contextweave 0.2.0__tar.gz

contextweave-0.2.0/.github/workflows/release.yml

@@ -0,0 +1,396 @@
+name: Build & Release
+
+on:
+  push:
+    tags:
+      - 'v*'
+  workflow_dispatch:
+
+permissions:
+  contents: write
+
+env:
+  PYTHON_VERSION: '3.12'
+  APP_NAME: 'ContextWeave'
+
+jobs:
+  build:
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - os: macos-14
+            platform_name: macos-arm64
+          - os: windows-latest
+            platform_name: windows-x86_64
+
+    runs-on: ${{ matrix.os }}
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v4
+
+      - name: Get version from tag
+        id: version
+        shell: bash
+        run: |
+          if [[ "${{ github.ref_type }}" == "tag" ]]; then
+            VERSION="${{ github.ref_name }}"
+          else
+            VERSION="v0.0.0-dev"
+          fi
+          echo "version=${VERSION}" >> "$GITHUB_OUTPUT"
+
+      - name: Install dependencies
+        run: uv sync --group dev
+
+      - name: Install CPU-only PyTorch (Windows)
+        if: runner.os == 'Windows'
+        run: uv pip install torch --index-url https://download.pytorch.org/whl/cpu --reinstall --no-deps
+
+      # ============================================================
+      # macOS: Import signing certificate
+      # ============================================================
+      - name: Import Apple signing certificate
+        if: runner.os == 'macOS'
+        env:
+          APPLE_CERTIFICATE_BASE64: ${{ secrets.APPLE_CERTIFICATE_BASE64 }}
+          APPLE_CERTIFICATE_PASSWORD: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
+        run: |
+          if [ -z "$APPLE_CERTIFICATE_BASE64" ]; then
+            echo "::notice::No Apple certificate configured — skipping code signing"
+            echo "CODESIGN_IDENTITY=" >> "$GITHUB_ENV"
+            exit 0
+          fi
+
+          # Create a temporary keychain
+          KEYCHAIN_PATH="$RUNNER_TEMP/app-signing.keychain-db"
+          KEYCHAIN_PASSWORD="$(openssl rand -hex 16)"
+          security create-keychain -p "$KEYCHAIN_PASSWORD" "$KEYCHAIN_PATH"
+          security set-keychain-settings -lut 21600 "$KEYCHAIN_PATH"
+          security unlock-keychain -p "$KEYCHAIN_PASSWORD" "$KEYCHAIN_PATH"
+
+          # Import the certificate
+          CERT_PATH="$RUNNER_TEMP/certificate.p12"
+          echo "$APPLE_CERTIFICATE_BASE64" | base64 --decode > "$CERT_PATH"
+          security import "$CERT_PATH" \
+            -P "$APPLE_CERTIFICATE_PASSWORD" \
+            -A -t cert -f pkcs12 -k "$KEYCHAIN_PATH"
+          security set-key-partition-list \
+            -S apple-tool:,apple: \
+            -k "$KEYCHAIN_PASSWORD" "$KEYCHAIN_PATH"
+          security list-keychain -d user -s "$KEYCHAIN_PATH"
+          rm -f "$CERT_PATH"
+
+          # Detect signing identity from the imported certificate
+          IDENTITY=$(security find-identity -v -p codesigning "$KEYCHAIN_PATH" \
+            | grep "Developer ID" | head -1 | sed 's/.*"\(.*\)"/\1/')
+          if [ -z "$IDENTITY" ]; then
+            echo "::warning::Certificate imported but no Developer ID identity found"
+            echo "CODESIGN_IDENTITY=" >> "$GITHUB_ENV"
+          else
+            echo "Found signing identity: $IDENTITY"
+            echo "CODESIGN_IDENTITY=$IDENTITY" >> "$GITHUB_ENV"
+          fi
+          echo "KEYCHAIN_PATH=$KEYCHAIN_PATH" >> "$GITHUB_ENV"
+
+      # ============================================================
+      # Build
+      # ============================================================
+      - name: Build with PyInstaller
+        run: uv run pyinstaller cat-ui.spec --clean --noconfirm
+
+      - name: Verify build output
