fs-ocr 1.0.1__tar.gz

fs_ocr-1.0.1/.github/workflows/release.yml

@@ -0,0 +1,214 @@
+name: Build & publish wheels
+
+# Produces the `fs-ocr` PyPI distribution (pure-Rust ocrs/rten backend) plus
+# standalone CLI binaries, all from this one codebase.
+#
+# HDF5 is built from source and statically linked (feature `static-hdf5`) so the
+# wheels carry no external libhdf5. There are no system OCR dependencies: text
+# recognition is fully embedded. Chinese custom names are read via the system
+# `tesseract` CLI when the user has it installed (detected at runtime), so
+# nothing here needs to build or bundle Tesseract.
+#
+# Triggers: tag push `v*` (build + publish) and manual dispatch / workflow PRs
+# (build only, no publish).
+
+on:
+  push:
+    tags: ["v*"]
+  workflow_dispatch:
+  pull_request:
+    paths: [".github/workflows/release.yml"]
+
+permissions:
+  contents: read
+
+# Friendly OS/arch names and archive extensions for the standalone CLI assets.
+env:
+  CLI_BIN: fs-ocr
+
+jobs:
+  # ===========================================================================
+  # WHEELS: fs-ocr (pure-Rust OCR, static HDF5, zero system libs)
+  # ===========================================================================
+  build-wheels:
+    name: wheels · ${{ matrix.platform.os }}-${{ matrix.platform.target }}
+    runs-on: ${{ matrix.platform.runner }}
+    strategy:
+      fail-fast: false
+      matrix:
+        platform:
+          - { runner: ubuntu-latest,  target: x86_64,  os: linux }
+          - { runner: windows-latest, target: x64,     os: windows }
+          # macOS is aarch64-only. The Intel macos-13 runner is being retired
+          # (jobs queue indefinitely), and x86_64 can't be cross-compiled on the
+          # arm64 runner because HDF5 is built from source and its CMake probes
+          # use try_run(), which can't execute x86_64 binaries on an arm64 host.
+          - { runner: macos-14,       target: aarch64, os: macos }
+    steps:
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v6
+        with:
+          python-version: "3.12"
+
+      # CMake is preinstalled on the Windows/macOS runners; Linux installs it
+      # inside the manylinux container via before-script-linux below.
+      - name: Build wheels
+        uses: PyO3/maturin-action@v1
+        with:
+          target: ${{ matrix.platform.target }}
+          # pyo3/extension-module is required so the wheel does NOT link libpython
+          # (abi3). Passing --features on the CLI overrides pyproject's
+          # [tool.maturin] features, so it must be repeated here explicitly.
+          args: --release --out dist --features static-hdf5,pyo3/extension-module
+          sccache: "true"
+          # manylinux_2_28 (AlmaLinux 8) ships a modern toolchain that can build
+          # HDF5 from source. before-script-linux runs INSIDE that container.
+          manylinux: "2_28"
+          before-script-linux: dnf install -y cmake gcc gcc-c++ make
+
+      # Install the freshly built wheel and import it to catch broken builds
+      # (missing symbols, bad abi3 tag, embedded model failures) before release.
+      - name: Smoke test wheel
+        shell: bash
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install --find-links dist --no-index fs-ocr
+          python -c "import fs_ocr; print('version', fs_ocr.__version__); print('backend', fs_ocr.OCR_BACKEND)"
+
+      - uses: actions/upload-artifact@v5
+        with:
+          name: wheels-${{ matrix.platform.os }}-${{ matrix.platform.target }}
+          path: dist/*.whl
+
+  build-sdist:
+    name: sdist
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+      - name: Build sdist
+        uses: PyO3/maturin-action@v1
+        with:
+          command: sdist
+          args: --out dist
+      - uses: actions/upload-artifact@v5
+        with:
+          name: sdist
+          path: dist/*.tar.gz
+
+  # ===========================================================================
+  # STANDALONE CLI — `fs-ocr` executable, no Python/libpython linkage.
+  # Built with `--no-default-features` (drops the `python` feature) so the
+  # binary links no libpython and runs on a machine without Python installed.
+  # HDF5 is statically linked (`static-hdf5`); zero system OCR libs.
+  # Each build is archived (.tar.gz on unix, .zip on Windows) and uploaded as a
+  # build artifact; the `release-cli` job attaches them to the GitHub Release.
+  # ===========================================================================
+  build-cli:
+    name: cli · ${{ matrix.platform.os }}-${{ matrix.platform.arch }}
+    runs-on: ${{ matrix.platform.runner }}
+    strategy:
+      fail-fast: false
+      matrix:
+        platform:
+          - { runner: ubuntu-latest,  arch: x86_64,  os: linux,   target: x86_64-unknown-linux-gnu }
+          - { runner: windows-latest, arch: x86_64,  os: windows, target: x86_64-pc-windows-msvc }
+          # macOS is aarch64-only (see build-wheels: x86_64 can't be cross-built
+          # because HDF5 is compiled from source with try_run-based CMake probes).
+          - { runner: macos-14,       arch: aarch64, os: macos,   target: aarch64-apple-darwin }
+    steps:
+      - uses: actions/checkout@v5
+      - uses: dtolnay/rust-toolchain@stable
+        with:
+          targets: ${{ matrix.platform.target }}
+      - uses: Swatinem/rust-cache@v2
+        with:
+          key: cli-${{ matrix.platform.target }}
+
+      - name: Build CLI (no python feature, static HDF5)
+        shell: bash
+        run: |
+          cargo build --release --locked \
+            --no-default-features --features static-hdf5 \
+            --bin "$CLI_BIN" --target ${{ matrix.platform.target }}
+
+      - name: Package
+        id: pkg
+        shell: bash
+        run: |
+          name="fs-ocr-${{ matrix.platform.os }}-${{ matrix.platform.arch }}"
+          src="target/${{ matrix.platform.target }}/release/$CLI_BIN"
+          if [ "${{ matrix.platform.os }}" = "windows" ]; then
+            7z a "$name.zip" "$src.exe"
+            echo "asset=$name.zip" >> "$GITHUB_OUTPUT"
+          else
+            tar -C "target/${{ matrix.platform.target }}/release" -czf "$name.tar.gz" "$CLI_BIN"
+            echo "asset=$name.tar.gz" >> "$GITHUB_OUTPUT"
+          fi
+
+      - uses: actions/upload-artifact@v5
+        with:
+          name: cli-${{ matrix.platform.os }}-${{ matrix.platform.arch }}
+          path: ${{ steps.pkg.outputs.asset }}
+
+  # Attach every standalone CLI archive to the GitHub Release (tag pushes only).
+  release-cli:
+    name: release · attach CLI binaries
+    needs: [build-cli]
+    if: startsWith(github.ref, 'refs/tags/v')
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/download-artifact@v5
+        with:
+          pattern: "cli-*"
+          path: cli
+          merge-multiple: true
+      - uses: softprops/action-gh-release@v2
+        with:
+          files: cli/*
+
+  # ===========================================================================
+  # VERSION GUARD (tag pushes only) — the published version comes from
+  # Cargo.toml, NOT the tag. The tag is only the trigger. Guard against the
+  # silent failure of bumping the tag but forgetting Cargo.toml: strip the
+  # leading `v` from the tag and require an exact match.
+  # ===========================================================================
+  check-version:
+    name: check · tag matches Cargo.toml
+    if: startsWith(github.ref, 'refs/tags/v')
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+      - name: Compare tag to Cargo.toml version
+        shell: bash
+        run: |
+          tag_version="${GITHUB_REF_NAME#v}"
+          cargo_version="$(grep -m1 '^version = ' Cargo.toml | sed -E 's/^version = "([^"]+)".*/\1/')"
+          echo "tag (stripped): $tag_version"
+          echo "Cargo.toml:     $cargo_version"
+          if [ "$tag_version" != "$cargo_version" ]; then
+            echo "::error::Tag $GITHUB_REF_NAME (-> $tag_version) does not match Cargo.toml version $cargo_version. Bump Cargo.toml or fix the tag." >&2
+            exit 1
+          fi
+
+  # ===========================================================================
+  # PUBLISH (tag pushes only) — trusted publishing via OIDC, no stored tokens.
+  # The PyPI project `fs-ocr` must list this repo+workflow as a trusted
+  # publisher for the `pypi` environment.
+  # ===========================================================================
+  publish:
+    name: publish · fs-ocr
+    needs: [check-version, build-wheels, build-sdist]
+    if: startsWith(github.ref, 'refs/tags/v')
+    runs-on: ubuntu-latest
+    environment: pypi
+    permissions:
+      id-token: write
+    steps:
+      - uses: actions/download-artifact@v5
+        with:
+          pattern: "{wheels-*,sdist}"
+          path: dist
+          merge-multiple: true
+      - uses: pypa/gh-action-pypi-publish@release/v1

