diff --git a/includes/mail.inc b/includes/mail.inc index be2df92..3b63459 100644 --- a/includes/mail.inc +++ b/includes/mail.inc @@ -267,7 +267,7 @@ interface MailSystemInterface { * @return * The formatted $message. */ - public function format(array $message); + public function format(array $message); /** * Send a message composed by drupal_mail(). @@ -294,7 +294,7 @@ interface MailSystemInterface { * @return * TRUE if the mail was successfully accepted for delivery, otherwise FALSE. */ - public function mail(array $message); + public function mail(array $message); } /** @@ -303,39 +303,41 @@ interface MailSystemInterface { * We use delsp=yes wrapping, but only break non-spaced languages when * absolutely necessary to avoid compatibility issues. * - * We deliberately use LF rather than CRLF, see drupal_mail(). + * We deliberately use MAIL_LINE_ENDINGS rather than CRLF. * * @param $text * The plain text to process. * @param $indent (optional) * A string to indent the text with. Only '>' characters are repeated on * subsequent wrapped lines. Others are replaced by spaces. + * + * @see drupal_mail() */ function drupal_wrap_mail($text, $indent = '') { - // Convert CRLF into LF. - $text = str_replace("\r", '', $text); + // Convert CRLF into MAIL_LINE_ENDINGS. + $text = preg_replace('/\r?\n/', MAIL_LINE_ENDINGS, $text); // See if soft-wrapping is allowed. $clean_indent = _drupal_html_to_text_clean($indent); $soft = strpos($clean_indent, ' ') === FALSE; // Check if the string has line breaks. - if (strpos($text, "\n") !== FALSE) { + if (strpos($text, MAIL_LINE_ENDINGS) !== FALSE) { // Remove trailing spaces to make existing breaks hard. - $text = preg_replace('/ +\n/m', "\n", $text); + $text = preg_replace('/ +\r?\n/m', MAIL_LINE_ENDINGS, $text); // Wrap each line at the needed width. - $lines = explode("\n", $text); - array_walk($lines, '_drupal_wrap_mail_line', array('soft' => $soft, 'length' => strlen($indent))); - $text = implode("\n", $lines); + $lines = explode(MAIL_LINE_ENDINGS, $text); + array_walk($lines, '_drupal_wrap_mail_line', array('soft' => $soft, 'length' => drupal_strlen($indent))); + $text = implode(MAIL_LINE_ENDINGS, $lines); } else { // Wrap this line. - _drupal_wrap_mail_line($text, 0, array('soft' => $soft, 'length' => strlen($indent))); + _drupal_wrap_mail_line($text, 0, array('soft' => $soft, 'length' => drupal_strlen($indent))); } // Empty lines with nothing but spaces. - $text = preg_replace('/^ +\n/m', "\n", $text); + $text = preg_replace('/^ +\r?\n/m', MAIL_LINE_ENDINGS, $text); // Space-stuff special lines. - $text = preg_replace('/^(>| |From)/m', ' $1', $text); + $text = preg_replace('/^(>|From)/m', ' $1', $text); // Apply indentation. We only include non-'>' indentation on the first line. - $text = $indent . substr(preg_replace('/^/m', $clean_indent, $text), strlen($indent)); + $text = $indent . drupal_substr(preg_replace('/^/m', $clean_indent, $text), drupal_strlen($indent)); return $text; } @@ -347,11 +349,18 @@ function drupal_wrap_mail($text, $indent = '') { * The output will be suitable for use as 'format=flowed; delsp=yes' text * (RFC 3676) and can be passed directly to drupal_mail() for sending. * - * We deliberately use LF rather than CRLF, see drupal_mail(). + * We deliberately use MAIL_LINE_ENDINGS rather than CRLF. * * This function provides suitable alternatives for the following tags: - *

