termux-tts 1.5.1 → 1.5.2

package/package.json

@@ -1,11 +1,19 @@
 {
   "name": "termux-tts",
-  "version": "1.5.1",
+  "version": "1.5.2",
   "description": "On-device Text-to-Speech framework utilizing device resources (DSP Formant Vocoder, ONNX Neural Runtime & Android Native Voice)",
   "main": "index.js",
   "bin": {
     "termux-tts": "bin/cli.js"
   },
+  "files": [
+    "index.js",
+    "bin",
+    "binding_node",
+    "README.md",
+    "README.pypi.md",
+    "LICENSE"
+  ],
   "scripts": {
     "test": "node binding_node/test_node.js",
     "doctor": "node bin/cli.js doctor"
@@ -54,4 +62,4 @@
   "engines": {
     "node": ">=16.0.0"
   }
-}
+}

package/.github/workflows/release.yml

@@ -1,98 +0,0 @@
-name: Release
-
-on:
-  push:
-    branches:
-      - main
-      - master
-    tags:
-      - 'v*'
-  workflow_dispatch:
-
-permissions:
-  contents: read
-
-jobs:
-  publish-pypi:
-    name: Publish to PyPI
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Check if PyPI project exists
-        id: check-py
-        run: |
-          if [ -f "pyproject.toml" ] || [ -f "setup.py" ]; then
-            echo "has_python=true" >> $GITHUB_OUTPUT
-          else
-            echo "has_python=false" >> $GITHUB_OUTPUT
-          fi
-
-      - name: Set up Python
-        if: steps.check-py.outputs.has_python == 'true'
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.11'
-
-      - name: Install build dependencies
-        if: steps.check-py.outputs.has_python == 'true'
-        run: |
-          python -m pip install --upgrade pip
-          pip install build twine
-
-      - name: Build distributions
-        if: steps.check-py.outputs.has_python == 'true'
-        run: python -m build
-
-      - name: Verify distribution archives
-        if: steps.check-py.outputs.has_python == 'true'
-        run: twine check dist/*
-
-      - name: Publish to PyPI
-        if: steps.check-py.outputs.has_python == 'true'
-        env:
-          TWINE_USERNAME: __token__
-          TWINE_PASSWORD: ${{ secrets.PYPI_API_TOKEN }}
-        run: |
-          twine upload --skip-existing dist/*
-
-  publish-npm:
-    name: Publish to NPM
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Set up Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: '20'
-          registry-url: 'https://registry.npmjs.org'
-
-      - name: Publish NPM package
-        env:
-          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}
-        run: |
-          if [ -f "package.json" ]; then
-            PKG_NAME=$(node -p "require('./package.json').name")
-            PKG_VER=$(node -p "require('./package.json').version")
-            if [ "$(npm view ${PKG_NAME}@${PKG_VER} version 2>/dev/null)" = "${PKG_VER}" ]; then
-              echo "Version ${PKG_VER} of ${PKG_NAME} is already published on NPM. Skipping."
-            else
-              echo "Publishing ${PKG_NAME}@${PKG_VER} to NPM..."
-              npm publish --access public
-            fi
-          elif [ -f "npm/package.json" ]; then
-            cd npm
-            PKG_NAME=$(node -p "require('./package.json').name")
-            PKG_VER=$(node -p "require('./package.json').version")
-            if [ "$(npm view ${PKG_NAME}@${PKG_VER} version 2>/dev/null)" = "${PKG_VER}" ]; then
-              echo "Version ${PKG_VER} of ${PKG_NAME} is already published on NPM. Skipping."
-            else
-              echo "Publishing ${PKG_NAME}@${PKG_VER} to NPM..."
-              npm publish --access public
-            fi
-          else
-            echo "No package.json found. Skipping NPM publish."
-          fi

