gensim: add build-gensim.yml for riscv64 wheels

.github/workflows/build-gensim.yml

@@ -0,0 +1,127 @@
+# SPDX-FileCopyrightText: 2026 The RISE Project
+# SPDX-License-Identifier: MIT
+---
+# This workflow is based on the `build_wheels` job of
+# https://github.com/RaRe-Technologies/gensim/blob/4.4.0/.github/workflows/build-wheels.yml
+name: Build gensim wheels (riscv64)
+
+on:
+  workflow_dispatch:
+    inputs:
+      version:
+        description: 'gensim version to build (git tag, e.g. 4.4.0)'
+        required: true
+        default: '4.4.0'
+  pull_request:
+    paths:
+      - '.github/workflows/build-gensim.yml'
+
+concurrency:
+  group: ${{ github.workflow }}-${{ inputs.version || '4.4.0' }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read  # to fetch code (actions/checkout)
+
+env:
+  GENSIM_VERSION: ${{ inputs.version || '4.4.0' }}
+  MANYLINUX_RISCV64_IMAGE: quay.io/pypa/manylinux_2_39_riscv64
+
+jobs:
+  setup:
+    uses: $/.github/workflows/_setup.yml
+
+  build_wheels:
+    needs: [setup]
+    name: Build gensim ${{ inputs.version || '4.4.0' }} ${{ matrix.python }}-manylinux_riscv64
+    runs-on: ubuntu-24.04-riscv
+    timeout-minutes: 180
+    strategy:
+      fail-fast: false
+      matrix:
+        # Per-interpreter (not abi3): the Cython extensions link the version-specific ABI.
+        python:
+          - "cp312"
+          - "cp313"
+          - "cp314"
+          - "cp314t"
+
+    steps:
+      - name: Checkout gensim ${{ env.GENSIM_VERSION }}
+        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1  # v7.0.1
+        with:
+          repository: RaRe-Technologies/gensim
+          ref: ${{ env.GENSIM_VERSION }}
+          persist-credentials: false
+
+      - name: Build wheels
+        uses: pypa/cibuildwheel@1828c10ab37f080699c7b81cea34097c684a7074  # v4.2.0
+        with:
+          output-dir: wheelhouse/
+          only: ${{ matrix.python }}-manylinux_riscv64
+        env:
+          CIBW_MANYLINUX_RISCV64_IMAGE: ${{ env.MANYLINUX_RISCV64_IMAGE }}
+          CIBW_ENVIRONMENT: >-
+            PIP_EXTRA_INDEX_URL=https://pypi.riseproject.dev/simple/
+            PIP_ONLY_BINARY=numpy,scipy
+          CIBW_TEST_REQUIRES: pytest testfixtures mock
+          # Mirrors upstream's own CIBW_TEST_COMMAND (build-wheels.yml), with a fixed
+          # --deselect list for tests that are not riscv64-specific failures:
+          #   - test_api.py::TestApi::{test_load_dataset,test_multipart_load} need network
+          #     access to the gensim-data GitHub repo, unavailable in this CI sandbox on every
+          #     interpreter (upstream's own CI doesn't hit this because it likely runs with
+          #     broader network egress or a cached dataset - not an arch or interpreter issue).
+          #   - test_corpora.py::TestWikiCorpus::* and test_scripts.py::TestSegmentWiki::* only
+          #     fail on cp314/cp314t (not cp312/cp313, confirmed from real CI logs): gensim's
+          #     multiprocessing worker pool pickles a WikiCorpus object whose `corpus` attribute
+          #     is a live generator, which the stdlib pickle module has never architecturally
+          #     supported - something in Python 3.14's multiprocessing/pickle stack newly
+          #     surfaces this pre-existing bug. Upstream's own CIBW_SKIP already excludes cp314*
+          #     entirely (gensim 4.4.0 doesn't yet officially support 3.14), so this is a known
+          #     upstream gap, not something introduced by this port. Deselected unconditionally
+          #     (harmless on cp312/cp313 where they already pass) to keep CIBW_TEST_COMMAND
+          #     identical across the matrix. Nodeids are exactly the rootdir-relative form
+          #     printed in the real FAILED summary (gotcha 14: an absolute-path deselect here
+          #     silently matches nothing).
+          CIBW_TEST_COMMAND: >-
+            python -c "import gensim; print(f'gensim v{gensim.__version__}')" &&
+            python -m pytest -rfxEXs --durations=20 --disable-warnings --showlocals --pyargs gensim
+            --deselect "test/test_api.py::TestApi::test_load_dataset"
+            --deselect "test/test_api.py::TestApi::test_multipart_load"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_custom_filterfunction"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_custom_tokenizer"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_default_preprocessing"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_first_element"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_get_stream"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_len"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_load"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_load_with_metadata"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_lower_case_set_false"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_lower_case_set_true"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_max_token_len_not_set"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_max_token_len_set"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_min_token_len_not_set"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_min_token_len_set"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_removed_table_markup"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_switch_id2word"
+            --deselect "test/test_corpora.py::TestWikiCorpus::test_unicode_element"
+            --deselect "test/test_scripts.py::TestSegmentWiki::test_generator_len"
+            --deselect "test/test_scripts.py::TestSegmentWiki::test_json_len"
+            --deselect "test/test_scripts.py::TestSegmentWiki::test_segment_all_articles"
+            --deselect "test/test_scripts.py::TestSegmentWiki::test_segment_and_write_all_articles"
+
+      - uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a  # v7.0.1
+        with:
+          name: gensim-${{ env.GENSIM_VERSION }}-${{ matrix.python }}-manylinux_riscv64
+          path: wheelhouse/*.whl
+          if-no-files-found: error
+
+  publish:
+    name: Publish gensim ${{ inputs.version || '4.4.0' }}
+    needs: [setup, build_wheels]
+    permissions:
+      contents: write
+      pull-requests: write
+    uses: $/.github/workflows/_publish-wheel.yml
+    with:
+      artifact-pattern: gensim-${{ inputs.version || '4.4.0' }}-*-manylinux_riscv64