pre-commit: Ignore time stamps in po files

.github/workflows/pre-commit.yml

@@ -8,23 +8,39 @@ jobs:
   pre-commit:
     runs-on: ubuntu-latest
     steps:
-    - uses: actions/checkout@v4
-    - uses: actions/setup-python@v5
-    - uses: pre-commit/[email protected]
-      env:
-        SKIP: rstcheck
+      - name: "Check out repository"
+        uses: actions/checkout@v4
+        with:
+          # Fetch the previous commit to be able to check for changes
+          fetch-depth: 2
 
-    - name: "Generate patch file"
-      if: failure()
-      run: |
-        git diff-index -p HEAD > "${PATCH_FILE}"
-        [ -s "${PATCH_FILE}" ] && echo "UPLOAD_PATCH_FILE=${PATCH_FILE}" >> "${GITHUB_ENV}"
-      env:
-        PATCH_FILE: pre-commit.patch
+      - name: "Detect code style issues (push)"
+        uses: pre-commit/[email protected]
+        if: github.event_name == 'push'
+        env:
+          SKIP: rstcheck,ignore-pot-creation-date
 
-    - name: "Upload patch artifact"
-      if: failure() && env.UPLOAD_PATCH_FILE != null
-      uses: actions/upload-artifact@v4
-      with:
-        name: ${{ env.UPLOAD_PATCH_FILE }}
-        path: ${{ env.UPLOAD_PATCH_FILE }}
+      - name: "Detect code style issues (pull_request)"
+        uses: pre-commit/[email protected]
+        if: github.event_name == 'pull_request'
+        env:
+          SKIP: rstcheck
+        with:
+          # HEAD is the not yet integrated PR merge commit +refs/pull/xxxx/merge
+          # HEAD^1 is the PR target branch and HEAD^2 is the HEAD of the source branch
+          extra_args: --from-ref HEAD^1 --to-ref HEAD
+
+      - name: "Generate patch file"
+        if: failure()
+        run: |
+          git diff-index -p HEAD > "${PATCH_FILE}"
+          [ -s "${PATCH_FILE}" ] && echo "UPLOAD_PATCH_FILE=${PATCH_FILE}" >> "${GITHUB_ENV}"
+        env:
+          PATCH_FILE: pre-commit.patch
+
+      - name: "Upload patch artifact"
+        if: failure() && env.UPLOAD_PATCH_FILE != null
+        uses: actions/upload-artifact@v4
+        with:
+          name: ${{ env.UPLOAD_PATCH_FILE }}
+          path: ${{ env.UPLOAD_PATCH_FILE }}

.pre-commit-config.yaml

@@ -15,12 +15,18 @@ repos:
     types: [rst]
     entry: python tools/fixup_gh_wiki_anchors.py
     language: python
+  - id: ignore-pot-creation-date
+    name: Ignore POT-Creation-Date
+    entry: python tools/ignore_pot_creation_date.py
+    language: python
+    files: ^.*\.po$
 - repo: https://github.com/pre-commit/pre-commit-hooks
-  rev: v3.1.0
+  rev: v5.0.0
   hooks:
   - id: check-byte-order-marker
   - id: check-case-conflict
   - id: end-of-file-fixer
+    exclude: ^.tx/config$
   - id: mixed-line-ending
   - id: trailing-whitespace
     exclude: ^source/locale/.*$

tools/ignore_pot_creation_date.py

@@ -0,0 +1,96 @@
+import argparse
+import logging
+import os
+import subprocess
+import sys
+import typing
+
+"""
+This script reverts changes in PO files when only the timestamp has changed.
+It helps to create meaningful commits by ignoring unnecessary changes.
+"""
+
+
+def get_git_changeset(from_ref, to_ref) -> str:
+    """
+    Constructs the changeset string for `git diff` based on from_ref and
+    to_ref.
+    """
+    from_ref = (
+        from_ref
+        or os.getenv("PRE_COMMIT_FROM_REF")
+        or os.getenv("PRE_COMMIT_SOURCE")
+        or "HEAD"
+    )
+    to_ref = (
+        to_ref
+        or os.getenv("PRE_COMMIT_TO_REF")
+        or os.getenv("PRE_COMMIT_ORIGIN")
+    )
+    return f"{from_ref}...{to_ref}" if to_ref else from_ref
+
+
+def count_meaningful_changes(changeset, po_file) -> int:
+    """
+    Counts meaningful changes in the diff for a given PO file.
+    """
+    cmd = ["git", "diff", "--patch", "--unified=0", changeset, "--", po_file]
+    proc = subprocess.run(cmd, capture_output=True, text=True)
+    proc.check_returncode()
+
+    diff_lines = proc.stdout.splitlines()
+    return sum(
+        1
+        for line in diff_lines
+        if (line.startswith("-") or line.startswith("+"))
+        and "POT-Creation-Date:" not in line
+        and "PO-Revision-Date:" not in line
+        and po_file not in line
+    )
+
+
+def revert_po_file(changeset, po_file, logger):
+    """
+    Reverts a PO file to its original state in the given changeset.
+    """
+    ref = changeset.split("...")[
+        0
+    ]  # Use the first part of the changeset as the reference
+    logger.info(f"{po_file} has no meaningful changes, reverting to {ref}")
+    cmd = ["git", "show", f"{ref}:{po_file}"]
+    proc = subprocess.run(cmd, capture_output=True, text=True)
+    proc.check_returncode()
+
+    with open(po_file, "w") as file:
+        file.write(proc.stdout)
+
+
+def main(argv: typing.Optional[typing.List[str]] = None) -> int:
+    logging.basicConfig(
+        format="[%(levelname)s] %(message)s", level=logging.INFO
+    )
+    logger = logging.getLogger(__name__)
+
+    parser = argparse.ArgumentParser(
+        description="Revert PO files with only timestamp changes."
+    )
+    parser.add_argument("--from-ref", help="Use changes since this commit.")
+    parser.add_argument("--to-ref", help="Use changes until this commit.")
+    parser.add_argument("files", nargs="*", help="Only check these files.")
+    args = parser.parse_args(argv)
+
+    # Remove the first entry which is the script file name itself
+    files_to_process = args.files[1:] if args.files else []
+
+    changeset = get_git_changeset(args.from_ref, args.to_ref)
+
+    for po_file in files_to_process:
+        count = count_meaningful_changes(changeset, po_file)
+        if count == 0:
+            revert_po_file(changeset, po_file, logger)
+
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())