Function readPdf

readPdf(
    vault: Vault,
    args: ReadPdfArgs,
    options?: ReadPdfOptions,
): Promise<ReadPdfResult>
Extract text from a PDF page-by-page, with optional page-range slicing and metadata.

Image-only / scanned PDFs surface has_text: false with complete: true; page failures surface complete: false and explicit failure evidence. Lazy-loads pdfjs-dist (optional dep) so markdown-only users pay zero cost. A valid upper bound beyond the document is clamped. Non-finite, fractional, non-positive, inverted, or wholly out-of-document ranges are rejected.
Parameters
- vault: Vault
  The vault.
- args: ReadPdfArgs
  ReadPdfArgs. path required.
- options: ReadPdfOptions = {}
  Internal narrowing options, excluded from the MCP input schema.
Returns Promise<ReadPdfResult>
A ReadPdfResult with per-page text, full-text join, metadata, and original total_page_count.
Throws
If path is empty, the file is missing or excluded, or pdfjs-dist is not installed.

Throws
If path resolves outside the vault.
Example
```
// Read pages 1-5 of a long paper
const r = await readPdf(vault, {
  path: "Papers/2024-rag-survey.pdf",
  pages: [1, 5],
  include_metadata: true
});
if (!r.has_text) console.log("Scanned PDF — try ocrPdf()");
console.log(r.metadata?.title, r.full_text.slice(0, 200));
```
- Defined in src/tools/media.ts:551