inpage-format (JavaScript / TypeScript)

TypeScript library for decoding InPage .INP Urdu/Arabic word processor files.

Works in Node.js 18+ and modern browsers (pure ESM, no native deps).

Install

npm install inpage-format

Usage

import * as CFB from 'cfb';
import { decodeV1V2, decodeV3, filterParagraphsWithMeta } from 'inpage-format';

const cfbFile = CFB.read(new Uint8Array(fileBuffer), { type: 'array' });
const entry = CFB.find(cfbFile, '/InPage200') ?? CFB.find(cfbFile, '/InPage300');
const stream = new Uint8Array(entry.content);
const version = entry.name.includes('300') ? 3 : 2;

const decoded = version === 3 ? decodeV3(stream) : decodeV1V2(stream);
const { paragraphs, filteredCount } = filterParagraphsWithMeta(
  decoded.paragraphs, decoded.paragraphMeta
);

paragraphs.forEach(p => console.log(p));

API

| Export | Description | |---|---| | decodeV1V2(data: Uint8Array) | Decode InPage v1/v2 stream | | decodeV3(data: Uint8Array) | Decode InPage v3 stream | | filterParagraphs(raw: string[]) | Remove noise paragraphs | | filterParagraphsWithMeta(raw, meta) | Filter + return metadata + count | | CHAR_MAP_URDU | 110+ byte→Unicode mappings | | CHAR_MAP_ARABIC | Arabic-mode overrides | | COMPOSITE_SEQUENCES | 4-byte composite sequence map | | isUrduChar(ch: string) | Test if char is Urdu/Arabic | | PAGE_BREAK_MARKER | Sentinel string for page breaks |

Full documentation: github.com/iamahsanmehmood/inpage-format