ffvoice 0.4.5__tar.gz

ffvoice-0.4.5/CMakeLists.txt

@@ -0,0 +1,412 @@
+# ffvoice-engine - Low-latency C++ voice engine
+cmake_minimum_required(VERSION 3.20)
+project(ffvoice-engine VERSION 0.1.0 LANGUAGES CXX C)
+
+# C++ Standard
+set(CMAKE_CXX_STANDARD 20)
+set(CMAKE_CXX_STANDARD_REQUIRED ON)
+set(CMAKE_CXX_EXTENSIONS OFF)
+set(CMAKE_EXPORT_COMPILE_COMMANDS ON)
+
+# Set architecture for macOS (allow override from setup.py for universal2 builds)
+if(APPLE AND NOT DEFINED CMAKE_OSX_ARCHITECTURES)
+    # Default to native architecture if not specified
+    set(CMAKE_OSX_ARCHITECTURES "${CMAKE_SYSTEM_PROCESSOR}" CACHE STRING "macOS architecture" FORCE)
+endif()
+
+# Build options
+option(BUILD_TESTS "Build unit tests" ON)
+option(BUILD_EXAMPLES "Build examples" ON)
+option(BUILD_PYTHON "Build Python bindings" OFF)
+option(ENABLE_WEBRTC_APM "Enable WebRTC Audio Processing Module" OFF)
+option(ENABLE_RNNOISE "Enable RNNoise deep learning noise suppression" OFF)
+option(ENABLE_WHISPER "Enable Whisper ASR (speech recognition)" OFF)
+
+# Compiler flags
+if(MSVC)
+    add_compile_options(/W4 /permissive-)
+else()
+    add_compile_options(-Wall -Wextra -Wpedantic)
+    if(CMAKE_BUILD_TYPE STREQUAL "Release")
+        # Use -march=native only for single-architecture builds on x86_64
+        # Skip for universal2 builds (multiple architectures) as -march=native is incompatible
+        if(APPLE)
+            # Check if building for multiple architectures (universal2)
+            string(FIND "${CMAKE_OSX_ARCHITECTURES}" ";" IS_UNIVERSAL)
+            if(IS_UNIVERSAL GREATER -1)
+                # Universal2 build - skip -march=native
+                add_compile_options(-O3)
+            elseif(CMAKE_OSX_ARCHITECTURES MATCHES "x86_64|AMD64")
+                # Single x86_64 build
+                add_compile_options(-O3 -march=native)
+            else()
+                # Single ARM64 build
+                add_compile_options(-O3)
+            endif()
+        elseif(CMAKE_SYSTEM_PROCESSOR MATCHES "x86_64|AMD64")
+            add_compile_options(-O3 -march=native)
+        else()
+            add_compile_options(-O3)
+        endif()
+    elseif(CMAKE_BUILD_TYPE STREQUAL "Debug")
+        add_compile_options(-g -O0)
+    endif()
+endif()
+
+# Find dependencies
+find_package(PkgConfig REQUIRED)
+
+# FFmpeg (required)
+pkg_check_modules(FFMPEG REQUIRED
+    libavcodec
+    libavformat
+    libavutil
+    libswresample
+)
+
+# PortAudio (required for audio capture)
+pkg_check_modules(PORTAUDIO REQUIRED portaudio-2.0)
+
+# FLAC (required for FLAC encoding)
+pkg_check_modules(FLAC REQUIRED flac)
+
+# WebRTC APM (optional)
+if(ENABLE_WEBRTC_APM)
+    pkg_check_modules(WEBRTC_APM webrtc-audio-processing-1)
+
+    if(NOT WEBRTC_APM_FOUND)
+        message(STATUS "")
+        message(STATUS "========================================")
+        message(STATUS "WebRTC APM library not found!")