+        shell: bash
+        run: |
+          echo "Working directory: $(pwd)"
+          echo "Contents of dist/:"
+          ls -la dist/ 2>/dev/null || { echo "::error::dist/ directory does not exist"; exit 1; }
+          if [ -d "dist/${{ env.APP_NAME }}" ]; then
+            echo "dist/${{ env.APP_NAME }}/ contents (first 20):"
+            ls dist/${{ env.APP_NAME }}/ | head -20
+          elif [ -d "dist/${{ env.APP_NAME }}.app" ]; then
+            echo "dist/${{ env.APP_NAME }}.app/ found (macOS bundle)"
+          else
+            echo "::error::Expected dist/${{ env.APP_NAME }} or dist/${{ env.APP_NAME }}.app not found"
+            exit 1
+          fi
+
+      - name: Verify OpenCC assets (Windows)
+        if: runner.os == 'Windows'
+        shell: bash
+        run: |
+          set -euo pipefail
+          APP_DIST="dist/${{ env.APP_NAME }}"
+
+          # PyInstaller layout may vary across versions/platforms:
+          # - dist/ContextWeave/_internal/opencc/...
+          # - dist/ContextWeave/opencc/...
+          OPENCC_BASE=""
+          for base in "$APP_DIST/_internal" "$APP_DIST"; do
+            if [ -f "$base/opencc/config/jp2s.json" ] && [ -d "$base/opencc/dictionary" ]; then
+              OPENCC_BASE="$base/opencc"
+              break
+            fi
+          done
+
+          if [ -z "$OPENCC_BASE" ]; then
+            echo "::group::OpenCC asset debug"
+            echo "Expected one of:"
+            echo "  $APP_DIST/_internal/opencc/config/jp2s.json"
+            echo "  $APP_DIST/opencc/config/jp2s.json"
+            echo "Actual opencc-related paths:"
+            find "$APP_DIST" -maxdepth 5 \( -type d -name opencc -o -type f -name 'jp2s.json' \) 2>/dev/null | sort || true
+            echo "::endgroup::"
+            echo "::error::Missing OpenCC config/dictionary assets under $APP_DIST"
+            exit 1
+          fi
+
+          echo "OpenCC assets verified at: $OPENCC_BASE"
+
+      # ============================================================
+      # macOS: Sign, package as DMG, notarize
+      # ============================================================
+      - name: Sign app bundle (macOS)
+        if: runner.os == 'macOS' && env.CODESIGN_IDENTITY != ''
+        run: |
+          echo "Signing with: $CODESIGN_IDENTITY"
+
+          # Sign all embedded binaries first (inside-out signing)
+          find "dist/${{ env.APP_NAME }}.app" -type f \
+            \( -name "*.so" -o -name "*.dylib" \) -print0 | \
+          while IFS= read -r -d '' lib; do
+            codesign --force --options runtime \
+              --entitlements installer/macos/entitlements.plist \
+              --sign "$CODESIGN_IDENTITY" "$lib"
+          done
+
+          # Sign executable binaries inside the bundle
+          find "dist/${{ env.APP_NAME }}.app/Contents/MacOS" -type f -perm +111 -print0 | \
+          while IFS= read -r -d '' bin; do
+            codesign --force --options runtime \
+              --entitlements installer/macos/entitlements.plist \
+              --sign "$CODESIGN_IDENTITY" "$bin"
+          done
+
+          # Sign the app bundle itself
+          codesign --force --options runtime \
+            --entitlements installer/macos/entitlements.plist \
+            --sign "$CODESIGN_IDENTITY" \
+            "dist/${{ env.APP_NAME }}.app"
+
+          # Verify nested signatures strictly and test Gatekeeper assessment
+          codesign --verify --deep --strict --verbose=2 "dist/${{ env.APP_NAME }}.app"
+          spctl --assess --type execute --verbose=4 "dist/${{ env.APP_NAME }}.app"
+          echo "Code signing verified successfully"
+
