extractTextFromPdf function exported ✗ 0.0%

export async function extractTextFromPdf(
  buffer: ArrayBuffer,
  maxChars = 12000,
): Promise<string> {
  log.debug(
    `extractTextFromPdf: buffer=${(buffer.byteLength / 1024).toFixed(1)}KB, maxChars=${maxChars}`,
  );
  await ensureWorker();

  const pdfjsLib = await import("pdfjs-dist");

  const loadingTask = pdfjsLib.getDocument({ data: buffer });
  const pdf = await loadingTask.promise;

  log.debug(`PDF carregado para extração: ${pdf.numPages} página(s)`);
  const textParts: string[] = [];

  for (let pageNum = 1; pageNum <= pdf.numPages; pageNum++) {
    try {
      const page = await pdf.getPage(pageNum);
      const content = await page.getTextContent();

      const pageText = (content.items as Array<{ str?: string }>)
        .filter((item) => typeof item.str === "string")
        .map((item) => item.str!)
        .join(" ")
        .replace(/ {2,}/g, " ")
        .trim();

      log.debug(`Página ${pageNum}: ${pageText.length} caracteres extraidos.`);
      if (pageText) textParts.push(pageText);
      page.cleanup();
    } catch (pageErr) {
      log.warn(`Falha ao extrair texto da página ${pageNum}:`, pageErr);
    }
  }

  await pdf.cleanup();

  const result = textParts.join("\n\n").slice(0, maxChars);
  log.info(
    `extractTextFromPdf concluído: ${result.length} chars (de ${textParts.length} página(s) com texto).`,
  );
  return result;
}