Save pages

save.php

<?php

  error_reporting(E_ALL);
  ini_set('display_errors', 1);


  class ExportHtml {

    const SAVE_DIR = '/home/ivan/Ubuntu One/saved';


    public function __construct() {
      $header = array(
        "User-Agent: Opera/9.62 (Windows NT 5.1; U; ru) Presto/2.1.1",
        "Accept: text/html, application/xml;q=0.9, application/xhtml+xml, image/png, image/jpeg, image/gif, image/x-xbitmap, */*;q=0.1",
        "Accept-Language: ru-RU,ru;q=0.9,en;q=0.8",
        "Accept-Charset: iso-8859-1, utf-8, utf-16, *;q=0.1",
        "Referer: http://translate.google.com/translate_t",
        "Connection: Keep-Alive"
      );
      $this->ch = curl_init(); // init curl
      curl_setopt($this->ch, CURLOPT_COOKIEJAR, "/tmp/cookie");
      curl_setopt($this->ch, CURLOPT_COOKIEFILE, "/tmp/cookie");
      curl_setopt($this->ch, CURLOPT_HTTPHEADER, $header); // set our headers information
      curl_setopt($this->ch, CURLOPT_HEADER, 0); // set headers (0 = no headers in result)
      curl_setopt($this->ch, CURLOPT_RETURNTRANSFER, 1); // type of transfer (1 = to string)
      curl_setopt($this->ch, CURLOPT_TIMEOUT, 10); // time to wait in
      curl_setopt($this->ch, CURLOPT_POST, 0);
    }


    public function prepareHtml($html) {
      $newHtml = $html;
      preg_match_all('!<img.*src\s*=\s*[\"\'](.*)[\"\'].*>!Uui', $html, $images);
      foreach ($images[0] as $index => $string) {
        $url = $images[1][$index];

        curl_setopt($this->ch, CURLOPT_URL, $url); // set the url to fetch
        $data = curl_exec($this->ch);
        $info = curl_getinfo($this->ch);

        if (strpos($info['content_type'], 'image/') === 0 and $info['http_code'] == 200) {
          $newHtml = str_replace($string, '<img src="data:' . $info['content_type'] . ';base64,' . base64_encode($data) . '" />', $newHtml);
        }

      }
      if (!empty($_SERVER['HTTP_REFERER'])) {
        $newHtml .= '<br><br><br><a href="' . $_SERVER['HTTP_REFERER'] . '">' . $_SERVER['HTTP_REFERER'] . '</a>';
      }

      return $newHtml;
    }

    public function saveHtml($html, $tags) {
      $fileDir = '/' . date('Y-m') . '/' . date('d');
      $file = '/' . md5(time()) . '.html';

      if (!is_dir(static::SAVE_DIR . $fileDir)) {
        mkdir(static::SAVE_DIR . $fileDir, 0755, true);
      }
      $html = '<div class="tags">' . $tags . '</div>' . $html;
      file_put_contents(static::SAVE_DIR . $fileDir . $file, $html);
      $partialFile = $fileDir . $file;

      $this->updateMainFile($tags, $partialFile);
      return $partialFile;
    }

    protected function updateMainFile($tags, $filePath) {
      $filePath = ltrim($filePath, '/');
      $yearFile = static::SAVE_DIR . '/' . date('Y') . '.html';
      if (is_file($yearFile)) {
        $data = file_get_contents($yearFile);
      } else {
        $data = '';
      }

      $data = '<div data-tags="' . $tags . '">
        <span>' . strftime("%F %T") . '</span> - <span>' . $tags . '</span>
        <br>
        <a href="' . $filePath . '">' . $filePath . '</a>
      </div>' . $data;

      file_put_contents($yearFile, $data);
    }
  }


  $exportHtml = new ExportHtml();
?>

<?
  if (!empty($_POST['html'])) {
    $tags = !empty($_POST['tags']) ? $_POST['tags'] : '';
    $html = $_POST['html'];

    $html = $exportHtml->prepareHtml($html);
    $file = $exportHtml->saveHtml($html, $tags);

    $yearFileUrl = 'file://' . ExportHtml::SAVE_DIR . '/' . date('Y') . '.html';
    ?>
    Done: <?= $file ?><br>
    <br>
    <?= $yearFileUrl ?>
    <? if (!empty($_SERVER['HTTP_REFERER'])) { ?>
      <br>
      <br>
      <a href="<?= $_SERVER['HTTP_REFERER'] ?>"><?= $_SERVER['HTTP_REFERER'] ?></a>
    <? } ?>
  <?
  }
?>