Add --reproduce

.github/workflows/ci.yml

@@ -37,9 +37,6 @@ jobs:
       - name: Check repository metadata is up-to-date
         run: |
           git config --global --add safe.directory '*'
-          shopt -s globstar
-          # Parse the value out of <revision></revision>
-          export SOURCE_DATE_EPOCH=$(grep -m 1 "revision" public/**/repomd.xml | cut -d '>' -f 2 | cut -d '<' -f 1)
-          ./tools/publish-real
+          ./tools/publish-real --reproduce
           git status
           git diff --exit-code

tools/publish-real

@@ -6,37 +6,54 @@ copied into public/ and metadata is generated there.
 import os
 import shutil
 import subprocess
+import sys
 from pathlib import Path
+import xml.etree.ElementTree as ET
+
+
+def fetch_reproduce_timestamp(public: Path) -> int:
+    repomd = next(public.glob("workstation/dom0/*/repodata/repomd.xml"))
+    tree = ET.parse(repomd)
+    root = tree.getroot()
+    revision = root.find('repo:revision', {'repo': 'http://linux.duke.edu/metadata/repo'})
+    print(f"Will use a timestamp of {revision.text} (from {repomd})")
+    return int(revision.text)
 
 
 def main():
     root = Path(__file__).parent.parent
     public = root / "public"
     workstation = root / "workstation"
-    source_date_epoch = int(os.environ.get("SOURCE_DATE_EPOCH", 0))
+    if "--reproduce" in sys.argv:
+        try:
+            timestamp = fetch_reproduce_timestamp(public)
+        except Exception as err:
+            raise RuntimeError("Failed to fetch timestamp from repomd.xml") from err
+    else:
+        timestamp = None
     # Reset public, copy the workstation/ tree into it
     print("Creating public/ (from scratch)")
     if public.exists():
         shutil.rmtree(public)
     public.mkdir()
     shutil.copytree(workstation, public / "workstation")
-    if source_date_epoch:
+    if timestamp:
         for rpm in public.glob("**/*.rpm"):
-            os.utime(rpm, (source_date_epoch, source_date_epoch))
+            os.utime(rpm, (timestamp, timestamp))
     # Folders are public/workstation/dom0/fXX, run createrepo_c in each one
     for folder in public.glob("*/*/*/"):
         if not folder.is_dir():
             continue
         print(f"Generating metadata for {folder}")
         args = ["createrepo_c"]
-        if source_date_epoch:
+        if timestamp:
             # The <revision> and <timestamp> fields are set to the current UNIX time
             # unless we explicitly override them. In most cases we want to use
             # the current time except when we're doing reproducibility testing.
             args.extend(
                 [
                     "--revision",
-                    str(source_date_epoch),
+                    str(timestamp),
                     "--set-timestamp-to-revision",
                 ]
             )