Workflow-App/Workflow/scripts/_cipc_service.js

// scripts/_cipc_service.js
//
// Long-lived wrapper for the frozen CIPC BizPortal .py scripts. Architecture
// A6 from the analysis: instead of Next.js `spawnSync`ing the .py per
// request (which hangs the Node thread when BizPortal auth stalls), this
// service owns one Python process at a time, keeps the agent-browser
// session warm via a watchdog, and answers TCP requests from the dev
// server. If the service is offline, the dev server falls back to the
// legacy `spawnSync` path — so this is fully backwards-compatible.
//
// PROTOCOL (line-delimited JSON, one request per connection):
//   client → service:  {"kind":"enterprise_no" | "director_id_no",
//                       "query":"K2018435959" | "8709225206083"}\n
//   service → client:  {"ok":true, "stdout":"<full .py stdout>",
//                       "exitCode":0,
//                       "duration_ms":73000}\n
//                    or {"ok":false, "error":"..."}\n
//
// The service never modifies the .py files — it's a thin executor that:
//   1. Ensures the agent-browser session is warm (logs in if not).
//   2. Spawns the matching .py with one positional arg.
//   3. Captures stdout/stderr.
//   4. Returns the captured output for the route to parse + enrich.
//
// A 90-second hard timeout wraps the .py spawn so even if BizPortal
// deadlocks the script, the client gets an error in <100s instead of
// hanging forever.
//
// Watchdog: every 30s, checks the agent-browser URL; if it has decayed
// away from bizprofile/user_profile, runs a login flow silently.

const net = require('net');
const { spawn } = require('node:child_process');
const path = require('node:path');
const fs = require('node:fs');

const SCRIPTS_DIR = path.resolve(__dirname);
const ENV_FILE = path.resolve(SCRIPTS_DIR, '..', '.env.local');
const VENV_PYTHON = path.join(SCRIPTS_DIR, '.venv', 'Scripts', 'python.exe');
const AGENT_BROWSER_EXE =
  'C:\\Users\\SJ\\AppData\\Roaming\\npm\\node_modules\\agent-browser\\bin\\agent-browser-win32-x64.exe';
const SESSION = 'biz-search';
const SERVICE_HOST = '127.0.0.1';
const SERVICE_PORT = 6002;
const PY_TIMEOUT_MS = 90_000;
const WATCHDOG_INTERVAL_MS = 30_000;
const LOG_PATH = path.join(SCRIPTS_DIR, '_cipc_service.log');

// Reads one key from the app's .env.local, so credentials stay out of source.
function envValue(key) {
  try {
    for (const line of fs.readFileSync(ENV_FILE, 'utf8').split(/\r?\n/)) {
      const t = line.trim();
      if (t.startsWith(`${key}=`)) return t.slice(key.length + 1).trim();
    }
  } catch (_) {
    // fall through to the explicit error below
  }
  throw new Error(`${key} not found in ${ENV_FILE}`);
}

// Login 3 — the one that was rotating successfully when Login 1 had its
// subscription expire. Code and password both come from .env.local.
const LOGIN_CREDS = {
  customer_code: envValue('CIPC_LOGIN3_CODE'),
  password: envValue('CIPC_LOGIN3_PASSWORD'),
};
const LOGIN_URL = 'https://www.bizportal.gov.za/login.aspx';

const SCRIPTS = {
  enterprise_no: 'Bizportal Enterprise No. Search.py',
  director_id_no: 'Bizportal Directopr id no search.py',
};

function log(level, msg, extra) {
  const line = `${new Date().toISOString()} [${level}] ${msg}` +
    (extra ? ' ' + JSON.stringify(extra) : '');
  try {
    fs.appendFileSync(LOG_PATH, line + '\n');
  } catch {}
  process.stdout.write(line + '\n');
}

