Depth = 4,
+    .HighNibble = true,
+};	
+
+struct GDS_Device* SH1122_Detect(char *Driver, struct GDS_Device* Device) {
+	if (!strcasestr(Driver, "SH1122")) return NULL;
+		
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+	*Device = SH1122;
+
+	return Device;
+}

lib/display/SSD1306.c

@@ -0,0 +1,174 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+
+static char TAG[] = "SSD1306";
+
+struct PrivateSpace {
+	uint8_t *Shadowbuffer;
+};
+
+// Functions are not declared to minimize # of lines
+
+static void SetColumnAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0x21 );
+	Device->WriteCommand( Device, Start );
+	Device->WriteCommand( Device, End );
+}
+static void SetPageAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0x22 );	
+	Device->WriteCommand( Device, Start );
+	Device->WriteCommand( Device, End );
+}
+
+static void Update( struct GDS_Device* Device ) {
+#ifdef SHADOW_BUFFER
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	// not sure the compiler does not have to redo all calculation in for loops, so local it is
+	int width = Device->Width, pages = Device->Height / 8;
+	uint8_t *optr = Private->Shadowbuffer, *iptr = Device->Framebuffer;
+	int CurrentPage = -1, FirstCol = -1, LastCol = -1;
+	
+	// by row, find first and last columns that have been updated
+	for (int p = 0; p < pages; p++) {
+		uint8_t first = 0, last;	
+		for (int c = 0; c < width; c++) {
+			if (*iptr != *optr) {
+				if (!first) first = c + 1;
+				last = c ;
+			}	
+			*optr++ = *iptr++;
+		}
+		
+		// now update the display by "byte rows"
+		if (first--) {
+			// only set column when useful, saves a fair bit of CPU
+			if (first > FirstCol && first <= FirstCol + 4 && last < LastCol && last >= LastCol - 4) {
+				first = FirstCol;
+				last = LastCol;
+			} else {	
+				SetColumnAddress( Device, first, last );
+				FirstCol = first;
+				LastCol = last;
+			}
+			
+			// Set row only when needed, otherwise let auto-increment work
+			if (p != CurrentPage) SetPageAddress( Device, p, Device->Height / 8 - 1 );
+			CurrentPage = p + 1;
+			
+			// actual write
+			Device->WriteData( Device, Private->Shadowbuffer + p*width + first, last - first + 1);
+		}
+	}	
+#else	
+	// automatic counter and end Page/Column (we assume Height % 8 == 0)
+	SetColumnAddress( Device, 0, Device->Width - 1);
+	SetPageAddress( Device, 0, Device->Height / 8 - 1);
+	Device->WriteData( Device, Device->Framebuffer, Device->FramebufferSize );
+#endif	
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+	Device->WriteCommand( Device, Layout->HFlip ? 0xA1 : 0xA0 );
+	Device->WriteCommand( Device, Layout->VFlip ? 0xC8 : 0xC0 );
+	Device->WriteCommand( Device, Layout->Invert ? 0xA7 : 0xA6 );
+}
+	
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAF ); }
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAE ); }
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+    Device->WriteCommand( Device, 0x81 );
+    Device->WriteCommand( Device, Contrast );
+}
+
+static bool Init( struct GDS_Device* Device ) {
+#ifdef SHADOW_BUFFER	
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+#ifdef USE_IRAM
+	if (Device->IF == GDS_IF_SPI) Private->Shadowbuffer = heap_caps_malloc( Device->FramebufferSize, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+	else 
+#endif
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	NullCheck( Private->Shadowbuffer, return false );
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#endif	
+		
+	// need to be off and disable display RAM
+	Device->DisplayOff( Device );
+    Device->WriteCommand( Device, 0xA5 );
+	
+	// charge pump regulator, do direct init
+	Device->WriteCommand( Device, 0x8D );
+	Device->WriteCommand( Device, 0x14 ); 
+			
+	// COM pins HW config (alternative:EN if 64, DIS if 32, remap:DIS) - some display might need something different
+	Device->WriteCommand( Device, 0xDA );
+	Device->WriteCommand( Device, ((Device->Height == 64 ? 1 : 0) << 4) | (0 < 5) );
+		
+	// MUX Ratio
+    Device->WriteCommand( Device, 0xA8 );
+    Device->WriteCommand( Device, Device->Height - 1);
+	// Display Offset
+    Device->WriteCommand( Device, 0xD3 );
+    Device->WriteCommand( Device, 0 );
+	// Display Start Line
+    Device->WriteCommand( Device, 0x40 + 0x00 );
+	Device->SetContrast( Device, 0x7F );
+	// set flip modes
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+	// no Display Inversion
+    Device->WriteCommand( Device, 0xA6 );
+	// set Clocks
+    Device->WriteCommand( Device, 0xD5 );
+    Device->WriteCommand( Device, ( 0x08 << 4 ) | 0x00 );
+	// set Adressing Mode Horizontal
+	Device->WriteCommand( Device, 0x20 );
+	Device->WriteCommand( Device, 0 );
+	
+	// gone with the wind
+	Device->WriteCommand( Device, 0xA4 );
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+	
+	return true;
+}	
+
+static const struct GDS_Device SSD1306 = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff, .SetContrast = SetContrast,
+	.SetLayout = SetLayout,
+	.Update = Update, .Init = Init,
+	.Mode = GDS_MONO, .Depth = 1,
+#if !defined SHADOW_BUFFER && defined USE_IRAM	
+	.Alloc = GDS_ALLOC_IRAM_SPI,
+#endif		
+};	
+
+struct GDS_Device* SSD1306_Detect(char *Driver, struct GDS_Device* Device) {
+	if (!strcasestr(Driver, "SSD1306")) return NULL;
+	
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+	*Device = SSD1306;	
+	
+	ESP_LOGI(TAG, "SSD1306 driver");
+	
+	return Device;
+}

lib/display/SSD1322.c

@@ -0,0 +1,198 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define PAGE_BLOCK	1024
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "SSD1322";
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	uint8_t ReMap, PageSize;
+	uint8_t Offset;
+};
+
+// Functions are not declared to minimize # of lines
+
+static void WriteDataByte( struct GDS_Device* Device, uint8_t Data ) {
+	Device->WriteData( Device, &Data, 1);
+}
+
+static void SetColumnAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0x15 );
+	Device->WriteData( Device, &Start, 1 );
+	Device->WriteData( Device, &End, 1 );
+}
+static void SetRowAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0x75 );
+	Device->WriteData( Device, &Start, 1 );
+	Device->WriteData( Device, &End, 1 );
+}
+
+static void Update( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+	// RAM is by columns of 4 pixels ...
+	SetColumnAddress( Device, Private->Offset, Private->Offset + Device->Width / 4 - 1);
+	
+#ifdef SHADOW_BUFFER
+	uint16_t *optr = (uint16_t*) Private->Shadowbuffer, *iptr = (uint16_t*) Device->Framebuffer;
+	bool dirty = false;
+	
+	for (int r = 0, page = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width always / by 2)
+		for (int c = Device->Width / 2 / 2; --c >= 0;) {
+			if (*optr != *iptr) {
+				dirty = true;
+				*optr = *iptr;
+			}
+			iptr++; optr++;
+		}
+		
+		// one line done, check for page boundary
+		if (++page == Private->PageSize) {
+			if (dirty) {
+				uint16_t *optr = (uint16_t*) Private->iRAM, *iptr = (uint16_t*) (Private->Shadowbuffer + (r - page + 1) * Device->Width / 2);
+				SetRowAddress( Device, r - page + 1, r );
+                // need byte swapping
+				for (int i = page * Device->Width / 2 / 2; --i >= 0; iptr++) *optr++ = (*iptr >> 8) | (*iptr << 8);
+				Device->WriteCommand( Device, 0x5c );
+				Device->WriteData( Device, Private->iRAM, Device->Width * page / 2 );
+				dirty = false;
+			}	
+			page = 0;
+		}	
+	}	
+#else
+	for (int r = 0; r < Device->Height; r += Private->PageSize) {
+		SetRowAddress( Device, r, r + Private->PageSize - 1 );
+		Device->WriteCommand( Device, 0x5c );
+        // need byte swapping
+		uint16_t *optr = (uint16_t*) Private->iRAM, *iptr = (uint16_t*) (Device->Framebuffer + r * Device->Width / 2);
+		for (int i = Private->PageSize * Device->Width / 2 / 2; --i >= 0; iptr++) *optr++ = (*iptr >> 8) | (*iptr << 8);
+		Device->WriteData( Device, Private->iRAM, Private->PageSize * Device->Width / 2 );
+	}	
+#endif	
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	Private->ReMap = Layout->HFlip ? (Private->ReMap & ~(1 << 1)) : (Private->ReMap | (1 << 1));
+	Private->ReMap = Layout->VFlip ? (Private->ReMap | (1 << 4)) : (Private->ReMap & ~(1 << 4));
+	Device->WriteCommand( Device, 0xA0 );
+	Device->WriteData( Device, &Private->ReMap, 1 );
+	WriteDataByte( Device, 0x11 );		
+	Device->WriteCommand( Device, Layout->Invert ? 0xA7 : 0xA6 );
+}	
+
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAF ); }
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAE ); }
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+    Device->WriteCommand( Device, 0xC1 );
+    Device->WriteData( Device, &Contrast, 1 );
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	
+	// these displays seems to be layout centered (1 column = 4 pixels of 4 bits each, little endian)
+	Private->Offset = (480 - Device->Width) / 4 / 2;
+	
+	// find a page size that is not too small is an integer of height
+	Private->PageSize = min(8, PAGE_BLOCK / (Device->Width / 2));
+	while (Private->PageSize && Device->Height != (Device->Height / Private->PageSize) * Private->PageSize) Private->PageSize--;
+	
+#ifdef SHADOW_BUFFER	
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#endif
+	Private->iRAM = heap_caps_malloc( Private->PageSize * Device->Width / 2, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+
+	ESP_LOGI(TAG, "SSD1322 with offset %u, page %u, iRAM %p", Private->Offset, Private->PageSize, Private->iRAM);
+			
+	// need to be off and disable display RAM
+	Device->DisplayOff( Device );
+    Device->WriteCommand( Device, 0xA5 );
+	
+	// Display Offset
+    Device->WriteCommand( Device, 0xA2 );
+    WriteDataByte( Device, 0 );
+	
+	// Display Start Line
+    Device->WriteCommand( Device, 0xA1 );
+	WriteDataByte( Device, 0x00 );
+	
+	// set flip modes
+	Private->ReMap = 0;
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+	
+	// set Display Enhancement
+    Device->WriteCommand( Device, 0xB4 );
+	WriteDataByte( Device, 0xA0 );
+	WriteDataByte( Device, 0xB5 );
+	
+	// set Clocks
+    Device->WriteCommand( Device, 0xB3 );
+	WriteDataByte( Device, 0xB2 ); // 0x91 seems to be common but is too slow for 5.5'
+	
+	// set MUX
+	Device->WriteCommand( Device, 0xCA );
+	WriteDataByte( Device, Device->Height - 1 );
+	
+	// phase 1 & 2 period
+	Device->WriteCommand( Device, 0xB1 );
+	WriteDataByte( Device, 0xE3 );	// 0xE2 was recommended
+	
+	// set pre-charge V 
+	Device->WriteCommand( Device, 0xBB );
+	WriteDataByte( Device, 0x0F); // 0x1F causes column interferences 
+	
+	// set COM deselect voltage
+	Device->WriteCommand( Device, 0xBE );
+	WriteDataByte( Device, 0x07 );
+	
+	// no Display Inversion
+    Device->WriteCommand( Device, 0xA6 );
+	
+	// gone with the wind
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+	
+	return true;
+}	
+
+static const struct GDS_Device SSD1322 = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff, .SetContrast = SetContrast,
+	.SetLayout = SetLayout,
+	.Update = Update, .Init = Init,
+	.Mode = GDS_GRAYSCALE, .Depth = 4,
+};	
+
+struct GDS_Device* SSD1322_Detect(char *Driver, struct GDS_Device* Device) {
+	if (!strcasestr(Driver, "SSD1322")) return NULL;
+		
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+	
+	*Device = SSD1322;	
+			
+	return Device;
+}