+        message(STATUS "========================================")
+        message(STATUS "")
+        message(STATUS "To install webrtc-audio-processing:")
+        message(STATUS "")
+        message(STATUS "Linux (Ubuntu/Debian):")
+        message(STATUS "  sudo apt-get install webrtc-audio-processing-dev")
+        message(STATUS "")
+        message(STATUS "macOS/Linux (from source):")
+        message(STATUS "  git clone https://gitlab.freedesktop.org/pulseaudio/webrtc-audio-processing.git")
+        message(STATUS "  cd webrtc-audio-processing")
+        message(STATUS "  meson setup build --prefix=/usr/local")
+        message(STATUS "  meson compile -C build")
+        message(STATUS "  sudo meson install -C build")
+        message(STATUS "")
+        message(STATUS "Or disable WebRTC APM and use other processors:")
+        message(STATUS "  cmake .. -DENABLE_WEBRTC_APM=OFF")
+        message(STATUS "")
+        message(STATUS "========================================")
+        message(FATAL_ERROR "WebRTC APM library required but not found")
+    endif()
+
+    message(STATUS "WebRTC APM found in system:")
+    message(STATUS "  Include dirs: ${WEBRTC_APM_INCLUDE_DIRS}")
+    message(STATUS "  Libraries: ${WEBRTC_APM_LIBRARIES}")
+    message(STATUS "  Library dirs: ${WEBRTC_APM_LIBRARY_DIRS}")
+endif()
+
+# Include directories
+include_directories(
+    ${CMAKE_CURRENT_SOURCE_DIR}/include
+    ${CMAKE_CURRENT_SOURCE_DIR}/src
+    ${FFMPEG_INCLUDE_DIRS}
+    ${PORTAUDIO_INCLUDE_DIRS}
+    ${FLAC_INCLUDE_DIRS}
+)
+
+if(ENABLE_WEBRTC_APM AND WEBRTC_APM_FOUND)
+    include_directories(${WEBRTC_APM_INCLUDE_DIRS})
+endif()
+
+# Source files
+set(FFVOICE_CORE_SOURCES
+    src/audio/audio_capture_device.cpp
+    src/audio/audio_processor.cpp
+    src/audio/vad_segmenter.cpp
+    src/media/audio_file_writer.cpp
+    src/media/wav_writer.cpp
+    src/media/flac_writer.cpp
+    src/utils/logger.cpp
+    src/utils/ring_buffer.cpp
+    src/utils/signal_generator.cpp
+)
+
+# WebRTC APM (conditional)
+if(ENABLE_WEBRTC_APM)
+    message(STATUS "WebRTC APM: Enabled")
+    add_compile_definitions(ENABLE_WEBRTC_APM)
+    list(APPEND FFVOICE_CORE_SOURCES
+        src/audio/webrtc_processor.cpp
+    )
+else()
+    message(STATUS "WebRTC APM: Disabled (use -DENABLE_WEBRTC_APM=ON to enable)")
+endif()
+
+# RNNoise (conditional)
+if(ENABLE_RNNOISE)
+    message(STATUS "RNNoise: Enabled")
+    add_compile_definitions(ENABLE_RNNOISE)
+
+    # Fetch RNNoise library via CMake FetchContent
+    include(FetchContent)
+
+    message(STATUS "Fetching RNNoise from GitHub...")
+
+    FetchContent_Declare(
+        rnnoise
+        GIT_REPOSITORY https://github.com/xiph/rnnoise.git
+        GIT_TAG master
+        GIT_SHALLOW TRUE
+        SOURCE_DIR ${CMAKE_BINARY_DIR}/_deps/rnnoise-src
+    )
+
+    # Download only, don't try to build (RNNoise uses autotools, not CMake)
+    FetchContent_GetProperties(rnnoise)
+    if(NOT rnnoise_POPULATED)
+        FetchContent_Populate(rnnoise)
+    endif()
+
+    # RNNoise source files (manual configuration since it uses autotools)
+    set(RNNOISE_SOURCES
+        ${rnnoise_SOURCE_DIR}/src/rnn.c
+        ${rnnoise_SOURCE_DIR}/src/rnn_data.c
+        ${rnnoise_SOURCE_DIR}/src/rnn_reader.c
+        ${rnnoise_SOURCE_DIR}/src/denoise.c
+        ${rnnoise_SOURCE_DIR}/src/celt_lpc.c
+        ${rnnoise_SOURCE_DIR}/src/kiss_fft.c
+        ${rnnoise_SOURCE_DIR}/src/pitch.c
+    )
+
+    # Create RNNoise static library
+    add_library(rnnoise STATIC ${RNNOISE_SOURCES})
+
+    target_include_directories(rnnoise PUBLIC
+        ${rnnoise_SOURCE_DIR}/include
+    )
+
+    # RNNoise is a C library
+    set_target_properties(rnnoise PROPERTIES
+        C_STANDARD 99
+        C_STANDARD_REQUIRED ON
+        POSITION_INDEPENDENT_CODE ON
+    )
+
+    message(STATUS "RNNoise configured successfully")
+    message(STATUS "  Source dir: ${rnnoise_SOURCE_DIR}")
+    message(STATUS "  Include dir: ${rnnoise_SOURCE_DIR}/include")
+
+    list(APPEND FFVOICE_CORE_SOURCES
+        src/audio/rnnoise_processor.cpp
+    )
+else()
+    message(STATUS "RNNoise: Disabled (use -DENABLE_RNNOISE=ON to enable)")
+endif()
+
+# Whisper ASR (conditional)
+if(ENABLE_WHISPER)
+    message(STATUS "Whisper ASR: Enabled")
+    add_compile_definitions(ENABLE_WHISPER)
+
+    # Fetch whisper.cpp library via CMake FetchContent
+    include(FetchContent)
+
+    message(STATUS "Fetching whisper.cpp from GitHub...")