  • Drupal' => "Drupal * Drupal", + '
    Drupal
    ' => "Drupal", + '
    Drupal

    Drupal

    ' => "Drupal\nDrupal", + '
    Drupal
    ' => "Drupal", + // Tests some unsupported HTML tags. + 'Drupal' => "Drupal", + '' => "Drupal", + ); + + foreach ($tests as $html => $text) { + $result = drupal_html_to_text($html); + $this->assertEqual($result, $text, + var_export($html, TRUE) + . '
    ' + . str_replace("\n", '\n', check_plain(var_export($result, TRUE))) + . '
    is equal to
    ' + . str_replace("\n", '\n', check_plain(var_export($text, TRUE))) + ); + } + } + + /** + * Test $allowed_tags argument of drupal_html_to_text(). + */ + function testDrupalHtmlToTextArgs() { + // The second parameter of drupal_html_to_text() overrules the allowed tags. + $result = drupal_html_to_text('Drupal Drupal Drupal', array('b')); + $this->assertEqual($result, 'Drupal *Drupal* Drupal', 'Allowed <b> tag found.'); + + $result = drupal_html_to_text('Drupal

    Drupal

    Drupal', array('b')); + $this->assertEqual($result, 'Drupal Drupal Drupal', 'Disallowed <h1> tag not found.'); + + $result = drupal_html_to_text('Drupal

    Drupal

    Drupal', array('a', 'br', 'h1')); + $this->assertEqual($result, 'Drupal Drupal Drupal', 'Disallowed <p>, <em>, and <b> tags not found.'); + + $result = drupal_html_to_text('Drupal', array('html', 'body')); + $this->assertEqual($result, 'Drupal', 'Unsupported <html> and <body> tags not found.'); + } + + /** + * Test that whitespace is collapsed, except within 

     tags.
+   */
+  function testDrupalHtmltoTextCollapsesWhitespace() {
+    $input = "\n\n \n  \n
    Drupal\n\n 
    Drupal\n\n  
    Drupal\n Drupal
    \n \n  \n\n";
+    $collapsed = "Drupal Drupal Drupal Drupal";
+    $preserved = "Drupal\n\n Drupal\n\n  Drupal\n Drupal";
+    $result = drupal_html_to_text($input, array('p'));
+    $this->assertEqual($result, $collapsed,
+      'Whitespace inside <pre> tags is collapsed if <pre> is disallowed:
    '
+      . "
    $input

    becomes
    $result
    "
+    );
+    $result = drupal_html_to_text($input);
+    $this->assertEqual($result, $preserved,
+      'Whitespace inside <pre> tags is preserved if <pre> is allowed:
    '
+      . "
    $input

    becomes
    $result
    "
+    );
+  }
+
+  /**
+   * Test that text separated by block-level tags in HTML get separated by
+   * (at least) a newline in the plaintext version.
+   */
+  function testDrupalHtmlToTextBlockTagToNewline() {
+    $input = '[text]'
+      . '
    [address]
    '
+      . '
    [blockquote]
    '
+      . '
    [br]'
+      . '[del]'
+      . '
    [div]
    '
+      . '
    [dl-dt]
    '
+      . '
    [dt]
    '
+      . '
    [dd]
    '
+      . '
    [dd-dl]
    '
+      . '
    [h1]
    '
+      . '
    [h2]
    '
+      . '
    [h3]
    '
+      . '
    [h4]
    '
+      . '
    [h5]
    '
+      . '
    [h6]
    '
+      . '
    [hr]'
+      . '[ins]'
+      . '
    1. [ol-li]
      2. '
+      . '
    2. [li]
      3. '
+      . '
    3. [li-ol]
    '
+      . '
    [p]
    '
+      . '
    [pre]
    '
+      . ''
+      . ''
+      . '
    [table-thead--tr-td]
    [tbody-tr-td]
    [tr-td]
    '
+      . '
    • [ul-li]
      • '
+      . '
    • [li-ul]
    '
+      . '[text]';
+    $output = drupal_html_to_text($input);
+    $this->assertFalse(
+      preg_match('/\][^\n]*\[/s', $output),
+      'Block-level HTML tags should force newlines: '
+      . nl2br(check_plain($output))
+    );
+    $output_upper = drupal_strtoupper($output);
+    $upper_input = drupal_strtoupper($input);
+    $upper_output = drupal_html_to_text($upper_input);
+    $this->assertEqual(
+      $upper_output,
+      $output_upper,
+      'Tag recognition should be case-insensitive:
    '
+      . $upper_output
+      . '
    should  be equal to 
    '
+      . $output_upper
+    );
+  }
+}