lib/display/SSD132x.c

@@ -0,0 +1,351 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+#define PAGE_BLOCK	1024
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "SSD132x";
+
+enum { SSD1326, SSD1327 };
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	uint8_t ReMap, PageSize;
+	uint8_t Model;
+};
+
+static const unsigned char BitReverseTable256[] = 
+{
+  0x00, 0x80, 0x40, 0xC0, 0x20, 0xA0, 0x60, 0xE0, 0x10, 0x90, 0x50, 0xD0, 0x30, 0xB0, 0x70, 0xF0, 
+  0x08, 0x88, 0x48, 0xC8, 0x28, 0xA8, 0x68, 0xE8, 0x18, 0x98, 0x58, 0xD8, 0x38, 0xB8, 0x78, 0xF8, 
+  0x04, 0x84, 0x44, 0xC4, 0x24, 0xA4, 0x64, 0xE4, 0x14, 0x94, 0x54, 0xD4, 0x34, 0xB4, 0x74, 0xF4, 
+  0x0C, 0x8C, 0x4C, 0xCC, 0x2C, 0xAC, 0x6C, 0xEC, 0x1C, 0x9C, 0x5C, 0xDC, 0x3C, 0xBC, 0x7C, 0xFC, 
+  0x02, 0x82, 0x42, 0xC2, 0x22, 0xA2, 0x62, 0xE2, 0x12, 0x92, 0x52, 0xD2, 0x32, 0xB2, 0x72, 0xF2, 
+  0x0A, 0x8A, 0x4A, 0xCA, 0x2A, 0xAA, 0x6A, 0xEA, 0x1A, 0x9A, 0x5A, 0xDA, 0x3A, 0xBA, 0x7A, 0xFA,
+  0x06, 0x86, 0x46, 0xC6, 0x26, 0xA6, 0x66, 0xE6, 0x16, 0x96, 0x56, 0xD6, 0x36, 0xB6, 0x76, 0xF6, 
+  0x0E, 0x8E, 0x4E, 0xCE, 0x2E, 0xAE, 0x6E, 0xEE, 0x1E, 0x9E, 0x5E, 0xDE, 0x3E, 0xBE, 0x7E, 0xFE,
+  0x01, 0x81, 0x41, 0xC1, 0x21, 0xA1, 0x61, 0xE1, 0x11, 0x91, 0x51, 0xD1, 0x31, 0xB1, 0x71, 0xF1,
+  0x09, 0x89, 0x49, 0xC9, 0x29, 0xA9, 0x69, 0xE9, 0x19, 0x99, 0x59, 0xD9, 0x39, 0xB9, 0x79, 0xF9, 
+  0x05, 0x85, 0x45, 0xC5, 0x25, 0xA5, 0x65, 0xE5, 0x15, 0x95, 0x55, 0xD5, 0x35, 0xB5, 0x75, 0xF5,
+  0x0D, 0x8D, 0x4D, 0xCD, 0x2D, 0xAD, 0x6D, 0xED, 0x1D, 0x9D, 0x5D, 0xDD, 0x3D, 0xBD, 0x7D, 0xFD,
+  0x03, 0x83, 0x43, 0xC3, 0x23, 0xA3, 0x63, 0xE3, 0x13, 0x93, 0x53, 0xD3, 0x33, 0xB3, 0x73, 0xF3, 
+  0x0B, 0x8B, 0x4B, 0xCB, 0x2B, 0xAB, 0x6B, 0xEB, 0x1B, 0x9B, 0x5B, 0xDB, 0x3B, 0xBB, 0x7B, 0xFB,
+  0x07, 0x87, 0x47, 0xC7, 0x27, 0xA7, 0x67, 0xE7, 0x17, 0x97, 0x57, 0xD7, 0x37, 0xB7, 0x77, 0xF7, 
+  0x0F, 0x8F, 0x4F, 0xCF, 0x2F, 0xAF, 0x6F, 0xEF, 0x1F, 0x9F, 0x5F, 0xDF, 0x3F, 0xBF, 0x7F, 0xFF
+};
+
+// Functions are not declared to minimize # of lines
+
+static void SetColumnAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0x15 );
+	Device->WriteCommand( Device, Start );
+	Device->WriteCommand( Device, End );
+}
+static void SetRowAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	// can be by row, not by page (see Update Optimization)
+	Device->WriteCommand( Device, 0x75 );
+	Device->WriteCommand( Device, Start );
+	Device->WriteCommand( Device, End );
+}
+
+static void Update4( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+	// always update by full lines
+	SetColumnAddress( Device, 0, Device->Width / 2 - 1);
+	
+#ifdef SHADOW_BUFFER
+	uint16_t *optr = (uint16_t*) Private->Shadowbuffer, *iptr = (uint16_t*) Device->Framebuffer;
+	bool dirty = false;
+	
+	for (int r = 0, page = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width always / by 2)
+		for (int c = Device->Width / 2 / 2; --c >= 0;) {
+			if (*optr != *iptr) {
+				dirty = true;
+				*optr = *iptr;
+			}
+			iptr++; optr++;
+		}
+		
+		// one line done, check for page boundary
+		if (++page == Private->PageSize) {
+			if (dirty) {
+				SetRowAddress( Device, r - page + 1, r );
+				// own use of IRAM has not proven to be much better than letting SPI do its copy
+				if (Private->iRAM) {
+					memcpy(Private->iRAM, Private->Shadowbuffer + (r - page + 1) * Device->Width / 2, page * Device->Width / 2 );
+					Device->WriteData( Device, Private->iRAM, Device->Width * page / 2 );
+				} else	{
+					Device->WriteData( Device, Private->Shadowbuffer + (r - page + 1) * Device->Width / 2, page * Device->Width / 2 );					
+				}	
+				dirty = false;
+			}	
+			page = 0;
+		}	
+	}	
+#else
+	for (int r = 0; r < Device->Height; r += Private->PageSize) {
+		SetRowAddress( Device, r, r + Private->PageSize - 1 );
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width / 2, Private->PageSize * Device->Width / 2 );
+			Device->WriteData( Device, Private->iRAM, Private->PageSize * Device->Width / 2 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width / 2, Private->PageSize * Device->Width / 2 );
+		}	
+	}	
+#endif	
+}
+
+/* 
+ We have to make a choice here: either we go by row one by one and send lots of 
+ small packets on the serial interface or we do a page of N rows once at least
+ a change has been detected. So far, choice is to go one-by-one
+*/ 
+static void Update1( struct GDS_Device* Device ) {
+#ifdef SHADOW_BUFFER
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	// not sure the compiler does not have to redo all calculation in for loops, so local it is
+	int width = Device->Width / 8, rows = Device->Height;
+	uint8_t *optr = Private->Shadowbuffer, *iptr = Device->Framebuffer;
+	int CurrentRow = -1, FirstCol = -1, LastCol = -1;
+	
+	// by row, find first and last columns that have been updated
+	for (int r = 0; r < rows; r++) {
+		uint8_t first = 0, last;	
+		for (int c = 0; c < width; c++) {
+			if (*iptr != *optr) {
+				if (!first) first = c + 1;
+				last = c ;
+			}	
+			*optr++ = *iptr++;
+		}
+		
+		// now update the display by "byte rows"
+		if (first--) {
+			// only set column when useful, saves a fair bit of CPU
+			if (first > FirstCol && first <= FirstCol + 4 && last < LastCol && last >= LastCol - 4) {
+				first = FirstCol;
+				last = LastCol;
+			} else {	
+				SetColumnAddress( Device, first, last );
+				FirstCol = first;
+				LastCol = last;
+			}
+			
+			// Set row only when needed, otherwise let auto-increment work
+			if (r != CurrentRow) SetRowAddress( Device, r, Device->Height - 1 );
+			CurrentRow = r + 1;
+			
+			// actual write
+			Device->WriteData( Device, Private->Shadowbuffer + r*width + first, last - first + 1 );
+		}
+	}	
+#else	
+	// automatic counter and end Row/Column
+	SetColumnAddress( Device, 0, Device->Width / 8 - 1);
+	SetRowAddress( Device, 0, Device->Height - 1);
+	Device->WriteData( Device, Device->Framebuffer, Device->FramebufferSize );
+#endif	
+}
+
+// in 1 bit mode, SSD1326 has a different memory map than SSD1306 and SH1106
+static void IRAM_ATTR _DrawPixel1Fast( struct GDS_Device* Device, int X, int Y, int Color ) {
+    uint32_t XBit = ( X & 0x07 );
+    uint8_t* FBOffset = Device->Framebuffer + ( ( Y * Device->Width + X ) >> 3 );
+
+    if ( Color == GDS_COLOR_XOR ) {
+        *FBOffset ^= BIT( 7 - XBit );
+    } else {
+		// we might be able to save the 7-Xbit using BitRemap (A0 bit 2)
+        *FBOffset = ( Color == GDS_COLOR_BLACK ) ?  *FBOffset & ~BIT( XBit ) : *FBOffset | BIT( XBit );
+    }
+}
+
+static void ClearWindow( struct GDS_Device* Device, int x1, int y1, int x2, int y2, int Color ) {
+	uint8_t _Color = Color == GDS_COLOR_BLACK ? 0: 0xff;
+	int Width = Device->Width >> 3;
+	uint8_t *optr = Device->Framebuffer;
+	
+	for (int r = y1; r <= y2; r++) {
+		int c = x1;
+		// for a row that is not on a boundary, not column opt can be done, so handle all columns on that line
+		while (c & 0x07 && c <= x2) _DrawPixel1Fast( Device, c++, r, Color );
+		// at this point we are aligned on column boundary
+		int chunk = (x2 - c + 1) >> 3;
+		memset(optr + Width * r + (c >> 3), _Color, chunk );
+		c += chunk * 8;
+		while (c <= x2) _DrawPixel1Fast( Device, c++, r, Color );
+	}
+}
+
+static void DrawBitmapCBR(struct GDS_Device* Device, uint8_t *Data, int Width, int Height, int Color ) {
+	if (!Height) Height = Device->Height;
+	if (!Width) Width = Device->Width;
+	int DWidth = Device->Width >> 3;
+	
+	// Two consecutive bits of source data are split over two different bytes of framebuffer
+	for (int c = 0; c < Width; c++) {
+		uint8_t shift = c & 0x07, bit = ~(1 << shift);
+		uint8_t *optr = Device->Framebuffer + (c >> 3);
+		
+		// we need to linearize code to let compiler better optimize
+		for (int r = Height >> 3; --r >= 0;) {
+			uint8_t Byte = BitReverseTable256[*Data++];
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth; Byte >>= 1;
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth; Byte >>= 1;
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth; Byte >>= 1;
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth; Byte >>= 1;
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth; Byte >>= 1;
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth; Byte >>= 1;
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth; Byte >>= 1;
+			*optr = (*optr & bit) | ((Byte & 0x01) << shift); optr += DWidth;
+		}	
+	}
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	if (Private->Model == SSD1326) {
+		Private->ReMap = Layout->HFlip ? (Private->ReMap | ((1 << 0) | (1 << 2))) : (Private->ReMap & ~((1 << 0) | (1 << 2)));
+		Private->ReMap = Layout->HFlip ? (Private->ReMap | (1 << 1)) : (Private->ReMap & ~(1 << 1));		
+	} else {
+		Private->ReMap = Layout->VFlip ? (Private->ReMap | ((1 << 0) | (1 << 1))) : (Private->ReMap & ~((1 << 0) | (1 << 1)));
+		Private->ReMap = Layout->VFlip ? (Private->ReMap | (1 << 4)) : (Private->ReMap & ~(1 << 4));
+	}	
+	Device->WriteCommand( Device, 0xA0 );
+	Device->WriteCommand( Device, Private->ReMap );
+    Device->WriteCommand( Device, Layout->Invert ? 0xA7 : 0xA4 );	
+}	
+
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAF ); }
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAE ); }
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+    Device->WriteCommand( Device, 0x81 );
+    Device->WriteCommand( Device, Contrast );
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	
+	// find a page size that is not too small is an integer of height
+	Private->PageSize = min(8, PAGE_BLOCK / (Device->Width / 2));
+	while (Private->PageSize && Device->Height != (Device->Height / Private->PageSize) * Private->PageSize) Private->PageSize--;
+	
+#ifdef SHADOW_BUFFER	
+#ifdef USE_IRAM
+	if (Device->IF == GDS_IF_SPI) {
+		if (Device->Depth == 1) {
+			Private->Shadowbuffer = heap_caps_malloc( Device->FramebufferSize, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+		} else {
+			Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+			Private->iRAM = heap_caps_malloc( Private->PageSize * Device->Width / 2, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+		}	
+	} else
+#endif
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#else
+#ifdef USE_IRAM	
+	if (Device->Depth == 4 && Device->IF == GDS_IF_SPI) Private->iRAM = heap_caps_malloc( Private->PageSize * Device->Width / 2, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+#endif	
+#endif
+
+	ESP_LOGI(TAG, "SSD1326/7 with bit depth %u, page %u, iRAM %p", Device->Depth, Private->PageSize, Private->iRAM);
+			
+	// need to be off and disable display RAM
+	Device->DisplayOff( Device );
+    Device->WriteCommand( Device, 0xA5 );
+	
+	// need COM split (6)
+	Private->ReMap = 1 << 6;
+	// MUX Ratio
+    Device->WriteCommand( Device, 0xA8 );
+    Device->WriteCommand( Device, Device->Height - 1);
+	// Display Offset
+    Device->WriteCommand( Device, 0xA2 );
+    Device->WriteCommand( Device, 0 );
+	// Display Start Line
+    Device->WriteCommand( Device, 0xA1 );
+	Device->WriteCommand( Device, 0x00 );
+	Device->SetContrast( Device, 0x7F );
+	// set flip modes
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+	// no Display Inversion
+    Device->WriteCommand( Device, 0xA6 );
+	// set Clocks
+    Device->WriteCommand( Device, 0xB3 );
+    Device->WriteCommand( Device, ( 0x08 << 4 ) | 0x00 );
+	// set Adressing Mode Horizontal
+	Private->ReMap |= (0 << 2);
+	// set monotchrome mode if required
+	if (Device->Depth == 1) Private->ReMap |= (1 << 4);
+	// write ReMap byte
+	Device->WriteCommand( Device, 0xA0 );
+	Device->WriteCommand( Device, Private->ReMap );		
+		
+	// gone with the wind
+	Device->WriteCommand( Device, 0xA4 );
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+	
+	return true;
+}	
+
+static const struct GDS_Device SSD132x = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff, .SetContrast = SetContrast,
+	.SetLayout = SetLayout,
+	.Update = Update4, .Init = Init,
+	.Mode = GDS_GRAYSCALE, .Depth = 4,
+};	
+
+struct GDS_Device* SSD132x_Detect(char *Driver, struct GDS_Device* Device) {
+	uint8_t Model;
+	int Depth;
+		
+	if (strcasestr(Driver, "SSD1326")) Model = SSD1326;
+	else if (strcasestr(Driver, "SSD1327")) Model = SSD1327;
+	else return NULL;
+	
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+	
+	*Device = SSD132x;	
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	Private->Model = Model;
+		
+	sscanf(Driver, "%*[^:]:%u", &Depth);
+		
+	if (Model == SSD1326 && Depth == 1) {
+		Device->Update = Update1;
+		Device->DrawPixelFast = _DrawPixel1Fast;
+		Device->DrawBitmapCBR = DrawBitmapCBR;
+		Device->ClearWindow = ClearWindow;
+		Device->Depth = 1;		
+		Device->Mode = GDS_MONO;
+#if !defined SHADOW_BUFFER && defined USE_IRAM	
+		Device->Alloc = GDS_ALLOC_IRAM_SPI;
+#endif	
+	}
+	
+	return Device;
+}

lib/display/SSD1351.c

@@ -0,0 +1,288 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+#define PAGE_BLOCK		2048
+#define ENABLE_WRITE	0x5c
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "SSD1351";
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	uint8_t ReMap, PageSize;
+};
+
+// Functions are not declared to minimize # of lines
+
+static void WriteByte( struct GDS_Device* Device, uint8_t Data ) {
+	Device->WriteData( Device, &Data, 1 );
+}
+
+static void SetColumnAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0x15 );
+	WriteByte( Device, Start );
+	WriteByte( Device, End );
+}
+static void SetRowAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	Device->WriteCommand( Device, 0x75 );
+	WriteByte( Device, Start );
+	WriteByte( Device, End );
+}
+
+static void Update16( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+#ifdef SHADOW_BUFFER
+	uint32_t *optr = (uint32_t*) Private->Shadowbuffer, *iptr = (uint32_t*) Device->Framebuffer;
+	int FirstCol = Device->Width / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+	
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width is always a multiple of 2)
+		for (int c = 0; c < Device->Width / 2; c++, iptr++, optr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+
+		// wait for a large enough window - careful that window size might increase by more than a line at once !
+		if (FirstRow < 0 || ((LastCol - FirstCol + 1) * (r - FirstRow + 1) * 4 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol *= 2;
+		LastCol = LastCol * 2 + 1;
+		SetRowAddress( Device, FirstRow, LastRow );
+		SetColumnAddress( Device, FirstCol, LastCol );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 2;
+			
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM < PAGE_BLOCK && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize );
+		}	
+
+		FirstCol = Device->Width / 2; LastCol = 0;
+		FirstRow = -1;
+	}	
+#else
+	// always update by full lines
+	SetColumnAddress( Device, 0, Device->Width - 1);
+	
+	for (int r = 0; r < Device->Height; r += min(Private->PageSize, Device->Height - r)) {
+		int Height = min(Private->PageSize, Device->Height - r);
+		
+		SetRowAddress( Device, r, r + Height - 1 );
+		Device->WriteCommand(Device, ENABLE_WRITE);		
+		
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width * 2, Height * Device->Width * 2 );
+			Device->WriteData( Device, Private->iRAM, Height * Device->Width * 2 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width * 2, Height * Device->Width * 2 );
+		}	
+	}	
+#endif	
+}
+
+static void Update24( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	int FirstCol = (Device->Width * 3) / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+		
+#ifdef SHADOW_BUFFER
+	uint16_t *optr = (uint16_t*) Private->Shadowbuffer, *iptr = (uint16_t*) Device->Framebuffer;
+	
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width always / by 2)
+		for (int c = 0; c < (Device->Width * 3) / 2; c++, optr++, iptr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+		
+		// do we have enough to send (cols are divided by 3/2)
+		if (FirstRow < 0 || ((((LastCol - FirstCol + 1) * 2 + 3 - 1) / 3) * (r - FirstRow + 1) * 3 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol = (FirstCol * 2) / 3;
+		LastCol = (LastCol * 2 + 1) / 3; 
+		SetRowAddress( Device, FirstRow, LastRow );
+		SetColumnAddress( Device, FirstCol, LastCol );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 3;
+					
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 3, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM < PAGE_BLOCK && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}	
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 3, ChunkSize );
+		}	
+
+		FirstCol = (Device->Width * 3) / 2; LastCol = 0;
+		FirstRow = -1;
+	}	
+#else
+	// always update by full lines
+	SetColumnAddress( Device, 0, Device->Width - 1);
+	
+	for (int r = 0; r < Device->Height; r += min(Private->PageSize, Device->Height - r)) {
+		int Height = min(Private->PageSize, Device->Height - r);
+		
+		SetRowAddress( Device, r, r + Height - 1 );
+		Device->WriteCommand(Device, ENABLE_WRITE);
+		
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width * 3, Height * Device->Width * 3 );
+			Device->WriteData( Device, Private->iRAM, Height * Device->Width * 3 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width * 3, Height * Device->Width * 3 );
+		}	
+	}	
+#endif	
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	Private->ReMap = Layout->HFlip ? (Private->ReMap & ~(1 << 1)) : (Private->ReMap | (1 << 1));
+	Private->ReMap = Layout->VFlip ? (Private->ReMap | (1 << 4)) : (Private->ReMap & ~(1 << 4));
+	Device->WriteCommand( Device, 0xA0 );
+	WriteByte( Device, Private->ReMap );
+    Device->WriteCommand( Device, Layout->Invert ? 0xA7 : 0xA6 );	
+}	
+
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAF ); }
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0xAE ); }
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+    Device->WriteCommand( Device, 0xC7 );
+	WriteByte( Device, Contrast >> 4);
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	int Depth = (Device->Depth + 8 - 1) / 8;
+	
+	Private->PageSize = min(8, PAGE_BLOCK / (Device->Width * Depth));
+	
+#ifdef SHADOW_BUFFER	
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#endif
+#ifdef USE_IRAM
+	Private->iRAM = heap_caps_malloc( (Private->PageSize + 1) * Device->Width * Depth, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+#endif
+
+	ESP_LOGI(TAG, "SSD1351 with bit depth %u, page %u, iRAM %p", Device->Depth, Private->PageSize, Private->iRAM);
+	
+	// unlock (specially 0xA2)
+	Device->WriteCommand( Device, 0xFD);
+	WriteByte(Device, 0xB1);
+		
+	// set clocks
+	/*
+    Device->WriteCommand( Device, 0xB3 );
+    WriteByte( Device, ( 0x08 << 4 ) | 0x00 );
+	*/
+
+	// need to be off and disable display RAM
+	Device->DisplayOff( Device );
+
+	// need COM split (5)
+	Private->ReMap = (1 << 5);
+	
+	// Display Offset
+    Device->WriteCommand( Device, 0xA2 );
+    WriteByte( Device, 0x00 );
+
+	// Display Start Line
+    Device->WriteCommand( Device, 0xA1 );
+	WriteByte( Device, 0x00 );
+	
+	// set flip modes & contrast
+	Device->SetContrast( Device, 0x7F );
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+	
+	// set Adressing Mode Horizontal
+	Private->ReMap |= (0 << 2);
+	// set screen depth (16/18)
+	if (Device->Depth == 24) Private->ReMap |= (0x02 << 6);
+	// write ReMap byte
+	Device->WriteCommand( Device, 0xA0 );
+	WriteByte( Device, Private->ReMap );		
+	
+	// no Display Inversion
+    Device->WriteCommand( Device, 0xA6 );	
+	
+	// gone with the wind
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+	
+	return true;
+}	
+
+static const struct GDS_Device SSD1351 = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff, .SetContrast = SetContrast,
+	.SetLayout = SetLayout,
+	.Update = Update16, .Init = Init,
+	.Mode = GDS_RGB565, .Depth = 16,
+};	
+
+struct GDS_Device* SSD1351_Detect(char *Driver, struct GDS_Device* Device) {
+	int Depth;
+	
+	if (!strcasestr(Driver, "SSD1351")) return NULL;
+	
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+	
+	*Device = SSD1351;	
+	sscanf(Driver, "%*[^:]:%u", &Depth);
+	
+	if (Depth == 18) {
+		Device->Mode = GDS_RGB666;
+		Device->Depth = 24;
+		Device->Update = Update24;
+	} 	
+	
+	return Device;
+}