+      - name: Create DMG (macOS)
+        if: runner.os == 'macOS'
+        run: |
+          mkdir -p release
+          brew install create-dmg || true
+          # create-dmg may return non-zero for cosmetic issues; DMG is still created
+          create-dmg \
+            --volname "ContextWeave" \
+            --window-pos 200 120 \
+            --window-size 600 400 \
+            --icon-size 100 \
+            --icon "${{ env.APP_NAME }}.app" 150 190 \
+            --app-drop-link 450 190 \
+            --no-internet-enable \
+            "release/${{ env.APP_NAME }}-${{ steps.version.outputs.version }}-${{ matrix.platform_name }}.dmg" \
+            "dist/${{ env.APP_NAME }}.app" \
+          || true
+          ls -lh release/*.dmg
+
+      - name: Sign and notarize DMG (macOS)
+        if: runner.os == 'macOS' && env.CODESIGN_IDENTITY != ''
+        env:
+          APPLE_ID: ${{ secrets.APPLE_ID }}
+          APPLE_ID_PASSWORD: ${{ secrets.APPLE_ID_PASSWORD }}
+          APPLE_TEAM_ID: ${{ secrets.APPLE_TEAM_ID }}
+        run: |
+          DMG_FILE=$(ls release/*.dmg)
+
+          # Sign the DMG
+          codesign --force --sign "$CODESIGN_IDENTITY" "$DMG_FILE"
+
+          # Notarize (requires Apple ID credentials)
+          if [ -n "$APPLE_ID" ] && [ -n "$APPLE_ID_PASSWORD" ] && [ -n "$APPLE_TEAM_ID" ]; then
+            echo "Submitting for notarization..."
+            xcrun notarytool submit "$DMG_FILE" \
+              --apple-id "$APPLE_ID" \
+              --password "$APPLE_ID_PASSWORD" \
+              --team-id "$APPLE_TEAM_ID" \
+              --wait
+
+            # Staple the notarization ticket to the DMG
+            xcrun stapler staple "$DMG_FILE"
+            echo "Notarization and stapling complete"
+          else
+            echo "::notice::Skipping notarization — Apple ID credentials not configured"
+          fi
+
+      - name: Cleanup macOS keychain
+        if: runner.os == 'macOS' && env.KEYCHAIN_PATH != ''
+        run: security delete-keychain "$KEYCHAIN_PATH"
+
+      # ============================================================
+      # Windows: Sign executable, package portable zip
+      # ============================================================
+      - name: Sign executable (Windows)
+        if: runner.os == 'Windows'
+        env:
+          WINDOWS_CERTIFICATE_BASE64: ${{ secrets.WINDOWS_CERTIFICATE_BASE64 }}
+          WINDOWS_CERTIFICATE_PASSWORD: ${{ secrets.WINDOWS_CERTIFICATE_PASSWORD }}
+        shell: powershell
+        run: |
+          if (-not $env:WINDOWS_CERTIFICATE_BASE64) {
+            Write-Host '::notice::No Windows certificate configured - skipping code signing'
+            exit 0
+          }
+
+          $certPath = Join-Path $env:RUNNER_TEMP 'certificate.pfx'
+          [IO.File]::WriteAllBytes($certPath,
+            [Convert]::FromBase64String($env:WINDOWS_CERTIFICATE_BASE64))
+
+          $signtool = (Resolve-Path 'C:\Program Files (x86)\Windows Kits\10\bin\*\x64\signtool.exe' |
+            Sort-Object -Descending | Select-Object -First 1).Path
+          Write-Host "Using signtool: $signtool"
+
+          $signArgs = @('sign', '/f', $certPath, '/p', $env:WINDOWS_CERTIFICATE_PASSWORD,
+            '/tr', 'http://timestamp.digicert.com', '/td', 'sha256', '/fd', 'sha256')
+
+          & $signtool @signArgs "dist\${{ env.APP_NAME }}\${{ env.APP_NAME }}.exe"
+
+          Get-ChildItem "dist\${{ env.APP_NAME }}" -Recurse -Include *.dll,*.pyd | ForEach-Object {
+            & $signtool @signArgs $_.FullName 2>$null
+          }
+
+          Remove-Item $certPath -Force
+          Write-Host 'Code signing complete'
+
