Symbolic growth, exact Pareto path search, and deterministic solver backends #1792
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: CI | |
| on: | |
| push: | |
| branches: [ main, develop ] | |
| pull_request: | |
| branches: [ main, develop ] | |
| # Cancel superseded runs on the same ref (e.g. rapid pushes to a PR). | |
| concurrency: | |
| group: ${{ github.workflow }}-${{ github.ref }} | |
| cancel-in-progress: true | |
| # Least privilege: jobs only read the repo; codecov auth is via token secret. | |
| permissions: | |
| contents: read | |
| env: | |
| CARGO_TERM_COLOR: always | |
| RUST_BACKTRACE: 1 | |
| # Incremental compilation hurts cold CI builds and bloats the cache; disable it | |
| # (recommended by Swatinem/rust-cache). | |
| CARGO_INCREMENTAL: 0 | |
| # Tolerate transient registry/network blips. | |
| CARGO_NET_RETRY: 10 | |
| RUSTUP_MAX_RETRIES: 10 | |
| jobs: | |
| # Formatting — cheap, fails fast, no build/cache needed. | |
| fmt: | |
| name: Rustfmt | |
| runs-on: ubuntu-latest | |
| steps: | |
| - uses: actions/checkout@v5 | |
| - uses: dtolnay/rust-toolchain@stable | |
| with: | |
| components: rustfmt | |
| - name: Check formatting | |
| run: cargo fmt --all --check | |
| # Lints. | |
| clippy: | |
| name: Clippy | |
| runs-on: ubuntu-latest | |
| steps: | |
| - uses: actions/checkout@v5 | |
| - uses: dtolnay/rust-toolchain@stable | |
| with: | |
| components: clippy | |
| - uses: Swatinem/rust-cache@v2 | |
| - name: Run clippy | |
| run: cargo clippy --all-targets --features ilp-highs -- -D warnings | |
| # Cross-compile the portable Rust surface to RISC-V Linux, then execute the | |
| # CLI under QEMU user-mode emulation to verify runtime behavior (not just | |
| # that the binary links). | |
| riscv: | |
| name: RISC-V build & run | |
| runs-on: ubuntu-latest | |
| env: | |
| FEATURES: "ilp-lp-solvers" | |
| steps: | |
| - uses: actions/checkout@v5 | |
| - uses: dtolnay/rust-toolchain@stable | |
| with: | |
| targets: riscv64gc-unknown-linux-gnu | |
| - uses: Swatinem/rust-cache@v2 | |
| - name: Install RISC-V cross tools and QEMU | |
| run: | | |
| sudo apt-get update | |
| sudo apt-get install -y gcc-riscv64-linux-gnu binutils-riscv64-linux-gnu qemu-user | |
| - name: Build workspace for RISC-V | |
| env: | |
| CARGO_TARGET_RISCV64GC_UNKNOWN_LINUX_GNU_LINKER: riscv64-linux-gnu-gcc | |
| run: cargo build --workspace --no-default-features --features "$FEATURES" --target riscv64gc-unknown-linux-gnu | |
| - name: Verify RISC-V executable | |
| run: | | |
| riscv64-linux-gnu-readelf -h target/riscv64gc-unknown-linux-gnu/debug/pred \ | |
| | grep 'Machine:.*RISC-V' | |
| - name: Run CLI smoke test under QEMU | |
| env: | |
| PRED: qemu-riscv64 -L /usr/riscv64-linux-gnu target/riscv64gc-unknown-linux-gnu/debug/pred | |
| run: | | |
| # Registry loads and the catalog renders on RISC-V. | |
| $PRED list | head -5 | |
| # End-to-end create + brute-force solve: MIS of a 5-cycle is 2. | |
| $PRED create MaximumIndependentSet --graph 0-1,1-2,2-3,3-4,4-0 -o mis.json | |
| $PRED solve mis.json --solver brute-force | tee solve.out | |
| grep -q '"evaluation": "Max(2)"' solve.out | |
| # Build, test (nextest), doc tests, and paper. | |
| test: | |
| name: Test | |
| runs-on: ubuntu-latest | |
| # Single feature set across compile + test + doctest so artifacts are reused | |
| # (no redundant full recompile between steps). | |
| env: | |
| FEATURES: "ilp-highs example-db" | |
| steps: | |
| - uses: actions/checkout@v5 | |
| - uses: dtolnay/rust-toolchain@stable | |
| - uses: taiki-e/install-action@v2 | |
| with: | |
| tool: nextest | |
| - uses: Swatinem/rust-cache@v2 | |
| - name: Install typst | |
| uses: typst-community/setup-typst@v5 | |
| - name: Compile tests | |
| run: cargo nextest run --no-run --workspace --features "$FEATURES" | |
| # The subprocess example tests (tests/suites/examples.rs) shell out to | |
| # `cargo run --example … --features ilp-highs`. Pre-build those example | |
| # binaries with that exact feature set so the subprocess reuses artifacts | |
| # instead of recompiling the whole crate mid-test (which otherwise adds | |
| # 60s+ to a single test's wall-clock and would trip the nextest timeout). | |
| - name: Build examples (for subprocess tests) | |
| run: cargo build --examples --features ilp-highs | |
| - name: Run tests | |
| run: cargo nextest run --workspace --features "$FEATURES" | |
| # nextest does not run doc tests; run them separately (reuses the build). | |
| - name: Run doc tests | |
| run: cargo test --doc --features "$FEATURES" --verbose | |
| - name: Build paper | |
| run: make paper | |
| # Coverage. Feature set intentionally matches the historical coverage gate | |
| # (ilp-highs only) to keep the codecov baseline stable. | |
| coverage: | |
| name: Code Coverage | |
| runs-on: ubuntu-latest | |
| steps: | |
| - uses: actions/checkout@v5 | |
| - uses: dtolnay/rust-toolchain@stable | |
| with: | |
| components: llvm-tools-preview | |
| - uses: taiki-e/install-action@v2 | |
| with: | |
| tool: cargo-llvm-cov,nextest | |
| - uses: Swatinem/rust-cache@v2 | |
| - name: Generate coverage | |
| run: cargo llvm-cov nextest --features ilp-highs --workspace --lcov --output-path lcov.info | |
| - name: Upload to codecov.io | |
| uses: codecov/codecov-action@v5 | |
| with: | |
| files: lcov.info | |
| fail_ci_if_error: false # Don't fail CI if upload fails | |
| token: ${{ secrets.CODECOV_TOKEN }} |