keyword-literature-download

What problem does it solve?

Researchers often spend days manually searching across multiple scholarly databases and then chasing down legally accessible full-text PDFs, which delays literature reviews, proposals, and gap-finding work.

Core Features & Use Cases

• Keyword-driven multi-source search: queries run against PubMed/PMC, Europe PMC, Crossref, and OpenAlex to assemble a broad candidate set for any English or Chinese topic.
• Priority scoring + candidate tables: ranks candidates using include/secondary/exclude term hits, publication-type signals, and metadata signals to produce high/medium-priority tables.
• Legally accessible PDF downloading with guardrails: parallel downloads save only true PDFs (verified by PDF signature) to a PDF-only folder while caching HTML/XML helper files separately for later second-pass PDF link discovery.
• Resume + second-pass PDF chasing: continues pending downloads and attempts to extract additional PDF links from cached auxiliary HTML pages.
• Deduplication for clean PDF collections: deduplicates the final PDF set using DOI, normalized title, and file hashes, producing a deduplicated PDF-only folder and manifest.

Quick Start

Request a keyword literature harvest by saying: “Use $keyword-literature-download with my topic keywords to search PubMed/PMC, Europe PMC, Crossref, and OpenAlex, rank candidates, download only legally accessible PDFs into my PDF output folder, chase missing PDFs from cached HTML, then deduplicate the final PDFs and produce the candidate and download reports.”