/**
* Test example messages
*/
public function testExampleMessages()
{
foreach (new DirectoryIterator(dirname(dirname(dirname(__DIR__))) . '/example_messages') as $file) {
if ($file->isDot() || $file->isDir()) {
continue;
}
if ($file->isFile() && $file->getExtension() == 'eml') {
$this->assertEquals('Email Reply', EmailReplyExtractor::extractReply($file->getPathname()));
}
}
}
/**
* Test example messages
*/
public function testExampleMessages()
{
foreach (new DirectoryIterator(dirname(dirname(dirname(__DIR__))) . '/example_messages') as $file) {
if ($file->isDot() || $file->isDir()) {
continue;
}
if ($file->isFile() && $file->getExtension() == 'eml') {
$expected_text = $this->setExpectedTextByFileName($file->getFilename());
$this->assertEquals($expected_text, EmailReplyExtractor::extractReplyEml($file->getPathname()));
// assert that text of the stripped emails match the original text even with <br /> instead of new row char
$this->assertEquals(nl2br($expected_text), nl2br(EmailReplyExtractor::extractReplyEml($file->getPathname())));
}
}
}
/**
* @param string $html
* @return string
*/
static function toPlainText($html)
{
$plain = (string) $html;
// strip slashes
$plain = (string) trim(stripslashes($plain));
// strip unnecessary characters
$plain = (string) preg_replace(["/\r/", "/<script[^>]*>.*?<\\/script>/si", "/<style[^>]*>.*?<\\/style>/is", "/style=\".*?\"/"], "", $plain);
// entities to convert (this is not a definite list)
$entities = [' ' => [' ', ' '], '"' => ['"', '”', '“', '“', '”', '“', '”'], '\'' => [''', '’', '‘', '‘', '’'], '>' => ['>'], '<' => ['<'], '&' => ['&', '&'], '(c)' => ['©', '©'], '(R)' => ['®', '®'], '(tm)' => ['™', '™', '™'], '--' => ['—', '—', '—'], '-' => ['–', '−', '–', '−'], '*' => ['•', '•', '•'], '�' => ['£', '£'], 'EUR' => ['€', '€']];
// convert specified entities
foreach ($entities as $character => $entity) {
$plain = (string) str_replace($entity, $character, $plain);
}
// strip other not previously converted entities
$plain = (string) preg_replace(['/&[^&;]+;/si'], "", $plain);
// <p> converts to 2 newlines
$plain = (string) preg_replace('/<p[^>]*>/i', "\n\n", $plain);
// <p>
// new line after div
$plain = (string) preg_replace('/<div[^>]*>/i', "\n", $plain);
// <div>
// uppercase html elements
$plain = (string) preg_replace_callback('/<h[123456][^>]*>(.*?)<\\/h[123456]>/i', function ($matches) {
return "\n\n" . mb_strtoupper($matches[1]) . "\n\n";
}, $plain);
// <h1-h6>
$plain = (string) preg_replace_callback(['/<b[^>]*>(.*?)<\\/b>/i', '/<strong[^>]*>(.*?)<\\/strong>/i'], function ($matches) {
return $matches[1];
}, $plain);
// <b> <strong>
// deal with italic elements
$plain = (string) preg_replace(array('/<i[^>]*>(.*?)<\\/i>/i', '/<em[^>]*>(.*?)<\\/em>/i'), '_\\1_', $plain);
// <i> <em>
// elements that convert to 2 newlines
$plain = (string) preg_replace(array('/(<ul[^>]*>|<\\/ul>)/i', '/(<ol[^>]*>|<\\/ol>)/i', '/(<table[^>]*>|<\\/table>)/i'), "\n\n", $plain);
// <ul> <ol> <table>
// elements that convert to single newline
$plain = (string) preg_replace(array('/<br[^>]*>/i', '/(<tr[^>]*>|<\\/tr>)/i'), "\n", $plain);
// <br> <tr>
// images
$plain = (string) preg_replace(array('/<img\\s+[^>]*src="([^"]*)"[^>]*>/i'), "[Image: \\1]", $plain);
// <br> <tr>
// <hr> converts to --------------------//---
$plain = (string) preg_replace('/<hr[^>]*>/i', "\n-------------------------\n", $plain);
// <hr>
// other table tags
$plain = (string) preg_replace('/<td[^>]*>(.*?)<\\/td>/i', "\t\\1\n", $plain);
// <td>
$plain = (string) preg_replace_callback('/<th[^>]*>(.*?)<\\/th>/i', function ($matches) {
return "\t\t" . mb_strtoupper($matches[0]) . "\n";
}, $plain);
// <th>
// list elements
$plain = (string) preg_replace('/<li[^>]*>(.*?)<\\/li>/i', "* \\1\n", $plain);
// <li>with content</li>
$plain = (string) preg_replace('/<li[^>]*>/i', "\n* ", $plain);
// <li />
// handle anchors
$plain = (string) preg_replace_callback('/<a [^>]*href="([^"]+)"[^>]*>(.*?)<\\/a>/i', function ($matches) {
$url = $matches[1];
$text = $matches[2];
if (EmailReplyExtractor::strStartsWith($url, 'http://') || EmailReplyExtractor::strStartsWith($url, 'https://')) {
return "{$text} [{$url}]";
} else {
if (EmailReplyExtractor::strStartsWith($url, 'mailto:')) {
return $text . ' [' . substr($url, 7) . ']';
} else {
return $text;
}
}
}, $plain);
// <a href="$url">$text</a>
// handle blockquotes
$plain = (string) preg_replace_callback('/<blockquote[^>]*>(.*?)<\\/blockquote>/is', function ($blockquote_content) {
$blockquote_content = isset($blockquote_content[1]) ? $blockquote_content[1] : '';
$lines = (array) explode("\n", $blockquote_content);
$return = array();
if (!empty($lines)) {
foreach ($lines as $line) {
$return[] = '> ' . $line;
}
}
return "\n\n" . implode("\n", $return) . "\n\n";
}, $plain);
$plain = (string) preg_replace('/<title[^>]*>(.*?)<\\/title>/i', "", $plain);
// remove unnecessary title tag
// strip other tags
$plain = (string) strip_tags($plain);
// clean up unnecessary newlines
$plain = (string) preg_replace("/\n\\s+\n/", "\n\n", $plain);
$plain = (string) preg_replace("/[\n]{3,}/", "\n\n", $plain);
return trim($plain);
}
/**
* Test if we can detect Hotmail
*/
public function testDetectHotmail()
{
$this->assertEquals(EmailReplyExtractor::HOTMAIL, EmailReplyExtractor::detectMailer(['received' => "from BAY174-W36 ([65.54.190.187]) by bay0-omc3-s10.bay0.hotmail.com with Microsoft SMTPSVC(6.0.3790.4675);\n Wed, 22 Jan 2014 18:02:13 -0800"]));
}
/**
* Test if we can detect MailRu
*/
public function testDetectMailRu()
{
$this->assertEquals(EmailReplyExtractor::MAIL_RU_MAIL, EmailReplyExtractor::detectMailer(['message-id' => '<*****@*****.**>']));
}