package/CHANGELOG.md

@@ -1,93 +0,0 @@
-# Changelog
-
-All notable changes to this project will be documented in this file.
-
-## [1.5.0] - 2026-09-15
-
-### Added & Architecture Overhaul
-- **100% Native Vulkan Hardware GPU Acceleration**:
-  - Eradicated Mesa CPU software emulation (`llvmpipe`) by enforcing direct vendor Android system ABI binding (`/system/lib64/libvulkan.so`), resolving previous GPU slowdowns and achieving true hardware shader acceleration.
-  - Empirically verified on physical Android 16 fleet:
-    - **Galaxy S21** (Exynos 2100 / ARM Mali-G78): **RTF 2.2055x** (9,371ms compute, 4.25s speech)
-    - **Galaxy S25** (Snapdragon 8 Elite / Qualcomm Adreno 830): **RTF 3.8145x** (16,208ms compute, 4.25s speech)
-    - **Galaxy S22** (Snapdragon 8 Gen 1 / Qualcomm Adreno 730): **RTF 8.9159x** (38,092ms compute, 4.27s speech)
-    - **Galaxy A35** (Exynos 1380 / ARM Mali-G68 MP5): **RTF 12.1505x** (51,912ms compute, 4.27s speech)
-- **Zero-Tolerance Anti-Deception & Fail-Fast Standard**:
-  - Permanently abolished all deceptive CPU offloading, chained fallback blocks, and unlogged exception catching in `VulkanNeuralEngine`.
-  - Introduced standardized Fail-Fast error codes: `[AMEVA-TTS-E001]` (Missing Binary/Model Weights), `[AMEVA-TTS-E002]` (Vulkan Runtime Execution Error), and `[AMEVA-TTS-E003]` (Buffer Truncation).
-  - Purged coupled legacy code (`engine_dsp.py`) under the AOSF Deletion-First protocol.
-- **Dynamic Path Resolution (Zero Hardcoded Paths)**:
-  - Eliminated all absolute file system assumptions (`/data/data/com.termux/files/home/...`), replacing them with dynamic candidate sets leveraging `$PREFIX`, `sys.prefix`, `$HOME`, and `$PATH`.
-- **MeloTTS Hardware Pipeline & Tokenizer**:
-  - Introduced standalone `MeloTokenizer` supporting phoneme, tone, and lexicon tensor construction (`x`, `tones`, `sid`, `length_scale`).
-  - Implemented dual C++ native ABI backends for HiFi-GAN NCNN Slicing (`melo-ncnn-cli`) and MNN Vulkan execution (`melo-mnn-cli`).
-  - Documented mobile GPU shader memory constraints (32MB `maxBufferSize` limit in mobile drivers vs. 52.4MB single-layer ConvTranspose).
-- **Test Suite Modernization**:
-  - Added dedicated unit tests (`test_vulkan_melo.py`, `test_nextgen_models.py`) with 100% passing test coverage (53 passed, 10 skipped, 0 failures).
-
----
-
-
-### Added
-- **Multilingual Neural Orchestrator**: Integrated `MultilingualNeuralEngine` capable of dynamic cross-language code-switching across 9 official languages (ko, en, ja, zh, hi, ru, es, fr, de).
-- **Resident C-API Acceleration**: Implemented `SherpaResidentManager` and `SherpaCapiSession` for in-memory model caching via direct C-API bindings, achieving sub-0.18x RTF and eliminating subprocess invocation overhead.
-- **Zero-Config CLI Ergonomics**: Promoted speech synthesis (`synth`) as the default top-level subcommand; allows direct positional text (`termux-tts "text" --play`); eliminated necessity of `-e multilingual` flag.
-- **Universal Unicode Script Classifier**: Built `MultilingualTokenizer` covering Hangul, Latin, Devanagari (Hindi), Cyrillic (Russian), CJK, and Arabic scripts with 50ms smooth pause padding.
-- **On-Demand Model Provisioning & Self-Healing**: Enhanced `termux-tts install` to provision English/Korean by default, with on-demand flags (`--models hi/ja/ru/zh/all`) and actionable English guidance with absolute paths for uninstalled models.
-- **Defensive Path Resolution**: Hardened path parsing across `AudioBuffer.save()` and CLI entrypoints with automatic directory creation and `Path.expanduser()` resolution.
-
----
-
-## [1.4.3] - 2026-09-07
-
-### Added
-- **Multi-Tier Dynamic Candidate Resolution**: Replaced single hardcoded `v1.0.0-vulkan` URL in `installer.py` with 4-tier candidate resolution (`TERMUX_TTS_RELEASE_TAG`, `v{__version__}`, `releases/latest/download`, `uno-km/ameva-runtime` SSOT fallback, and companion fallback).
-- **Dual-Path Installation**: Installs and links precompiled Vulkan binary `sherpa-ncnn-offline-tts` to both `~/.local/bin` and `$PREFIX/bin` for instant global PATH resolution.
-- **Dynamic SSOT User-Agent**: Injects dynamic package version into download requests.
-
----
-
-## [1.4.2] - 2026-09-07
-
-### Added
-- Complete 12-tier enterprise English documentation overhaul for PyPI and GitHub/NPM.
-- Detailed empirical mobile hardware benchmarks (Snapdragon 8 Elite / Adreno 830, Exynos 1380 / Mali-G68 MP5).
-- Full GPU interconnect architecture documentation with SPIR-V compute shader details.
-- Comprehensive CPU vs. GPU thermal dissipation and latency trade-off analysis.
-- 3-stage 24/7 unattended background execution guide (Termux wake-lock, battery optimization, ADB phantom process killer).
-- Expanded technical SEO metadata keywords (34 keywords).
-
----
-
-## [1.4.1] - 2026-09-07
-
-### Fixed
-- Fixed npm package.json bin path specification (removed './' prefix).
-- Integrated GitHub Actions automated CI/CD release workflow.
-
----
-
-## [1.4.0] - 2026-09-07
-
-### Added
-- Direct integration with `TtsAdapter` from `ameva_runtime.adapters` SSOT.
-- Dual-tier VITS Vulkan neural engine routing with strict Fail-Fast error semantics.
-- English localization for all diagnostics, logs, and exception messages.
-
----
-
-## [1.1.5] - 2026-09-05
-
-### Changed
-- Synchronized install.sh hardware diagnostics binding to ameva-runtime.
-- Modernized 12-stage hardware diagnostic bridge and documentation architecture diagrams.
-
----
-
-## [1.1.4] - 2026-09-05
-
-### Changed
-- Migrated hardware acceleration dependency to unified `ameva-runtime>=2.0.0` and `@ameva/runtime>=2.0.0`.
-- Synchronized Python and npm package versions to v1.1.4.
-
-