8000
Skip to content

fix: retrieval test and knowledge retrieval node failed in multimodal mode#30210

Merged
JohnJyong merged 9 commits intomainfrom
fix/multible-retrieval-test
Dec 26, 2025
Merged

fix: retrieval test and knowledge retrieval node failed in multimodal mode#30210
JohnJyong merged 9 commits intomainfrom
fix/multible-retrieval-test

Conversation

@JohnJyong
Copy link
Copy Markdown
Contributor
@JohnJyong JohnJyong commented Dec 26, 2025

Important

  1. Make sure you have read our contribution guidelines
  2. Ensure there is an associated issue and you have been assigned to it
  3. Use the correct syntax to link this PR: Fixes #<issue number>.

Summary

Fix: retrieval test failed in multimodal mode

Screenshots

Before After
... ...

Checklist

  • This change requires a documentation update, included: Dify Document
  • I understand that this PR may be closed in case there was no previous discussion or issues. (This doesn't apply to typos!)
  • I've added a test for each change that was introduced, and I tried as much as possible to make a single atomic change.
  • I've updated the documentation accordingly.
  • I ran dev/reformat(backend) and cd web && npx lint-staged(frontend) to appease the lint gods

@dosubot dosubot bot added size:L This PR changes 100-499 lines, ignoring generated files. 👻 feat:rag Embedding related issue, like qdrant, weaviate, milvus, vector database. labels Dec 26, 2025
@gemini-code-assist
Copy link
Copy Markdown
Contributor
gemini-code-assist bot commented Dec 26, 2025

Summary of Changes

Hello @JohnJyong, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request addresses a critical bug that caused retrieval tests to fail in multimodal mode. The changes primarily involve a significant refactoring of the format_retrieval_documents function within the retrieval service. The update ensures that the system accurately processes and aggregates information from multiple child chunks and attachments linked to a single document segment, which is crucial for robust multimodal data handling. This fix improves the reliability and accuracy of document retrieval, especially in complex data environments.

Highlights

  • Refactored Multimodal Retrieval Logic: The core retrieval service has been updated to correctly handle scenarios where multiple child chunks and attachments are associated with a single document segment, resolving previous issues in multimodal retrieval.
  • Improved Score Aggregation: The scoring mechanism now accurately aggregates scores from all associated child chunks and attachments, ensuring that the highest relevant score is used for parent segments.
  • Enhanced Data Structure Handling: Internal maps like attachment_map, child_chunk_map, and doc_segment_map have been modified to store lists of associated items, preventing data loss when multiple entities link to the same segment.
  • Consistent Result Sorting: The final retrieval results are now consistently sorted by their aggregated score in descending order, presenting the most relevant information first.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature Command Description
Code Review /gemini review Performs a code review for the current pull request in its current state.
Pull Request Summary /gemini summary Provides a summary of the current pull request in its current state.
Comment @gemini-code-assist Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help /gemini help Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

  1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist[bot]
gemini-code-assist bot reviewed Dec 26, 2025
View reviewed changes
Copy link
Copy Markdown
Contributor
@gemini-code-assist gemini-code-assist bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This pull request refactors the document retrieval formatting logic to address a bug in multimodal retrieval. The changes simplify the code, particularly in handling parent-child index structures with multiple child chunks and attachments. Overall, the refactoring improves clarity and correctness. I've identified a critical runtime bug in the sorting logic and some potentially dead code with a logical flaw that should be addressed to improve maintainability.

JohnJyong and others added 3 commits December 26, 2025 17:25
@JohnJyong @gemini-code-assist
Update api/core/rag/datasource/retrieval_service.py
b8e1ba8 
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
@JohnJyong
8000
fix multimodal embedding retrival test
771b93f
@JohnJyong
Merge remote-tracking branch 'origin/fix/multible-retrieval-test' int…
4797f3f 
…o fix/multible-retrieval-test
@JohnJyong JohnJyong changed the title Fix: retrieval test failed in multimodal mode Fix: retrieval test and knowledge retrieval node failed in multimodal mode Dec 26, 2025
@JohnJyong JohnJyong changed the title Fix: retrieval test and knowledge retrieval node failed in multimodal mode Retrieval test and knowledge retrieval node failed in multimodal mode Dec 26, 2025
@JohnJyong JohnJyong changed the title Retrieval test and knowledge retrieval node failed in multimodal mode fix: retrieval test and knowledge retrieval node failed in multimodal mode Dec 26, 2025
@dosubot dosubot bot added the lgtm This PR has been approved by a maintainer label Dec 26, 2025
@JohnJyong JohnJyong merged commit f610f68 into main Dec 26, 2025
16 of 18 checks passed
@JohnJyong JohnJyong deleted the fix/multible-retrieval-test branch December 26, 2025 13:42
Eric-Guo pushed a commit to Eric-Guo/dify that referenced this pull request Dec 30, 2025
@JohnJyong @Eric-Guo
Revert "fix: retrieval test and knowledge retrieval node failed in mu…
ad0a096 
…ltimodal mode (langgenius#30210)"

This reverts commit f610f68.
@dosubot dosubot bot mentioned this pull request Jan 23, 2026
Open
6 tasks
@dosubot dosubot bot mentioned this pull request Mar 2, 2026
Closed
6 tasks
@dosubot dosubot bot mentioned this pull request Mar 13, 2026
Open
6 tasks
@dosubot dosubot bot mentioned this pull request Apr 7, 2026
Open
6 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@laipz8200 laipz8200 laipz8200 approved these changes

+1 more reviewer

@gemini-code-assist gemini-code-assist[bot] gemini-code-assist[bot] left review comments

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

👻 feat:rag Embedding related issue, like qdrant, weaviate, milvus, vector database. lgtm This PR has been approved by a maintainer size:L This PR changes 100-499 lines, ignoring generated files.

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@JohnJyong @laipz8200 @hyoban
0