mozilla/pdf.js · error · Error

Invalid pageIndex request.

Error message

Invalid pageIndex request.

What it means

Thrown by PDFDocumentProxy.getPageIndex(ref) when the supplied reference does not look like a PDF object reference. The guard isRefProxy requires an object with integer properties num>=0 and gen>=0 (matching an indirect object identifier). It exists because the worker message GetPageIndex needs a real reference; anything else cannot be resolved to a page.

Source

Thrown at src/display/api.js:3059

        }

        const page = new PDFPageProxy(
          pageIndex,
          pageInfo,
          this,
          this.pagesMapper,
          this._params.pdfBug
        );
        this.#pageCache.set(pageIndex, page);
        return page;
      });
    this.#pagePromises.set(pageIndex, promise);
    return promise;
  }

  async getPageIndex(ref) {
    if (!isRefProxy(ref)) {
      throw new Error("Invalid pageIndex request.");
    }
    const index = await this.messageHandler.sendWithPromise("GetPageIndex", {
      num: ref.num,
      gen: ref.gen,
    });
    const pageNumber = this.pagesMapper.getPageNumber(index + 1);
    if (pageNumber === 0) {
      throw new Error("GetPageIndex: page has been removed.");
    }
    return pageNumber - 1;
  }

  getAnnotations(pageIndex, intent) {
    return this.messageHandler.sendWithPromise("GetAnnotations", {
      pageIndex: this.pagesMapper.getPageId(pageIndex + 1) - 1,
      intent,
    });
  }

View on GitHub (pinned to 5903d58d58)

Solutions

1. Validate the value with the same shape as isRefProxy before calling: object with integer num>=0 and gen>=0.
2. If the source is a destination array, check Array.isArray(dest) && typeof dest[0] === 'object' before calling getPageIndex(dest[0]).
3. Use pdfDocument.getPageIndex only after pdfDocument.getPage has already resolved a ref, or cache refs from getPage output.

Example fix

// before
const idx = await pdfDocument.getPageIndex(dest[0]);

// after
const ref = dest?.[0];
if (ref && Number.isInteger(ref.num) && ref.num >= 0 && Number.isInteger(ref.gen) && ref.gen >= 0) {
  const idx = await pdfDocument.getPageIndex(ref);
} else {
  // handle string/null destination via getDestination
}

function isRefLike(v) {
  return v != null &&
    typeof v === 'object' &&
    Number.isInteger(v.num) && v.num >= 0 &&
    Number.isInteger(v.gen) && v.gen >= 0;
}
// before calling:
if (!isRefLike(ref)) return null;

const isRefLike = (v): v is { num: number; gen: number } =>
  !!v && typeof v === 'object' &&
  Number.isInteger((v as any).num) && (v as any).num >= 0 &&
  Number.isInteger((v as any).gen) && (v as any).gen >= 0;

try {
  const idx = await pdf.getPageIndex(ref);
} catch (e) {
  if (e.message === 'Invalid pageIndex request.') { /* skip bad dest */ }
  else throw e;
}

When it happens

Trigger: Calling pdfDocument.getPageIndex(ref) with undefined, null, a plain number, a string, or an object missing/overflowing num or gen. Common when chaining from getPage(dest[0]) where dest is an array that does not start with a ref, or when passing a deserialized ref that lost its fields.

Common situations: Resolving a named/remote GoTo destination whose target array begins with a page number instead of a ref; passing an outline destination straight through without validating; ref objects built from JSON without Number-casting their fields.

Related errors

• Invalid updateScale options: either `steps` or `scaleFactor`
• BinaryCMapReader.process: Invalid dataSize.
• Page count in top-level pages dictionary is not an integer.
• Invalid type in PageLabel dictionary.
• Invalid style in PageLabel dictionary.

AI-assisted analysis of mozilla/pdf.js@5903d58d58 (2026-08-13). Data as JSON: /api/errors/4bd6f98cc5b498fe. Report an issue: GitHub.