homelab/stack
1
0
Fork 0
Code Issues 262 Pull Requests 2 Actions Packages Projects 1 Releases Wiki Activity

fix(llm): fold fingerprint/hash skip counts into index_docs skipped (refs #615)

Browse Source
This commit is contained in:
kert
2026-09-08 14:04:31 -04:00
parent 414b416bd6
commit 74d72f7d93
2 changed files with 6 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/llm/index.py
View File

@@ -238,4 +238,8 @@ def index_docs(
stats = index_refs( stats = index_refs(
refs, collection=collection, cfg=cfg, pool=pool, force=force, sealed=None refs, collection=collection, cfg=cfg, pool=pool, force=force, sealed=None
) )
return {k: stats[k] for k in ("indexed", "skipped", "chunks")} # Fold the fingerprint/hash skip counters into "skipped" so the 3-key
# contract still means "not embedded" — index_refs itself keeps them
# separate; this is only a local projection.
skipped = stats["skipped"] + stats["fingerprint_skipped"] + stats["hash_skipped"]
return {"indexed": stats["indexed"], "skipped": skipped, "chunks": stats["chunks"]}

5
tests/llm/test_index.py
View File

@@ -56,12 +56,9 @@ class TestIndexDocs:
assert kwargs["ids"][0].startswith("K1:") assert kwargs["ids"][0].startswith("K1:")
def test_unchanged_doc_skipped(self): def test_unchanged_doc_skipped(self):
"""A hash match lands in the internal ``hash_skipped`` bucket, which
index_docs (indexed/skipped/chunks only) doesn't surface — but it
must not be embedded or counted as indexed."""
h = content_hash(DOC.text) h = content_hash(DOC.text)
stats, store, _, _ensure_hnsw = _run([DOC], state_rows=[("K1", h, "")]) stats, store, _, _ensure_hnsw = _run([DOC], state_rows=[("K1", h, "")])
assert stats["indexed"] == 0 assert stats["skipped"] == 1
store.add_embeddings.assert_not_called() store.add_embeddings.assert_not_called()
def test_changed_doc_deletes_old_chunks_first(self): def test_changed_doc_deletes_old_chunks_first(self):