@effortlessmotion/pdf-to-md
    Preparing search index...

    Function analyze

    • Analyze a PDF (or replay an already-recognized OCR word dump) into Analysis: plain, JSON-serializable per-page words/blocks plus the same QA report the other entry points return. Render with toMarkdown()/toText(), or hand-edit/splice it (no merge API — see toMarkdown's doc comment).

      Parameters

      • input: string | { words: PageWord[] }

        a bare pdf path, or { words } to replay a dump (always analyzed as OCR) — unlike extractOcr, never { pdfPath }

      • opts: AnalyzeOptions = {}

        see AnalyzeOptions; for a pdf path, opts.path picks 'text' (default) or 'ocr' — analyze() makes no auto-fallback decision (that's pdfToMarkdown's job)

      Returns Promise<Analysis>

    Settings
    Member Visibility