apunkt/PageIndex
1
0
Fork 0
mirror of https://github.com/VectifyAI/PageIndex.git synced 2026-06-30 20:39:43 +02:00
Code Issues Projects Releases Packages Wiki Activity

refactor(filesystem): unify pifs metadata state

Browse source
This commit is contained in:
BukeLy 2026-05-26 17:12:59 +08:00
parent fc0be1aeee
commit 7c021a7dd0
8 changed files with 479 additions and 324 deletions
Download patch file Download diff file Expand all files Collapse all files

3
pageindex/filesystem/projection_indexing.py
View file

@ -68,7 +68,7 @@ class SummaryProjectionIndexer:
)
def upsert_summary(self, record: dict[str, Any]) -> dict[str, Any]:
summary = str((record.get("derived_metadata") or {}).get("summary") or "").strip()
summary = str((record.get("metadata") or {}).get("summary") or "").strip()
if not summary:
return {"status": "skipped", "reason": "missing_summary"}
vector = self.embedding_cache.embed_texts(
@ -79,7 +79,6 @@ class SummaryProjectionIndexer:
batch_size=1,
)[0]
metadata = dict(record.get("metadata") or {})
metadata.update(record.get("derived_metadata") or {})
count = self.index.upsert_many(
[
SemanticIndexRecord(