@emdzej/ncsx-text-tables

Parsers for the text-format companion files NCSEXPER ships alongside its binary DATEN files under C:\NCSEXPER\DATEN\<BR>\. These are all line-oriented, ISO-8859-1, with their own legacy syntax conventions (semicolon-prefixed data rows, positional whitespace columns, etc.).

Spec: ../../docs/daten-format.md §2.

Files parsed

| File | Parser | Purpose | |-----------------------|-------------------|------------------------------------------------------------| | <BR>SGFAM.DAT | parseSgfam | Logical SG short-name → (CABD, SGBD, ZCS-flag, FA-flag) | | <BR>AT.000 | parseAt | FA-token dictionary: type-letter + code + FSW list | | <BR>AT.M00 | parseAtM00 | Compact M-list (Z / E / W / S category records) | | <BR>AT.ZUS | parseAtZus | AT companion (same lexical form, mostly change log) | | <BR>ZST.000 | parseZst | SA-bit / ASW-bit / FA-bit master with masks + FSW keywords |

Install

pnpm add @emdzej/ncsx-text-tables
# or:
"@emdzej/ncsx-text-tables": "workspace:*"

Encoding matters. These files are Latin-1, not UTF-8. Decode accordingly:

const raw = readFileSync(path, 'latin1');

Quick start

import { readFileSync } from 'node:fs';
import {
  parseSgfam,
  parseAt,
  parseAtM00,
  parseZst,
} from '@emdzej/ncsx-text-tables';

const root = '/Users/me/NCSEXPER/DATEN/e46';

const sgfam = parseSgfam(readFileSync(`${root}/E46SGFAM.DAT`, 'latin1'));
console.log(sgfam.rows.find((r) => r.sgName === 'EWS'));
// { sgName: 'EWS', cabd: 'A_EWS3', sgbd: 'C_EWS3', zcsFlag: …, faFlag: … }

const at = parseAt(readFileSync(`${root}/E46AT.000`, 'latin1'));
const w502 = at.records.find((r) => r.category === 'W' && r.code === '502');
console.log(w502?.fsws);   // ['SWA', …] — FSWs implied by FA code 502

const zst = parseZst(readFileSync(`${root}/E46ZST.000`, 'latin1'));
console.log(zst.records.filter((r) => r.fsw === 'DAUERTON').length);

API

| Export | Returns | |-----------------------|-----------------------------------------------| | parseSgfam(text) | SgfamFile { rows: SgfamRow[] } | | parseAt(text) | AtFile { records: AtRecord[] } | | parseAtM00(text) | AtM00File { entries: AtM00Entry[] } | | parseAtZus(text) | Same shape as parseAt | | parseZst(text) | ZstFile { header, records, unparsed } | | iterLines(text) | Low-level line iterator (handles comments) | | tokens(line) | Whitespace-tokenise a row line |

Key concepts

SGFAM — the SG family table

Per-SG declaration:

• sgName — logical short name like EWS, KMB, LSZ_E46
• cabd — coding-bundle DATEN file name to load (e.g. A_EWS3 → opens A_EWS3.C07)
• sgbd — EDIABAS SGBD file the cable speaks (e.g. C_EWS3)
• zcsFlag / faFlag — whether this SG participates in ZCS / FA scopes

AT — FA token dictionary

Each record is (category, code, fsws[]):

• category — one of W (Werks-/order option), Z (Zwang/forced), E (Entfall/excluded), S (Sonderfall/special)
• code — 4-digit short code, e.g. 0205
• fsws[] — function-keyword names this FA code implies (used by fa-asw to build the ASW bit set)

AT.M00 — compact M-list

A variant of AT used for chassis-wide forced-inclusion rules. parseAtM00 exposes the same (category, code, fsws[]) shape.

ZST — the master SA/FA/FSW table

The grand reference for SA/VN/FA-code ↔ ASW-bit-mask + FA-bit-mask + FSW keyword. Notably, the file format uses lines that start with ; to look like comments to legacy editors, but the parser knows to treat them as data when they match the right shape.

Related

• @emdzej/ncsx-chassis — loads + indexes all of these into a Chassis bundle.
• @emdzej/ncsx-fa-asw — uses parseAt output to map FA strings → ASW bit set.
• Spec: ../../docs/daten-format.md §2.