Method parseTable

src/parser/xref-parser.ts:139–182

* Parse traditional xref table format. * Scanner must be positioned at "xref" keyword.

()

   * Scanner must be positioned at "xref" keyword.
   */
  parseTable(): XRefData {
    const entries = new Map<number, XRefEntry>();

    // Consume "xref" keyword
    this.expectKeyword("xref");
    this.skipWhitespaceFromCurrent();

    // Read subsections until we hit "trailer"
    // Skip whitespace before each check (some PDFs have extra newlines between entries and trailer)
    while (true) {
      this.skipWhitespaceFromCurrent();

      if (this.peekKeyword("trailer")) {
        break;
      }

      this.parseSubsection(entries);
    }

    // Consume "trailer" keyword
    this.expectKeyword("trailer");
    this.skipWhitespaceFromCurrent();

    // Parse trailer dictionary
    const reader = new TokenReader(this.scanner);
    const parser = new ObjectParser(reader);
    const result = parser.parseObject();

    if (result === null || result.object.type !== "dict") {
      throw new XRefParseError("Invalid trailer dictionary");
    }

    const trailer = result.object;

    // Extract /Prev if present
    const prevNum = trailer.getNumber("Prev");
    const prev = prevNum?.value;

    return {
      entries,
      trailer,
      prev,
    };
  }

  /**
   * Parse xref stream format (PDF 1.5+).