Scan old imports

git-importer.py

@@ -7,6 +7,8 @@ import sys
 
 import osc.core
 
+from lib.db import DB
+from lib.db_revision import DBRevision
 from lib.git_exporter import GitExporter
 from lib.importer import Importer
 from lib.test_exporter import TestExporter
@@ -100,6 +102,56 @@ def main():
             requests_log.setLevel(logging.DEBUG)
             requests_log.propagate = True
 
+    def check_old_package(db: DB, dir: pathlib.Path):
+        md5file = dir / "MD5SUMS"
+        print(md5file)
+        valid_revisions = None
+        with open(md5file, "rb") as f:
+            for line in f.readlines():
+                try:
+                    md5, file = line.decode("utf-8").strip().split("  ")
+                except UnicodeDecodeError:
+                    logging.error(f"Corrupt MD5 file: {md5file}")
+                    return
+                if file == "ready":
+                    continue
+                if len(md5) != 32:
+                    logging.error(f"Corrupt MD5 file: {md5file}")
+                    return
+                with db.cursor() as cur:
+                    cur.execute(
+                        "SELECT revision_id FROM files WHERE md5=%s AND name=%s",
+                        (md5, file),
+                    )
+                    nrevs = set([row[0] for row in cur.fetchall()])
+                    if valid_revisions is None:
+                        valid_revisions = nrevs
+                    else:
+                        valid_revisions = valid_revisions & nrevs
+                    if not valid_revisions:
+                        break
+
+        with db.cursor() as cur:
+            cur.execute(
+                "SELECT * FROM revisions WHERE id = ANY(%s) AND project=%s",
+                (list(valid_revisions), "openSUSE:Factory"),
+            )
+            for row in cur.fetchall():
+                r = DBRevision(db, row)
+                print("Valid", r, r.files_hash)
+                return True
+
+    if False:
+        import os
+
+        db = DB()
+        basedir = pathlib.Path(
+            f"/mounts/work/SAVE/oldpackages/stable/{args.packages[0]}"
+        )
+        for subdir in sorted(os.listdir(basedir)):
+            if check_old_package(db, basedir / subdir):
+                break
+
     if args.export:
         if len(args.packages) != 1:
             print("Can only export one package")