lib/display/SSD1675.c

@@ -0,0 +1,251 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+#include "driver/gpio.h"
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+static char TAG[] = "SSD1675";
+
+const unsigned char EPD_lut_full_update[] = {
+    0x80,0x60,0x40,0x00,0x00,0x00,0x00,             //LUT0: BB:     VS 0 ~7
+    0x10,0x60,0x20,0x00,0x00,0x00,0x00,             //LUT1: BW:     VS 0 ~7
+    0x80,0x60,0x40,0x00,0x00,0x00,0x00,             //LUT2: WB:     VS 0 ~7
+    0x10,0x60,0x20,0x00,0x00,0x00,0x00,             //LUT3: WW:     VS 0 ~7
+    0x00,0x00,0x00,0x00,0x00,0x00,0x00,             //LUT4: VCOM:   VS 0 ~7
+    0x03,0x03,0x00,0x00,0x02,                       // TP0 A~D RP0
+    0x09,0x09,0x00,0x00,0x02,                       // TP1 A~D RP1
+    0x03,0x03,0x00,0x00,0x02,                       // TP2 A~D RP2
+    0x00,0x00,0x00,0x00,0x00,                       // TP3 A~D RP3
+    0x00,0x00,0x00,0x00,0x00,                       // TP4 A~D RP4
+    0x00,0x00,0x00,0x00,0x00,                       // TP5 A~D RP5
+    0x00,0x00,0x00,0x00,0x00,                       // TP6 A~D RP6
+    0x15,0x41,0xA8,0x32,0x30,0x0A,
+};
+
+const unsigned char EPD_lut_partial_update[]= { //20 bytes
+    0x00,0x00,0x00,0x00,0x00,0x00,0x00,             //LUT0: BB:     VS 0 ~7
+    0x80,0x00,0x00,0x00,0x00,0x00,0x00,             //LUT1: BW:     VS 0 ~7
+    0x40,0x00,0x00,0x00,0x00,0x00,0x00,             //LUT2: WB:     VS 0 ~7
+    0x00,0x00,0x00,0x00,0x00,0x00,0x00,             //LUT3: WW:     VS 0 ~7
+    0x00,0x00,0x00,0x00,0x00,0x00,0x00,             //LUT4: VCOM:   VS 0 ~7
+    0x0A,0x00,0x00,0x00,0x00,                       // TP0 A~D RP0
+    0x00,0x00,0x00,0x00,0x00,                       // TP1 A~D RP1
+    0x00,0x00,0x00,0x00,0x00,                       // TP2 A~D RP2
+    0x00,0x00,0x00,0x00,0x00,                       // TP3 A~D RP3
+    0x00,0x00,0x00,0x00,0x00,                       // TP4 A~D RP4
+    0x00,0x00,0x00,0x00,0x00,                       // TP5 A~D RP5
+    0x00,0x00,0x00,0x00,0x00,                       // TP6 A~D RP6
+    0x15,0x41,0xA8,0x32,0x30,0x0A,
+};
+
+struct PrivateSpace {
+	int	ReadyPin;
+	uint16_t Height;
+};
+
+// Functions are not declared to minimize # of lines
+
+void WaitReady( struct GDS_Device* Device) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	if (Private->ReadyPin >= 0) {
+		int count = 4*1000;
+		while (gpio_get_level( Private->ReadyPin ) && count) {
+			vTaskDelay( pdMS_TO_TICKS(100) );
+			count -= 100;
+		}	
+	} else {
+		vTaskDelay( pdMS_TO_TICKS(2000) );
+	}	
+}	
+
+static void WriteByte( struct GDS_Device* Device, uint8_t Data ) {
+	Device->WriteData( Device, &Data, 1 );
+}
+
+static void SetColumnAddress( struct GDS_Device* Device, uint8_t Start, uint8_t End ) {
+	// start might be greater than end if we decrement
+	Device->WriteCommand( Device, 0x44 );
+	Device->WriteData( Device, &Start, 1 );
+	Device->WriteData( Device, &End, 1 );
+	
+	// we obviously want to start ... from the start
+	Device->WriteCommand( Device, 0x4e );	
+	WriteByte( Device, Start );
+}
+static void SetRowAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	// start might be greater than end if we decrement
+	Device->WriteCommand( Device, 0x45 );	
+	WriteByte( Device, Start );
+	WriteByte( Device, Start >> 8 );
+	WriteByte( Device, End );
+	WriteByte( Device, End >> 8 );
+	
+	// we obviously want to start ... from the start
+	Device->WriteCommand( Device, 0x4f );	
+	WriteByte( Device, Start );
+	WriteByte( Device, Start >> 8 );
+}
+
+static void Update( struct GDS_Device* Device ) {
+	uint8_t *iptr = Device->Framebuffer;
+	
+	Device->WriteCommand( Device, 0x24 );
+	
+	// this is awfully slow, but e-ink are slow anyway ...
+	for (int i = Device->FramebufferSize; --i >= 0;) {
+		WriteByte( Device, ~*iptr++ );
+	}	
+	
+	Device->WriteCommand( Device, 0x22 ); 
+    WriteByte( Device, 0xC7);
+    Device->WriteCommand( Device, 0X20 );
+
+	WaitReady( Device );
+}
+
+// remember that for these ELD drivers W and H are "inverted"
+static inline void _DrawPixel( struct GDS_Device* Device, int X, int Y, int Color ) {
+    uint32_t YBit = ( Y & 0x07 );
+    Y>>= 3;
+
+    uint8_t* FBOffset = Device->Framebuffer + ( ( Y * Device->Width ) + X );
+    *FBOffset = ( Color == GDS_COLOR_BLACK ) ? *FBOffset & ~BIT( 7-YBit ) : *FBOffset | BIT( 7-YBit );
+}	
+
+static void ClearWindow( struct GDS_Device* Device, int x1, int y1, int x2, int y2, int Color ) {
+	for (int r = y1; r <= y2; r++) {
+		for (int c = x1; c <= x2; c++) {
+			_DrawPixel( Device, c, r, Color );
+		}	
+	}
+}
+
+static void DrawBitmapCBR(struct GDS_Device* Device, uint8_t *Data, int Width, int Height, int Color ) {
+	if (!Height) Height = Device->Height;
+	if (!Width) Width = Device->Width;
+	
+	// just do row/column swap
+	for (int r = 0; r < Height; r++) {
+		uint8_t *optr = Device->Framebuffer + r*Device->Width, *iptr = Data + r;
+		for (int c = Width; --c >= 0;) {
+			*optr++ = *iptr;
+			iptr += Height;
+		}	
+	}
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	
+	// need to re-adjust framebuffer because these are non % 8 
+	Private->Height = Device->Height;
+	if (Device->Height & 0x07) Device->Height = ((Device->Height >> 3) + 1) << 3;
+	
+	Device->FramebufferSize = Device->Width * Device->Height / 8;
+	Device->Framebuffer = calloc(1, Device->FramebufferSize);
+	NullCheck( Device->Framebuffer, return false );
+	
+	if (Private->ReadyPin >= 0) {
+		gpio_pad_select_gpio( Private->ReadyPin );
+		gpio_set_pull_mode( Private->ReadyPin, GPIO_PULLUP_ONLY);
+		gpio_set_direction( Private->ReadyPin, GPIO_MODE_INPUT );
+	}
+	
+	// soft reset	
+	vTaskDelay(pdMS_TO_TICKS( 2000 ));
+    Device->WriteCommand( Device, 0x12 ); 	
+	WaitReady( Device );
+	
+	Device->WriteCommand( Device, 0x74 ); 			
+    WriteByte( Device, 0x54 );
+	Device->WriteCommand( Device, 0x7e ); 			
+    WriteByte( Device, 0x3B );
+	
+	Device->WriteCommand( Device, 0x3c );	
+    WriteByte( Device, 0x03 );
+	
+	Device->WriteCommand( Device, 0x2c );	
+    WriteByte( Device, 0x55 );
+
+	Device->WriteCommand( Device, 0x03 );	
+	WriteByte( Device, EPD_lut_full_update[70] );
+	
+	Device->WriteCommand( Device, 0x04 );	
+	WriteByte( Device, EPD_lut_full_update[71] );
+	WriteByte( Device, EPD_lut_full_update[72] );
+	WriteByte( Device, EPD_lut_full_update[73] );
+	
+	Device->WriteCommand( Device, 0x3a );	
+	WriteByte( Device, EPD_lut_full_update[74] );
+	Device->WriteCommand( Device, 0x3b );	
+	WriteByte( Device, EPD_lut_full_update[75] );
+	
+	Device->WriteCommand( Device, 0X32 );	
+	for (int i = 0; i < 70; i++) {
+		WriteByte( Device, EPD_lut_full_update[i] );
+	}	
+
+	// now deal with funny X/Y layout (W and H are "inverted")
+	Device->WriteCommand( Device, 0x01 );
+    WriteByte( Device, Device->Width - 1 );
+    WriteByte( Device, (Device->Width - 1) >> 8 );
+    WriteByte( Device, (0 << 0) );		
+
+	/* 
+	 Start from 0, Ymax, incX, decY. Starting from X=Height would be difficult
+	 as we would hit the extra bits added because height % 8 != 0 and they are
+	 not written by the DrawPixel. By starting from X=0 we are aligned and 
+	 doing incY is like a clockwise 90° rotation (vs otherwise we would virtually 
+	 do a counter-clockwise rotation but again have the "border" issue.
+	*/ 
+	Device->WriteCommand( Device, 0x11 ); 			
+	WriteByte( Device, (1 << 2) | (0 << 1) | (1 << 0));
+		
+	// must be in order with increment/decrement i.e start might be > end if we decrement
+	SetColumnAddress( Device, 0x0, (Device->Height >> 3) - 1 );
+	SetRowAddress ( Device, Device->Width - 1, 0 );
+
+	WaitReady( Device );
+	
+	Update( Device );
+	
+	return true;
+}	
+
+static const struct GDS_Device SSD1675 = {
+	.DrawBitmapCBR = DrawBitmapCBR, .ClearWindow = ClearWindow,
+	.DrawPixelFast = _DrawPixel,
+	.Update = Update, .Init = Init,
+	.Mode = GDS_MONO, .Depth = 1,
+	.Alloc = GDS_ALLOC_NONE,
+};	
+
+struct GDS_Device* SSD1675_Detect(char *Driver, struct GDS_Device* Device) {
+	if (!strcasestr(Driver, "SSD1675")) return NULL;
+	
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+	*Device = SSD1675;	
+	
+	char *p;
+	struct PrivateSpace* Private = (struct PrivateSpace*) Device->Private;
+	Private->ReadyPin = -1;
+	if ((p = strcasestr(Driver, "ready")) && (p = strchr(p, '='))) Private->ReadyPin = atoi(p + 1);
+	
+	ESP_LOGI(TAG, "SSD1675 driver with ready GPIO %d", Private->ReadyPin);
+	
+	return Device;
+}