+      - name: Create portable zip (Windows)
+        if: runner.os == 'Windows'
+        shell: bash
+        run: |
+          mkdir -p release
+          python - <<'PY'
+          from pathlib import Path
+          import zipfile
+
+          source_dir = Path("dist/${{ env.APP_NAME }}")
+          archive_path = Path("release/${{ env.APP_NAME }}-${{ steps.version.outputs.version }}-${{ matrix.platform_name }}.zip")
+
+          if not source_dir.is_dir():
+              raise SystemExit(f"Missing source directory: {source_dir}")
+
+          if archive_path.exists():
+              archive_path.unlink()
+
+          with zipfile.ZipFile(
+              archive_path,
+              mode="w",
+              compression=zipfile.ZIP_DEFLATED,
+              compresslevel=6,
+              allowZip64=True,
+          ) as archive:
+              for path in sorted(source_dir.rglob("*")):
+                  if path.is_file():
+                      archive.write(path, arcname=path.relative_to(source_dir.parent))
+
+          print(f"Portable archive created: {archive_path}")
+          PY
+
+      - name: Smoke test portable zip (Windows)
+        if: runner.os == 'Windows'
+        env:
+          CONTEXTWEAVE_SMOKE_TEST_STARTUP: '1'
+        shell: powershell
+        run: |
+          $archivePath = Resolve-Path "release\${{ env.APP_NAME }}-${{ steps.version.outputs.version }}-${{ matrix.platform_name }}.zip"
+          $extractRoot = Join-Path $env:RUNNER_TEMP 'portable-smoke'
+          $appRoot = Join-Path $extractRoot '${{ env.APP_NAME }}'
+          $exePath = Join-Path $appRoot '${{ env.APP_NAME }}.exe'
+          $openccCandidates = @(
+            (Join-Path $appRoot '_internal\opencc\config\jp2s.json'),
+            (Join-Path $appRoot 'opencc\config\jp2s.json')
+          )
+
+          if (Test-Path $extractRoot) {
+            Remove-Item $extractRoot -Recurse -Force
+          }
+
+          Expand-Archive -Path $archivePath -DestinationPath $extractRoot -Force
+
+          $topLevelEntries = @(Get-ChildItem -LiteralPath $extractRoot)
+          if ($topLevelEntries.Count -ne 1 -or -not $topLevelEntries[0].PSIsContainer -or $topLevelEntries[0].Name -ne '${{ env.APP_NAME }}') {
+            throw "Unexpected archive layout after extraction: $($topLevelEntries.Name -join ', ')"
+          }
+
+          if (-not (Test-Path -LiteralPath $exePath -PathType Leaf)) {
+            throw "Extracted executable not found: $exePath"
+          }
+
+          if (-not ($openccCandidates | Where-Object { Test-Path -LiteralPath $_ -PathType Leaf })) {
+            throw 'Extracted portable zip is missing expected OpenCC assets'
+          }
+
+          $process = Start-Process -FilePath $exePath -WorkingDirectory $appRoot -PassThru
+          try {
+            if (-not $process.WaitForExit(30000)) {
+              Stop-Process -Id $process.Id -Force
+              throw 'Portable app smoke test timed out after 30 seconds'
+            }
+            if ($process.ExitCode -ne 0) {
+              throw "Portable app smoke test exited with code $($process.ExitCode)"
+            }
+          }
+          finally {
+            if (-not $process.HasExited) {
+              Stop-Process -Id $process.Id -Force
+            }
+          }
+
+          Write-Host "Portable zip smoke test passed: $archivePath"
+
+      # ============================================================
+      # Upload
+      # ============================================================
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: ${{ env.APP_NAME }}-${{ matrix.platform_name }}
+          path: release/*
+          if-no-files-found: error
