| | |
|---|
| | | |
|---|
| | | $p += array('safe' => false, 'inline_html' => true); |
|---|
| | | |
|---|
| | | // special replacements (not properly handled by washtml class) |
|---|
| | | $html_search = array( |
|---|
| | | '/(<\/nobr>)(\s+)(<nobr>)/i', // space(s) between <NOBR> |
|---|
| | | '/<title[^>]*>[^<]*<\/title>/i', // PHP bug #32547 workaround: remove title tag |
|---|
| | | '/^(\0\0\xFE\xFF|\xFF\xFE\0\0|\xFE\xFF|\xFF\xFE|\xEF\xBB\xBF)/', // byte-order mark (only outlook?) |
|---|
| | | '/<html\s[^>]+>/i', // washtml/DOMDocument cannot handle xml namespaces |
|---|
| | | ); |
|---|
| | | $html_replace = array( |
|---|
| | | '\\1'.' '.'\\3', |
|---|
| | | '', |
|---|
| | | '', |
|---|
| | | '<html>', |
|---|
| | | ); |
|---|
| | | $html = preg_replace($html_search, $html_replace, trim($html)); |
|---|
| | | |
|---|
| | | // PCRE errors handling (#1486856), should we use something like for every preg_* use? |
|---|
| | | if ($html === null && ($preg_error = preg_last_error()) != PREG_NO_ERROR) { |
|---|
| | | $errstr = "Could not clean up HTML message! PCRE Error: $preg_error."; |
|---|
| | | |
|---|
| | | if ($preg_error == PREG_BACKTRACK_LIMIT_ERROR) |
|---|
| | | $errstr .= " Consider raising pcre.backtrack_limit!"; |
|---|
| | | if ($preg_error == PREG_RECURSION_LIMIT_ERROR) |
|---|
| | | $errstr .= " Consider raising pcre.recursion_limit!"; |
|---|
| | | |
|---|
| | | raise_error(array('code' => 620, 'type' => 'php', |
|---|
| | | 'line' => __LINE__, 'file' => __FILE__, |
|---|
| | | 'message' => $errstr), true, false); |
|---|
| | | return ''; |
|---|
| | | } |
|---|
| | | |
|---|
| | | // fix (unknown/malformed) HTML tags before "wash" |
|---|
| | | $html = preg_replace_callback('/(<[\/]*)([^\s>]+)/', 'rcmail_html_tag_callback', $html); |
|---|
| | | |
|---|
| | | // charset was converted to UTF-8 in rcube_storage::get_message_part(), |
|---|
| | | // change/add charset specification in HTML accordingly, |
|---|
| | | // washtml cannot work without that |
|---|
| | |
|---|
| | | if (!$rcount) { |
|---|
| | | $html = '<head>' . $meta . '</head>' . $html; |
|---|
| | | } |
|---|
| | | |
|---|
| | | // turn relative into absolute urls |
|---|
| | | $html = rcmail_resolve_base($html); |
|---|
| | | |
|---|
| | | // clean HTML with washhtml by Frederic Motte |
|---|
| | | $wash_opts = array( |
|---|
| | |
|---|
| | | $wash_opts['html_attribs'] = $p['html_attribs']; |
|---|
| | | |
|---|
| | | // initialize HTML washer |
|---|
| | | $washer = new washtml($wash_opts); |
|---|
| | | $washer = new rcube_washtml($wash_opts); |
|---|
| | | |
|---|
| | | if (!$p['skip_washer_form_callback']) |
|---|
| | | $washer->add_callback('form', 'rcmail_washtml_callback'); |
|---|
| | |
|---|
| | | } |
|---|
| | | |
|---|
| | | return $out; |
|---|
| | | } |
|---|
| | | |
|---|
| | | |
|---|
| | | /** |
|---|
| | | * Callback function for HTML tags fixing |
|---|
| | | */ |
|---|
| | | function rcmail_html_tag_callback($matches) |
|---|
| | | { |
|---|
| | | $tagname = $matches[2]; |
|---|
| | | |
|---|
| | | $tagname = preg_replace(array( |
|---|
| | | '/:.*$/', // Microsoft's Smart Tags <st1:xxxx> |
|---|
| | | '/[^a-z0-9_\[\]\!-]/i', // forbidden characters |
|---|
| | | ), '', $tagname); |
|---|
| | | |
|---|
| | | return $matches[1].$tagname; |
|---|
| | | } |
|---|
| | | |
|---|
| | | |
|---|
| | |
|---|
| | | ) { |
|---|
| | | return $types[$extension]; |
|---|
| | | } |
|---|
| | | } |
|---|
| | | |
|---|
| | | /** |
|---|
| | | * Convert all relative URLs according to a <base> in HTML |
|---|
| | | */ |
|---|
| | | function rcmail_resolve_base($body) |
|---|
| | | { |
|---|
| | | // check for <base href=...> |
|---|
| | | if (preg_match('!(<base.*href=["\']?)([hftps]{3,5}://[a-z0-9/.%-]+)!i', $body, $regs)) { |
|---|
| | | $replacer = new rcube_base_replacer($regs[2]); |
|---|
| | | $body = $replacer->replace($body); |
|---|
| | | } |
|---|
| | | |
|---|
| | | return $body; |
|---|
| | | } |
|---|
| | | |
|---|
| | | |
|---|
