build-daft.yml

# SPDX-FileCopyrightText: 2026 The RISE Project
# SPDX-License-Identifier: MIT
---
# This workflow is based on the `build` job of
# https://github.com/Eventual-Inc/Daft/blob/v0.7.24/.github/workflows/build-wheel.yml
name: Build daft wheels (riscv64)

on:
  workflow_dispatch:
    inputs:
      version:
        description: 'Version glob to (re)build; empty builds every version of docs/packages/daft.yaml not released yet'
        required: false
        default: ''
  pull_request:
    branches: [main]
    paths:
      - '.github/workflows/build-daft.yml'
      - 'docs/packages/daft.yaml'
  push:
    branches: [main]
    paths:
      - '.github/workflows/build-daft.yml'
      - 'docs/packages/daft.yaml'

concurrency:
  group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
  cancel-in-progress: true

permissions:
  contents: read  # to fetch code (actions/checkout)

env:
  CARGO_INCREMENTAL: 0
  CARGO_NET_RETRY: 10
  RUSTUP_MAX_RETRIES: 10
  # Limit parallel rustc invocations to avoid resource exhaustion on the
  # riscv64 runner (build-lancedb.yml/build-polars-runtime.yml need the same).
  CARGO_BUILD_JOBS: 2

jobs:
  setup:
    uses: $/.github/workflows/_setup.yml
    with:
      package: daft
      version: ${{ inputs.version }}

  build_wheels:
    needs: [setup]
    if: needs.setup.outputs.versions != '[]'
    strategy:
      fail-fast: false
      matrix:
        version: ${{ fromJSON(needs.setup.outputs.versions) }}
    name: Build daft ${{ matrix.version }} cp310-abi3-manylinux_riscv64
    runs-on: ubuntu-24.04-riscv
    timeout-minutes: 1440

    env:
      DAFT_VERSION: ${{ matrix.version }}

    steps:
      - name: Checkout Daft v${{ env.DAFT_VERSION }}
        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1  # v7.0.1
        with:
          repository: Eventual-Inc/Daft
          ref: v${{ env.DAFT_VERSION }}
          persist-credentials: false

      # Upstream derives this from `setuptools_scm` and writes it into
      # Cargo.toml before building (build-wheel.yml's "Patch package version"
      # step) because their CI also builds untagged/nightly commits. We only
      # ever build an exact release tag, so the version is already known; the
      # workspace-wide placeholder is the only place it needs to land.
      - name: Patch package version
        run: |
          awk -v ver="${DAFT_VERSION}" '
            /^\[workspace\.package\]/ { in_block=1 }
            /^\[/ && !/^\[workspace\.package\]/ { in_block=0 }
            in_block && /^version = / { print "version = \"" ver "\""; next }
            { print }
          ' Cargo.toml > Cargo.toml.new
          mv Cargo.toml.new Cargo.toml
          grep -qx "version = \"${DAFT_VERSION}\"" Cargo.toml

      - name: Free disk space
        uses: jlumbroso/free-disk-space@54081f138730dfa15788a46383842cd2f914a1be  # v1.3.1

      # Same OOM guard build-lancedb.yml/build-deltalake.yml need on these
      # runners for a comparably sized Rust workspace.
      - name: Set swap space
        uses: pierotofy/set-swap-space@fc79b3f67fa8a838184ce84a674ca12238d2c761  # master
        with:
          swap-size-gb: 16

