youngkingdom/dify
1
0
Fork 0
mirror of https://github.com/langgenius/dify.git synced 2026-03-25 08:18:02 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix: fix session problem.

Browse Source
This commit is contained in:
FFXN
2026-01-28 19:03:01 +08:00
parent 28af5d522b
commit 1ef11e1849
2 changed files with 21 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
api/services/summary_index_service.py
View File

@ -219,10 +219,13 @@ class SummaryIndexService:
session.query(DocumentSegmentSummary).filter_by(id=summary_record_id).first()
)
if summary_record_in_session:
# Update all fields including summary_content (in case it was updated in outer session)
summary_record_in_session.summary_index_node_id = summary_index_node_id
summary_record_in_session.summary_index_node_hash = summary_hash
summary_record_in_session.tokens = embedding_tokens # Save embedding tokens
summary_record_in_session.status = "completed"
# Ensure summary_content is preserved (use the latest from summary_record parameter)
summary_record_in_session.summary_content = summary_content
# Explicitly update updated_at to ensure it's refreshed even if other fields haven't changed
summary_record_in_session.updated_at = datetime.now(UTC).replace(tzinfo=None)
session.add(summary_record_in_session)
@ -232,6 +235,7 @@ class SummaryIndexService:
summary_record.summary_index_node_hash = summary_hash
summary_record.tokens = embedding_tokens
summary_record.status = "completed"
summary_record.summary_content = summary_content
summary_record.updated_at = summary_record_in_session.updated_at
# Success, exit function
return
@ -428,6 +432,9 @@ class SummaryIndexService:
# Update summary content
summary_record_in_session.summary_content = summary_content
session.add(summary_record_in_session)
# Flush to ensure summary_content is saved before vectorize_summary queries it
session.flush()
# Log LLM usage for summary generation
if llm_usage and llm_usage.total_tokens > 0:
@ -442,6 +449,7 @@ class SummaryIndexService:
# Vectorize summary (will delete old vector if exists before creating new one)
# Pass the session-managed record to vectorize_summary
# vectorize_summary will update status to "completed" and tokens in its own session
# vectorize_summary will also ensure summary_content is preserved
SummaryIndexService.vectorize_summary(summary_record_in_session, segment, dataset)
# Refresh the object from database to get the updated status and tokens from vectorize_summary
@ -834,7 +842,8 @@ class SummaryIndexService:
summary_record.status = "generating"
summary_record.error = None # Clear any previous errors
session.add(summary_record)
# Don't flush here - wait until after vectorization succeeds
# Flush to ensure summary_content is saved before vectorize_summary queries it
session.flush()
# Delete old vector if exists (before vectorization)
if old_summary_node_id:
@ -849,6 +858,7 @@ class SummaryIndexService:
)
# Re-vectorize summary (this will update status to "completed" and tokens in its own session)
# vectorize_summary will also ensure summary_content is preserved
try:
SummaryIndexService.vectorize_summary(summary_record, segment, dataset)
# Refresh the object from database to get the updated status and tokens from vectorize_summary

13
api/tasks/document_indexing_task.py
View File

@ -124,12 +124,17 @@ def _document_indexing(dataset_id: str, document_ids: Sequence[str]):
)
if document:
logger.info(
"Checking document %s for summary generation: status=%s, doc_form=%s",
"Checking document %s for summary generation: status=%s, doc_form=%s, need_summary=%s",
document_id,
document.indexing_status,
document.doc_form,
document.need_summary,
)
if document.indexing_status == "completed" and document.doc_form != "qa_model":
if (
document.indexing_status == "completed"
and document.doc_form != "qa_model"
and document.need_summary is True
):
try:
generate_summary_index_task.delay(dataset.id, document_id, None)
logger.info(
@ -146,10 +151,12 @@ def _document_indexing(dataset_id: str, document_ids: Sequence[str]):
# Don't fail the entire indexing process if summary task queuing fails
else:
logger.info(
"Skipping summary generation for document %s: status=%s, doc_form=%s",
"Skipping summary generation for document %s: "
"status=%s, doc_form=%s, need_summary=%s",
document_id,
document.indexing_status,
document.doc_form,
document.need_summary,
)
else:
logger.warning("Document %s not found after indexing", document_id)