+
+    FetchContent_Declare(
+        whisper
+        GIT_REPOSITORY https://github.com/ggerganov/whisper.cpp.git
+        GIT_TAG v1.5.4
+        GIT_SHALLOW TRUE
+    )
+
+    # Configure whisper.cpp build options
+    set(WHISPER_BUILD_TESTS OFF CACHE BOOL "" FORCE)
+    set(WHISPER_BUILD_EXAMPLES OFF CACHE BOOL "" FORCE)
+    set(BUILD_SHARED_LIBS OFF CACHE BOOL "" FORCE)
+
+    # Disable AVX/AVX2/SIMD for Rosetta 2 compatibility
+    set(WHISPER_NO_AVX ON CACHE BOOL "" FORCE)
+    set(WHISPER_NO_AVX2 ON CACHE BOOL "" FORCE)
+    set(WHISPER_NO_FMA ON CACHE BOOL "" FORCE)
+    set(WHISPER_NO_F16C ON CACHE BOOL "" FORCE)
+    set(GGML_NATIVE OFF CACHE BOOL "" FORCE)
+
+    # Enable Position Independent Code for Python bindings
+    set(CMAKE_POSITION_INDEPENDENT_CODE ON CACHE BOOL "" FORCE)
+
+    FetchContent_MakeAvailable(whisper)
+
+    message(STATUS "whisper.cpp fetched successfully")
+    message(STATUS "  Source dir: ${whisper_SOURCE_DIR}")
+
+    # Download Whisper tiny model (39MB)
+    set(WHISPER_MODEL_DIR "${CMAKE_BINARY_DIR}/models")
+    set(WHISPER_TINY_MODEL "${WHISPER_MODEL_DIR}/ggml-tiny.bin")
+
+    if(NOT EXISTS ${WHISPER_TINY_MODEL})
+        message(STATUS "Downloading whisper tiny model (39MB)...")
+        file(MAKE_DIRECTORY ${WHISPER_MODEL_DIR})
+
+        file(DOWNLOAD
+            "https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-tiny.bin"
+            ${WHISPER_TINY_MODEL}
+            SHOW_PROGRESS
+            STATUS DOWNLOAD_STATUS
+        )
+
+        list(GET DOWNLOAD_STATUS 0 DOWNLOAD_CODE)
+        if(NOT DOWNLOAD_CODE EQUAL 0)
+            message(WARNING "Failed to download whisper model. Please download manually from:")
+            message(WARNING "  https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-tiny.bin")
+            message(WARNING "  Save to: ${WHISPER_TINY_MODEL}")
+        else()
+            message(STATUS "Model downloaded successfully")
+        endif()
+    else()
+        message(STATUS "Whisper tiny model already exists at ${WHISPER_TINY_MODEL}")
+    endif()
+
+    # Add compile definition for model path
+    add_compile_definitions(WHISPER_MODEL_PATH="${WHISPER_TINY_MODEL}")
+
+    list(APPEND FFVOICE_CORE_SOURCES
+        src/audio/whisper_processor.cpp
+        src/utils/subtitle_generator.cpp
+        src/utils/audio_converter.cpp
+    )
+else()
+    message(STATUS "Whisper ASR: Disabled (use -DENABLE_WHISPER=ON to enable)")
+endif()
+
+# Python bindings (optional)
+if(BUILD_PYTHON)
+    message(STATUS "Python bindings: Enabled")
+
+    # Find Python - use PYTHON_EXECUTABLE if provided (from setup.py)
+    if(DEFINED PYTHON_EXECUTABLE)
+        set(Python_EXECUTABLE ${PYTHON_EXECUTABLE})
+        message(STATUS "Using Python from setup.py: ${PYTHON_EXECUTABLE}")
+    endif()
+    find_package(Python COMPONENTS Interpreter Development REQUIRED)
+
+    # Find pybind11
+    find_package(pybind11 CONFIG)
+
+    if(NOT pybind11_FOUND)
+        message(STATUS "pybind11 not found in system, fetching from GitHub...")