+          retention-days: 5
+
+  release:
+    needs: build
+    runs-on: ubuntu-latest
+    if: startsWith(github.ref, 'refs/tags/v')
+
+    steps:
+      - name: Download all artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: artifacts
+          merge-multiple: true
+
+      - name: List release files
+        run: ls -lhR artifacts/
+
+      - name: Create GitHub Release
+        uses: softprops/action-gh-release@v2
+        with:
+          files: artifacts/*
+          generate_release_notes: true
+          draft: true

contextweave-0.2.0/.gitignore

@@ -0,0 +1,59 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*.pyo
+*.so
+*.dylib
+*.dll
+
+# Virtual environments
+.venv/
+venv/
+.env
+.env.local
+.python-version
+
+# Packaging
+build/
+dist/
+release/
+wheels/
+*.egg-info/
+.eggs/
+pip-wheel-metadata/
+
+# Caches and tooling
+.mypy_cache/
+.pytest_cache/
+.ruff_cache/
+.nox/
+.cache/
+*.log
+*.pid
+
+# Coverage
+.coverage
+.coverage.*
+htmlcov/
+
+# Editors/IDE
+.idea/
+.vscode/
+
+# macOS / system
+.DS_Store
+
+# uv
+.uv/
+
+.config.yaml
+output
+output.txt
+output.epub
+delete-me/
+delete-me-output/
+.sisyphus/*
+.omc
+.claude
+evaluation/output/
+MagicMock/

contextweave-0.2.0/AGENTS.md

@@ -0,0 +1,58 @@
+<!-- Generated: 2026-02-26 -->
+
+# context-aware-translation
+
+## Purpose
+A Python 3.12+ document translation system that uses LLMs with context-aware glossary management. Translates text, PDF, scanned books, and manga with hierarchical context trees that reduce token usage by 99%+.
+
+## Key Files
+
+| File | Description |
+|------|-------------|
+| `pyproject.toml` | Project config: dependencies, tooling (ruff, mypy, pytest) |
+| `Makefile` | Build commands: `make check`, `make test`, `make build-ui` |
+| `cat-ui.spec` | PyInstaller spec for standalone UI builds |
+| `uv.lock` | Locked dependency versions (managed by uv) |
+
+## Subdirectories
+
+| Directory | Purpose |
+|-----------|---------|
+| `context_aware_translation/` | Main package: core logic, LLM, documents, storage, UI (see `context_aware_translation/AGENTS.md`) |
+| `tests/` | Test suites mirroring main package structure (see `tests/AGENTS.md`) |
+| `scripts/` | Build and evaluation scripts (see `scripts/AGENTS.md`) |
+| `installer/` | Platform-specific packaging configs (see `installer/AGENTS.md`) |
+
+## For AI Agents
+
+### Working In This Directory
+- Use `uv` for dependency management (`uv sync`, `uv run`)
+- Entry points: `cat-ui` (PySide6 GUI via `context_aware_translation.ui.main:main`)
+- Run `make check` before committing (lint, format, typecheck, lupdate)
+- All translation strings must have zh_CN translations; `make lupdate` will fail on unfinished entries
+
+### Testing Requirements
+- Run `make test` or `uv run pytest tests/`
+- Tests run in parallel via pytest-xdist (`-n auto`)
+- Async tests use pytest-asyncio with `auto` mode
+
+### Common Patterns
+- Config models in `config.py` using dataclasses
+- SQLite with WAL mode for all storage (registry.db global, book.db per-book)
+- 5 required config sections: extractor, summarizor, translator, glossary, review
+- Ruff for linting+formatting, mypy for type checking (strict, excludes ui/ and tests/)
+- PySide6 i18n with `.ts`/`.qm` translation files
+
+## Dependencies
+
+### External
+- `pyside6` - Qt-based GUI framework
+- `openai`, `google-genai` - LLM API clients
+- `torch`, `transformers` - ML models for embeddings/tokenization
+- `faiss-cpu` - Vector similarity search
+- `pikepdf`, `pypdfium2` - PDF handling
+- `pypandoc-binary` - Document format conversion
+- `semchunk` - Semantic text chunking
+- `tenacity` - Retry logic for API calls
+
+<!-- MANUAL: -->