// ---------- Agent-browser helpers ----------
function abCmdParts(args) {
  return [AGENT_BROWSER_EXE, '--session', SESSION, ...args].map((c) =>
    /[\s"]/.test(c) ? `"${c}"` : c
  ).join(' ');
}

function runAb(args, timeoutMs = 30_000) {
  // Runs a single agent-browser command by spawning the .exe directly
  // (no cmd.exe wrapper), which avoids quoting surprises on Windows.
  // Returns trimmed stdout; throws on timeout or spawn error.
  return new Promise((resolve, reject) => {
    let stdout = '', stderr = '', killed = false;
    let proc;
    try {
      proc = spawn(AGENT_BROWSER_EXE, ['--session', SESSION, ...args], {
        windowsHide: true,
      });
    } catch (e) { return reject(e); }
    const timer = setTimeout(() => {
      killed = true;
      try { proc.kill('SIGTERM'); } catch {}
    }, timeoutMs);
    proc.stdout.on('data', (d) => (stdout += d));
    proc.stderr.on('data', (d) => (stderr += d));
    proc.on('error', (e) => {
      clearTimeout(timer);
      reject(e);
    });
    proc.on('close', (code) => {
      clearTimeout(timer);
      if (killed) return reject(new Error('agent-browser timeout'));
      // agent-browser returns 0 even on bad JSON-eval results — the
      // actual error message is encoded in the JSON-stringified stdout.
      resolve({ stdout: stdout.trim(), stderr: stderr.trim(), code });
    });
  });
}

async function abEval(js, timeoutMs = 30_000) {
  const b64 = Buffer.from(js, 'utf-8').toString('base64');
  const r = await runAb(['eval', '-b', b64], timeoutMs);
  // Strip surrounding JSON-string quotes if present.
  return r.stdout.replace(/^"|"$/g, '');
}

// ---------- Login / warm-state management ----------
async function getUrl() {
  return abEval('window.location.href');
}

async function isWarm() {
  try {
    const url = await getUrl();
    return url.includes('bizprofile.aspx') || url.includes('user_profile.aspx');
  } catch (e) {
    log('WARN', 'isWarm probe failed', { err: e.message });
    return false;
  }
}

async function login() {
  log('INFO', 'logging in');
  await runAb(['open', LOGIN_URL], 30_000);
  await sleep(4_000);
  await abEval(
    "document.getElementById('cntMain_chkId_ToggleButton').click(); 'T'"
  );
  await sleep(2_000);
  const chk = await abEval(
    "!!document.getElementById('cntMain_txtCustCode') ? 'OK' : 'MISSING'"
  );
  if (!chk.includes('OK')) {
    await abEval(
      "document.getElementById('cntMain_chkId_ToggleButton').click(); 'T2'"
    );
    await sleep(2_000);
  }
  const js = [
    "var el=document.getElementById('cntMain_txtCustCode');",
    `el.value='${LOGIN_CREDS.customer_code}';`,
    "el.dispatchEvent(new Event('input',{bubbles:true}));",
    "var el2=document.getElementById('cntMain_txtPassword');",
    `el2.value='${LOGIN_CREDS.password}';`,
    "el2.dispatchEvent(new Event('input',{bubbles:true}));",
    "document.getElementById('cntMain_btnLogin').click();",
    "'L'",
  ].join('');
  await abEval(js);
  await sleep(18_000);
  const url = await getUrl();
  log('INFO', 'login flow complete', { url: url.replace(/^"|"$/g, '') });
  return url;
}

// ---------- Serialised mutex so watchdog + requests don't race ----------
let inFlight = false;
const queue = [];
async function serialised(fn) {
  return new Promise((resolve, reject) => {
    queue.push({ fn, resolve, reject });
    drain();
  });
}
async function drain() {
  if (inFlight || queue.length === 0) return;
  inFlight = true;
  const { fn, resolve, reject } = queue.shift();
  try { resolve(await fn()); }
  catch (e) { reject(e); }
  finally {
    inFlight = false;
    setImmediate(drain);
  }
}

async function ensureWarm() {
  return serialised(async () => {
    if (await isWarm()) return { warmed: false };
    log('INFO', 'session cold, logging in');
    await login();
    return { warmed: true };
  });
}

