Reset branch if the devel branch is based on Factory

This happens in packages that change their devel project over time. Then
the commit in the devel project no longer has the parent in the devel branch
but is based on factory

git-importer.py

@@ -105,7 +105,9 @@ def main():
 
     importer = Importer(URL_OBS, "openSUSE:Factory", args.package)
     importer.import_into_db()
-    exporter = GitExporter(URL_OBS, "openSUSE:Factory", args.package, args.repodir, args.cachedir)
+    exporter = GitExporter(
+        URL_OBS, "openSUSE:Factory", args.package, args.repodir, args.cachedir
+    )
     exporter.set_gc_interval(args.gc)
     exporter.export_as_git()
 

lib/git.py

@@ -184,6 +184,9 @@ class Git:
     def branch_head(self, branch):
         return self.repo.references["refs/heads/" + branch].target
 
+    def set_branch_head(self, branch, commit):
+        self.repo.references["refs/heads/" + branch].set_target(commit)
+
     def gc(self):
         logging.info(f"Garbage recollect and repackage {self.path}")
         subprocess.run(

lib/git_exporter.py

@@ -28,22 +28,6 @@ class GitExporter:
         self.gc_interval = 200
         self.cachedir = cachedir
 
-    def download(self, revision):
-        obs_files = self.obs.files(revision.project, revision.package, revision.srcmd5)
-        git_files = {
-            (f.name, f.stat().st_size, md5(f))
-            for f in self.git.path.iterdir()
-            if f.is_file() and f.name not in (".gitattributes")
-        }
-
-        # Overwrite ".gitattributes" with the
-        self.git.add_default_lfs_gitattributes(force=True)
-
-        # Download each file in OBS if it is not a binary (or large)
-        # file
-        for name in obs_files:
-            self.git.add(name)
-
     def set_gc_interval(self, gc):
         self.gc_interval = gc
 
@@ -85,44 +69,58 @@ class GitExporter:
             logging.debug(f"Committing {flat}")
             self.commit_flat(flat, branch_state)
 
+    def commit_file(self, flat, file, size, md5):
+        # have such files been detected as text mimetype before?
+        is_text = self.proxy_sha256.is_text(file.name)
+        if not is_text and is_binary_or_large(file.name, size):
+            file_sha256 = self.proxy_sha256.get_or_put(
+                flat.commit.project,
+                flat.commit.package,
+                file.name,
+                flat.commit.expanded_srcmd5,
+                md5,
+                size,
+            )
+            self.git.add_lfs(file.name, file_sha256["sha256"], size)
+        else:
+            self.obs.download(
+                flat.commit.project,
+                flat.commit.package,
+                file.name,
+                flat.commit.expanded_srcmd5,
+                self.git.path,
+                self.cachedir,
+                file_md5=md5,
+            )
+            self.git.add(file)
+
+    def branch_fits_parent1(self, flat, branch_state):
+        if branch_state[flat.branch] is None:
+            # everything fits nothing
+            return True
+        return flat.parent1 == branch_state[flat.branch]
+
     def commit_flat(self, flat, branch_state):
         parents = []
         self.git.checkout(flat.branch)
 
+        if flat.parent1:
+            if not self.branch_fits_parent1(flat, branch_state):
+                logging.info(f"Reset {flat.branch} onto {flat.parent1.short_string()}")
+                self.git.set_branch_head(flat.branch, flat.parent1.git_commit)
+                self.git.checkout(flat.branch)
+            parents.append(flat.parent1.git_commit)
+            if flat.parent2:
+                parents.append(flat.parent2.git_commit)
+
         # Overwrite ".gitattributes" with the
         self.git.add_default_lfs_gitattributes(force=True)
 
-        if flat.parent1:
-            parents.append(flat.parent1.git_commit)
-        if flat.parent2:
-            parents.append(flat.parent2.git_commit)
         to_download, to_delete = flat.commit.calc_delta(branch_state[flat.branch])
         for file in to_delete:
             self.git.remove(file)
         for file, size, md5 in to_download:
-            # have such files been detected as text mimetype before?
+            self.commit_file(flat, file, size, md5)
-            is_text = self.proxy_sha256.is_text(file.name)
-            if not is_text and is_binary_or_large(file.name, size):
-                file_sha256 = self.proxy_sha256.get_or_put(
-                    flat.commit.project,
-                    flat.commit.package,
-                    file.name,
-                    flat.commit.expanded_srcmd5,
-                    md5,
-                    size,
-                )
-                self.git.add_lfs(file.name, file_sha256["sha256"], size)
-            else:
-                self.obs.download(
-                    flat.commit.project,
-                    flat.commit.package,
-                    file.name,
-                    flat.commit.expanded_srcmd5,
-                    self.git.path,
-                    self.cachedir,
-                    file_md5=md5,
-                )
-                self.git.add(file)
 
         commit = self.git.commit(
             f"OBS User {flat.commit.userid}",

lib/obs.py

@@ -5,11 +5,10 @@ import time
 import urllib.parse
 import xml.etree.ElementTree as ET
 from urllib.error import HTTPError
-from pathlib import Path
-from lib.proxy_sha256 import md5
 
 import osc.core
 
+from lib.proxy_sha256 import md5
 from lib.request import Request
 from lib.user import User
 
@@ -164,7 +163,7 @@ class OBS:
         cachedir: str,
         file_md5: str,
     ) -> None:
-        
+
         cached_file = self._path_from_md5(name, cachedir, file_md5)
         if not self.in_cache(name, cachedir, file_md5):
             with (dirpath / name).open("wb") as f:

lib/proxy_sha256.py

@@ -40,7 +40,7 @@ class ProxySHA256:
             self.texts = set(["_config", "_service"])
             self.hashes = dict()
             return
-        logging.info("Retrieve all previously defined SHA256")
+        logging.debug("Retrieve all previously defined SHA256")
         response = requests.get(f"http://source.dyn.cloud.suse.de/package/{package}")
         if response.status_code == 200:
             json = response.json()