#!/usr/bin/env node

/**
 * Discovery script: reads all recipe .md files, finds count measurements,
 * extracts the preceding word(s), and reports items not in the pluralization dict.
 *
 * Usage: node scripts/find-missing-plurals.js
 */

const fs = require('fs');
const path = require('path');
const { findAllMeasurements } = require('../lib/measurements/matcher');
const { matchPlural } = require('../lib/measurements/plurals');

const recipesDir = path.join(__dirname, '..', 'recipes');

if (!fs.existsSync(recipesDir)) {
  console.error('recipes/ directory not found');
  process.exit(1);
}

const files = fs.readdirSync(recipesDir).filter(f => f.endsWith('.md'));
const missing = new Map(); // noun -> [{ file, line }]

for (const file of files) {
  const content = fs.readFileSync(path.join(recipesDir, file), 'utf8');
  const lines = content.split('\n');

  for (let i = 0; i < lines.length; i++) {
    const line = lines[i];
    const measurements = findAllMeasurements(line);

    for (const m of measurements) {
      if (m.type !== 'count') continue;

      const textBefore = line.slice(0, m.index);
      const result = matchPlural(textBefore);

      if (!result) {
        // Extract the last 1-3 words before the count as a candidate noun
        const trimmed = textBefore.trimEnd();
        const words = trimmed.split(/\s+/).filter(Boolean);
        if (words.length === 0) continue;

        // Skip checkbox markers and list markers
        const filtered = words.filter(w => !/^[-*\[\]x]$/.test(w));
        if (filtered.length === 0) continue;

        const candidate = filtered.slice(-2).join(' ').toLowerCase();
        if (!candidate || /^\d/.test(candidate)) continue;

        if (!missing.has(candidate)) {
          missing.set(candidate, []);
        }
        missing.get(candidate).push({ file, line: i + 1 });
      }
    }
  }
}

if (missing.size === 0) {
  console.log('All count nouns are covered by the pluralization dictionary.');
} else {
  console.log('Count measurements with unrecognized preceding nouns:\n');
  const sorted = [...missing.entries()].sort((a, b) => b[1].length - a[1].length);
  for (const [noun, locations] of sorted) {
    console.log(`  "${noun}" (${locations.length} occurrence${locations.length > 1 ? 's' : ''})`);
    for (const loc of locations) {
      console.log(`    - ${loc.file}:${loc.line}`);
    }
  }
  console.log(`\nTotal: ${missing.size} unrecognized noun(s)`);
}