Fix nested dependency depth tracking in PyPI cache worker

When the cache worker downloaded a package through the proxy, dependencies
were always queued with depth=0 instead of depth+1. This meant depth limits
weren't properly enforced for nested dependencies.

Changes:
- Add cache-depth query parameter to pypi_download_file endpoint
- Worker now passes its current depth when fetching packages
- Dependencies are queued at cache_depth+1 instead of hardcoded 0
- Add tests for depth tracking behavior
This commit is contained in:
Mondo Diaz
2026-02-02 13:47:22 -06:00
parent 8edb45879f
commit 3bdeade7ca
3 changed files with 120 additions and 4 deletions

View File

@@ -516,6 +516,7 @@ async def pypi_download_file(
package_name: str,
filename: str,
upstream: Optional[str] = None,
cache_depth: int = Query(default=0, ge=0, le=100, alias="cache-depth"),
db: Session = Depends(get_db),
storage: S3Storage = Depends(get_storage),
):
@@ -526,6 +527,7 @@ async def pypi_download_file(
package_name: The package name
filename: The filename to download
upstream: URL-encoded upstream URL to fetch from
cache_depth: Current cache recursion depth (used by cache worker for nested deps)
"""
if not upstream:
raise HTTPException(
@@ -772,17 +774,19 @@ async def pypi_download_file(
db.add(dep)
# Proactively cache dependencies via task queue
# Dependencies are queued at cache_depth + 1 to track recursion
if unique_deps:
next_depth = cache_depth + 1
for dep_name, dep_version in unique_deps:
enqueue_cache_task(
db,
package_name=dep_name,
version_constraint=dep_version,
parent_task_id=None, # Top-level, triggered by user download
depth=0,
depth=next_depth,
triggered_by_artifact=sha256,
)
logger.info(f"PyPI proxy: queued {len(unique_deps)} dependencies for caching")
logger.info(f"PyPI proxy: queued {len(unique_deps)} dependencies for caching (depth={next_depth})")
db.commit()