luvul_crawl/sort-logs.js

const fs = require('fs');
const path = require('path');

function usage() {
  console.log('Usage: node sort-logs.js <file-or-directory> [--write]');
  console.log('  <file-or-directory> : .txt file or directory containing .txt logs');
  console.log('  --write             : rewrite files in chronological order');
}

function parseDateFromFileName(filePath) {
  const match = path.basename(filePath).match(/(\d{4})_(\d{2})_(\d{2})/);
  if (!match) {
    return null;
  }

  return {
    year: Number(match[1]),
    month: Number(match[2]),
    day: Number(match[3]),
  };
}

function parseTimestamp(line, fileDate) {
  const match = line.match(/\((?:(\d{4})\/)?(\d{1,2})\/(\d{1,2})\s+(\d{1,2}):(\d{2}):(\d{2})\)\s*$/);
  if (!match) {
    return null;
  }

  const year = match[1] ? Number(match[1]) : fileDate?.year;
  if (!year) {
    return null;
  }

  const month = Number(match[2]);
  const day = Number(match[3]);
  const hour = Number(match[4]);
  const minute = Number(match[5]);
  const second = Number(match[6]);

  return new Date(year, month - 1, day, hour, minute, second).getTime();
}

function getTargetFiles(targetPath) {
  const resolvedPath = path.resolve(targetPath);
  const stat = fs.statSync(resolvedPath);

  if (stat.isFile()) {
    return [resolvedPath];
  }

  if (stat.isDirectory()) {
    return fs.readdirSync(resolvedPath)
      .filter((entry) => entry.endsWith('.txt'))
      .map((entry) => path.join(resolvedPath, entry))
      .sort();
  }

  throw new Error(`Unsupported target: ${resolvedPath}`);
}

function sortFile(filePath, writeBack) {
  const content = fs.readFileSync(filePath, 'utf8');
  const hasTrailingNewline = content.endsWith('\n');
  const lines = content.split(/\r?\n/).filter((line) => line.length > 0);
  const fileDate = parseDateFromFileName(filePath);

  const decorated = lines.map((line, index) => ({
    line,
    index,
    timestamp: parseTimestamp(line, fileDate),
  }));

  const sortable = decorated.filter((entry) => entry.timestamp !== null);
  if (sortable.length === 0) {
    return { changed: false, reason: 'no timestamps found' };
  }

  const sorted = [...decorated].sort((left, right) => {
    if (left.timestamp === null && right.timestamp === null) {
      return left.index - right.index;
    }
    if (left.timestamp === null) {
      return 1;
    }
    if (right.timestamp === null) {
      return -1;
    }
    if (left.timestamp !== right.timestamp) {
      return left.timestamp - right.timestamp;
    }
    return left.index - right.index;
  });

  const sortedContent = sorted.map((entry) => entry.line).join('\n') + (hasTrailingNewline ? '\n' : '');
  if (sortedContent === content) {
    return { changed: false, reason: 'already sorted' };
  }

  if (writeBack) {
    fs.writeFileSync(filePath, sortedContent, 'utf8');
  }

  return {
    changed: true,
    reason: writeBack ? 'rewritten' : 'would rewrite',
    firstBefore: decorated[0]?.line ?? '',
    firstAfter: sorted[0]?.line ?? '',
    lastBefore: decorated[decorated.length - 1]?.line ?? '',
    lastAfter: sorted[sorted.length - 1]?.line ?? '',
  };
}

function main() {
  const args = process.argv.slice(2);
  const writeBack = args.includes('--write');
  const target = args.find((arg) => !arg.startsWith('-'));

  if (!target) {
    usage();
    process.exitCode = 1;
    return;
  }

  let files;
  try {
    files = getTargetFiles(target);
  } catch (error) {
    console.error(error.message);
    process.exitCode = 1;
    return;
  }

  let changedCount = 0;
  for (const filePath of files) {
    const result = sortFile(filePath, writeBack);
    if (result.changed) {
      changedCount += 1;
    }
    console.log(`${path.relative(process.cwd(), filePath)}: ${result.reason}`);
  }

  console.log(`${files.length} file(s) checked, ${changedCount} file(s) ${writeBack ? 'rewritten' : 'need sorting'}.`);
}

main();