datasketches: add build-datasketches.yml for riscv64 wheels

Author: luhenry

.github/workflows/build-datasketches.yml

@@ -0,0 +1,107 @@
+# SPDX-FileCopyrightText: 2026 The RISE Project
+# SPDX-License-Identifier: MIT
+---
+# Mirrors the Linux core of upstream's .github/workflows/build_wheels.yml.
+name: Build datasketches wheels (riscv64)
+
+on:
+  workflow_dispatch:
+    inputs:
+      version:
+        description: 'datasketches version to build (git tag, e.g. 5.2.0)'
+        required: true
+        default: '5.2.0'
+  pull_request:
+    paths:
+      - '.github/workflows/build-datasketches.yml'
+      - 'patches/datasketches/**'
+
+concurrency:
+  group: ${{ github.workflow }}-${{ inputs.version || '5.2.0' }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read  # to fetch code (actions/checkout)
+
+env:
+  DATASKETCHES_VERSION: ${{ inputs.version || '5.2.0' }}
+  MANYLINUX_RISCV64_IMAGE: quay.io/pypa/manylinux_2_39_riscv64
+  MUSLLINUX_RISCV64_IMAGE: quay.io/pypa/musllinux_1_2_riscv64
+
+jobs:
+  setup:
+    uses: $/.github/workflows/_setup.yml
+
+  build_wheels:
+    needs: [setup]
+    name: Build datasketches ${{ inputs.version || '5.2.0' }} ${{ matrix.python }}-${{ matrix.libc }}_riscv64
+    runs-on: ubuntu-24.04-riscv
+    timeout-minutes: 120
+    strategy:
+      fail-fast: false
+      matrix:
+        python: ["cp312", "cp313", "cp314", "cp314t"]
+        libc: [manylinux, musllinux]
+
+    steps:
+      - name: Checkout apache/datasketches-python ${{ env.DATASKETCHES_VERSION }}
+        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1  # v7.0.1
+        with:
+          repository: apache/datasketches-python
+          ref: ${{ env.DATASKETCHES_VERSION }}
+          persist-credentials: false
+
+      - name: Checkout python-wheels
+        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1  # v7.0.1
+        with:
+          path: python-wheels
+          persist-credentials: false
+
+      # vector_of_kll_sketches's own operator= assigns to its const k_/d_ members,
+      # which GCC 13 rejects when instantiating the nanobind bindings (not riscv64-specific,
+      # reproduces on any manylinux arch; gotcha 23 shape).
+      - name: Patch datasketches source
+        run: git apply python-wheels/patches/datasketches/${{ env.DATASKETCHES_VERSION }}/00*.patch
+
+      - uses: pypa/cibuildwheel@1828c10ab37f080699c7b81cea34097c684a7074  # v4.2.0
+        env:
+          CIBW_ARCHS: riscv64
+          CIBW_BUILD: ${{ matrix.python }}-${{ matrix.libc }}_riscv64
+          CIBW_MANYLINUX_RISCV64_IMAGE: ${{ env.MANYLINUX_RISCV64_IMAGE }}
+          CIBW_MUSLLINUX_RISCV64_IMAGE: ${{ env.MUSLLINUX_RISCV64_IMAGE }}
+          CIBW_ENVIRONMENT: PIP_EXTRA_INDEX_URL=https://pypi.riseproject.dev/simple/
+          CIBW_TEST_REQUIRES: pytest
+          # tests/ has __init__.py, and the checkout root's own datasketches/ package
+          # (source, no compiled _datasketches) would shadow the installed wheel if
+          # pytest ran from the checkout (gotcha 25); stage only tests/ into an empty cwd.
+          CIBW_TEST_SOURCES: tests
+          CIBW_TEST_COMMAND: python -m pytest tests
+
+      - name: Check wheel contents
+        run: |
+          python3 - wheelhouse/*.whl <<'EOF'
+          import sys, zipfile
+          path = sys.argv[1]
+          names = zipfile.ZipFile(path).namelist()
+          exts = [n for n in names if n.endswith(".so")]
+          assert exts and all("_datasketches" in n for n in exts), exts
+          licences = {n.split(".dist-info/licenses/", 1)[1] for n in names
+                      if ".dist-info/licenses/" in n and not n.endswith("/")}
+          assert "LICENSE" in licences, licences
+          EOF
+
+      - uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a  # v7.0.1
+        with:
+          name: datasketches-${{ env.DATASKETCHES_VERSION }}-${{ matrix.python }}-${{ matrix.libc }}_riscv64
+          path: ./wheelhouse/*.whl
+          if-no-files-found: error
+
+  publish:
+    name: Publish datasketches ${{ inputs.version || '5.2.0' }}
+    needs: [setup, build_wheels]
+    permissions:
+      contents: write
+      pull-requests: write
+    uses: $/.github/workflows/_publish-wheel.yml
+    with:
+      artifact-pattern: datasketches-${{ inputs.version || '5.2.0' }}-*riscv64

patches/datasketches/5.2.0/0001-Drop-const-from-vector_of_kll_sketches-k_-d_-so-ass.patch

@@ -0,0 +1,47 @@
+From 3a3ab48c3cae8ccb6c07db0a728cec4060821f7c Mon Sep 17 00:00:00 2001
+From: Ludovic Henry <git@ludovic.dev>
+Date: Tue, 8 Sep 2026 18:46:43 +0200
+Subject: [PATCH] Drop const from vector_of_kll_sketches k_/d_ so assignment
+ operators compile
+
+vector_of_kll_sketches declares k_ and d_ const, but its own out-of-line
+operator= (copy) and operator= (move) both assign to them
+(k_ = copy.k_; d_ = copy.d_;), which is ill-formed - assigning to a
+const member outside of initialization. GCC 13 (the manylinux_riscv64
+image's compiler) instantiates and rejects both operators while
+building the vector_of_kll_ints_sketches/vector_of_kll_floats_sketches
+nanobind bindings:
+
+  src/vector_of_kll.cpp:146:6: error: assignment of read-only member
+  'datasketches::vector_of_kll_sketches<T, C>::k_'
+
+Reproduces identically with the same compiler on manylinux_2_39_aarch64
+(no riscv64 involvement) and is unrelated to nanobind's version: this
+tag's CMakeLists.txt pins no nanobind ceiling, main is unchanged since
+this file was last touched, and it is not riscv64-specific.
+
+Upstream-Status: To upstream [python-wheels only files issues/PRs against riseproject-dev/python-wheels; not yet reported to apache/datasketches-python]
+
+Signed-off-by: Ludovic Henry <git@ludovic.dev>
+---
+ src/vector_of_kll.cpp | 4 ++--
+ 1 file changed, 2 insertions(+), 2 deletions(-)
+
+diff --git a/src/vector_of_kll.cpp b/src/vector_of_kll.cpp
+index 3bbe17b..7013915 100644
+--- a/src/vector_of_kll.cpp
++++ b/src/vector_of_kll.cpp
+@@ -104,8 +104,8 @@ class vector_of_kll_sketches {
+     template<typename TT>
+     Array2D<TT> make_ndarray(size_t rows, size_t cols) const;
+ 
+-    const uint32_t k_; // kll sketch k parameter
+-    const uint32_t d_; // number of dimensions (here: sketches) to hold
++    uint32_t k_; // kll sketch k parameter
++    uint32_t d_; // number of dimensions (here: sketches) to hold
+     std::vector<kll_sketch<T, C>> sketches_;
+ };
+ 
+-- 
+2.50.1 (Apple Git-155)
+