lib/display/ST77xx.c

@@ -0,0 +1,304 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <esp_heap_caps.h>
+#include <esp_log.h>
+
+#include "gds.h"
+#include "gds_private.h"
+
+#define SHADOW_BUFFER
+#define USE_IRAM
+#define PAGE_BLOCK		2048
+#define ENABLE_WRITE	0x2c
+
+#define min(a,b) (((a) < (b)) ? (a) : (b))
+
+static char TAG[] = "ST77xx";
+
+enum { ST7735, ST7789 };
+
+struct PrivateSpace {
+	uint8_t *iRAM, *Shadowbuffer;
+	struct {
+		uint16_t Height, Width;
+	} Offset;
+	uint8_t MADCtl, PageSize;
+	uint8_t Model;
+};
+
+// Functions are not declared to minimize # of lines
+
+static void WriteByte( struct GDS_Device* Device, uint8_t Data ) {
+	Device->WriteData( Device, &Data, 1 );
+}
+
+static void SetColumnAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2A );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void SetRowAddress( struct GDS_Device* Device, uint16_t Start, uint16_t End ) {
+	uint32_t Addr = __builtin_bswap16(Start) | (__builtin_bswap16(End) << 16);
+	Device->WriteCommand( Device, 0x2B );
+	Device->WriteData( Device, (uint8_t*) &Addr, 4 );
+}
+
+static void Update16( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+#ifdef SHADOW_BUFFER
+	uint32_t *optr = (uint32_t*) Private->Shadowbuffer, *iptr = (uint32_t*) Device->Framebuffer;
+	int FirstCol = Device->Width / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+	
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width is always a multiple of 2)
+		for (int c = 0; c < Device->Width / 2; c++, iptr++, optr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+
+		// wait for a large enough window - careful that window size might increase by more than a line at once !
+		if (FirstRow < 0 || ((LastCol - FirstCol + 1) * (r - FirstRow + 1) * 4 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol *= 2;
+		LastCol = LastCol * 2 + 1;
+		SetRowAddress( Device, FirstRow + Private->Offset.Height, LastRow + Private->Offset.Height);
+		SetColumnAddress( Device, FirstCol + Private->Offset.Width, LastCol + Private->Offset.Width );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 2;
+			
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM <= (PAGE_BLOCK - ChunkSize) && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 2, ChunkSize );
+		}	
+
+		FirstCol = Device->Width / 2; LastCol = 0;
+		FirstRow = -1;
+	}	
+#else
+	// always update by full lines
+	SetColumnAddress( Device, Private->Offset.Width, Device->Width - 1);
+	
+	for (int r = 0; r < Device->Height; r += min(Private->PageSize, Device->Height - r)) {
+		int Height = min(Private->PageSize, Device->Height - r);
+		
+		SetRowAddress( Device, Private->Offset.Height + r, Private->Offset.Height + r + Height - 1 );
+		Device->WriteCommand(Device, ENABLE_WRITE);
+		
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width * 2, Height * Device->Width * 2 );
+			Device->WriteData( Device, Private->iRAM, Height * Device->Width * 2 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width * 2, Height * Device->Width * 2 );
+		}	
+	}	
+#endif	
+}
+
+static void Update24( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+		
+#ifdef SHADOW_BUFFER
+	uint16_t *optr = (uint16_t*) Private->Shadowbuffer, *iptr = (uint16_t*) Device->Framebuffer;
+	int FirstCol = (Device->Width * 3) / 2, LastCol = 0, FirstRow = -1, LastRow = 0;  
+	
+	for (int r = 0; r < Device->Height; r++) {
+		// look for change and update shadow (cheap optimization = width always / by 2)
+		for (int c = 0; c < (Device->Width * 3) / 2; c++, optr++, iptr++) {
+			if (*optr != *iptr) {
+				*optr = *iptr;
+				if (c < FirstCol) FirstCol = c;	
+				if (c > LastCol) LastCol = c;
+				if (FirstRow < 0) FirstRow = r;
+				LastRow = r;
+			}
+		}
+		
+		// do we have enough to send (cols are divided by 3/2)
+		if (FirstRow < 0 || ((((LastCol - FirstCol + 1) * 2 + 3 - 1) / 3) * (r - FirstRow + 1) * 3 < PAGE_BLOCK && r != Device->Height - 1)) continue;
+		
+		FirstCol = (FirstCol * 2) / 3;
+		LastCol = (LastCol * 2 + 1) / 3; 
+		SetRowAddress( Device, FirstRow + Private->Offset.Height, LastRow + Private->Offset.Height);
+		SetColumnAddress( Device, FirstCol + Private->Offset.Width, LastCol + Private->Offset.Width );
+		Device->WriteCommand( Device, ENABLE_WRITE );
+			
+		int ChunkSize = (LastCol - FirstCol + 1) * 3;
+					
+		// own use of IRAM has not proven to be much better than letting SPI do its copy
+		if (Private->iRAM) {
+			uint8_t *optr = Private->iRAM;
+			for (int i = FirstRow; i <= LastRow; i++) {
+				memcpy(optr, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 3, ChunkSize);
+				optr += ChunkSize;
+				if (optr - Private->iRAM <= (PAGE_BLOCK - ChunkSize) && i < LastRow) continue;
+				Device->WriteData(Device, Private->iRAM, optr - Private->iRAM);
+				optr = Private->iRAM;
+			}	
+		} else for (int i = FirstRow; i <= LastRow; i++) {
+			Device->WriteData( Device, Private->Shadowbuffer + (i * Device->Width + FirstCol) * 3, ChunkSize );
+		}	
+
+		FirstCol = (Device->Width * 3) / 2; LastCol = 0;
+		FirstRow = -1;
+	}	
+#else
+	// always update by full lines
+	SetColumnAddress( Device, Private->Offset.Width, Device->Width - 1);
+	
+	for (int r = 0; r < Device->Height; r += min(Private->PageSize, Device->Height - r)) {
+		int Height = min(Private->PageSize, Device->Height - r);
+		
+		SetRowAddress( Device, Private->Offset.Height + r, Private->Offset.Height + r + Height - 1 );
+		Device->WriteCommand(Device, ENABLE_WRITE);
+		
+		if (Private->iRAM) {
+			memcpy(Private->iRAM, Device->Framebuffer + r * Device->Width * 3, Height * Device->Width * 3 );
+			Device->WriteData( Device, Private->iRAM, Height * Device->Width * 3 );
+		} else	{
+			Device->WriteData( Device, Device->Framebuffer + r * Device->Width * 3, Height * Device->Width * 3 );
+		}	
+	}	
+#endif	
+}
+
+static void SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { 
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	
+	if (Private->Model == ST7789) {
+		if (Layout->Rotate) Private->Offset.Width += Layout->HFlip ? 320 - Device->Width : 0;
+		else Private->Offset.Height += Layout->HFlip ? 320 - Device->Height : 0;
+		Device->WriteCommand( Device, Layout->Invert ? 0x20 : 0x21 );			
+	} else {
+		Device->WriteCommand( Device, Layout->Invert ? 0x21 : 0x20 );	
+	}		
+
+	Private->MADCtl = Layout->HFlip ? (Private->MADCtl | (1 << 7)) : (Private->MADCtl & ~(1 << 7));
+	Private->MADCtl = Layout->VFlip ? (Private->MADCtl | (1 << 6)) : (Private->MADCtl & ~(1 << 6));
+	Private->MADCtl = Layout->Rotate ? (Private->MADCtl | (1 << 5)) : (Private->MADCtl & ~(1 << 5));
+	Private->MADCtl = Layout->ColorSwap ? (Private->MADCtl & ~(1 << 3)) : (Private->MADCtl | (1 << 3));	
+
+	Device->WriteCommand( Device, 0x36 );
+	WriteByte( Device, Private->MADCtl );
+
+#ifdef SHADOW_BUFFER
+	// force a full refresh (almost ...)
+	memset(Private->Shadowbuffer, 0xAA, Device->FramebufferSize);
+#endif	
+}	
+
+static void DisplayOn( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0x29 ); }
+static void DisplayOff( struct GDS_Device* Device ) { Device->WriteCommand( Device, 0x28 ); }
+
+static void SetContrast( struct GDS_Device* Device, uint8_t Contrast ) {
+	Device->WriteCommand( Device, 0x51 );
+	WriteByte( Device, Contrast );
+	
+	Device->SetContrast = NULL;
+	GDS_SetContrast( Device, Contrast );
+	Device->SetContrast = SetContrast;
+}
+
+static bool Init( struct GDS_Device* Device ) {
+	struct PrivateSpace *Private = (struct PrivateSpace*) Device->Private;
+	int Depth = (Device->Depth + 8 - 1) / 8;
+	
+	Private->PageSize = min(8, PAGE_BLOCK / (Device->Width * Depth));
+
+#ifdef SHADOW_BUFFER	
+	Private->Shadowbuffer = malloc( Device->FramebufferSize );	
+	memset(Private->Shadowbuffer, 0xFF, Device->FramebufferSize);
+#endif
+#ifdef USE_IRAM
+	Private->iRAM = heap_caps_malloc( (Private->PageSize + 1) * Device->Width * Depth, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+#endif
+
+	ESP_LOGI(TAG, "ST77xx with bit depth %u, offsets %hu:%hu, page %u, iRAM %p", Device->Depth, Private->Offset.Height, Private->Offset.Width, Private->PageSize, Private->iRAM);
+	
+	// Sleepout + Booster
+	Device->WriteCommand( Device, 0x11 );
+	
+	// need BGR & Address Mode
+	Private->MADCtl = 1 << 3;
+	Device->WriteCommand( Device, 0x36 );
+	WriteByte( Device, Private->MADCtl );		
+	
+	// set flip modes & contrast
+	GDS_SetContrast( Device, 0x7f );
+	struct GDS_Layout Layout = { };
+	Device->SetLayout( Device, &Layout );
+	
+	// set screen depth (16/18)
+	Device->WriteCommand( Device, 0x3A );
+	if (Private->Model == ST7789) WriteByte( Device, Device->Depth == 24 ? 0x066 : 0x55 );
+	else WriteByte( Device, Device->Depth == 24 ? 0x06 : 0x05 );
+	
+	// gone with the wind
+	Device->DisplayOn( Device );
+	Device->Update( Device );
+
+	return true;
+}	
+
+static const struct GDS_Device ST77xx = {
+	.DisplayOn = DisplayOn, .DisplayOff = DisplayOff,
+	.SetLayout = SetLayout,
+	.Update = Update16, .Init = Init,
+	.Mode = GDS_RGB565, .Depth = 16,
+};		
+
+struct GDS_Device* ST77xx_Detect(char *Driver, struct GDS_Device* Device) {
+	uint8_t Model;
+	int Depth;
+		
+	if (strcasestr(Driver, "ST7735")) Model = ST7735;
+	else if (strcasestr(Driver, "ST7789")) Model = ST7789;
+	else return NULL;
+		
+	if (!Device) Device = calloc(1, sizeof(struct GDS_Device));
+		
+	*Device = ST77xx;	
+	sscanf(Driver, "%*[^:]:%u", &Depth);
+
+	struct PrivateSpace* Private = (struct PrivateSpace*) Device->Private;
+	Private->Model = Model;
+
+	sscanf(Driver, "%*[^:]%*[^x]%*[^=]=%hu", &Private->Offset.Height);		
+	sscanf(Driver, "%*[^:]%*[^y]%*[^=]=%hu", &Private->Offset.Width);		
+	
+	if (Depth == 18) {
+		Device->Mode = GDS_RGB666;
+		Device->Depth = 24;
+		Device->Update = Update24;
+	} 	
+	
+	if (Model == ST7789) Device->SetContrast = SetContrast;
+
+	return Device;
+}

lib/display/component.mk

@@ -0,0 +1,12 @@
+#
+# Component Makefile
+#
+# This Makefile should, at the very least, just include $(SDK_PATH)/Makefile. By default,
+# this will take the sources in the src/ directory, compile them and link them into
+# lib(subdirectory_name).a in the build directory. This behaviour is entirely configurable,
+# please read the SDK documents if you need to do this.
+#
+
+COMPONENT_SRCDIRS := . core core/ifaces fonts 
+COMPONENT_ADD_INCLUDEDIRS := .
+COMPONENT_ADD_INCLUDEDIRS += ./core

