Method encodeTextForFont

src/api/pdf-page.ts:2755–2790

* Encode text to a PDF string for the given font. * * Standard 14 fonts use WinAnsiEncoding (or SymbolEncoding/ZapfDingbatsEncoding). * Unencodable characters are substituted with .notdef (byte 0x00). * Embedded fonts use Identity-H encoding with glyph IDs.

(text: string, font: FontInput)

   * Embedded fonts use Identity-H encoding with glyph IDs.
   */
  private encodeTextForFont(text: string, font: FontInput): PdfString {
    if (typeof font === "string") {
      // Standard 14 font - use the appropriate encoding
      const encoding = getEncodingForStandard14(font);
      const codes: number[] = [];

      for (const char of text) {
        if (encoding.canEncode(char)) {
          // biome-ignore lint/style/noNonNullAssertion: canEncode guarantees getCode succeeds
          codes.push(encoding.getCode(char.codePointAt(0)!)!);
        } else {
          // Substitute unencodable characters with .notdef (byte 0x00)
          codes.push(0x00);
        }
      }

      const bytes = new Uint8Array(codes);

      // Use hex format for defense-in-depth: hex strings are pure ASCII
      // and immune to any string encoding transformation
      return PdfString.fromBytes(bytes);
    }

    // Embedded font - use Identity-H encoding with GIDs
    // With CIDToGIDMap /Identity, the content stream must contain glyph IDs
    const gids = font.encodeTextToGids(text);
    const bytes = new Uint8Array(gids.length * 2);

    for (let i = 0; i < gids.length; i++) {
      const gid = gids[i];
      bytes[i * 2] = (gid >> 8) & 0xff;
      bytes[i * 2 + 1] = gid & 0xff;
    }

    return PdfString.fromBytes(bytes);
  }

  // ─────────────────────────────────────────────────────────────────────────────
  // Text Extraction