Function readRawContent

src/pkg/utils/encoding.ts:293–349

(
  raw: Blob | File | Response | Uint8Array<ArrayBuffer>,
  contentType: string | null
)

 * @returns {Promise<string>}
 */
export const readRawContent = async (
  raw: Blob | File | Response | Uint8Array<ArrayBuffer>,
  contentType: string | null
): Promise<string> => {
  let uint8: Uint8Array<ArrayBuffer>;
  if (ArrayBuffer.isView(raw)) {
    uint8 = new Uint8Array<ArrayBuffer>(raw.buffer, raw.byteOffset, raw.byteLength);
  } else {
    const buffer = await raw.arrayBuffer();
    uint8 = new Uint8Array<ArrayBuffer>(buffer);
  }

  if (uint8.length === 0) {
    return "";
  }

  // 优先尝试使用 Content-Type header 中的 charset
  const headerCharset = parseCharsetFromContentType(contentType);
  if (headerCharset) {
    try {
      // 验证 charset 是否有效
      return bytesDecode(headerCharset, uint8);
    } catch (e: any) {
      console.warn(`Invalid charset from Content-Type header: ${headerCharset}, error: ${e.message}`);
    }
  }

  // BOM detection (after Content-Type header)
  const bomEncoding = detectBOM(uint8);
  if (bomEncoding) return bytesDecode(bomEncoding, uint8);

  const checkSize = Math.min(uint8.length, CHECK_SIZE);

  // Heuristic detection (first 16 KB)
  const heuristicEncoding = guessByNullPattern(uint8, checkSize);
  if (heuristicEncoding && validatesHeuristicEncoding(heuristicEncoding, uint8)) {
    try {
      return bytesDecode(heuristicEncoding, uint8);
    } catch {
      // Invalid full decode despite a valid sample: fall through to UTF-8/legacy.
    }
  }

  // UTF-8 validation → legacy detection
  let encoding = "utf-8";
  try {
    assertLikelyUtf8(uint8);
  } catch {
    const mostlyUtf8 = decodeMostlyUtf8(uint8);
    if (mostlyUtf8 !== null) return mostlyUtf8;

    // Invalid UTF-8 → use chardet-based legacy detection
    encoding = legacyDetectEncoding(uint8);
  }

  return bytesDecode(encoding, uint8);
};