lib/display/core/gds.c

@@ -0,0 +1,319 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G. 
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <string.h>
+#include <ctype.h>
+#include <stdint.h>
+#include <math.h>
+#include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+#include "driver/gpio.h"
+#include "driver/ledc.h"
+#include "esp_log.h"
+
+#include "gds.h"
+#include "gds_private.h"
+
+#ifdef CONFIG_IDF_TARGET_ESP32S3
+#define LEDC_SPEED_MODE LEDC_LOW_SPEED_MODE
+#else
+#define LEDC_SPEED_MODE LEDC_HIGH_SPEED_MODE
+#endif                
+
+static struct GDS_Device Display;
+static struct GDS_BacklightPWM PWMConfig;
+
+static char TAG[] = "gds";
+
+struct GDS_Device* GDS_AutoDetect( char *Driver, GDS_DetectFunc* DetectFunc[], struct GDS_BacklightPWM* PWM ) {
+	if (!Driver) return NULL;
+	if (PWM) PWMConfig = *PWM;
+	
+	for (int i = 0; DetectFunc[i]; i++) {
+		if (DetectFunc[i](Driver, &Display)) {
+			if (PWM && PWM->Init) {
+				ledc_timer_config_t PWMTimer = {
+						.duty_resolution = LEDC_TIMER_13_BIT,
+						.freq_hz = 5000,                   
+						.speed_mode = LEDC_SPEED_MODE,
+						.timer_num = PWMConfig.Timer,
+					};
+				ledc_timer_config(&PWMTimer);
+			}	
+			ESP_LOGD(TAG, "Detected driver %p with PWM %d", &Display, PWM ? PWM->Init : 0);			
+			return &Display;
+		}	
+	}
+	
+	return NULL;
+}
+
+void GDS_ClearExt(struct GDS_Device* Device, bool full, ...) {
+	bool commit = true;
+	
+	if (full) {
+		GDS_Clear( Device, GDS_COLOR_BLACK ); 
+	} else {
+		va_list args;
+		va_start(args, full);
+		commit = va_arg(args, int);
+		int x1 = va_arg(args, int), y1 = va_arg(args, int), x2 = va_arg(args, int), y2 = va_arg(args, int);
+		if (x2 < 0) x2 = Device->Width - 1;
+		if (y2 < 0) y2 = Device->Height - 1;
+		GDS_ClearWindow( Device, x1, y1, x2, y2, GDS_COLOR_BLACK );
+		va_end(args);
+	}
+	
+	Device->Dirty = true;
+	if (commit)	GDS_Update(Device);		
+}	
+
+void GDS_Clear( struct GDS_Device* Device, int Color ) {
+	if (Color == GDS_COLOR_BLACK) memset( Device->Framebuffer, 0, Device->FramebufferSize );
+	else if (Device->Depth == 1) memset( Device->Framebuffer, 0xff, Device->FramebufferSize );
+	else if (Device->Depth == 4) memset( Device->Framebuffer, Color | (Color << 4), Device->FramebufferSize );
+	else if (Device->Depth == 8) memset( Device->Framebuffer, Color, Device->FramebufferSize );
+	else GDS_ClearWindow(Device, 0, 0, -1, -1, Color);
+	Device->Dirty = true;
+}
+
+#define CLEAR_WINDOW(x1,y1,x2,y2,F,W,C,T,N)				\
+	for (int y = y1; y <= y2; y++) {					\
+		T *Ptr = (T*) F + (y * W + x1)*N;				\
+		for (int c = (x2 - x1)*N; c-- >= 0; *Ptr++ = C);	\
+	}
+
+void GDS_ClearWindow( struct GDS_Device* Device, int x1, int y1, int x2, int y2, int Color ) {
+	// -1 means up to width/height
+	if (x2 < 0) x2 = Device->Width - 1;
+	if (y2 < 0) y2 = Device->Height - 1;
+	
+	// driver can provide own optimized clear window
+	if (Device->ClearWindow) {
+		Device->ClearWindow( Device, x1, y1, x2, y2, Color );
+	} else if (Device->Depth == 1) {
+		// single shot if we erase all screen
+		if (x2 - x1 == Device->Width - 1 && y2 - y1 == Device->Height - 1) {
+			memset( Device->Framebuffer, Color == GDS_COLOR_BLACK ? 0 : 0xff, Device->FramebufferSize );
+		} else {
+			uint8_t _Color = Color == GDS_COLOR_BLACK ? 0: 0xff;
+			uint8_t Width = Device->Width >> 3;
+			uint8_t *optr = Device->Framebuffer;
+			// try to do byte processing as much as possible
+			for (int r = y1; r <= y2;) {
+				int c = x1;
+				// for a row that is not on a boundary, no optimization possible
+				while (r & 0x07 && r <= y2) {
+					for (c = x1; c <= x2; c++) Device->DrawPixelFast( Device, c, r, Color );
+					r++;
+				}
+				// go fast if we have more than 8 lines to write
+				if (r + 8 <= y2) {
+					memset(optr + Width * r + x1, _Color, x2 - x1 + 1);
+					r += 8;
+				} else while (r <= y2) {
+					for (c = x1; c <= x2; c++) Device->DrawPixelFast( Device, c, r, Color );
+					r++;
+				}
+			}
+		}
+	} if (Device->Depth == 4) {
+		if (x2 - x1 == Device->Width - 1 && y2 - y1 == Device->Height - 1) {
+			// we assume color is 0..15
+			memset( Device->Framebuffer, Color | (Color << 4), Device->FramebufferSize );
+		} else {
+			uint8_t _Color = Color | (Color << 4);
+			int Width = Device->Width;
+			uint8_t *optr = Device->Framebuffer;
+			// try to do byte processing as much as possible
+			for (int r = y1; r <= y2; r++) {
+				int c = x1;
+				if (c & 0x01) Device->DrawPixelFast( Device, c++, r, Color);
+				int chunk = (x2 - c + 1) >> 1;
+				memset(optr + ((r * Width + c)  >> 1), _Color, chunk);
+				if (c + chunk <= x2) Device->DrawPixelFast( Device, x2, r, Color);
+			}
+		}	
+	} else if (Device->Depth == 8) {
+		CLEAR_WINDOW(x1,y1,x2,y2,Device->Framebuffer,Device->Width,Color,uint8_t,1);
+	} else if (Device->Depth == 16) {
+		CLEAR_WINDOW(x1,y1,x2,y2,Device->Framebuffer,Device->Width,Color,uint16_t,1);
+	} else if (Device->Depth == 24) {
+		CLEAR_WINDOW(x1,y1,x2,y2,Device->Framebuffer,Device->Width,Color,uint8_t,3);
+	} else {
+		for (int y = y1; y <= y2; y++) {
+			for (int x = x1; x <= x2; x++) {
+				Device->DrawPixelFast( Device, x, y, Color);
+			}
+		}
+	}
+	
+	// make sure diplay will do update
+	Device->Dirty = true;
+}
+
+void GDS_Update( struct GDS_Device* Device ) {
+	if (Device->Dirty) Device->Update( Device );
+	Device->Dirty = false;
+}
+
+bool GDS_Reset( struct GDS_Device* Device ) {
+	if ( Device->RSTPin >= 0 ) {
+		gpio_set_level( Device->RSTPin, 0 );
+		vTaskDelay( pdMS_TO_TICKS( 100 ) );
+        gpio_set_level( Device->RSTPin, 1 );
+    }
+    return true;
+}
+
+static void IRAM_ATTR DrawPixel1Fast( struct GDS_Device* Device, int X, int Y, int Color ) {
+    uint32_t YBit = ( Y & 0x07 );
+    uint8_t* FBOffset;
+
+    /* 
+     * We only need to modify the Y coordinate since the pitch
+     * of the screen is the same as the width.
+     * Dividing Y by 8 gives us which row the pixel is in but not
+     * the bit position.
+     */
+    Y>>= 3;
+
+    FBOffset = Device->Framebuffer + ( ( Y * Device->Width ) + X );
+
+    if ( Color == GDS_COLOR_XOR ) {
+        *FBOffset ^= BIT( YBit );
+    } else {
+        *FBOffset = ( Color == GDS_COLOR_BLACK ) ? *FBOffset & ~BIT( YBit ) : *FBOffset | BIT( YBit );
+    }
+}
+
+static void IRAM_ATTR DrawPixel4Fast( struct GDS_Device* Device, int X, int Y, int Color ) {
+	uint8_t* FBOffset = Device->Framebuffer + ( (Y * Device->Width >> 1) + (X >> 1));
+	*FBOffset = X & 0x01 ? (*FBOffset & 0x0f) | ((Color & 0x0f) << 4) : ((*FBOffset & 0xf0) | (Color & 0x0f));
+}
+
+static void IRAM_ATTR DrawPixel4FastHigh( struct GDS_Device* Device, int X, int Y, int Color ) {
+    uint8_t* FBOffset = Device->Framebuffer + ( (Y * Device->Width >> 1) + (X >> 1));
+	*FBOffset = X & 0x01 ? ((*FBOffset & 0xf0) | (Color & 0x0f)) : (*FBOffset & 0x0f) | ((Color & 0x0f) << 4);
+}
+
+static void IRAM_ATTR DrawPixel8Fast( struct GDS_Device* Device, int X, int Y, int Color ) {
+	Device->Framebuffer[Y * Device->Width + X] = Color;
+}
+
+// assumes that Color is 16 bits R..RG..GB..B from MSB to LSB and FB wants 1st serialized byte to start with R
+static void IRAM_ATTR DrawPixel16Fast( struct GDS_Device* Device, int X, int Y, int Color ) {
+	uint16_t* FBOffset = (uint16_t*) Device->Framebuffer + Y * Device->Width + X;
+	*FBOffset = __builtin_bswap16(Color);
+}
+
+// assumes that Color is 18 bits RGB from MSB to LSB RRRRRRGGGGGGBBBBBB, so byte[0] is B 
+// FB is 3-bytes packets and starts with R for serialization so 0,1,2 ... = xxRRRRRR xxGGGGGG xxBBBBBB 
+static void IRAM_ATTR DrawPixel18Fast( struct GDS_Device* Device, int X, int Y, int Color ) {
+	uint8_t* FBOffset = Device->Framebuffer + (Y * Device->Width + X) * 3;
+	*FBOffset++ = Color >> 12; *FBOffset++ = (Color >> 6) & 0x3f; *FBOffset = Color & 0x3f;
+}
+
+// assumes that Color is 24 bits RGB from MSB to LSB RRRRRRRRGGGGGGGGBBBBBBBB, so byte[0] is B 
+// FB is 3-bytes packets and starts with R for serialization so 0,1,2 ... = RRRRRRRR GGGGGGGG BBBBBBBB 
+static void IRAM_ATTR DrawPixel24Fast( struct GDS_Device* Device, int X, int Y, int Color ) {
+	uint8_t* FBOffset = Device->Framebuffer + (Y * Device->Width + X) * 3;
+	*FBOffset++ = Color >> 16; *FBOffset++ = Color >> 8; *FBOffset = Color;
+}
+
+bool GDS_Init( struct GDS_Device* Device ) {
+	
+	if (Device->Depth > 8) Device->FramebufferSize = Device->Width * Device->Height * ((8 + Device->Depth - 1) / 8);
+	else Device->FramebufferSize = (Device->Width * Device->Height) / (8 / Device->Depth);
+    
+    // set the proper DrawPixel function if not already set by driver
+    if (!Device->DrawPixelFast) {
+        if (Device->Depth == 1) Device->DrawPixelFast = DrawPixel1Fast;
+        else if (Device->Depth == 4 && Device->HighNibble) Device->DrawPixelFast = DrawPixel4FastHigh;
+        else if (Device->Depth == 4) Device->DrawPixelFast = DrawPixel4Fast;
+        else if (Device->Depth == 8) Device->DrawPixelFast = DrawPixel8Fast;	
+        else if (Device->Depth == 16) Device->DrawPixelFast = DrawPixel16Fast;	
+        else if (Device->Depth == 24 && Device->Mode == GDS_RGB666) Device->DrawPixelFast = DrawPixel18Fast;	
+        else if (Device->Depth == 24 && Device->Mode == GDS_RGB888) Device->DrawPixelFast = DrawPixel24Fast;	
+    }	
+	
+	// allocate FB unless explicitely asked not to
+	if (!(Device->Alloc & GDS_ALLOC_NONE)) {
+		if ((Device->Alloc & GDS_ALLOC_IRAM) || ((Device->Alloc & GDS_ALLOC_IRAM_SPI) && Device->IF == GDS_IF_SPI)) {
+			Device->Framebuffer = heap_caps_calloc( 1, Device->FramebufferSize, MALLOC_CAP_INTERNAL | MALLOC_CAP_DMA );
+		} else {
+			Device->Framebuffer = calloc( 1, Device->FramebufferSize );
+		}	
+		NullCheck( Device->Framebuffer, return false );
+	}	
+	
+	if (Device->Backlight.Pin >= 0) {
+		Device->Backlight.Channel = PWMConfig.Channel++;
+		Device->Backlight.PWM = PWMConfig.Max - 1;
+
+		ledc_channel_config_t PWMChannel = {
+            .channel    = Device->Backlight.Channel,
+            .duty       = Device->Backlight.PWM,
+            .gpio_num   = Device->Backlight.Pin,
+            .speed_mode = LEDC_SPEED_MODE,
+            .hpoint     = 0,
+            .timer_sel  = PWMConfig.Timer,
+        };
+		
+		ledc_channel_config(&PWMChannel);
+	}
+	
+	bool Res = Device->Init( Device );
+	if (!Res && Device->Framebuffer) free(Device->Framebuffer);
+	return Res;
+}
+
+int GDS_GrayMap( struct GDS_Device* Device, uint8_t Level) {
+	switch(Device->Mode) {
+		case GDS_MONO: return Level;
+		case GDS_GRAYSCALE: return Level >> (8 - Device->Depth);
+		case GDS_RGB332:
+			Level >>= 5;	
+			return (Level << 6) | (Level << 3) | (Level >> 1);
+		case GDS_RGB444:	
+			Level >>= 4;
+			return (Level << 8) | (Level << 4) | Level;
+		case GDS_RGB555:	
+			Level >>= 3;
+			return (Level << 10) | (Level << 5) | Level;			
+		case GDS_RGB565:	
+			Level >>= 2;
+			return ((Level & ~0x01) << 10) | (Level << 5) | (Level >> 1);						
+		case GDS_RGB666:	
+			Level >>= 2;
+			return (Level << 12) | (Level << 6) | Level;									
+		case GDS_RGB888:	
+			return (Level << 16) | (Level << 8) | Level;												
+	}
+	
+	return -1;
+}
+
+void GDS_SetContrast( struct GDS_Device* Device, uint8_t Contrast ) { 
+	if (Device->SetContrast) Device->SetContrast( Device, Contrast ); 
+	else if (Device->Backlight.Pin >= 0) {
+		Device->Backlight.PWM = PWMConfig.Max * powf(Contrast / 255.0, 3);
+		ledc_set_duty( LEDC_SPEED_MODE, Device->Backlight.Channel, Device->Backlight.PWM );
+		ledc_update_duty( LEDC_SPEED_MODE, Device->Backlight.Channel );		
+	}
+}
+
+void GDS_SetLayout( struct GDS_Device* Device, struct GDS_Layout *Layout ) { if (Device->SetLayout) Device->SetLayout( Device, Layout ); }
+void GDS_SetDirty( struct GDS_Device* Device ) { Device->Dirty = true; }
+int	 GDS_GetWidth( struct GDS_Device* Device ) { return Device ? Device->Width : 0; }
+void GDS_SetTextWidth( struct GDS_Device* Device, int TextWidth ) { Device->TextWidth = Device && TextWidth && TextWidth < Device->Width ? TextWidth : Device->Width; }
+int	 GDS_GetHeight( struct GDS_Device* Device ) { return Device ? Device->Height : 0; }
+int	 GDS_GetDepth( struct GDS_Device* Device ) { return Device ? Device->Depth : 0; }
+int	 GDS_GetMode( struct GDS_Device* Device ) { return Device ? Device->Mode : 0; }
+void GDS_DisplayOn( struct GDS_Device* Device ) { if (Device->DisplayOn) Device->DisplayOn( Device ); }
+void GDS_DisplayOff( struct GDS_Device* Device ) { if (Device->DisplayOff) Device->DisplayOff( Device ); }

lib/display/core/gds.h

@@ -0,0 +1,56 @@
+#ifndef _GDS_H_
+#define _GDS_H_
+
+#include <stdint.h>
+#include <stdbool.h>
+
+/* NOTE for drivers:
+ The build-in DrawPixel(Fast), DrawCBR and ClearWindow have optimized for 1 bit 
+ and 4 bits grayscale screen depth and 8, 16, 24 color. For any other type of screen, 
+ DrawCBR and ClearWindow default to use DrawPixel, which is very sub-optimal. For 
+ other depth, you  must supply the DrawPixelFast. The built-in 1 bit depth function 
+ are only for screen with vertical framing (1 byte = 8 lines). For example SSD1326 in 
+ monochrome mode is not such type of screen, SH1106 and SSD1306 are
+*/ 
+
+// this is an ordered enum, do not change!
+enum { GDS_MONO = 0, GDS_GRAYSCALE, GDS_RGB332, GDS_RGB444, GDS_RGB555, GDS_RGB565, GDS_RGB666, GDS_RGB888 };
+		
+#define GDS_COLOR_BLACK (0)
+#define GDS_COLOR_WHITE (-1)
+#define GDS_COLOR_XOR 	(256)
+
+struct GDS_Device;
+struct GDS_FontDef;
+struct GDS_BacklightPWM { 
+	int Channel, Timer, Max;
+	bool Init;
+};
+struct GDS_Layout {
+	bool HFlip, VFlip;
+	bool Rotate;
+	bool Invert; 
+	bool ColorSwap;
+};
+
+typedef struct GDS_Device* GDS_DetectFunc(char *Driver, struct GDS_Device *Device);
+
+struct GDS_Device*	GDS_AutoDetect( char *Driver, GDS_DetectFunc* DetectFunc[], struct GDS_BacklightPWM *PWM );
+
+void 	GDS_SetContrast( struct GDS_Device* Device, uint8_t Contrast );
+void 	GDS_DisplayOn( struct GDS_Device* Device );
+void 	GDS_DisplayOff( struct GDS_Device* Device ); 
+void 	GDS_Update( struct GDS_Device* Device );
+void 	GDS_SetLayout( struct GDS_Device* Device, struct GDS_Layout* Layout);
+void 	GDS_SetDirty( struct GDS_Device* Device );
+int 	GDS_GetWidth( struct GDS_Device* Device );
+void 	GDS_SetTextWidth( struct GDS_Device* Device, int TextWidth );
+int 	GDS_GetHeight( struct GDS_Device* Device );
+int 	GDS_GetDepth( struct GDS_Device* Device );
+int 	GDS_GetMode( struct GDS_Device* Device );
+int 	GDS_GrayMap( struct GDS_Device* Device, uint8_t Level );
+void 	GDS_ClearExt( struct GDS_Device* Device, bool full, ...);
+void 	GDS_Clear( struct GDS_Device* Device, int Color );
+void 	GDS_ClearWindow( struct GDS_Device* Device, int x1, int y1, int x2, int y2, int Color );
+
+#endif

lib/display/core/gds_default_if.h

@@ -0,0 +1,24 @@
+#ifndef _GDS_DEFAULT_IF_H_
+#define _GDS_DEFAULT_IF_H_
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+struct GDS_Device;
+
+bool GDS_I2CInit( int PortNumber, int SDA, int SCL, int speed );
+bool GDS_I2CAttachDevice( struct GDS_Device* Device, int Width, int Height, int I2CAddress, int RSTPin, int BacklightPin );
+
+bool GDS_SPIInit( int SPI, int DC );
+bool GDS_SPIAttachDevice( struct GDS_Device* Device, int Width, int Height, int CSPin, int RSTPin, int Speed, int BacklightPin, int Mode );
+
+bool GDS_QSPIInit( int QSPI, int DC );
+bool GDS_QSPIAttachDevice( struct GDS_Device* Device, int Width, int Height, int CSPin, int RSTPin, int Speed, int BacklightPin, int Mode );
+bool GDS_QSPIBusInit( int MOSIPin, int MISOPin, int CLKPin, int Host );
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

