Script to manipulate subtitles in .srt files or SRT embedded in .mkv

fix_subs.php

<?php

try {
  $usageException = new \RuntimeException("Usage: php {$argv[0]} file [file [...]]");

  if ($argc < 2) throw $usageException;
  array_shift($argv);
  $filenames = $argv;

  foreach ($filenames as $filename) {
    if (!file_exists($filename)) throw new \RuntimeException('ERROR: file not found');

    if (preg_match('/\.srt$/', $filename)) {
      fixSrtFile($filename);
      echo 'Processed ' . basename($filename) . PHP_EOL;
    } else if (preg_match('/\.mkv$/', $filename)) {
      fixMkvFile($filename);
      echo 'Written new ' . basename($filename) . PHP_EOL;
    } else {
      throw new \RuntimeException("ERROR: file '{$filename}' is not an SRT or MKV file");
    }
  }
} catch (\Exception $exception) {
  fprintf(STDERR, $exception->getMessage() . PHP_EOL);
  exit(1);
}

function fixSrtFile($filename) {
  $srt = file_get_contents($filename);
  $srt = fixSrt($srt);
  file_put_contents($filename, $srt);
}

function fixMkvFile($filename) {
  $cmd = 'mkvmerge -i -F verbose-text ' . escapeshellarg($filename);
  unset($trackOutput);
  exec($cmd, $trackOutput, $status);
  if ($status != 0) throw new \RuntimeException('ERROR: could not inspect MKV tracks');

  $tracks = [];
  $tmpFilename = null;
  try {
    foreach ($trackOutput as $line) {
      if (!preg_match('/^Track ID (\d+): subtitles \(SubRip\/SRT\) \[.*language:(\S+).*\]$/', $line, $matches)) continue;
      $track = [
        'number' => $matches[1],
        'language' => $matches[2],
        'srtFile' => tempnam('/tmp', 'srt'),
      ];
      $tracks[] = $track;

      $cmd = 'mkvextract tracks ' . escapeshellarg($filename) . ' ' . $track['number'] . ':' . escapeshellarg($track['srtFile']);
      exec($cmd, $output, $status);
      if ($status != 0) throw new \RuntimeException('ERROR: could not extract SRT track from MKV');
      fixSrtFile($track['srtFile']);          
      echo 'Processed track ' . $track['number'] . ' (' . $track['language'] . ') of ' . basename($filename) . PHP_EOL;
    }

    $tmpFilename = $filename . '~';
    $cmd = 'mkvmerge -o ' . escapeshellarg($tmpFilename) . ' -S ' . escapeshellarg($filename);
    foreach ($tracks as $track) {
      $cmd .= " --language 0:{$track['language']} {$track['srtFile']}";
    }
    exec($cmd, $output, $status);
    foreach ($tracks as $track) @unlink($track['srtFile']);
    rename($tmpFilename, $filename);
  } catch (\Exception $exception) {
    foreach ($tracks as $track) @unlink($track['srtFile']);
    if (!is_null($tmpFilename)) @unlink($tmpFilename);
    throw $exception;
  }
}

function fixSrt($srt) {
  $subs = parseSrt($srt);

  $n = count($subs);
  $lastUnique = 0;
  for ($i = 1; $i < $n; $i++) {
    list($time, $text) = $subs[$i];
    if ($time == $subs[$lastUnique][0]) {
      $subs[$lastUnique][1] .= PHP_EOL . $text;
      unset($subs[$i]);
    } else {
      $lastUnique = $i;
    }
  }
  $subs = array_values($subs);

  return getSrt($subs);
}

function parseSrt($srt) {
  $subs = [];
  $blocks = preg_split('/\n\n/', trim($srt));
  foreach ($blocks as $block) {
    $lines = explode(PHP_EOL, trim($block));
    if ($lines < 3) throw new \RuntimeException('Invalid SRT');
    $id = array_shift($lines);
    $time = array_shift($lines);
    $text = implode(PHP_EOL, $lines);
    $subs[] = [$time, $text];
  }
  return $subs;
}

function getSrt($subs) {
  $srt = '';
  $i = 1;
  foreach ($subs as $sub) {
    list($time, $text) = $sub;
    $srt .= $i . PHP_EOL;
    $srt .= $time . PHP_EOL;
    $srt .= $text . PHP_EOL . PHP_EOL;
    $i++;
  }
  return $srt;
}