+        include(FetchContent)
+
+        FetchContent_Declare(
+            pybind11
+            GIT_REPOSITORY https://github.com/pybind/pybind11.git
+            GIT_TAG v2.11.1
+            GIT_SHALLOW TRUE
+        )
+
+        FetchContent_MakeAvailable(pybind11)
+        message(STATUS "pybind11 fetched successfully")
+    else()
+        message(STATUS "pybind11 found: ${pybind11_DIR}")
+    endif()
+
+    # Python module
+    pybind11_add_module(_ffvoice src/python/bindings.cpp)
+
+    target_link_libraries(_ffvoice PRIVATE ffvoice-core)
+
+    target_include_directories(_ffvoice PRIVATE
+        ${CMAKE_CURRENT_SOURCE_DIR}/include
+        ${CMAKE_CURRENT_SOURCE_DIR}/src
+    )
+
+    # Install Python module
+    install(TARGETS _ffvoice LIBRARY DESTINATION python/ffvoice)
+    install(DIRECTORY python/ffvoice/ DESTINATION python/ffvoice
+            FILES_MATCHING PATTERN "*.py")
+else()
+    message(STATUS "Python bindings: Disabled (use -DBUILD_PYTHON=ON to enable)")
+endif()
+
+# Core library
+add_library(ffvoice-core STATIC ${FFVOICE_CORE_SOURCES})
+
+# Enable Position Independent Code for Python bindings on Linux
+set_target_properties(ffvoice-core PROPERTIES POSITION_INDEPENDENT_CODE ON)
+
+target_link_libraries(ffvoice-core
+    PUBLIC
+        ${FFMPEG_LINK_LIBRARIES}
+        ${PORTAUDIO_LIBRARIES}
+        ${FLAC_LIBRARIES}
+)
+
+if(ENABLE_WEBRTC_APM AND WEBRTC_APM_FOUND)
+    target_link_libraries(ffvoice-core
+        PUBLIC ${WEBRTC_APM_LIBRARIES}
+    )
+endif()
+
+if(ENABLE_RNNOISE)
+    target_link_libraries(ffvoice-core
+        PUBLIC rnnoise
+    )
+    target_include_directories(ffvoice-core
+        PUBLIC ${rnnoise_SOURCE_DIR}/include
+    )
+endif()
+
+if(ENABLE_WHISPER)
+    target_link_libraries(ffvoice-core
+        PUBLIC whisper
+    )
+    target_include_directories(ffvoice-core
+        PUBLIC ${whisper_SOURCE_DIR}
+    )
+endif()
+
+target_link_directories(ffvoice-core
+    PUBLIC
+        ${FFMPEG_LIBRARY_DIRS}
+        ${PORTAUDIO_LIBRARY_DIRS}
+        ${FLAC_LIBRARY_DIRS}
+)
+
+if(ENABLE_WEBRTC_APM AND WEBRTC_APM_FOUND)
+    target_link_directories(ffvoice-core
+        PUBLIC ${WEBRTC_APM_LIBRARY_DIRS}
+    )
+endif()
+
+target_include_directories(ffvoice-core
+    PUBLIC
+        $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/include>
+        $<INSTALL_INTERFACE:include>
+    PRIVATE
+        ${CMAKE_CURRENT_SOURCE_DIR}/src
+)
+
+# CLI application
+add_executable(ffvoice apps/cli/main.cpp)
+target_link_libraries(ffvoice PRIVATE ffvoice-core)
+
+# Tests
+if(BUILD_TESTS)
+    enable_testing()
+    add_subdirectory(tests)
+endif()
+
+# Examples
+if(BUILD_EXAMPLES)
+    if(EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/examples/CMakeLists.txt)
+        add_subdirectory(examples)
+    endif()
+endif()
+
+# Install
+install(TARGETS ffvoice ffvoice-core
+    RUNTIME DESTINATION bin
+    LIBRARY DESTINATION lib
+    ARCHIVE DESTINATION lib
+)
+
+install(DIRECTORY include/ffvoice DESTINATION include)

ffvoice-0.4.5/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 chico
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ffvoice-0.4.5/MANIFEST.in

@@ -0,0 +1,30 @@
+# Include the license and readme
+include LICENSE
+include README.md
+include pyproject.toml
+include setup.py
+include CMakeLists.txt
+
+# Include all C++ source and header files
+recursive-include src *.cpp *.h *.hpp
+recursive-include include *.h *.hpp
+
+# Include CMake modules
+recursive-include cmake *.cmake
+
+# Include Python bindings source
+recursive-include python *.cpp *.py
+
+# Include third-party dependencies that are bundled
+recursive-include third_party *.cpp *.c *.h *.hpp *.cmake CMakeLists.txt
+
+# Exclude build artifacts
+global-exclude *.pyc
+global-exclude *.pyo
+global-exclude __pycache__
+global-exclude *.so
+global-exclude *.dylib
+global-exclude .DS_Store
+prune build
+prune dist
+prune *.egg-info