How opendataloader-project/opendataloader-pdf works
OpenDataLoader PDF converts PDFs into structured, layout-aware outputs and can auto-tag PDFs for accessibility. Users invoke a Python, Node.js, Java API, CLI, or MCP tool; conversion is handled by the Java core, with optional hybrid processing for complex pages, then emitted in formats such as Markdown, JSON, HTML, and Tagged PDF. The README establishes these product workflows and capabilities; sampled source confirms selected CLI, Node-to-JVM, and hybrid HTTP behavior, but not the internal processor call sequence.
Main entry points: DocumentProcessor.java Document Processing · CLIMain.java Java CLI · OpenDataLoaderPDF.java Java API · HybridDocumentProcessor.java Hybrid Routing
Components 22
User Interfaces
- Python SDK
python/opendataloader-pdf/src/opendataloader_pdf/runner.py - Node SDK
node/opendataloader-pdf/src/index.ts - Java CLI
java/opendataloader-pdf-cli/src/main/java/org/opendataloader/pdf/cli/CLIMain.java - MCP Server
python/opendataloader-pdf-mcp/src/opendataloader_pdf_mcp/server.py
PDF Processing
- Java API
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/api/OpenDataLoaderPDF.java - Conversion Config
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/api/Config.java - Document Processing
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Reading Order
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/readingorder/XYCutPlusPlusSorter.java - List Detection
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/ListProcessor.java - Paragraph Analysis
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/ParagraphProcessor.java - Safety Filters
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/utils/ContentSanitizer.java
Hybrid Analysis
- Hybrid Routing
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/HybridDocumentProcessor.java - Page Triage
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/hybrid/TriageProcessor.java - Hybrid Client
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/hybrid/HybridClient.java - Docling Client
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/hybrid/DoclingFastServerClient.java - Hybrid Backend
Outputs and Accessibility
- Markdown Output
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/markdown/MarkdownGenerator.java - JSON Output
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/json/JsonWriter.java - HTML Output
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/html/HtmlGenerator.java - Auto-tagging
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/AutoTaggingProcessor.java - Tagged PDF
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/pdf/PDFWriter.java
Other components
- PDF User
Connections 22
- PDF User → Python SDK: calls
README.md - PDF User → Node SDK: calls
README.md - PDF User → Java CLI: runs
README.md - PDF User → MCP Server: requests conversion
python/opendataloader-pdf-mcp/src/opendataloader_pdf_mcp/server.py - Node SDK → Java CLI: spawns JVM CLI
node/opendataloader-pdf/src/index.ts - Java CLI → Java API: invokes
java/opendataloader-pdf-cli/src/main/java/org/opendataloader/pdf/cli/CLIMain.java - Java API → Document Processing: processes PDF
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/api/OpenDataLoaderPDF.java - Document Processing → Reading Order: orders content
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Document Processing → List Detection: detects lists
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Document Processing → Paragraph Analysis: analyzes paragraphs
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Document Processing → Safety Filters: filters content
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Document Processing → Markdown Output: generates
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Document Processing → JSON Output: generates
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Document Processing → HTML Output: generates
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Document Processing → Auto-tagging: auto-tags
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/processors/DocumentProcessor.java - Auto-tagging → Tagged PDF: writes
- Hybrid Routing → Page Triage: triages pages
- Hybrid Routing → Hybrid Client: requests analysis
- Hybrid Client → Docling Client: uses client
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/hybrid/DoclingFastServerClient.java - Docling Client → Hybrid Backend: sends HTTP request
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/hybrid/DoclingFastServerClient.java - MCP Server → Python SDK: calls conversion
python/opendataloader-pdf-mcp/src/opendataloader_pdf_mcp/server.py - Conversion Config → Java API: configures
java/opendataloader-pdf-core/src/main/java/org/opendataloader/pdf/api/OpenDataLoaderPDF.java