opensource/dify
2
0
Fork 0
mirror of https://github.com/langgenius/dify.git synced 2026-06-16 22:11:09 +08:00
Code Issues Packages Projects Releases Wiki Activity

chore: example to color the session (#37402)

Browse Source 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
This commit is contained in:
Asuka Minato 2026-06-16 03:06:24 +09:00 committed by GitHub
parent 2b7f5ab982
commit de2ec990d8
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 21 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
api/core/rag/index_processor/processor/paragraph_index_processor.py
View File

@ -5,6 +5,8 @@ import re
import uuid
from typing import Any, TypedDict, cast, override
from sqlalchemy.orm import scoped_session
logger = logging.getLogger(__name__)
from sqlalchemy import select
@ -411,11 +413,13 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
if supports_vision:
# First, try to get images from SegmentAttachmentBinding (preferred method)
if segment_id:
image_files = ParagraphIndexProcessor._extract_images_from_segment_attachments(tenant_id, segment_id)
image_files = ParagraphIndexProcessor._extract_images_from_segment_attachments(
tenant_id, segment_id, db.session
)
# If no images from attachments, fall back to extracting from text
if not image_files:
image_files = ParagraphIndexProcessor._extract_images_from_text(tenant_id, text)
image_files = ParagraphIndexProcessor._extract_images_from_text(tenant_id, text, db.session)
# Build prompt messages
prompt_messages = []
@ -469,7 +473,7 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
return summary_content, usage
@staticmethod
def _extract_images_from_text(tenant_id: str, text: str) -> list[File]:
def _extract_images_from_text(tenant_id: str, text: str, session: scoped_session) -> list[File]:
"""
Extract images from markdown text and convert them to File objects.
@ -518,7 +522,7 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
# Get unique IDs for database query
unique_upload_file_ids = list(set(upload_file_id_list))
upload_files = db.session.scalars(
upload_files = session.scalars(
select(UploadFile).where(UploadFile.id.in_(unique_upload_file_ids), UploadFile.tenant_id == tenant_id)
).all()
@ -549,7 +553,9 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
return file_objects
@staticmethod
def _extract_images_from_segment_attachments(tenant_id: str, segment_id: str) -> list[File]:
def _extract_images_from_segment_attachments(
tenant_id: str, segment_id: str, session: scoped_session
) -> list[File]:
"""
Extract images from SegmentAttachmentBinding table (preferred method).
This matches how DatasetRetrieval gets segment attachments.
@ -564,7 +570,7 @@ class ParagraphIndexProcessor(BaseIndexProcessor):
from sqlalchemy import select
# Query attachments from SegmentAttachmentBinding table
attachments_with_bindings = db.session.execute(
attachments_with_bindings = session.execute(
select(SegmentAttachmentBinding, UploadFile)
.join(UploadFile, UploadFile.id == SegmentAttachmentBinding.attachment_id)
.where(

18
api/tests/unit_tests/core/rag/indexing/processor/test_paragraph_index_processor.py
View File

@ -528,21 +528,24 @@ class TestParagraphIndexProcessor:
session.scalars.return_value = scalars_result
with (
patch("core.rag.index_processor.processor.paragraph_index_processor.db.session", session),
patch(
"core.rag.index_processor.processor.paragraph_index_processor.build_from_mapping",
return_value=SimpleNamespace(id="file-1"),
) as mock_builder,
patch("core.rag.index_processor.processor.paragraph_index_processor.logger") as mock_logger,
):
files = ParagraphIndexProcessor._extract_images_from_text("tenant-1", text)
files = ParagraphIndexProcessor._extract_images_from_text("tenant-1", text, session)
assert len(files) == 1
assert mock_builder.call_count == 1
mock_logger.warning.assert_not_called()
def test_extract_images_from_text_returns_empty_when_no_matches(self) -> None:
assert ParagraphIndexProcessor._extract_images_from_text("tenant-1", "no images here") == []
scalars_result = Mock()
scalars_result.all.return_value = []
session = Mock()
session.scalars.return_value = scalars_result
assert ParagraphIndexProcessor._extract_images_from_text("tenant-1", "no images here", session) == []
def test_extract_images_from_text_logs_when_build_fails(self) -> None:
text = "![img](/files/11111111-1111-1111-1111-111111111111/image-preview)"
@ -562,14 +565,13 @@ class TestParagraphIndexProcessor:
session.scalars.return_value = scalars_result
with (
patch("core.rag.index_processor.processor.paragraph_index_processor.db.session", session),
patch(
"core.rag.index_processor.processor.paragraph_index_processor.build_from_mapping",
side_effect=RuntimeError("build failed"),
),
patch("core.rag.index_processor.processor.paragraph_index_processor.logger") as mock_logger,
):
files = ParagraphIndexProcessor._extract_images_from_text("tenant-1", text)
files = ParagraphIndexProcessor._extract_images_from_text("tenant-1", text, session)
assert files == []
mock_logger.warning.assert_called_once()
@ -608,10 +610,9 @@ class TestParagraphIndexProcessor:
session.execute.return_value = execute_result
with (
patch("core.rag.index_processor.processor.paragraph_index_processor.db.session", session),
patch("core.rag.index_processor.processor.paragraph_index_processor.logger") as mock_logger,
):
files = ParagraphIndexProcessor._extract_images_from_segment_attachments("tenant-1", "seg-1")
files = ParagraphIndexProcessor._extract_images_from_segment_attachments("tenant-1", "seg-1", session)
assert len(files) == 1
mock_logger.warning.assert_called_once()
@ -622,7 +623,6 @@ class TestParagraphIndexProcessor:
session = Mock()
session.execute.return_value = execute_result
with patch("core.rag.index_processor.processor.paragraph_index_processor.db.session", session):
empty_files = ParagraphIndexProcessor._extract_images_from_segment_attachments("tenant-1", "seg-1")
empty_files = ParagraphIndexProcessor._extract_images_from_segment_attachments("tenant-1", "seg-1", session)
assert empty_files == []