#!/usr/bin/env node
// scripts/bench_pipeline.mjs
// Quick micro-benchmark for the pipeline using mock providers.
// Measures throughput (questions/sec) over a limited run.

import { performance } from 'perf_hooks';
import path from 'path';
import os from 'os';
import { fileURLToPath } from 'url';
import { runPipelineBatch } from '../src/pipeline/pipeline.mjs';

const __filename = fileURLToPath(import.meta.url);
const __dirname = path.dirname(__filename);
const PROJECT_ROOT = path.join(__dirname, '..');

function parseArgs(argv) {
  const args = argv.slice(2);
  let limit = 50;
  let chunkLimit;
  let cacheDir;
  let randomWalk = false;
  for (let i = 0; i < args.length; i++) {
    const a = args[i];
    if (a === '--limit' || a === '-n') {
      const v = Number(args[i + 1]);
      if (!Number.isNaN(v)) limit = v;
      i++;
    } else if (a === '--chunk-limit') {
      const v = Number(args[i + 1]);
      if (!Number.isNaN(v)) chunkLimit = v;
      i++;
    } else if (a === '--cache-dir') {
      cacheDir = args[i + 1];
      i++;
    } else if (a === '--random-walk') {
      randomWalk = true;
    }
  }
  return { limit, chunkLimit, cacheDir, randomWalk };
}

function bar(label, fraction, width = 30) {
  const clamped = Math.max(0, Math.min(1, fraction));
  const filled = Math.round(clamped * width);
  const empty = width - filled;
  return `${label} [${'#'.repeat(filled)}${'.'.repeat(empty)}] ${(clamped * 100).toFixed(1)}%`;
}

async function main() {
  const { limit, chunkLimit, cacheDir, randomWalk } = parseArgs(process.argv);

  // Force mock providers for speed and determinism
  process.env.GENERATOR_PROVIDER = 'mock';
  process.env.VERIFIER_PROVIDER = 'mock';
  process.env.REWARD_PROVIDER = 'mock';
  process.env.QUESTION_PROVIDER = 'mock';
  process.env.PROVIDER_TYPE = 'mock';

  // Seed mode: question-first avoids ES by using rag chunks JSONL
  process.env.PIPELINE_SEED_MODE = 'question-first';

  // Optional random walk over chunks
  if (randomWalk) process.env.PIPELINE_RANDOM_WALK = '1';

  // Isolate cache/output
  const cachePath =
    cacheDir ||
    path.join(os.tmpdir(), `distill-cache-bench-${Date.now()}`);
  process.env.PIPELINE_CACHE_DIR = cachePath;
  const outPath = path.join(
    os.tmpdir(),
    `pipeline_gold_bench_${Date.now()}.jsonl`,
  );

  console.log('🏎️  Benchmarking pipeline (mock providers)');
  console.log(`   limit:       ${limit}`);
  console.log(`   chunkLimit:  ${chunkLimit ?? 'default'}`);
  console.log(`   randomWalk:  ${randomWalk ? 'yes' : 'no'}`);
  console.log(`   cache:       ${cachePath}`);
  console.log(`   out:         ${outPath}`);
  console.log('');

  const start = performance.now();
  const silentLogger = { log: () => {}, error: console.error };
  const result = await runPipelineBatch({
    limit,
    chunkLimit,
    verbose: false,
    outPath,
    seedMode: 'question-first',
    logger: silentLogger,
  });
  const end = performance.now();
  const ms = end - start;

  const qps = result.processed > 0 ? (result.processed / ms) * 1000 : 0;
  const acceptRatio = result.processed > 0 ? result.accepted / result.processed : 0;

  console.log('🎯 Benchmark complete');
  console.log(`   mode:        ${result.mode}`);
  console.log(`   processed:   ${result.processed}`);
  console.log(`   accepted:    ${result.accepted}`);
  console.log(`   duration:    ${ms.toFixed(1)} ms`);
  console.log(`   throughput:  ${qps.toFixed(2)} q/s`);
  console.log(`   ${bar('accept rate ', acceptRatio)}`);
  console.log(`   ${bar('throughput ', Math.min(1, qps / 50))} (normalized vs 50 q/s)`);
  console.log(`   cache dir:   ${cachePath}`);
  console.log(`   out file:    ${outPath}`);
}

main().catch((err) => {
  console.error('Benchmark error:', err);
  process.exit(1);
});