      - name: Build wheel
        uses: PyO3/maturin-action@e83996d129638aa358a18fbd1dfb82f0b0fb5d3b  # v1.51.0
        with:
          command: build
          target: riscv64gc-unknown-linux-gnu
          # pyo3 carries `abi3-py310` unconditionally in Cargo.toml, so
          # upstream ships one cp310-abi3 wheel and no free-threaded variant;
          # build on the floor interpreter the tag claims (gotcha 181).
          # `rust-toolchain.toml` pins a nightly channel (daft-core and
          # friends use unstable `#![feature(...)]`); maturin-action picks it
          # up on its own, same as upstream's job does.
          args: -i python3.10 --profile release --strip --out dist
          manylinux: '2_39'
          before-script-linux: |
            git config --global --add safe.directory "*"
            # The dashboard's `build.rs` requires Node/npm for release builds
            # unless this is set, in which case it embeds an empty asset
            # bundle instead; avoids needing a riscv64 Node toolchain for a
            # bundled web dashboard nothing in the test suite exercises.
            export CI=true
            export DAFT_DASHBOARD_SKIP_BUILD=1
        env:
          DAFT_ANALYTICS_ENABLED: '0'

      - name: Install Python
        uses: astral-sh/setup-uv@20cfd1bf945f4377ade1205e4dbc17946fc9a30d  # v10.0.1
        with:
          # cp312 is the floor our registry currently publishes riscv64
          # numpy/pandas/pyarrow builds compatible with daft's own extras
          # bounds for (the abi3 wheel under test installs fine on any
          # interpreter >= the cp310 floor it claims).
          python-version: '3.12'
          activate-environment: true
          enable-cache: false

      # The system `pip` on this runner predates riscv64 manylinux tag
      # support and rejects the wheel as "not a supported wheel on this
      # platform"; uv's resolver knows the tag (same reason
      # build-polars-runtime.yml's "Test wheel" step uses uv too).
      # Upstream's test-wheels job also runs this suite under `DAFT_RUNNER=ray`;
      # that pass is dropped here because the only riscv64 ray build published
      # anywhere is 2.58.0 (cp312/cp313/cp314 only), which exceeds the
      # `ray[data, client]<2.58.0,>=2.11.0` upper bound daft's own `ray`
      # extra declares, so there is no ray build daft supports to test with.
      - name: Test wheel
        run: |
          uv pip install pytest pandas==2.3.3 pytz numpy==2.3.4 pyarrow==25.0.1 dist/*.whl --force-reinstall
          # `daft/` at the checkout root would otherwise shadow the installed
          # wheel (same reason upstream's own test-wheels job removes it).
          rm -rf daft
          DAFT_ANALYTICS_ENABLED=0 DAFT_RUNNER=native python -m pytest tests/dataframe --durations=50
        env:
          # `uv pip install` (unlike cibuildwheel's pip-based test step) only
          # honors its own UV_* env vars, not PIP_*; the PIP_-prefixed forms
          # silently no-op, so uv skipped our registry, resolved pyarrow only
          # against PyPI (no riscv64 wheel there) and source-built it, which
          # needs the real Arrow C++ CMake package the manylinux image
          # doesn't have (gotcha 244). `unsafe-best-match` is needed on top
          # of the extra index because uv's default "first-index" strategy
          # stops at the first index that lists the package name at all,
          # never reaching our registry's riscv64 build (build-matplotlib.yml
          # uses the same pair for the same reason).
          UV_EXTRA_INDEX_URL: https://pypi.riseproject.dev/simple/
          UV_INDEX_STRATEGY: unsafe-best-match
          UV_ONLY_BINARY: numpy,pandas,pyarrow

      - uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a  # v7.0.1
        with:
          name: daft-${{ env.DAFT_VERSION }}-cp310-abi3-manylinux_riscv64
          path: dist/*.whl
          if-no-files-found: error

  publish:
    name: Publish daft ${{ matrix.version }}
    needs: [setup, build_wheels]
    if: needs.setup.outputs.versions != '[]'
    strategy:
      fail-fast: false
      matrix:
        version: ${{ fromJSON(needs.setup.outputs.versions) }}
    permissions:
      contents: write
      pull-requests: write
    uses: $/.github/workflows/_publish-wheel.yml
    secrets:
      app-private-key: ${{ secrets.RISEPROJECT_APP_PRIVATE_KEY }}
    with:
      artifact-pattern: daft-${{ matrix.version }}-*-manylinux_riscv64