inbox-parser

Extracting text and metadata from PDF papers is done by processing files placed in a specified 'inbox/' directory. The system parses the content, classifies it against an active literature map, and outputs the structured data into 'literature/papers/' and 'literature/bibliography.bib'.

You can integrate locally stored PDF files by placing them in the 'inbox/' directory and activating the extraction process. The system automatically extracts text, classifies the papers, and moves the structured outputs to 'literature/papers/' and 'literature/bibliography.bib'.

Classification works by parsing extracted text and metadata from new PDF documents, then comparing those features against your project's active literature map. This categorizes the incoming papers so they can be filed correctly into the literature pipeline for downstream use.

Yes, a specific structured directory environment is required. The workflow expects incoming PDFs in an 'inbox/' folder and outputs the extracted text, metadata, and classification data directly into 'literature/papers/' and 'literature/bibliography.bib'.

You should run PDF text extraction before starting a literature review or ideation pass. Activating the parser ensures all newly downloaded papers in your inbox are incorporated into the project's knowledge base and are fully prepared for downstream analysis.