Document Parser

pnpm add @aicr/document-parser

import { parseDocument, parsePDF, parseDOCX, parseXLSX } from '@aicr/document-parser';
 
// Auto-detect format
const result = await parseDocument(buffer, { filename: 'report.pdf' });
 
console.log(result.text);      // Full text content
console.log(result.pages);     // Page-by-page content
console.log(result.tables);    // Extracted tables
console.log(result.metadata);  // Document metadata

interface ParseResult {
  format: 'pdf' | 'docx' | 'xlsx' | 'text' | 'markdown' | 'html';
  text: string;
  metadata: DocumentMetadata;
  pages: PageContent[];
  tables: TableData[];
  images: ImageData[];
  stats: ParseStats;
}

const result = await parsePDF(pdfBuffer, { maxPages: 10 });

const result = await parseDOCX(docxBuffer);

const result = await parseXLSX(xlsxBuffer);
 
// Access sheet data
result.tables.forEach(table => {
  console.log(`Sheet ${table.pageNumber}:`, table.headers);
  table.rows.forEach(row => console.log(row));
});

const result = await parseText(textContent, { format: 'markdown' });

interface DocumentMetadata {
  pageCount: number;
  wordCount: number;
  charCount: number;
  title?: string;
  author?: string;
  createdAt?: Date;
  modifiedAt?: Date;
  extra?: Record<string, unknown>;
}

interface TableData {
  pageNumber: number;
  tableIndex: number;
  headers: string[];
  rows: string[][];
  cells?: CellData[][];
}
 
interface CellData {
  value: string;
  type: 'string' | 'number' | 'boolean' | 'date' | 'formula' | 'empty';
  raw?: unknown;
}

interface ParseStats {
  durationMs: number;
  fileSizeBytes: number;
  pagesParsed: number;
  tablesExtracted: number;
  imagesFound: number;
  warnings: string[];
}

import { parseDocument, ParserError } from '@aicr/document-parser';
 
try {
  const result = await parseDocument(buffer);
} catch (error) {
  if (error instanceof ParserError) {
    switch (error.code) {
      case 'UNSUPPORTED_FORMAT':
        console.error('File format not supported');
        break;
      case 'ENCRYPTED_DOCUMENT':
        console.error('Document is password protected');
        break;
      case 'CORRUPTED_FILE':
        console.error('File appears to be corrupted');
        break;
      case 'PARSE_FAILED':
        console.error('Parse failed:', error.message);
        break;
    }
  }
}