ydy0615/llm-in-text
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(copilot): enhance OCR handling with inline tags and document serializer

Browse Source 
- Replace HTML comment OCR metadata with inline `<OCR:...>` tags
- Implement serializer-based markdown conversion for prefix/suffix content
- Add extractTextFromOCR utility function for text extraction
- Enable Table, Diagram, and ListCheck features in MilkdownEditor
- Add periodic debug logging for document state analysis
This commit is contained in:
ydy0615
2026-02-14 23:53:26 +08:00
parent 794fbf8493
commit 03bb21d5c6
5 changed files with 3496 additions and 225 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/prompt.py
+4 -4
View File
@@ -38,10 +38,10 @@ Your job:
- Avoid overly short outputs with little information value.
Important context:
- PREFIX may contain hidden OCR metadata in HTML comments such as <!--OCR:...-->.
- These comments are non-visible context only.
- Never copy, rewrite, or emit HTML comments in output.
- Never output <!-- or -->.
- PREFIX may contain OCR metadata inline after images, e.g. ![alt](url) <OCR:description>.
- The <OCR:...> is hidden context describing image content.
- Never copy, rewrite, or emit OCR tags in output.
- Never output <OCR: or >.
Hard rules:
1. Seamless join: