diff options
-rw-r--r-- | phpBB/develop/regex_idn.php | 62 | ||||
-rw-r--r-- | phpBB/includes/functions.php | 24 |
2 files changed, 52 insertions, 34 deletions
diff --git a/phpBB/develop/regex_idn.php b/phpBB/develop/regex_idn.php index 4124829a11..d871695c50 100644 --- a/phpBB/develop/regex_idn.php +++ b/phpBB/develop/regex_idn.php @@ -8,7 +8,6 @@ // die("Please read the first lines of this script for instructions on how to enable it"); - // IP regular expressions $dec_octet = '(?:\d{1,2}|1\d\d|2[0-4]\d|25[0-5])'; @@ -111,33 +110,42 @@ $add_deviations = '\x{00DF}\x{03C2}\x{200C}\x{200D}'; $remove_chars = "$no_cc$no_symbol$no_hangul$no_cdm$no_musical$no_ancient_greek_musical$no_certain_exceptions"; $add_chars = "$add_certain_exceptions$add_deviations"; -$pct_encoded = "%[\dA-F]{2}"; -$unreserved = "$add_chars\pL0-9\-._~"; -$sub_delims = '!$&\'()*+,;='; -$pchar = "(?:[^$remove_chars]*[$unreserved$sub_delims:@|]+|$pct_encoded)"; // rfc: no "|" +// Initialize inline mode +$inline = false; + +do +{ + $inline = !$inline; -$scheme = '[a-z][a-z\d+\-.]*'; -$reg_name = "(?:[^$remove_chars]*[$unreserved$sub_delims:@|]+|$pct_encoded)+"; // rfc: * instead of + and no "|" and no "@" and no ":" (included instead of userinfo) -//$userinfo = "(?:(?:[$unreserved$sub_delims:]+|$pct_encoded))*"; -$ipv4_simple = '[0-9.]+'; -$ipv6_simple = '\[[a-z0-9.]+:[a-z0-9.]+:[a-z0-9.:]+\]'; -$host = "(?:$reg_name|$ipv4_simple|$ipv6_simple)"; -$port = '\d*'; -//$authority = "(?:$userinfo@)?$host(?::$port)?"; -$authority = "$host(?::$port)?"; -$segment = "$pchar*"; -$path_abempty = "(?:/$segment)*"; -$hier_part = "/{2}$authority$path_abempty"; -$query = "(?:[^$remove_chars]*[$unreserved$sub_delims:@/?|]+|$pct_encoded)*"; // pchar | "/" | "?", rfc: no "|" -$fragment = $query; + $pct_encoded = "%[\dA-F]{2}"; + $unreserved = "$add_chars\pL0-9\-._~"; + $sub_delims = ($inline) ? '!$&\'(*+,;=' : '!$&\'()*+,;='; + $scheme = ($inline) ? '[a-z][a-z\d+]*': '[a-z][a-z\d+\-.]*' ; // avoid automatic parsing of "word" in "last word.http://..." + $pchar = "(?:[^$remove_chars]*[$unreserved$sub_delims:@|]+|$pct_encoded)"; // rfc: no "|" -$url = "$scheme:$hier_part(?:\?$query)?(?:\#$fragment)?"; -echo 'URL: ' . $url . "<br /><br />\n\n"; + $reg_name = "(?:[^$remove_chars]*[$unreserved$sub_delims:@|]+|$pct_encoded)+"; // rfc: * instead of + and no "|" and no "@" and no ":" (included instead of userinfo) + //$userinfo = "(?:(?:[$unreserved$sub_delims:]+|$pct_encoded))*"; + $ipv4_simple = '[0-9.]+'; + $ipv6_simple = '\[[a-z0-9.]+:[a-z0-9.]+:[a-z0-9.:]+\]'; + $host = "(?:$reg_name|$ipv4_simple|$ipv6_simple)"; + $port = '\d*'; + //$authority = "(?:$userinfo@)?$host(?::$port)?"; + $authority = "$host(?::$port)?"; + $segment = "$pchar*"; + $path_abempty = "(?:/$segment)*"; + $hier_part = "/{2}$authority$path_abempty"; + $query = "(?:[^$remove_chars]*[$unreserved$sub_delims:@/?|]+|$pct_encoded)*"; // pchar | "/" | "?", rfc: no "|" + $fragment = $query; -// no scheme, shortened authority, but host has to start with www. -$www_url = "www\.$reg_name(?::$port)?$path_abempty(?:\?$query)?(?:\#$fragment)?"; -echo 'www.URL: ' . $www_url . "<br /><br />\n\n"; + $url = "$scheme:$hier_part(?:\?$query)?(?:\#$fragment)?"; + echo (($inline) ? 'URL inline: ' : 'URL: ') . $url . "<br /><br />\n\n"; -// no schema and no authority -$relative_url = "$segment$path_abempty(?:\?$query)?(?:\#$fragment)?"; -echo 'relative URL: ' . $relative_url . "<br /><br />\n\n"; + // no scheme, shortened authority, but host has to start with www. + $www_url = "www\.