lib/display/core/gds_draw.c

@@ -0,0 +1,379 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ *				 2020 Philippe G.
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdbool.h>
+#include <stdlib.h>
+#include <math.h>
+#include <esp_attr.h>
+
+#include "gds_private.h"
+#include "gds.h"
+#include "gds_draw.h"
+
+static const unsigned char BitReverseTable256[] = 
+{
+  0x00, 0x80, 0x40, 0xC0, 0x20, 0xA0, 0x60, 0xE0, 0x10, 0x90, 0x50, 0xD0, 0x30, 0xB0, 0x70, 0xF0, 
+  0x08, 0x88, 0x48, 0xC8, 0x28, 0xA8, 0x68, 0xE8, 0x18, 0x98, 0x58, 0xD8, 0x38, 0xB8, 0x78, 0xF8, 
+  0x04, 0x84, 0x44, 0xC4, 0x24, 0xA4, 0x64, 0xE4, 0x14, 0x94, 0x54, 0xD4, 0x34, 0xB4, 0x74, 0xF4, 
+  0x0C, 0x8C, 0x4C, 0xCC, 0x2C, 0xAC, 0x6C, 0xEC, 0x1C, 0x9C, 0x5C, 0xDC, 0x3C, 0xBC, 0x7C, 0xFC, 
+  0x02, 0x82, 0x42, 0xC2, 0x22, 0xA2, 0x62, 0xE2, 0x12, 0x92, 0x52, 0xD2, 0x32, 0xB2, 0x72, 0xF2, 
+  0x0A, 0x8A, 0x4A, 0xCA, 0x2A, 0xAA, 0x6A, 0xEA, 0x1A, 0x9A, 0x5A, 0xDA, 0x3A, 0xBA, 0x7A, 0xFA,
+  0x06, 0x86, 0x46, 0xC6, 0x26, 0xA6, 0x66, 0xE6, 0x16, 0x96, 0x56, 0xD6, 0x36, 0xB6, 0x76, 0xF6, 
+  0x0E, 0x8E, 0x4E, 0xCE, 0x2E, 0xAE, 0x6E, 0xEE, 0x1E, 0x9E, 0x5E, 0xDE, 0x3E, 0xBE, 0x7E, 0xFE,
+  0x01, 0x81, 0x41, 0xC1, 0x21, 0xA1, 0x61, 0xE1, 0x11, 0x91, 0x51, 0xD1, 0x31, 0xB1, 0x71, 0xF1,
+  0x09, 0x89, 0x49, 0xC9, 0x29, 0xA9, 0x69, 0xE9, 0x19, 0x99, 0x59, 0xD9, 0x39, 0xB9, 0x79, 0xF9, 
+  0x05, 0x85, 0x45, 0xC5, 0x25, 0xA5, 0x65, 0xE5, 0x15, 0x95, 0x55, 0xD5, 0x35, 0xB5, 0x75, 0xF5,
+  0x0D, 0x8D, 0x4D, 0xCD, 0x2D, 0xAD, 0x6D, 0xED, 0x1D, 0x9D, 0x5D, 0xDD, 0x3D, 0xBD, 0x7D, 0xFD,
+  0x03, 0x83, 0x43, 0xC3, 0x23, 0xA3, 0x63, 0xE3, 0x13, 0x93, 0x53, 0xD3, 0x33, 0xB3, 0x73, 0xF3, 
+  0x0B, 0x8B, 0x4B, 0xCB, 0x2B, 0xAB, 0x6B, 0xEB, 0x1B, 0x9B, 0x5B, 0xDB, 0x3B, 0xBB, 0x7B, 0xFB,
+  0x07, 0x87, 0x47, 0xC7, 0x27, 0xA7, 0x67, 0xE7, 0x17, 0x97, 0x57, 0xD7, 0x37, 0xB7, 0x77, 0xF7, 
+  0x0F, 0x8F, 0x4F, 0xCF, 0x2F, 0xAF, 0x6F, 0xEF, 0x1F, 0x9F, 0x5F, 0xDF, 0x3F, 0xBF, 0x7F, 0xFF
+};
+
+__attribute__( ( always_inline ) ) static inline void SwapInt( int* a, int* b ) {
+    int Temp = *b;
+
+    *b = *a;
+    *a = Temp;
+}
+
+void IRAM_ATTR GDS_DrawPixelFast( struct GDS_Device* Device, int X, int Y, int Color ) {
+	Device->DrawPixelFast( Device, X, Y, Color );
+}
+
+void IRAM_ATTR GDS_DrawPixel( struct GDS_Device* Device, int X, int Y, int Color ) {
+	DrawPixel( Device, X, Y, Color );
+}
+
+void GDS_DrawHLine( struct GDS_Device* Device, int x, int y, int Width, int Color ) {
+    int XEnd = x + Width;
+
+	Device->Dirty = true;
+	
+	if (x < 0) x = 0;
+	if (XEnd >= Device->Width) XEnd = Device->Width - 1;
+	
+	if (y < 0) y = 0;
+	else if (y >= Device->Height) y = Device->Height - 1;
+
+    for ( ; x < XEnd; x++ ) Device->DrawPixelFast( Device, x, y, Color );
+}
+
+void GDS_DrawVLine( struct GDS_Device* Device, int x, int y, int Height, int Color ) {
+    int YEnd = y + Height;
+
+	Device->Dirty = true;
+	
+	if (x < 0) x = 0;
+	if (x >= Device->Width) x = Device->Width - 1;
+	
+	if (y < 0) y = 0;
+	else if (YEnd >= Device->Height) YEnd = Device->Height - 1;
+
+    for ( ; y < YEnd; y++ ) DrawPixel( Device, x, y, Color );
+}
+
+static inline void DrawWideLine( struct GDS_Device* Device, int x0, int y0, int x1, int y1, int Color ) {
+    int dx = ( x1 - x0 );
+    int dy = ( y1 - y0 );
+    int Error = 0;
+    int Incr = 1;
+    int x = x0;
+    int y = y0;
+
+    if ( dy < 0 ) {
+        Incr = -1;
+        dy = -dy;
+    }
+
+    Error = ( dy * 2 ) - dx;
+
+    for ( ; x < x1; x++ ) {
+        if ( IsPixelVisible( Device, x, y ) == true ) {
+            Device->DrawPixelFast( Device, x, y, Color );
+        }
+
+        if ( Error > 0 ) {
+            Error-= ( dx * 2 );
+            y+= Incr;
+        }
+
+        Error+= ( dy * 2 );
+    }
+}
+
+static inline void DrawTallLine( struct GDS_Device* Device, int x0, int y0, int x1, int y1, int Color ) {
+    int dx = ( x1 - x0 );
+    int dy = ( y1 - y0 );
+    int Error = 0;
+    int Incr = 1;
+    int x = x0;
+    int y = y0;
+
+    if ( dx < 0 ) {
+        Incr = -1;
+        dx = -dx;
+    }
+
+    Error = ( dx * 2 ) - dy;
+
+    for ( ; y < y1; y++ ) {
+        if ( IsPixelVisible( Device, x, y ) == true ) {
+            Device->DrawPixelFast( Device, x, y, Color );
+        }
+
+        if ( Error > 0 ) {
+            Error-= ( dy * 2 );
+            x+= Incr;
+        }
+
+        Error+= ( dx * 2 );
+    }
+}
+
+void GDS_DrawLine( struct GDS_Device* Device, int x0, int y0, int x1, int y1, int Color ) {
+    if ( x0 == x1 ) {
+        GDS_DrawVLine( Device, x0, y0, ( y1 - y0 ), Color );
+    } else if ( y0 == y1 ) {
+        GDS_DrawHLine( Device, x0, y0, ( x1 - x0 ), Color );
+    } else {
+		Device->Dirty = true;
+        if ( abs( x1 - x0 ) > abs( y1 - y0 ) ) {
+            /* Wide ( run > rise ) */
+            if ( x0 > x1 ) {
+                SwapInt( &x0, &x1 );
+                SwapInt( &y0, &y1 );
+            }
+
+            DrawWideLine( Device, x0, y0, x1, y1, Color );
+        } else {
+            /* Tall ( rise > run ) */
+            if ( y0 > y1 ) {
+                SwapInt( &y0, &y1 );
+                SwapInt( &x0, &x1 );
+            }
+
+            DrawTallLine( Device, x0, y0, x1, y1, Color );
+        }
+    }
+}
+
+void GDS_DrawBox( struct GDS_Device* Device, int x1, int y1, int x2, int y2, int Color, bool Fill ) {
+    int Width = ( x2 - x1 );
+    int Height = ( y2 - y1 );
+
+	Device->Dirty = true;
+	
+    if ( Fill == false ) {
+        /* Top side */
+        GDS_DrawHLine( Device, x1, y1, Width, Color );
+
+        /* Bottom side */
+        GDS_DrawHLine( Device, x1, y1 + Height, Width, Color );
+
+        /* Left side */
+        GDS_DrawVLine( Device, x1, y1, Height, Color );
+
+        /* Right side */
+        GDS_DrawVLine( Device, x1 + Width, y1, Height, Color );
+    } else {
+        /* Fill the box by drawing horizontal lines */
+        for ( ; y1 <= y2; y1++ ) {
+            GDS_DrawHLine( Device, x1, y1, Width, Color );
+        }
+    }
+}
+
+
+/****************************************************************************************
+ * Process graphic display data from column-oriented data (MSbit first)
+ */
+void GDS_DrawBitmapCBR(struct GDS_Device* Device, uint8_t *Data, int Width, int Height, int Color ) {
+	if (!Height) Height = Device->Height;
+	if (!Width) Width = Device->Width;
+		
+	if (Device->DrawBitmapCBR) {
+		Device->DrawBitmapCBR( Device, Data, Width, Height, Color );
+	} else if (Device->Depth == 1) {
+		
+		Height >>= 3;
+		
+		// need to do row/col swap and bit-reverse
+		for (int r = 0; r < Height; r++) {
+			uint8_t *optr = Device->Framebuffer + r*Device->Width, *iptr = Data + r;
+			for (int c = Width; --c >= 0;) {
+				*optr++ = BitReverseTable256[*iptr];
+				iptr += Height;
+			}	
+		}
+	} else if (Device->Depth == 4) {
+		uint8_t *optr = Device->Framebuffer;
+		int LineLen = Device->Width >> 1;
+		
+		Height >>= 3;
+		Color &= 0x0f;
+        
+        if (Device->HighNibble) {
+            for (int i = Width * Height, r = 0, c = 0; --i >= 0;) {
+                uint8_t Byte = BitReverseTable256[*Data++];
+                // we need to linearize code to let compiler better optimize
+                if (c & 0x01) {
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; 
+                } else {
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen;
+                }	
+            // end of a column, move to next one
+                if (++r == Height) { c++; r = 0; optr = Device->Framebuffer + (c >> 1); }		
+            }
+        } else {
+            for (int i = Width * Height, r = 0, c = 0; --i >= 0;) {
+                uint8_t Byte = BitReverseTable256[*Data++];
+                // we need to linearize code to let compiler better optimize
+                if (c & 0x01) {
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0x0f) | (((Byte & 0x01)*Color)<<4); optr += LineLen; 
+                } else {
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen; Byte >>= 1;
+                    *optr = (*optr & 0xf0) | (((Byte & 0x01)*Color)); optr += LineLen;
+                }	
+                // end of a column, move to next one
+                if (++r == Height) { c++; r = 0; optr = Device->Framebuffer + (c >> 1); }		
+            }
+		}
+	} else if (Device->Depth == 8) {
+		uint8_t *optr = Device->Framebuffer;
+		int LineLen = Device->Width;
+	
+		Height >>= 3;
+			
+		for (int i = Width * Height, r = 0, c = 0; --i >= 0;) {
+			uint8_t Byte = BitReverseTable256[*Data++];
+		
+			// we need to linearize code to let compiler better optimize
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; 
+
+			// end of a column, move to next one
+			if (++r == Height) { c++; r = 0; optr = Device->Framebuffer + c; }		
+		}	
+	} else if (Device->Depth == 16) {
+		uint16_t *optr = (uint16_t*) Device->Framebuffer;
+		int LineLen = Device->Width;
+	
+		Height >>= 3;
+		Color = __builtin_bswap16(Color);
+	
+		for (int i = Width * Height, r = 0, c = 0; --i >= 0;) {
+			uint8_t Byte = BitReverseTable256[*Data++];
+		
+			// we need to linearize code to let compiler better optimize
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			*optr = ((Byte & 0x01) * Color); optr += LineLen; 
+
+			// end of a column, move to next one
+			if (++r == Height) { c++; r = 0; optr = (uint16_t*) Device->Framebuffer + c; }		
+		}
+	} else if (Device->Depth == 24) {
+		uint8_t *optr = Device->Framebuffer;
+		int LineLen = Device->Width * 3;
+		
+		Height >>= 3;
+		if (Device->Mode == GDS_RGB666) Color = ((Color << 4) & 0xff0000) | ((Color << 2) & 0xff00) | (Color & 0x00ff);
+			
+		for (int i = Width * Height, r = 0, c = 0; --i >= 0;) {
+			uint8_t Byte = BitReverseTable256[*Data++];
+		
+			// we need to linearize code to let compiler better optimize		
+			#define SET24(O,D) O[0]=(D)>>16; O[1]=(D)>>8; O[2]=(D);
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen; Byte >>= 1;
+			SET24(optr,(Byte & 0x01) * Color); optr += LineLen;
+
+			// end of a column, move to next one
+			if (++r == Height) { c++; r = 0; optr = Device->Framebuffer + c * 3; }		
+		}
+	} else {
+		Height >>= 3;
+		
+		// don't know bitdepth, use brute-force solution
+		for (int i = Width * Height, r = 0, c = 0; --i >= 0;) {
+			uint8_t Byte = *Data++;
+			Device->DrawPixelFast( Device, c, (r << 3) + 7, (Byte & 0x01) * Color ); Byte >>= 1;
+			Device->DrawPixelFast( Device, c, (r << 3) + 6, (Byte & 0x01) * Color ); Byte >>= 1;
+			Device->DrawPixelFast( Device, c, (r << 3) + 5, (Byte & 0x01) * Color ); Byte >>= 1;
+			Device->DrawPixelFast( Device, c, (r << 3) + 4, (Byte & 0x01) * Color ); Byte >>= 1;
+			Device->DrawPixelFast( Device, c, (r << 3) + 3, (Byte & 0x01) * Color ); Byte >>= 1;
+			Device->DrawPixelFast( Device, c, (r << 3) + 2, (Byte & 0x01) * Color ); Byte >>= 1;
+			Device->DrawPixelFast( Device, c, (r << 3) + 1, (Byte & 0x01) * Color ); Byte >>= 1;
+			Device->DrawPixelFast( Device, c, (r << 3) + 0, (Byte & 0x01) * Color ); 
+			if (++r == Height) { c++; r = 0; }			
+		}
+		/* for better understanding, here is the mundane version 
+		for (int x = 0; x < Width; x++) {
+			for (int y = 0; y < Height; y++) {
+				uint8_t Byte = *Data++;
+				GDSDrawPixel4Fast( Device, x, y * 8 + 0, ((Byte >> 7) & 0x01) * Color );
+				GDSDrawPixel4Fast( Device, x, y * 8 + 1, ((Byte >> 6) & 0x01) * Color );
+				GDSDrawPixel4Fast( Device, x, y * 8 + 2, ((Byte >> 5) & 0x01) * Color );
+				GDSDrawPixel4Fast( Device, x, y * 8 + 3, ((Byte >> 4) & 0x01) * Color );
+				GDSDrawPixel4Fast( Device, x, y * 8 + 4, ((Byte >> 3) & 0x01) * Color );
+				GDSDrawPixel4Fast( Device, x, y * 8 + 5, ((Byte >> 2) & 0x01) * Color );
+				GDSDrawPixel4Fast( Device, x, y * 8 + 6, ((Byte >> 1) & 0x01) * Color );
+				GDSDrawPixel4Fast( Device, x, y * 8 + 7, ((Byte >> 0) & 0x01) * Color );
+			}
+		}
+		*/
+	}
+	
+	Device->Dirty = true;
+}

lib/display/core/gds_draw.h

@@ -0,0 +1,33 @@
+/* 
+ * (c) Philippe G. 2019, philippe_44@outlook.com
+ *
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ * 
+ */
+ 
+#ifndef _GDS_DRAW_H_
+#define _GDS_DRAW_H_
+
+#include <stdint.h>
+#include <stdbool.h>
+#include "esp_attr.h"
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+void IRAM_ATTR GDS_DrawPixelFast( struct GDS_Device* Device, int X, int Y, int Color );
+void IRAM_ATTR GDS_DrawPixel( struct GDS_Device* Device, int X, int Y, int Color );
+void GDS_DrawHLine( struct GDS_Device* Device, int x, int y, int Width, int Color );
+void GDS_DrawVLine( struct GDS_Device* Device, int x, int y, int Height, int Color );
+void GDS_DrawLine( struct GDS_Device* Device, int x0, int y0, int x1, int y1, int Color );
+void GDS_DrawBox( struct GDS_Device* Device, int x1, int y1, int x2, int y2, int Color, bool Fill );
+// draw a bitmap with source 1-bit depth organized in column and col0 = bit7 of byte 0 
+void GDS_DrawBitmapCBR( struct GDS_Device* Device, uint8_t *Data, int Width, int Height, int Color);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

lib/display/core/gds_err.h

@@ -0,0 +1,39 @@
+#ifndef _GDS_ERR_H_
+#define _GDS_ERR_H_
+
+#include <esp_log.h>
+
+#define GDS_DoAbort( )
+
+#if ! defined NullCheck
+    #define NullCheck( ptr, retexpr ) { \
+        if ( ptr == NULL ) { \
+            ESP_LOGE( __FUNCTION__, "%s == NULL", #ptr ); \
+            GDS_DoAbort( ); \
+            retexpr; \
+        } \
+    }
+#endif
+
+#if ! defined ESP_ERROR_CHECK_NONFATAL
+    #define ESP_ERROR_CHECK_NONFATAL( expr, retexpr ) { \
+        esp_err_t __err_rc = ( expr ); \
+        if ( __err_rc != ESP_OK ) { \
+            ESP_LOGE( __FUNCTION__, "%s != ESP_OK, result: %d", #expr, __err_rc ); \
+            GDS_DoAbort( ); \
+            retexpr; \
+        } \
+    }
+#endif
+
+#if ! defined CheckBounds
+    #define CheckBounds( expr, retexpr ) { \
+        if ( expr ) { \
+            ESP_LOGE( __FUNCTION__, "Line %d: %s", __LINE__, #expr ); \
+            GDS_DoAbort( ); \
+            retexpr; \
+        } \
+    }
+#endif
+
+#endif

