fix(core): preserve multimodal content parts in DashScope tool result…#1213
Merged
Conversation
…s instead of merging into text
|
|
Codecov Report❌ Patch coverage is
📢 Thoughts on this report? Let us know! |
…s instead of merging into text
…x/DashScope-Formatter
Member
|
Contributor
Author
|
@LearningGp done，DeploymentPropertiesTest.java没有引入assertNull导致main分支流水线报错 |
liangxingguang
pushed a commit
to liangxingguang/agentscope-java
that referenced
this pull request
May 21, 2026
…s instead of merging into text. (agentscope-ai#1213) ## AgentScope-Java Version >1.0.12-SNAPSHOT ## Description fix [issue#1153](agentscope-ai#1153) Problem reproduction: From the image, we can see that the current processing method converts all modalities into text format and then concatenates them. This causes an issue where image URLs cannot be extracted correctly from the same modality. <img width="1386" height="393" alt="debug2" src="https://github.com/user-attachments/assets/2b68b693-2254-498e-81ee-b96f1940295c" /> Comparison with the OpenAI approach: In the OpenAI approach, a modality check is added before extraction. As shown in the image, different types can be separated, which makes it possible to extract image URLs. <img width="1422" height="510" alt="debug1" src="https://github.com/user-attachments/assets/506edbc1-4704-4bd8-ab87-3f495525e7ce" /> Implementation and idea: The implementation borrows from OpenAI’s approach by adding modality detection and separation before processing. Due to inconsistencies between the DTO formats and OpenAI, the final output is structured according to the DashScopeMessage DTO format. <img width="707" height="462" alt="image" src="https://github.com/user-attachments/assets/0d2f44cc-5c11-40ec-b837-85d0a2408d3e" /> <img width="1480" height="565" alt="debug3" src="https://github.com/user-attachments/assets/15e5591c-4b2b-4b87-bef8-4e0e7f614af7" /> Edge Cases Handled： Due to structural adjustments, the input format of Tool result has been modified in DashScopeChatFormatterGroundTruthTest and DashScopeMultiAgentFormatterGroundTruthTest. ## Checklist Please check the following items before code is ready to be reviewed. - [x] Code has been formatted with `mvn spotless:apply` - [x] All tests are passing (`mvn test`) - [x] Javadoc comments are complete and follow project conventions - [x] Related documentation has been updated (e.g. links, examples, etc.) - [x] Code is ready for review
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
AgentScope-Java Version
Description
fix issue#1153
Problem reproduction:
From the image, we can see that the current processing method converts all modalities into text format and then concatenates them. This causes an issue where image URLs cannot be extracted correctly from the same modality.
Comparison with the OpenAI approach:
In the OpenAI approach, a modality check is added before extraction. As shown in the image, different types can be separated, which makes it possible to extract image URLs.
Implementation and idea:
The implementation borrows from OpenAI’s approach by adding modality detection and separation before processing. Due to inconsistencies between the DTO formats and OpenAI, the final output is structured according to the DashScopeMessage DTO format.
Due to structural adjustments, the input format of Tool result has been modified in DashScopeChatFormatterGroundTruthTest and DashScopeMultiAgentFormatterGroundTruthTest.
Checklist
Please check the following items before code is ready to be reviewed.
mvn spotless:applymvn test)