fix(gemini): strip query string and fragment before extracting file e…#1666
fix(gemini): strip query string and fragment before extracting file e…#1666hangweizhang wants to merge 1 commit into
Conversation
…xtension GeminiMediaConverter.extractExtension() parsed the entire URL suffix after the last dot as the file extension, which included query strings and fragments. This caused URLs like http://localhost/cat.png?token=abc123 to produce extension "png?token=abc123" instead of "png", resulting in an IllegalArgumentException. Similarly, readFileAsBytes() failed for local file paths that contained query strings or fragments because Paths.get() treated them as part of the filename. Changes: - Add stripQueryAndFragment() utility to strip ?... and #... from URLs - Refactor extractExtension() to use stripQueryAndFragment() before extracting the extension - Refactor readFileAsBytes() to strip query/fragment from local file paths before resolving them on disk - Add test cases for URLs with query strings, fragments, and both Closes agentscope-ai#1645
Codecov Report❌ Patch coverage is
📢 Thoughts on this report? Let us know! |
AgentScopeJavaBot
added
bug
AgentScopeJavaBot
left a comment
AgentScopeJavaBot
left a comment
There was a problem hiding this comment.
🤖 AI Review
Clean, focused bug fix for Issue #1645. The new stripQueryAndFragment() utility correctly handles URLs and local paths containing query strings (?...) and fragments (#...), which previously caused extractExtension() to produce invalid extensions like png?token=abc123 and readFileAsBytes() to fail on local paths with appended query parameters. The implementation is minimal, well-documented with updated Javadoc, applied consistently in both affected methods, and includes three test cases covering query strings, fragments, and both combined. No issues found — this is a textbook bug fix.
| /** | ||
| * Strips query string ({@code ?...}) and fragment ({@code #...}) from a URL/path string. | ||
| */ | ||
| private static String stripQueryAndFragment(String url) { |
There was a problem hiding this comment.
[praise] Clean, focused utility method. The fix is applied consistently in both extractExtension() and readFileAsBytes(), the stripping order (fragment first, then query) correctly handles all URL forms including malformed ones like path#frag?query, and the error messages still reference the original URL for debugging. Good test coverage for the three main cases.
2 participants
…xtension
GeminiMediaConverter.extractExtension() parsed the entire URL suffix after the last dot as the file extension, which included query strings and fragments. This caused URLs like
http://localhost/cat.png?token=abc123 to produce extension "png?token=abc123" instead of "png", resulting in an IllegalArgumentException.
Similarly, readFileAsBytes() failed for local file paths that contained query strings or fragments because Paths.get() treated them as part of the filename.
Changes:
Closes #1645
AgentScope-Java Version
[The version of AgentScope-Java you are working on, e.g. 1.0.12, check your pom.xml dependency version or run
mvn dependency:tree | grep agentscope-parent:pom(only mac/linux)]Description
[Please describe the background, purpose, changes made, and how to test this PR]
Checklist
Please check the following items before code is ready to be reviewed.
mvn spotless:applymvn test)