public_git/bambuddy
1
0
Fork 0
mirror of https://github.com/maziggy/bambuddy.git synced 2026-10-18 21:47:39 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix(backup): halve the provider round-trips, and stop losing commit metadata (#2656)

Browse Source 
The three remaining review items, all in the read path.

E1 — four provider calls where two would do. preview() called list_commits
twice: once inside _resolve_ref to turn HEAD into a SHA, once more at limit=20
purely to find the entry describing that same SHA. And list_tree's recursive
tree GET was thrown away, so fetch_files immediately fetched the identical tree
again to map path -> blob SHA. _resolve_ref now returns the entry it already
has, and list_tree returns its blob_shas map for fetch_files to take as an
optional argument. GitLab reads files by path and ignores it.

E2 — `commit: null` for a ref outside the 20 most recent. Two causes, and the
second is the one that actually bit: REF_PATTERN accepts a 7-character ref while
providers return the full 40, so the exact `==` in the scan never matched an
abbreviated SHA *even when the commit was in the window*. Fixed by prefix
comparison, plus a get_commit(ref) on the GitHub and GitLab backends for the
genuinely-outside-the-window case. Gitea and Forgejo inherit GitHub's. Still
best-effort: it is a subject line and a date, so a failed lookup renders the
preview without them rather than failing it.

E7 — the two tree readers disagreed, and each was wrong in the other's
direction. GitHub's recursive trees endpoint is not paginated and signals
overflow with truncated=true, which _blob_shas_at hard-fails on. Gitea and
Forgejo *do* page that endpoint, and inherited that single GET unchanged — so a
large backup repo returned only the first page and every category beyond it
looked absent from the commit. GiteaBackend now has its own paging
_blob_shas_at. GitLab had the mirror-image bug the review did not name: at its
50-page cap it exited through the while condition and returned success: True
with a silently partial path list. Both now fail loudly, which is what the
GitHub version was always doing.

Both halves of E7 are the same failure the module already refuses to allow: a
restore that skips categories and calls it "not present in this backup commit".

24 new or changed tests, all failing against this commit's parent.
This commit is contained in:
jmoore-skild
2026-08-04 08:57:34 -04:00
parent 945f50a1ab
commit 158301ac8a
7 changed files with 554 additions and 34 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/app/services/github_restore.py
+56 -15
View File
@@ -307,21 +307,54 @@ class GitHubRestoreService:
result["branch"] = config.branch
return result
async def _resolve_ref(self, config: GitHubBackupConfig, ref: str) -> tuple[str | None, str]:
async def _resolve_ref(self, config: GitHubBackupConfig, ref: str) -> tuple[str | None, str, dict | None]:
"""Turn ``HEAD`` into a concrete commit SHA.
Done once up front so a preview and the restore that follows it act on
the same commit even if a scheduled backup lands in between.
The third element is the commit entry, when resolving already fetched
one. ``preview`` displays it, and taking it from here means the ``HEAD``
case — by far the common one — costs one ``list_commits`` call rather
than two.
"""
if ref and ref.upper() != "HEAD":
return ref, ""
return ref, "", None
result = await self.list_commits(config, limit=1)
if not result.get("success"):
return None, result.get("message") or "Could not read the backup repository"
return None, result.get("message") or "Could not read the backup repository", None
commits = result.get("commits") or []
if not commits:
return None, f"Branch '{config.branch}' has no commits to restore from"
return commits[0]["sha"], ""
return None, f"Branch '{config.branch}' has no commits to restore from", None
return commits[0]["sha"], "", commits[0]
async def _describe_commit(self, config: GitHubBackupConfig, resolved: str) -> dict | None:
"""Find the display metadata for one commit SHA.
Two things used to leave ``commit: null`` in a preview, and the second is
the one that bit in practice:
* the commit is older than the 20 the picker lists, so it is not in the
scan at all — that is what ``get_commit`` is for;
* ``REF_PATTERN`` accepts a 7-character ref while providers return the
full 40, so an exact ``==`` never matched an abbreviated SHA *even when
the commit was in the window*. Hence the prefix comparison.
Best-effort throughout: this is a subject line and a date, so a failure
returns None and the preview renders without them rather than failing.
"""
commits = (await self.list_commits(config, limit=20)).get("commits") or []
for entry in commits:
sha = entry.get("sha") or ""
if sha == resolved or sha.startswith(resolved) or resolved.startswith(sha):
return entry
backend = get_provider_backend(config.provider)
client = await self._get_client()
result = await backend.get_commit(
repo_url=config.repository_url, token=config.access_token, ref=resolved, client=client
)
return result.get("commit") if result.get("success") else None
def _category_paths(self, category: RestoreCategory, available: list[str]) -> list[str]:
"""Return the paths in ``available`` that belong to ``category``."""
@@ -429,7 +462,7 @@ class GitHubRestoreService:
Takes a session because the settings count depends on local state — see
``_plan_settings``. ``ref`` stays keyword-friendly for callers.
"""
resolved, error = await self._resolve_ref(config, ref)
resolved, error, commit_info = await self._resolve_ref(config, ref)
if resolved is None:
return {"success": False, "message": error, "ref": ref, "categories": []}
@@ -449,7 +482,14 @@ class GitHubRestoreService:
wanted.extend(self._category_paths(category, available))
fetched = await backend.fetch_files(
repo_url=config.repository_url, token=config.access_token, ref=resolved, paths=wanted, client=client
repo_url=config.repository_url,
token=config.access_token,
ref=resolved,
paths=wanted,
client=client,
# The listing above already built this map; without it the GitHub
# family would GET the same recursive tree a second time.
blob_shas=tree.get("blob_shas") or None,
)
if not fetched.get("success"):
return {
@@ -485,12 +525,8 @@ class GitHubRestoreService:
count, detail = await self._count_items(db, category, parsed)
categories.append(self._category_entry(category, True, count, detail))
commit_info = None
commits = (await self.list_commits(config, limit=20)).get("commits") or []
for entry in commits:
if entry["sha"] == resolved:
commit_info = entry
break
if commit_info is None:
commit_info = await self._describe_commit(config, resolved)
return {
"success": True,
@@ -622,7 +658,7 @@ class GitHubRestoreService:
return {"success": False, "message": "Configuration not found", "results": {}}
self._progress = "Resolving commit..."
resolved, error = await self._resolve_ref(config, ref)
resolved, error, _ = await self._resolve_ref(config, ref)
if resolved is None:
return {"success": False, "message": error, "results": {}}
@@ -706,7 +742,12 @@ class GitHubRestoreService:
self._progress = "Downloading backup files..."
fetched = await backend.fetch_files(
repo_url=config.repository_url, token=config.access_token, ref=ref, paths=wanted, client=client
repo_url=config.repository_url,
token=config.access_token,
ref=ref,
paths=wanted,
client=client,
blob_shas=tree.get("blob_shas") or None,
)
if not fetched.get("success"):
return {}, fetched.get("message") or "Could not read the commit contents"