Function readPDF

source code/utils/pdf.ts:36–115

(filePath: string)

 * @returns Result containing PDF data or a structured error
 */
export async function readPDF(filePath: string): Promise<
  PDFResult<{
    type: 'pdf'
    file: {
      filePath: string
      base64: string
      originalSize: number
    }
  }>
> {
  try {
    const fs = getFsImplementation()
    const stats = await fs.stat(filePath)
    const originalSize = stats.size

    // Check if file is empty
    if (originalSize === 0) {
      return {
        success: false,
        error: { reason: 'empty', message: `PDF file is empty: ${filePath}` },
      }
    }

    // Check if PDF exceeds maximum size
    // The API has a 32MB total request limit. After base64 encoding (~33% larger),
    // a PDF must be under ~20MB raw to leave room for conversation context.
    if (originalSize > PDF_TARGET_RAW_SIZE) {
      return {
        success: false,
        error: {
          reason: 'too_large',
          message: `PDF file exceeds maximum allowed size of ${formatFileSize(PDF_TARGET_RAW_SIZE)}.`,
        },
      }
    }

    const fileBuffer = await readFile(filePath)

    // Validate PDF magic bytes — reject files that aren't actually PDFs
    // (e.g., HTML files renamed to .pdf) before they enter conversation context.
    // Once an invalid PDF document block is in the message history, every subsequent
    // API call fails with 400 "The PDF specified was not valid" and the session
    // becomes unrecoverable without /clear.
    const header = fileBuffer.subarray(0, 5).toString('ascii')
    if (!header.startsWith('%PDF-')) {
      return {
        success: false,
        error: {
          reason: 'corrupted',
          message: `File is not a valid PDF (missing %PDF- header): ${filePath}`,
        },
      }
    }

    const base64 = fileBuffer.toString('base64')

    // Note: We cannot check page count here without parsing the PDF
    // The API will enforce the 100-page limit and return an error if exceeded