Tags: SylphxAI/pdf-reader-mcp
Tags
feat: preserve text and image order based on Y-coordinates - Extract Y-coordinates from PDF.js transform matrices - Sort content items by vertical position within each page - Return ordered content parts for optimal AI consumption - Text and images interleaved as they appear in document - Automatic line grouping for text on same Y-coordinate - Improved error handling with descriptive messages BREAKING CHANGE: Content parts now include individual text blocks instead of concatenated page text when images are enabled Resolves user requirement for exact document ordering like: page 1 [text, image, text, image, image, text]
PreviousNext