Build: Add build step to deduplicate release wheels by architecture.

.github/workflows/wheels.yml

@@ -134,10 +134,25 @@ jobs:
           path: ./wheelhouse/*.whl
           name: wheels-${{ matrix.only }}
 
+  merge_wheels:
+    name: Merge wheel archives
+    needs: build_wheels
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Merge wheels
+      uses: actions/upload-artifact/merge@v4
+      with:
+        name: all_wheels
+        pattern: wheels-*
+        delete-merged: true
+        compression-level: 9
+
   upload_release_assets:
     name: Upload packages
-    needs: [ sdist, build_wheels ]
+    needs: [ sdist, merge_wheels ]
     runs-on: ubuntu-latest
+    if: github.ref_type == 'tag'
 
     permissions:
       contents: write
@@ -152,13 +167,10 @@ jobs:
       - name: List downloaded artifacts
         run: ls -la ./dist_downloads
 
-      - uses: actions/upload-artifact@694cdabd8bdb0f10b2cea11669e1bf5453eed0a6 # v4.2.0
-        with:
-          path: ./dist_downloads/*.whl
-          name: all_wheels
+      - name: Deduplicate wheels
+        run: python3 dedup_wheels.py -d ./dist_downloads
 
       - name: Release
-        if: github.ref_type == 'tag'
         uses: softprops/action-gh-release@v2
         with:
-          files: ./dist_downloads/*.whl
+          files: ./dist_downloads/*

dedup_wheels.py

@@ -0,0 +1,80 @@
+"""
+Split wheels by set of architectures and delete redundant ones.
+"""
+
+import argparse
+import os
+import pathlib
+import re
+import sys
+
+from collections import defaultdict
+from typing import Iterable
+
+
+def list_wheels(wheel_dir: pathlib.Path):
+    return (wheel.name for wheel in wheel_dir.glob("*.whl"))
+
+
+def dedup_wheels(wheel_names: Iterable[str]):
+    split_wheelname = re.compile(r"(?P<project>\w+-[0-9a-z.]+)-(?P<python>[^-]+-[^-]+)-(?P<archs>(?:[^.]+[.])+)whl").match
+
+    keep: set = set()
+    seen: dict[tuple[str,str],list[set[str]]] = defaultdict(list)
+    best: dict[tuple[str,str],list[str]] = defaultdict(list)
+
+    for wheel in sorted(wheel_names, key=len, reverse=True):
+        parts = split_wheelname(wheel)
+        if not parts:
+            keep.add(wheel)
+            continue
+
+        archs = set(parts['archs'].rstrip('.').split('.'))
+        key = (parts['project'], parts['python'])
+        for archs_seen in seen[key]:
+            if not (archs - archs_seen):
+                break
+        else:
+            seen[key].append(archs)
+            best[key].append(wheel)
+
+    keep.update(wheel for wheel_list in best.values() for wheel in wheel_list)
+    return sorted(keep)
+
+
+def print_wheels(wheel_names: Iterable[str]):
+    for wheel in sorted(wheel_names):
+        print(f"  {wheel}")
+
+
+def main(wheel_dir: str, delete=True):
+    wheel_path = pathlib.Path(wheel_dir)
+    all_wheels = list(list_wheels(wheel_path))
+    wheels = dedup_wheels(all_wheels)
+    redundant = set(all_wheels).difference(wheels)
+    if not redundant:
+        return
+
+    print("Redundant wheels found:")
+    print_wheels(redundant)
+
+    if delete:
+        for wheel in redundant:
+            print(f"deleting {wheel}")
+            os.unlink(wheel_path / wheel)
+
+
+def parse_args(args):
+    parser = argparse.ArgumentParser()
+    parser.add_argument("directory")
+    parser.add_argument("-d", "--delete", action="store_true")
+
+    return parser.parse_args(args)
+
+
+if __name__ == "__main__":
+    if len(sys.argv) > 1:
+        args = parse_args(sys.argv[1:])
+        main(args.directory, delete=args.delete)
+    else:
+        print(f"Usage: {sys.argv[0]} WHEEL_DIR", file=sys.stderr)