fs_ocr-1.0.1/.gitignore

@@ -0,0 +1,32 @@
+# Build artifacts
+/target/
+
+# Local Cargo config (machine-specific CPU tuning; not for CI/consumers)
+.cargo/
+
+# Debug images
+*.png
+
+# Python
+__pycache__/
+*.pyc
+*.pyo
+*.egg-info/
+.eggs/
+dist/
+build/
+
+# Compiled native extension built by `maturin develop` (platform-specific artifact)
+python/fs_ocr/*.so
+python/fs_ocr/*.pyd
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# Environment
+.env
+.venv/
+venv/

fs_ocr-1.0.1/.pre-commit-config.yaml

@@ -0,0 +1,20 @@
+# Formatting-only pre-commit hooks. Any committed code is auto-formatted; if a
+# hook reformats a file the commit aborts so you can re-stage the result.
+# One-time setup per clone:  pre-commit install
+repos:
+  - repo: local
+    hooks:
+      - id: cargo-fmt
+        name: cargo fmt (Rust)
+        # cargo fmt formats the whole crate (it ignores per-file paths), which
+        # is why the repo is kept fully formatted — no unrelated-file churn.
+        entry: cargo fmt
+        language: system
+        types: [rust]
+        pass_filenames: false
+
+      - id: ruff-format
+        name: ruff format (Python)
+        entry: ruff format
+        language: system
+        types_or: [python, pyi]