lib/display/core/gds_font.c

@@ -0,0 +1,272 @@
+/**
+ * Copyright (c) 2017-2018 Tara Keeling
+ * 
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <math.h>
+#include "gds_private.h"
+#include "gds.h"
+#include "gds_font.h"
+#include "gds_draw.h"
+#include "gds_err.h"
+
+static int RoundUpFontHeight( const struct GDS_FontDef* Font ) {
+    int Height = Font->Height;
+
+    if ( ( Height % 8 ) != 0 ) {
+        return ( ( Height + 7 ) / 8 ) * 8;
+    }
+
+    return Height;
+}
+
+static const uint8_t* GetCharPtr( const struct GDS_FontDef* Font, char Character ) {
+    return &Font->FontData[ ( Character - Font->StartChar ) * ( ( Font->Width * ( RoundUpFontHeight( Font ) / 8 ) ) + 1 ) ];
+}
+
+void GDS_FontDrawChar( struct GDS_Device* Device, char Character, int x, int y, int Color ) {
+    const uint8_t* GlyphData = NULL;
+    int GlyphColumnLen = 0;
+    int CharStartX =  0;
+    int CharStartY = 0;
+    int CharWidth = 0;
+    int CharHeight = 0;
+    int CharEndX = 0;
+    int CharEndY = 0;
+    int OffsetX = 0;
+    int OffsetY = 0;
+    int YByte = 0;
+    int YBit = 0;
+    int i = 0;
+
+    NullCheck( ( GlyphData = GetCharPtr( Device->Font, Character ) ), return );
+
+    if ( Character >= Device->Font->StartChar && Character <= Device->Font->EndChar ) {
+        /* The first byte in the glyph data is the width of the character in pixels, skip over */
+        GlyphData++;
+        GlyphColumnLen = RoundUpFontHeight( Device->Font ) / 8;
+        
+        CharWidth = GDS_FontGetCharWidth( Device, Character );
+        CharHeight = GDS_FontGetHeight( Device );
+
+        CharStartX = x;
+        CharStartY = y;
+        
+        CharEndX = CharStartX + CharWidth;
+        CharEndY = CharStartY + CharHeight;
+
+        /* If the character is partially offscreen offset the end by
+        * distance between (coord) and 0.
+        */
+        OffsetX = ( CharStartX < 0 ) ? abs( CharStartX ) : 0;
+        OffsetY = ( CharStartY < 0 ) ? abs( CharStartY ) : 0;
+
+        /* This skips into the proper column within the glyph data */
+        GlyphData+= ( OffsetX * GlyphColumnLen );
+
+        CharStartX+= OffsetX;
+        CharStartY+= OffsetY;
+
+        /* Do not attempt to draw if this character is entirely offscreen */
+        if ( CharEndX < 0 || CharStartX >= Device->TextWidth || CharEndY < 0 || CharStartY >= Device->Height ) {
+            ClipDebug( x, y );
+            return;
+        }
+
+        /* Do not attempt to draw past the end of the screen */
+        CharEndX = ( CharEndX >= Device->TextWidth ) ? Device->TextWidth - 1 : CharEndX;
+        CharEndY = ( CharEndY >= Device->Height ) ? Device->Height - 1 : CharEndY;
+		Device->Dirty = true;
+
+        for ( x = CharStartX; x < CharEndX; x++ ) {
+            for ( y = CharStartY, i = 0; y < CharEndY && i < CharHeight; y++, i++ ) {
+                YByte = ( i + OffsetY ) / 8;
+                YBit = ( i + OffsetY ) & 0x07;
+
+                if ( GlyphData[ YByte ] & BIT( YBit ) ) {
+                    DrawPixel( Device, x, y, Color );
+                }            
+            }
+
+            GlyphData+= GlyphColumnLen;
+        }
+    }
+}
+
+const struct GDS_FontDef* GDS_SetFont( struct GDS_Device* Display, const struct GDS_FontDef* Font ) {
+	const struct GDS_FontDef* OldFont = Display->Font;
+
+    Display->FontForceProportional = false;
+    Display->FontForceMonospace = false;
+    Display->Font = Font;
+
+    return OldFont;
+}
+
+void GDS_FontForceProportional( struct GDS_Device* Display, bool Force ) {
+    Display->FontForceProportional = Force;
+}
+
+void GDS_FontForceMonospace( struct GDS_Device* Display, bool Force ) {
+    Display->FontForceMonospace = Force;
+}
+
+int GDS_FontGetWidth( struct GDS_Device* Display ) {
+    return Display->Font->Width;
+}
+
+int GDS_FontGetHeight( struct GDS_Device* Display ) {
+    return Display->Font->Height;
+}
+
+int GDS_FontGetCharWidth( struct GDS_Device* Display, char Character ) {
+    const uint8_t* CharPtr = NULL;
+    int Width = 0;
+
+    if ( Character >= Display->Font->StartChar && Character <= Display->Font->EndChar ) {
+        CharPtr = GetCharPtr( Display->Font, Character );
+
+        Width = ( Display->Font->Monospace == true ) ? Display->Font->Width : *CharPtr;
+
+        if ( Display->FontForceMonospace == true ) {
+            Width = Display->Font->Width;
+        }
+
+        if ( Display->FontForceProportional == true ) {
+            Width = *CharPtr;
+        }
+    }
+
+    return Width;
+}
+
+int GDS_FontGetMaxCharsPerRow( struct GDS_Device* Display ) {
+    return Display->TextWidth / Display->Font->Width;
+}
+
+int GDS_FontGetMaxCharsPerColumn( struct GDS_Device* Display ) {
+    return Display->Height / Display->Font->Height;    
+}
+
+int GDS_FontGetCharHeight( struct GDS_Device* Display ) {
+    return Display->Font->Height;
+}
+
+int GDS_FontMeasureString( struct GDS_Device* Display, const char* Text ) {
+    int Width = 0;
+    int Len = 0;
+
+    NullCheck( Text, return 0 );
+
+    for ( Len = strlen( Text ); Len >= 0; Len--, Text++ ) {
+        if ( *Text >= Display->Font->StartChar && *Text <= Display->Font->EndChar ) {
+            Width+= GDS_FontGetCharWidth( Display, *Text );
+        }
+    }
+
+    return Width;
+}
+
+void GDS_FontDrawString( struct GDS_Device* Display, int x, int y, const char* Text, int Color ) {
+    int Len = 0;
+    int i = 0;
+
+    NullCheck( Text, return );
+
+    for ( Len = strlen( Text ), i = 0; i < Len; i++ ) {
+        GDS_FontDrawChar( Display, *Text, x, y, Color );
+
+        x+= GDS_FontGetCharWidth( Display, *Text );
+        Text++;
+    }
+}
+
+void GDS_FontDrawAnchoredString( struct GDS_Device* Display, TextAnchor Anchor, const char* Text, int Color ) {
+    int x = 0;
+    int y = 0;
+
+    NullCheck( Text, return );
+
+    GDS_FontGetAnchoredStringCoords( Display, &x, &y, Anchor, Text );
+    GDS_FontDrawString( Display, x, y, Text, Color );
+}
+
+void GDS_FontGetAnchoredStringCoords( struct GDS_Device* Display, int* OutX, int* OutY, TextAnchor Anchor, const char* Text ) {
+    int StringWidth = 0;
+    int StringHeight = 0;
+
+    NullCheck( OutX, return );
+    NullCheck( OutY, return );
+    NullCheck( Text, return );
+
+    StringWidth = GDS_FontMeasureString( Display, Text );
+    StringHeight = GDS_FontGetCharHeight( Display );
+
+    switch ( Anchor ) {
+        case TextAnchor_East: {
+            *OutY = ( Display->Height / 2 ) - ( StringHeight / 2 );
+            *OutX = ( Display->TextWidth - StringWidth );
+
+            break;
+        }
+        case TextAnchor_West: {
+            *OutY = ( Display->Height / 2 ) - ( StringHeight / 2 );
+            *OutX = 0;
+
+            break;
+        }
+        case TextAnchor_North: {
+            *OutX = ( Display->TextWidth / 2 ) - ( StringWidth / 2 );
+            *OutY = 0;
+
+            break;
+        }
+        case TextAnchor_South: {
+            *OutX = ( Display->TextWidth / 2 ) - ( StringWidth / 2 );
+            *OutY = ( Display->Height - StringHeight );
+            
+            break;
+        }
+        case TextAnchor_NorthEast: {
+            *OutX = ( Display->TextWidth - StringWidth );
+            *OutY = 0;
+
+            break;
+        }
+        case TextAnchor_NorthWest: {
+            *OutY = 0;
+            *OutX = 0;
+
+            break;
+        }
+        case TextAnchor_SouthEast: {
+            *OutY = ( Display->Height - StringHeight );
+            *OutX = ( Display->TextWidth - StringWidth );
+
+            break;
+        }
+        case TextAnchor_SouthWest: {
+            *OutY = ( Display->Height - StringHeight );
+            *OutX = 0;
+
+            break;
+        }
+        case TextAnchor_Center: {
+            *OutY = ( Display->Height / 2 ) - ( StringHeight / 2 );
+            *OutX = ( Display->TextWidth / 2 ) - ( StringWidth / 2 );
+
+            break;
+        }
+        default: {
+            *OutX = 128;
+            *OutY = 64;
+            
+            break;
+        }
+    };
+}

lib/display/core/gds_font.h

@@ -0,0 +1,92 @@
+#ifndef _GDS_FONT_H_
+#define _GDS_FONT_H_
+
+#include <stdint.h>
+#include <stdbool.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+struct GDS_Device;
+
+/* 
+ * X-GLCD Font format:
+ *
+ * First byte of glyph is it's width in pixels.
+ * Each data byte represents 8 pixels going down from top to bottom.
+ * 
+ * Example glyph layout for a 16x16 font
+ * 'a': [Glyph width][Pixel column 0][Pixel column 1] where the number of pixel columns is the font height divided by 8
+ * 'b': [Glyph width][Pixel column 0][Pixel column 1]...
+ * 'c': And so on...
+ */
+ 
+struct GDS_FontDef {
+    const uint8_t* FontData;
+
+    int Width;
+    int Height;
+
+    int StartChar;
+    int EndChar;
+
+    bool Monospace;
+};
+
+typedef enum {
+    TextAnchor_East = 0,
+    TextAnchor_West,
+    TextAnchor_North,
+    TextAnchor_South,
+    TextAnchor_NorthEast,
+    TextAnchor_NorthWest,
+    TextAnchor_SouthEast,
+    TextAnchor_SouthWest,
+    TextAnchor_Center
+} TextAnchor;
+
+const struct GDS_FontDef* GDS_SetFont( struct GDS_Device* Display, const struct GDS_FontDef* Font );
+
+void GDS_FontForceProportional( struct GDS_Device* Display, bool Force );
+void GDS_FontForceMonospace( struct GDS_Device* Display, bool Force );
+
+int GDS_FontGetWidth( struct GDS_Device* Display );
+int GDS_FontGetHeight( struct GDS_Device* Display );
+
+int GDS_FontGetMaxCharsPerRow( struct GDS_Device* Display );
+int GDS_FontGetMaxCharsPerColumn( struct GDS_Device* Display );
+
+int GDS_FontGetCharWidth( struct GDS_Device* Display, char Character );
+int GDS_FontGetCharHeight( struct GDS_Device* Display );
+int GDS_FontMeasureString( struct GDS_Device* Display, const char* Text );
+int GDS_FontMeasureStringLine( struct GDS_Device* Display, int Line, const char* Text );
+
+void GDS_FontDrawChar( struct GDS_Device* Display, char Character, int x, int y, int Color );
+void GDS_FontDrawString( struct GDS_Device* Display, int x, int y, const char* Text, int Color );
+void GDS_FontDrawAnchoredString( struct GDS_Device* Display, TextAnchor Anchor, const char* Text, int Color );
+void GDS_FontGetAnchoredStringCoords( struct GDS_Device* Display, int* OutX, int* OutY, TextAnchor Anchor, const char* Text );
+
+extern const struct GDS_FontDef Font_droid_sans_fallback_11x13;
+extern const struct GDS_FontDef Font_droid_sans_fallback_15x17;
+extern const struct GDS_FontDef Font_droid_sans_fallback_24x28;
+
+extern const struct GDS_FontDef Font_droid_sans_mono_7x13;
+extern const struct GDS_FontDef Font_droid_sans_mono_13x24;
+extern const struct GDS_FontDef Font_droid_sans_mono_16x31;
+
+extern const struct GDS_FontDef Font_liberation_mono_9x15;
+extern const struct GDS_FontDef Font_liberation_mono_13x21;
+extern const struct GDS_FontDef Font_liberation_mono_17x30;
+
+extern const struct GDS_FontDef Font_Tarable7Seg_16x32;
+extern const struct GDS_FontDef Font_Tarable7Seg_32x64;
+
+extern const struct GDS_FontDef Font_line_1;
+extern const struct GDS_FontDef Font_line_2;
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

