@emdzej/nfsx-data-files

Plaintext parsers for BMW NFS / WinKFP data files — the inputs to the flash planner.

All file formats are line-oriented plaintext. The .DA2 / .AS2 extensions are versioned-format naming (not binary v2). See ../../docs/architecture.md §9.2 for the full format reference.

Status

| File | Parser | Status | |---|---|---| | kmm_SIT.txt | parseKmmSit | ✅ | | HWNR.DA2 | parseHwnr | ✅ | | KFCONF10.DA2 | parseKfConf | ✅ | | SGIDC.AS2, SGIDD.AS2 | parseSgId (shared) | ✅ | | npv.dat | parseNpv | ✅ | | prgifsel.dat | parsePrgIfSel | ✅ | | kmm_SG.txt, kmm_SGK.txt, kmm_SWT.txt, … (15 others) | — | ⏳ pending non-E46 SP-Daten samples |

Usage

import { parseKmmSit } from '@emdzej/nfsx-data-files';
import { readFileSync } from 'node:fs';

const content = readFileSync('kmm_SIT.txt', { encoding: 'latin1' });
const file = parseKmmSit(content);

for (const row of file.rows) {
  console.log(`0x${row.diagAddr.toString(16)} ${row.shortName} ${row.transport}`);
}

for (const err of file.unparsed) {
  console.warn(`line ${err.lineNo}: ${err.reason}`);
}

Conventions

• Encoding: BMW ships files as ISO-8859-1 / Latin-1 (umlauts in comments). Read the file with encoding: 'latin1' upstream; the parsers take JS strings and don't care about encoding.
• Error model: rows that fail to parse land in unparsed rather than throwing. A single malformed row in a 3000-row file shouldn't sink the whole load.
• Raw preserved: every parsed row carries raw: string[] for forward-compat — unknown enum values stay round-trippable.