function sleep(ms) { return new Promise((r) => setTimeout(r, ms)); }

// ---------- .py execution ----------
function runPy(scriptName, arg) {
  const scriptPath = path.join(SCRIPTS_DIR, scriptName);
  // Spawn the python binary directly with the .py and query as positional
  // args — no cmd.exe shell, no quoting dance. This sidesteps the Windows
  // cmd-line escaping surprises we'd hit with `cmd /d /s /c "python.exe ..."`.
  return new Promise((resolve) => {
    const start = Date.now();
    let stdout = '', stderr = '', killedByTimeout = false;
    let proc;
    try {
      proc = spawn(VENV_PYTHON, [scriptPath, arg], {
        cwd: SCRIPTS_DIR,
        windowsHide: true,
      });
    } catch (e) {
      return resolve({ ok: false, error: `spawn failed: ${e.message}` });
    }
    const timer = setTimeout(() => {
      killedByTimeout = true;
      try { proc.kill('SIGTERM'); } catch {}
    }, PY_TIMEOUT_MS);
    proc.stdout.on('data', (d) => (stdout += d));
    proc.stderr.on('data', (d) => (stderr += d));
    proc.on('error', (e) => {
      clearTimeout(timer);
      resolve({ ok: false, error: `spawn error: ${e.message}` });
    });
    proc.on('close', (code) => {
      clearTimeout(timer);
      const duration = Date.now() - start;
      if (killedByTimeout) {
        resolve({
          ok: false,
          error: `python timeout (${PY_TIMEOUT_MS}ms)`,
          duration_ms: duration,
          stdout,
          stderr,
        });
      } else {
        resolve({
          ok: code === 0,
          exitCode: code,
          stdout,
          stderr,
          duration_ms: duration,
        });
      }
    });
  });
}

// ---------- TCP server ----------
async function handleRequest(req) {
  const { kind, query } = req || {};
  if (!SCRIPTS[kind]) return { ok: false, error: `unknown kind: ${kind}` };
  if (typeof query !== 'string' || !query.trim()) {
    return { ok: false, error: 'missing query' };
  }
  const warmResult = await ensureWarm();
  const scriptName = SCRIPTS[kind];
  const pyResult = await runPy(scriptName, query);
  if (!pyResult.ok) {
    return {
      ok: false,
      error: pyResult.error || `python exited ${pyResult.exitCode}`,
      stdout: pyResult.stdout,
      stderr: pyResult.stderr,
      duration_ms: pyResult.duration_ms,
      warmed_up: warmResult.warmed,
    };
  }
  return {
    ok: true,
    stdout: pyResult.stdout,
    exitCode: pyResult.exitCode,
    duration_ms: pyResult.duration_ms,
    warmed_up: warmResult.warmed,
  };
}

function singleConnection(socket) {
  let buf = '';
  socket.on('data', (chunk) => {
    buf += chunk.toString('utf-8');
    let idx;
    while ((idx = buf.indexOf('\n')) >= 0) {
      const line = buf.slice(0, idx);
      buf = buf.slice(idx + 1);
      if (!line) continue;
      handleRequestLine(line, socket);
      return; // one request per connection
    }
  });
  socket.on('error', () => {});
  socket.on('close', () => {});
}

async function handleRequestLine(line, socket) {
  let req;
  try { req = JSON.parse(line); }
  catch (e) {
    socket.write(JSON.stringify({ ok: false, error: 'invalid json' }) + '\n');
    return socket.end();
  }
  try {
    const res = await handleRequest(req);
    socket.write(JSON.stringify(res) + '\n');
  } catch (e) {
    socket.write(JSON.stringify({ ok: false, error: e.message || String(e) }) + '\n');
  } finally {
    socket.end();
  }
}

net.createServer(singleConnection).listen(SERVICE_PORT, SERVICE_HOST, () => {
  log('INFO', `service listening on ${SERVICE_HOST}:${SERVICE_PORT}`);
});

// Watchdog: keep the session warm silently.
setInterval(() => {
  serialised(async () => {
    if (!(await isWarm())) {
      log('INFO', 'watchdog: session decayed, re-logging in');
      try { await login(); log('INFO', 'watchdog: re-login complete'); }
      catch (e) { log('ERROR', 'watchdog login failed', { err: e.message }); }
    }
  }).catch((e) => log('ERROR', 'watchdog tick crashed', { err: e.message }));
}, WATCHDOG_INTERVAL_MS);

log('INFO', 'service starting', {
  python: VENV_PYTHON,
  scripts_dir: SCRIPTS_DIR,
  watchdog_ms: WATCHDOG_INTERVAL_MS,
  py_timeout_ms: PY_TIMEOUT_MS,
});

// Hard refuse to run twice — concurrency would corrupt the BizPortal session.
// (Linux would use a flock; on Windows we use a TCP-port-already-bound check
// via net.createServer().listen — calling .listen twice yields EADDRINUSE.)