lib/display/core/gds_image.c

@@ -0,0 +1,442 @@
+/* 
+ * (c) Philippe G. 2019, philippe_44@outlook.com
+ *
+ * This software is released under the MIT License.
+ * https://opensource.org/licenses/MIT
+ * 
+ */
+ 
+#include <string.h>
+#include "math.h"
+#ifdef TJPGD_ROM
+#include "esp32/rom/tjpgd.h"
+#else
+#include "tjpgd.h"
+#endif
+#include "esp_log.h"
+
+#include "gds.h"
+#include "gds_private.h"
+#include "gds_image.h"
+
+const static char TAG[] = "ImageDec";
+
+#define SCRATCH_SIZE	3100
+
+//Data that is passed from the decoder function to the infunc/outfunc functions.
+typedef struct {
+    const unsigned char *InData;	// Pointer to jpeg data
+    int InPos;						// Current position in jpeg data
+	int Width, Height;	
+	uint8_t Mode;
+	union {
+		void *OutData;
+		struct {						// DirectDraw
+			struct GDS_Device *Device;
+			int XOfs, YOfs;
+			int XMin, YMin;
+			int Depth;
+		};	
+	};	
+} JpegCtx;
+
+/****************************************************************************************
+ * RGB conversion (24 bits 888: RRRRRRRRGGGGGGGGBBBBBBBB and 16 bits 565: RRRRRGGGGGGBBBBB = B31..B0)
+ * so in other words for an array of 888 bytes: [0]=B, [1]=G, [2]=R, ...
+ * monochrome (0.2125 * color.r) + (0.7154 * color.g) + (0.0721 * color.b)
+ * grayscale (0.3 * R) + (0.59 * G) + (0.11 * B) )
+ */
+static inline int Scaler332(uint8_t *Pixels) {
+	return (Pixels[2] & ~0x1f) | ((Pixels[1] & ~0x1f) >> 3) | (Pixels[0] >> 6);
+}
+
+static inline int Scaler444(uint8_t *Pixels) {
+	return ((Pixels[2] & ~0x0f) << 4) | (Pixels[1] & ~0x0f) | (Pixels[0] >> 4);
+}
+
+static inline int Scaler555(uint8_t *Pixels) {
+	return ((Pixels[2] & ~0x07) << 7) | ((Pixels[1] & ~0x07) << 2) | (Pixels[0] >> 3);
+}
+
+static inline int Scaler565(uint8_t *Pixels) {
+	return ((Pixels[2] & ~0x07) << 8) | ((Pixels[1] & ~0x03) << 3) | (Pixels[0] >> 3);
+}
+
+static inline int Scaler666(uint8_t *Pixels) {
+	return ((Pixels[2] & ~0x03) << 10) | ((Pixels[1] & ~0x03) << 4) | (Pixels[0] >> 2);
+}
+
+static inline int Scaler888(uint8_t *Pixels) {
+	return (Pixels[2] << 16) | (Pixels[1] << 8) | Pixels[0];
+}
+
+static inline int ScalerGray(uint8_t *Pixels) {
+	return (Pixels[2] * 14 + Pixels[1] * 76 + Pixels[0] * 38) >> 7;
+}
+
+static unsigned InHandler(JDEC *Decoder, uint8_t *Buf, unsigned Len) {
+    JpegCtx *Context = (JpegCtx*) Decoder->device;
+    if (Buf) memcpy(Buf, Context->InData +  Context->InPos, Len);
+    Context->InPos += Len;
+    return Len;
+}
+
+#define OUTHANDLER(F)										\
+	for (int y = Frame->top; y <= Frame->bottom; y++) {		\
+		for (int x = Frame->left; x <= Frame->right; x++) {	\
+			OutData[Context->Width * y + x] = F(Pixels);	\
+			Pixels += 3;									\
+		}													\
+	}	
+	
+#define OUTHANDLER24(F)													\
+	for (int y = Frame->top; y <= Frame->bottom; y++) {					\
+		uint8_t *p = OutData + (Context->Width * y + Frame->left) * 3;	\
+		for (int c = Frame->right - Frame->left; c-- >= 0;) {			\
+			uint32_t v = F(Pixels);										\
+			*p++ = v; *p++ = v >> 8; *p++ = v >> 16;					\
+			Pixels += 3;												\
+		}																\
+	}		
+
+static unsigned OutHandler(JDEC *Decoder, void *Bitmap, JRECT *Frame) {
+	JpegCtx *Context = (JpegCtx*) Decoder->device;
+    uint8_t *Pixels = (uint8_t*) Bitmap;
+
+	// decoded image is RGB888
+	if (Context->Mode == GDS_RGB888) {
+		uint8_t *OutData = (uint8_t*) Context->OutData;		
+		OUTHANDLER24(Scaler888);
+	} else if (Context->Mode == GDS_RGB666) {
+		uint8_t *OutData = (uint8_t*) Context->OutData;		
+		OUTHANDLER24(Scaler666);		
+	} else if (Context->Mode == GDS_RGB565) {
+		uint16_t *OutData = (uint16_t*) Context->OutData;
+		OUTHANDLER(Scaler565);		
+	} else if (Context->Mode == GDS_RGB555) {
+		uint16_t *OutData = (uint16_t*) Context->OutData;
+		OUTHANDLER(Scaler555);				
+	} else if (Context->Mode == GDS_RGB444) {
+		uint16_t *OutData = (uint16_t*) Context->OutData;
+		OUTHANDLER(Scaler444);						
+	} else if (Context->Mode == GDS_RGB332) {
+		uint8_t *OutData = (uint8_t*) Context->OutData;
+		OUTHANDLER(Scaler332);						
+	} else if (Context->Mode <= GDS_GRAYSCALE) { 	 
+		uint8_t *OutData = (uint8_t*) Context->OutData;		
+		OUTHANDLER(ScalerGray);
+	}
+    
+    return 1;
+}
+
+// Convert the RGB888 to destination color plane, use DrawPixel and not "fast" 
+// version as X,Y may be beyond screen				
+#define OUTHANDLERDIRECT(F,S)																		\
+	for (int y = Frame->top; y <= Frame->bottom; y++) {												\
+		if (y < Context->YMin) continue;															\
+		for (int x = Frame->left; x <= Frame->right; x++) {											\
+			if (x < Context->XMin) continue;														\
+			DrawPixel( Context->Device, x + Context->XOfs, y + Context->YOfs, F(Pixels) >> S);	\
+			Pixels += 3;																			\
+		}																							\
+	}
+	
+static unsigned OutHandlerDirect(JDEC *Decoder, void *Bitmap, JRECT *Frame) {
+	JpegCtx *Context = (JpegCtx*) Decoder->device;
+    uint8_t *Pixels = (uint8_t*) Bitmap;
+	int Shift = 8 - Context->Depth;
+	
+	// decoded image is RGB888, shift only make sense for grayscale
+	if (Context->Mode == GDS_RGB888) {
+		OUTHANDLERDIRECT(Scaler888, 0);
+	} else if (Context->Mode == GDS_RGB666) {
+		OUTHANDLERDIRECT(Scaler666, 0);		
+	} else if (Context->Mode == GDS_RGB565) {
+		OUTHANDLERDIRECT(Scaler565, 0);		
+	} else if (Context->Mode == GDS_RGB555) {
+		OUTHANDLERDIRECT(Scaler555, 0);				
+	} else if (Context->Mode == GDS_RGB444) {
+		OUTHANDLERDIRECT(Scaler444, 0);						
+	} else if (Context->Mode == GDS_RGB332) {
+		OUTHANDLERDIRECT(Scaler332, 0);						
+	} else if (Context->Mode <= GDS_GRAYSCALE) { 	 
+		OUTHANDLERDIRECT(ScalerGray, Shift);
+	}
+    
+    return 1;
+}
+
+//Decode the embedded image into pixel lines that can be used with the rest of the logic.
+static void* DecodeJPEG(uint8_t *Source, int *Width, int *Height, float Scale, bool SizeOnly, int RGB_Mode) {
+    JDEC Decoder;
+    JpegCtx Context;
+	char *Scratch = malloc(SCRATCH_SIZE);
+	
+    if (!Scratch) {
+        ESP_LOGE(TAG, "Cannot allocate workspace");
+        return NULL;
+    }
+
+	Context.OutData = NULL;
+    Context.InData = Source;
+    Context.InPos = 0;
+	        
+    //Prepare and decode the jpeg.
+    int Res = jd_prepare(&Decoder, InHandler, Scratch, SCRATCH_SIZE, (void*) &Context);
+	if (Width) *Width = Decoder.width;
+	if (Height) *Height = Decoder.height;
+	Decoder.scale = Scale;
+
+    if (Res == JDR_OK && !SizeOnly) {
+		if (RGB_Mode <= GDS_RGB332) Context.OutData = malloc(Decoder.width * Decoder.height);
+		else if (RGB_Mode < GDS_RGB666) Context.OutData = malloc(Decoder.width * Decoder.height * 2);
+		else if (RGB_Mode <= GDS_RGB888) Context.OutData = malloc(Decoder.width * Decoder.height * 3);
+		
+		// find the scaling factor
+		uint8_t N = 0, ScaleInt =  ceil(1.0 / Scale);
+		ScaleInt--; ScaleInt |= ScaleInt >> 1; ScaleInt |= ScaleInt >> 2; ScaleInt++;
+		while (ScaleInt >>= 1) N++;
+		if (N > 3) {
+			ESP_LOGW(TAG, "Image will not fit %dx%d", Decoder.width, Decoder.height);
+			N = 3;
+		}	
+		
+		// ready to decode		
+		if (Context.OutData) {
+			Context.Width = Decoder.width / (1 << N);
+			Context.Height = Decoder.height / (1 << N);
+			Context.Mode = RGB_Mode;
+			if (Width) *Width = Context.Width;
+			if (Height) *Height = Context.Height;
+			Res = jd_decomp(&Decoder, OutHandler, N);
+			if (Res != JDR_OK) {
+				ESP_LOGE(TAG, "Image decoder: jd_decode failed (%d)", Res);
+			}	
+		} else {
+			ESP_LOGE(TAG, "Can't allocate bitmap %dx%d or invalid mode %d", Decoder.width, Decoder.height, RGB_Mode);			
+		}	
+	} else if (!SizeOnly) {
+        ESP_LOGE(TAG, "Image decoder: jd_prepare failed (%d)", Res);
+    }    
+
+	// free scratch area
+    if (Scratch) free(Scratch);
+    return Context.OutData;
+}
+
+void* GDS_DecodeJPEG(uint8_t *Source, int *Width, int *Height, float Scale, int RGB_Mode) {
+	return DecodeJPEG(Source, Width, Height, Scale, false, RGB_Mode);
+}	
+
+void GDS_GetJPEGSize(uint8_t *Source, int *Width, int *Height) {
+	DecodeJPEG(Source, Width, Height, 1, true, -1);
+}	
+
+/****************************************************************************************
+ * RGB conversion (24 bits: RRRRRRRRGGGGGGGGBBBBBBBB and 16 bits 565: RRRRRGGGGGGBBBBB = B31..B0)
+ * so in other words for an array of 888 bytes: [0]=B, [1]=G, [2]=R, ...
+ * monochrome (0.2125 * color.r) + (0.7154 * color.g) + (0.0721 * color.b)
+ * grayscale (0.3 * R) + (0.59 * G) + (0.11 * B) )
+ */
+ 
+static inline int ToGray888(uint8_t **Pixel) {
+	uint32_t v = *(*Pixel)++; v |= *(*Pixel)++ << 8; v |= *(*Pixel)++ << 16; 
+	return (((v & 0xff) * 14) + ((v >> 8) & 0xff) * 76 + ((v >> 16) * 38) + 1) >> 7;
+} 
+ 
+static inline int ToGray666(uint8_t **Pixel) {
+	uint32_t v = *(*Pixel)++; v |= *(*Pixel)++ << 8; v |= *(*Pixel)++ << 16; 
+	return (((v & 0x3f) * 14) + ((v >> 6) & 0x3f) * 76 + ((v >> 12) * 38) + 1) >> 7;
+}
+
+static inline int ToGray565(uint16_t **Pixel) {
+	uint16_t v = *(*Pixel)++; 
+	return ((((v & 0x1f) * 14) << 1) + ((v >> 5) & 0x3f) * 76 + (((v >> 11) * 38) << 1) + 1) >> 7;
+}
+
+static inline int ToGray555(uint16_t **Pixel) {
+	uint16_t v = *(*Pixel)++; 
+	return ((v & 0x1f) * 14 + ((v >> 5) & 0x1f) * 76 + (v >> 10) * 38) >> 7;
+}
+
+static inline int ToGray444(uint16_t **Pixel) {
+	uint16_t v = *(*Pixel)++; 
+	return ((v & 0x0f) * 14 + ((v >> 4) & 0x0f) * 76 + (v >> 8) * 38) >> 7;
+}
+
+static inline int ToGray332(uint8_t **Pixel) {
+	uint8_t v = *(*Pixel)++; 
+	return ((((v & 0x3) * 14) << 1) + ((v >> 2) & 0x7) * 76 + (v >> 5) * 38 + 1) >> 7;
+}
+
+static inline int ToSelf(uint8_t **Pixel) {
+	return *(*Pixel)++; 
+}
+	
+#define DRAW_GRAYRGB(S,F)													\
+	if (Scale > 0) {														\
+		for (int r = 0; r < Height; r++) {									\
+			for (int c = 0; c < Width; c++) {								\
+				DrawPixel( Device, c + x, r + y, F(S) >> Scale);		\
+			}																\
+		}																	\
+	} else {																\
+		for (int r = 0; r < Height; r++) {									\
+			for (int c = 0; c < Width; c++) {								\
+				DrawPixel( Device, c + x, r + y, F(S) << -Scale);	\
+			}																\
+		}																	\
+	}									
+	
+#define DRAW_RGB(T)										\
+	T *S = (T*) Image;									\
+	for (int r = 0; r < Height; r++) {					\
+		for (int c = 0; c < Width; c++) {				\
+			DrawPixel(Device, c + x, r + y, *S++);	\
+		}												\
+	}																	
+	
+#define DRAW_RGB24													\
+	uint8_t *S = (uint8_t*) Image;									\
+	for (int r = 0; r < Height; r++) {								\
+		for (int c = 0; c < Width; c++) {							\
+			uint32_t v = *S++; v |= *S++ << 8; v |= *S++ << 16;		\
+			DrawPixel(Device, c + x, r + y, v);					\
+		}															\
+	}	
+
+/****************************************************************************************
+ *  Decode the embedded image into pixel lines that can be used with the rest of the logic.
+ */
+void GDS_DrawRGB( struct GDS_Device* Device, uint8_t *Image, int x, int y, int Width, int Height, int RGB_Mode ) {
+
+	// don't do anything if driver supplies a draw function
+	if (Device->DrawRGB) {
+		Device->DrawRGB( Device, Image, x, y, Width, Height, RGB_Mode );
+		Device->Dirty = true;	
+		return;
+	}
+	
+	// RGB type displays
+	if (Device->Mode > GDS_GRAYSCALE) {
+		// image must match the display mode!
+		if (Device->Mode != RGB_Mode) {
+			ESP_LOGE(TAG, "non-matching display & image mode %u %u", Device->Mode, RGB_Mode);
+			return;
+		}	
+		
+		if (RGB_Mode == GDS_RGB332) {
+			DRAW_RGB(uint8_t);
+		} else if (RGB_Mode < GDS_RGB666) {
+			DRAW_RGB(uint16_t);
+		} else {
+			DRAW_RGB24;
+		}	
+		
+		Device->Dirty = true;
+		return;
+	}
+	
+	// set the right scaler when displaying grayscale
+	if (RGB_Mode <= GDS_GRAYSCALE) {
+		int Scale = 8 - Device->Depth;
+		DRAW_GRAYRGB(&Image,ToSelf);
+	} else if (RGB_Mode == GDS_RGB332) {
+		int Scale = 3 - Device->Depth;		
+		DRAW_GRAYRGB(&Image,ToGray332);
+	} else if (RGB_Mode < GDS_RGB666)	{
+		if (RGB_Mode == GDS_RGB565) {
+			int Scale = 6 - Device->Depth;
+			DRAW_GRAYRGB((uint16_t**)&Image,ToGray565);
+		} else if (RGB_Mode == GDS_RGB555) {
+			int Scale = 5 - Device->Depth;
+			DRAW_GRAYRGB((uint16_t**)&Image,ToGray555);
+		} else if (RGB_Mode == GDS_RGB444) {
+			int Scale = 4 - Device->Depth; 
+			DRAW_GRAYRGB((uint16_t**)&Image,ToGray444)
+		}	
+	} else {
+		if (RGB_Mode == GDS_RGB666) {
+			int Scale = 6 - Device->Depth;
+			DRAW_GRAYRGB(&Image,ToGray666);
+		} else if (RGB_Mode == GDS_RGB888) {
+			int Scale = 8 - Device->Depth;
+			DRAW_GRAYRGB(&Image,ToGray888);
+		}	
+	} 
+	
+	Device->Dirty = true;	
+}
+
+/****************************************************************************************
+ *  Decode the embedded image into pixel lines that can be used with the rest of the logic.
+ */
+bool GDS_DrawJPEG(struct GDS_Device* Device, uint8_t *Source, int x, int y, int Fit) {
+    JDEC Decoder;
+    JpegCtx Context;
+	bool Ret = false;
+	char *Scratch = malloc(SCRATCH_SIZE);
+	
+    if (!Scratch) {
+        ESP_LOGE(TAG, "Cannot allocate workspace");
+        return NULL;
+    }
+
+    // Populate fields of the JpegCtx struct.
+    Context.InData = Source;
+    Context.InPos = 0;
+	Context.XOfs = x;
+	Context.YOfs = y;
+	Context.Device = Device;
+	Context.Depth = Device->Depth;
+        
+    //Prepare and decode the jpeg.
+    int Res = jd_prepare(&Decoder, InHandler, Scratch, SCRATCH_SIZE, (void*) &Context);
+	Context.Width = Decoder.width;
+	Context.Height = Decoder.height;
+	
+    if (Res == JDR_OK) {
+		uint8_t N = 0;
+		
+		// do we need to fit the image
+		if (Fit & GDS_IMAGE_FIT) {
+			float XRatio = (Device->Width - x) / (float) Decoder.width, YRatio = (Device->Height - y) / (float) Decoder.height;
+			uint8_t Ratio = XRatio < YRatio ? ceil(1/XRatio) : ceil(1/YRatio);
+			Ratio--; Ratio |= Ratio >> 1; Ratio |= Ratio >> 2; Ratio++;
+			while (Ratio >>= 1) N++;
+			if (N > 3) {
+				ESP_LOGW(TAG, "Image will not fit %dx%d", Decoder.width, Decoder.height);
+				N = 3;
+			}	
+			Context.Width /= 1 << N;
+			Context.Height /= 1 << N;
+		} 
+		
+		// then place it
+		if (Fit & GDS_IMAGE_CENTER_X) Context.XOfs = (Device->Width + x - Context.Width) / 2;
+		else if (Fit & GDS_IMAGE_RIGHT) Context.XOfs = Device->Width - Context.Width;
+		if (Fit & GDS_IMAGE_CENTER_Y) Context.YOfs = (Device->Height + y - Context.Height) / 2;
+		else if (Fit & GDS_IMAGE_BOTTOM) Context.YOfs = Device->Height - Context.Height;
+
+		Context.XMin = x - Context.XOfs;
+		Context.YMin = y - Context.YOfs;
+		Context.Mode = Device->Mode;
+					
+		// do decompress & draw
+		Res = jd_decomp(&Decoder, OutHandlerDirect, N);
+		if (Res == JDR_OK) {
+			Device->Dirty = true;
+			Ret = true;
+		} else {	
+			ESP_LOGE(TAG, "Image decoder: jd_decode failed (%d)", Res);
+		}	
+	} else {	
+        ESP_LOGE(TAG, "Image decoder: jd_prepare failed (%d)", Res);
+    }    
+      
+	// free scratch area
+    if (Scratch) free(Scratch);
+	return Ret;
+}
+