X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=program%2Finclude%2Frcube_string_replacer.php;h=b69e9cf50b8814afd6ebe3a5ec06777666957eaf;hb=76507f7c63a660742e76889ad6e3919f3dde3bb0;hp=b1c1fbbbeb6644202857d3bbda230b92fde8d78e;hpb=e8a0682b96f5b7f297e58d101735ba20a0cc3a89;p=roundcube.git diff --git a/program/include/rcube_string_replacer.php b/program/include/rcube_string_replacer.php index b1c1fbb..b69e9cf 100644 --- a/program/include/rcube_string_replacer.php +++ b/program/include/rcube_string_replacer.php @@ -4,8 +4,8 @@ +-----------------------------------------------------------------------+ | program/include/rcube_string_replacer.php | | | - | This file is part of the RoundCube Webmail client | - | Copyright (C) 2009, RoundCube Dev. - Switzerland | + | This file is part of the Roundcube Webmail client | + | Copyright (C) 2009, The Roundcube Dev Team | | Licensed under the GNU GPL | | | | PURPOSE: | @@ -15,7 +15,7 @@ | Author: Thomas Bruederli | +-----------------------------------------------------------------------+ - $Id: $ + $Id: rcube_string_replacer.php 5481 2011-11-24 07:53:00Z alec $ */ @@ -35,11 +35,18 @@ class rcube_string_replacer function __construct() { - $url_chars = 'a-z0-9_\-\+\*\$\/&%=@#:;'; - $url_chars_within = '\?\.~,!'; - - $this->link_pattern = "/([\w]+:\/\/|\Wwww\.)([a-z0-9\-\.]+[a-z]{2,4}([$url_chars$url_chars_within]*[$url_chars])?)/i"; - $this->mailto_pattern = "/([a-z0-9][a-z0-9\-\.\+\_]*@([a-z0-9]([-a-z0-9]*[a-z0-9])?\\.)+[a-z]{2,5})/i"; + // Simplified domain expression for UTF8 characters handling + // Support unicode/punycode in top-level domain part + $utf_domain = '[^?&@"\'\\/()\s\r\t\n]+\\.([^\\x00-\\x2f\\x3b-\\x40\\x5b-\\x60\\x7b-\\x7f]{2,}|xn--[a-z0-9]{2,})'; + $url1 = '.:;,'; + $url2 = 'a-z0-9%=#@+?!&\\/_~\\[\\]{}-'; + + $this->link_pattern = "/([\w]+:\/\/|\Wwww\.)($utf_domain([$url1]?[$url2]+)*)/i"; + $this->mailto_pattern = "/(" + ."[-\w!\#\$%&\'*+~\/^`|{}=]+(?:\.[-\w!\#\$%&\'*+~\/^`|{}=]+)*" // local-part + ."@$utf_domain" // domain-part + ."(\?[$url1$url2]+)?" // e.g. ?subject=test... + .")/i"; } /** @@ -76,11 +83,19 @@ class rcube_string_replacer if (preg_match('!^(http|ftp|file)s?://!', $scheme)) { $url = $matches[1] . $matches[2]; - $i = $this->add(html::a(array('href' => $url, 'target' => '_blank'), Q($url))); } else if (preg_match('/^(\W)www\.$/', $matches[1], $m)) { - $url = 'www.' . $matches[2]; - $i = $this->add($m[1] . html::a(array('href' => 'http://' . $url, 'target' => '_blank'), Q($url))); + $url = 'www.' . $matches[2]; + $url_prefix = 'http://'; + $prefix = $m[1]; + } + + if ($url) { + $suffix = $this->parse_url_brackets($url); + $i = $this->add($prefix . html::a(array( + 'href' => $url_prefix . $url, + 'target' => '_blank' + ), Q($url)) . $suffix); } // Return valid link for recognized schemes, otherwise, return the unmodified string for unrecognized schemes. @@ -95,11 +110,13 @@ class rcube_string_replacer */ public function mailto_callback($matches) { + $href = $matches[1]; + $suffix = $this->parse_url_brackets($href); + $i = $this->add(html::a(array( - 'href' => 'mailto:' . $matches[1], - 'onclick' => "return ".JS_OBJECT_NAME.".command('compose','".JQ($matches[1])."',this)", - ), - Q($matches[1]))); + 'href' => 'mailto:' . $href, + 'onclick' => "return ".JS_OBJECT_NAME.".command('compose','".JQ($href)."',this)", + ), Q($href)) . $suffix); return $i >= 0 ? $this->get_replacement($i) : ''; } @@ -124,4 +141,38 @@ class rcube_string_replacer return preg_replace_callback(self::$pattern, array($this, 'replace_callback'), $str); } -} \ No newline at end of file + /** + * Fixes bracket characters in URL handling + */ + public static function parse_url_brackets(&$url) + { + // #1487672: special handling of square brackets, + // URL regexp allows [] characters in URL, for example: + // "http://example.com/?a[b]=c". However we need to handle + // properly situation when a bracket is placed at the end + // of the link e.g. "[http://example.com]" + if (preg_match('/(\\[|\\])/', $url)) { + $in = false; + for ($i=0, $len=strlen($url); $i<$len; $i++) { + if ($url[$i] == '[') { + if ($in) + break; + $in = true; + } + else if ($url[$i] == ']') { + if (!$in) + break; + $in = false; + } + } + + if ($i<$len) { + $suffix = substr($url, $i); + $url = substr($url, 0, $i); + } + } + + return $suffix; + } + +}