Files
COMERCIAL-GTI/scripts/extrair-base-cnpj.mjs
T
DESKTOP-F2O4A80\UserandClaude b8d10b210c Commit inicial: GTI Maquina Comercial
App local (Node + SQLite) para captura, tratamento e exportacao de leads
comerciais, com dashboard de gestao, login/usuarios e integracao Z-PRO.

Co-Authored-By: Claude <noreply@anthropic.com>
2026-07-12 14:22:21 -03:00

77 lines
2.6 KiB
JavaScript

import path from 'node:path';
import { loadDotEnv, paths } from '../src/env.mjs';
import { openDb, ensureSeeds } from '../src/db.mjs';
import { extractUf } from '../src/extract/rfb-extractor.mjs';
// Extrai a base CNPJ de uma UF inteira a partir dos zips da Receita Federal:
// node scripts/extrair-base-cnpj.mjs --uf=BA
// Pre-requisito: zips baixados em prospeccao/cnpj/rfb-raw (scripts/baixar-base-rfb.sh).
await loadDotEnv();
const args = new Map(
process.argv.slice(2).map((arg) => {
const [key, value = 'true'] = arg.replace(/^--/, '').split('=');
return [key, value];
})
);
const uf = (args.get('uf') || '').toUpperCase();
if (!/^[A-Z]{2}$/.test(uf)) {
console.error('Uso: node scripts/extrair-base-cnpj.mjs --uf=BA');
process.exit(1);
}
const rawDir = args.get('raw-dir') || paths.rfbRaw;
const outCsv = path.join(paths.cnpjDir, `estabelecimentos-${uf}.csv`);
const checkpointPath = path.join(paths.cnpjDir, `checkpoint-estabelecimentos-${uf}.ndjson`);
async function main() {
const db = openDb();
await ensureSeeds(db, { log: (m) => console.log(`[extrair] ${m}`) });
const inserted = db.prepare(
"INSERT INTO extractions (uf, status, started_at) VALUES (?, 'running', datetime('now'))"
).run(uf);
const extractionId = Number(inserted.lastInsertRowid);
db.prepare('UPDATE extractions SET pid = ? WHERE id = ?').run(process.pid, extractionId);
try {
const result = await extractUf({
uf,
rawDir,
outCsv,
checkpointPath,
db,
log: (m) => console.log(`[extrair] ${m}`),
onProgress: (p) => {
console.log(`[extrair] ${p.pct}% ${p.phase}: ${p.message}`);
db.prepare('UPDATE extractions SET progress_json = ? WHERE id = ?')
.run(JSON.stringify(p), extractionId);
}
});
if (result.cancelled) {
db.prepare("UPDATE extractions SET status = 'cancelled', finished_at = datetime('now') WHERE id = ?")
.run(extractionId);
} else {
db.prepare(
`UPDATE extractions SET status = 'done', csv_path = ?, total_rows = ?, with_phone = ?,
municipios_count = ?, finished_at = datetime('now') WHERE id = ?`
).run(result.csvPath, result.totalRows, result.withPhone, result.municipiosCount, extractionId);
}
} catch (error) {
db.prepare(
"UPDATE extractions SET status = 'error', error = ?, finished_at = datetime('now') WHERE id = ?"
).run(String(error.message || error), extractionId);
throw error;
} finally {
db.close();
}
}
main().catch((error) => {
console.error(`[extrair] erro: ${error.stack || error.message}`);
process.exit(1);
});