commit ba40d01131a94614deca71197688d61ec9052529
from: Antoine Lambert <anlambert@softwareheritage.org>
date: Mon Feb 26 13:48:49 2024 UTC

dumb: Handle HEAD file legacy format

Some dumb git servers can send a HEAD file in a legacy format that
contains a commit id instead of the string: "ref: <ref_name>".

So handle that edge case to avoid an error when loading such repository.

commit - b7d16897c24c7d5a0e28d952b027401b37417f92
commit + ba40d01131a94614deca71197688d61ec9052529
blob - 6e3006e035fa06720e40ef205e8e7ef151f1863e
blob + e372fa28d82e4ff65cdf6292701e454bf6a4e43a
--- swh/loader/git/dumb.py
+++ swh/loader/git/dumb.py
@@ -195,7 +195,13 @@ class GitObjectsFetcher:
 
     def _get_head(self) -> Dict[bytes, HexBytes]:
         head_resp_bytes = self._http_get("HEAD")
-        _, head_target = head_resp_bytes.readline().replace(b"\n", b"").split(b" ")
+        head_split = head_resp_bytes.readline().replace(b"\n", b"").split(b" ")
+        head_target = head_split[1] if len(head_split) > 1 else head_split[0]
+        # handle HEAD legacy format containing a commit id instead of a ref name
+        for ref_name, ret_target in self.refs.items():
+            if ret_target == head_target:
+                head_target = ref_name
+                break
         return {b"HEAD": head_target}
 
     def _get_pack_data(self, pack_name: str) -> Callable[[], PackData]:
blob - 1858a71402a362f799ebe7a8390325d4dca7a6f2
blob + c38c9efca9cce23733ee2eb12bac100936febf8e
--- swh/loader/git/tests/test_loader.py
+++ swh/loader/git/tests/test_loader.py
@@ -944,8 +944,20 @@ class DumbGitLoaderTestBase(FullGitLoaderTests):
             "skipped_content": 0,
             "snapshot": 1,
         }
+
+    def test_load_head_legacy_format(self, requests_mock):
+        # mock a single request and let the others pass through as real ones
+        requests_mock.real_http = True
+        requests_mock.get(
+            self.repo_url + "/HEAD", content=self.repo.refs[b"refs/heads/master"]
+        )
 
+        res = self.loader.load()
+        assert res == {"status": "eventful"}
+        assert b"HEAD" in self.loader.snapshot.branches
+        assert self.loader.snapshot.branches[b"HEAD"].target == b"refs/heads/master"
 
+
 class TestDumbGitLoaderWithPack(DumbGitLoaderTestBase):
     @classmethod
     def setup_class(cls):