$reg_name(?::$port)?$path_abempty(?:\?$query)?(?:\#$fragment)?"; + echo (($inline) ? 'www.URL_inline: ' : 'www.URL: ') . $www_url . "<br /><br />\n\n"; + + // no schema and no authority + $relative_url = "$segment$path_abempty(?:\?$query)?(?:\#$fragment)?"; + echo (($inline) ? 'relative URL inline: ' : 'relative URL: ') . $relative_url . "<br /><br />\n\n"; +} +while ($inline); diff --git a/phpBB/includes/functions.php b/phpBB/includes/functions.php index d62e2111c3..9cd43765a4 100644 --- a/phpBB/includes/functions.php +++ b/phpBB/includes/functions.php @@ -3327,23 +3327,33 @@ function get_preg_expression($mode) break; case 'url': + // generated with regex_idn.php file in the develop folder + return "[a-z][a-z\d+\-.]*:/{2}(?:(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})+|[0-9.]+|\[[a-z0-9.]+:[a-z0-9.]+:[a-z0-9.:]+\])(?::\d*)?(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?"; + break; + case 'url_inline': - $inline = ($mode == 'url') ? ')' : ''; - $scheme = ($mode == 'url') ? '[a-z\d+\-.]' : '[a-z\d+]'; // avoid automatic parsing of "word" in "last word.http://..." // generated with regex_idn.php file in the develop folder - return "[a-z]$scheme*:/{2}(?:(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@|]+|%[\dA-F]{2})+|[0-9.]+|\[[a-z0-9.]+:[a-z0-9.]+:[a-z0-9.:]+\])(?::\d*)?(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@/?|]+|%[\dA-F]{2})*)?"; + return "[a-z][a-z\d+]*:/{2}(?:(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@|]+|%[\dA-F]{2})+|[0-9.]+|\[[a-z0-9.]+:[a-z0-9.]+:[a-z0-9.:]+\])(?::\d*)?(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@/?|]+|%[\dA-F]{2})*)?"; break; case 'www_url': + // generated with regex_idn.php file in the develop folder + return "www\.(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})+(?::\d*)?(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?"; + break; + case 'www_url_inline': - $inline = ($mode == 'www_url') ? ')' : ''; - return "www\.(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@|]+|%[\dA-F]{2})+(?::\d*)?(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@/?|]+|%[\dA-F]{2})*)?"; + // generated with regex_idn.php file in the develop folder + return "www\.(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@|]+|%[\dA-F]{2})+(?::\d*)?(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@/?|]+|%[\dA-F]{2})*)?"; break; case 'relative_url': + // generated with regex_idn.php file in the develop folder + return "(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'()*+,;=:@/?|]+|%[\dA-F]{2})*)?"; + break; + case 'relative_url_inline': - $inline = ($mode == 'relative_url') ? ')' : ''; - return "(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@|]+|%[\dA-F]{2})*(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'($inline*+,;=:@/?|]+|%[\dA-F]{2})*)?"; + // generated with regex_idn.php file in the develop folder + return "(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@|]+|%[\dA-F]{2})*(?:/(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@|]+|%[\dA-F]{2})*)*(?:\?(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@/?|]+|%[\dA-F]{2})*)?(?:\#(?:[^\p{C}\p{Z}\p{S}\p{P}\p{Nl}\p{No}\p{Me}\x{1100}-\x{115F}\x{A960}-\x{A97C}\x{1160}-\x{11A7}\x{D7B0}-\x{D7C6}\x{20D0}-\x{20FF}\x{1D100}-\x{1D1FF}\x{1D200}-\x{1D24F}\x{0640}\x{07FA}\x{302E}\x{302F}\x{3031}-\x{3035}\x{303B}]*[\x{00B7}\x{0375}\x{05F3}\x{05F4}\x{30FB}\x{002D}\x{06FD}\x{06FE}\x{0F0B}\x{3007}\x{00DF}\x{03C2}\x{200C}\x{200D}\pL0-9\-._~!$&'(*+,;=:@/?|]+|%[\dA-F]{2})*)?"; break; case 'table_prefix': |