X-Git-Url: http://git.osdn.net/view?a=blobdiff_plain;f=spam.php;h=c59d798976097ac08ea805180e684dcba90acc9e;hb=25e665f54c6af5a91543c76c63784a8b95ffee44;hp=b3e7128c2995d9c1ef4d65e184074fac2284acc6;hpb=0fff3da81d47b28be2b88db8d7de3b6809a7bb22;p=pukiwiki%2Fpukiwiki_sandbox.git diff --git a/spam.php b/spam.php index b3e7128..c59d798 100644 --- a/spam.php +++ b/spam.php @@ -1,9 +1,44 @@ = 4.3.0): preg_match_all(PREG_OFFSET_CAPTURE): $method['uri_XXX'] related feature + +if (! defined('SPAM_INI_FILE')) define('SPAM_INI_FILE', 'spam.ini.php'); + +// --------------------- +// Compat etc + +// (PHP 4 >= 4.2.0): var_export(): mail-reporting and dump related +if (! function_exists('var_export')) { + function var_export() { + return 'var_export() is not found on this server' . "\n"; + } +} + +// (PHP 4 >= 4.2.0): preg_grep() enables invert option +function preg_grep_invert($pattern = '//', $input = array()) +{ + static $invert; + if (! isset($invert)) $invert = defined('PREG_GREP_INVERT'); + + if ($invert) { + return preg_grep($pattern, $input, PREG_GREP_INVERT); + } else { + $result = preg_grep($pattern, $input); + if ($result) { + return array_diff($input, preg_grep($pattern, $input)); + } else { + return $input; + } + } +} + +// --------------------- +// URI pickup // Return an array of URIs in the $string // [OK] http://nasty.example.org#nasty_string @@ -18,76 +53,258 @@ function uri_pickup($string = '', $normalize = TRUE, preg_match_all( // scheme://userinfo@host:port/path/or/pathinfo/maybefile.and?query=string#fragment // Refer RFC3986 (Regex below is not strict) - '#(\b[a-z][a-z0-9.+-]{1,8})://' . // 1: Scheme + '#(\b[a-z][a-z0-9.+-]{1,8}):/+' . // 1: Scheme '(?:' . '([^\s<>"\'\[\]/\#?@]*)' . // 2: Userinfo (Username) '@)?' . '(' . // 3: Host '\[[0-9a-f:.]+\]' . '|' . // IPv6([colon-hex and dot]): RFC2732 - '(?:[0-9]{1-3}\.){3}[0-9]{1-3}' . '|' . // IPv4(dot-decimal): 001.22.3.44 - '[^\s<>"\'\[\]:/\#?]+' . // FQDN: foo.example.org + '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' . // IPv4(dot-decimal): 001.22.3.44 + '[a-z0-9.-]+' . // hostname(FQDN) : foo.example.org ')' . '(?::([0-9]*))?' . // 4: Port '((?:/+[^\s<>"\'\[\]/\#]+)*/+)?' . // 5: Directory path or path-info - '([^\s<>"\'\[\]\#]+)?' . // 6: File and query string - '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 7: Fragment + '([^\s<>"\'\[\]\#?]+)?' . // 6: File? + '(?:\?([^\s<>"\'\[\]\#]+))?' . // 7: Query string + '(?:\#([a-z0-9._~%!$&\'()*+,;=:@-]*))?' . // 8: Fragment '#i', $string, $array, PREG_SET_ORDER | PREG_OFFSET_CAPTURE ); - //var_dump(recursive_map('htmlspecialchars', $array)); // Shrink $array static $parts = array( 1 => 'scheme', 2 => 'userinfo', 3 => 'host', 4 => 'port', - 5 => 'path', 6 => 'file', 7 => 'fragment' + 5 => 'path', 6 => 'file', 7 => 'query', 8 => 'fragment' ); $default = array(''); foreach(array_keys($array) as $uri) { - array_rename_keys($array[$uri], $parts, TRUE, $default); - $offset = $array[$uri]['scheme'][1]; // Scheme's offset + $_uri = & $array[$uri]; + array_rename_keys($_uri, $parts, TRUE, $default); - foreach(array_keys($array[$uri]) as $part) { + $offset = $_uri['scheme'][1]; // Scheme's offset + foreach(array_keys($_uri) as $part) { // Remove offsets for each part - $array[$uri][$part] = & $array[$uri][$part][0]; + $_uri[$part] = & $_uri[$part][0]; } if ($normalize) { - $array[$uri]['scheme'] = scheme_normalize($array[$uri]['scheme']); - //if ($array[$uri]['scheme'] === '') { - // // Ignore - // unset ($array[$uri]); - // continue; - //} - - $array[$uri]['host'] = strtolower($array[$uri]['host']); - $array[$uri]['port'] = port_normalize($array[$uri]['port'], $array[$uri]['scheme'], FALSE); - $array[$uri]['path'] = path_normalize($array[$uri]['path']); - //$array[$uri]['uri'] = uri_array_implode($array[$uri]); - if ($preserve_rawuri) $array[$uri]['rawuri'] = & $array[$uri][0]; + $_uri['scheme'] = scheme_normalize($_uri['scheme']); + if ($_uri['scheme'] === '') { + unset($array[$uri]); + continue; + } + $_uri['host'] = host_normalize($_uri['host']); + $_uri['port'] = port_normalize($_uri['port'], $_uri['scheme'], FALSE); + $_uri['path'] = path_normalize($_uri['path']); + if ($preserve_rawuri) $_uri['rawuri'] = & $_uri[0]; + + // DEBUG + //$_uri['uri'] = uri_array_implode($_uri); } else { - $array[$uri]['uri'] = & $array[$uri][0]; // Raw + $_uri['uri'] = & $_uri[0]; // Raw } - unset($array[$uri][0]); // Matched string itself + unset($_uri[0]); // Matched string itself if (! $preserve_chunk) { unset( - $array[$uri]['scheme'], - $array[$uri]['userinfo'], - $array[$uri]['host'], - $array[$uri]['port'], - $array[$uri]['path'], - $array[$uri]['file'], - $array[$uri]['fragment'] + $_uri['scheme'], + $_uri['userinfo'], + $_uri['host'], + $_uri['port'], + $_uri['path'], + $_uri['file'], + $_uri['query'], + $_uri['fragment'] ); } - $array[$uri]['offset'] = $offset; - $array[$uri]['area'] = 0; + // Area offset for area_measure() + $_uri['area']['offset'] = $offset; } return $array; } +// Destructive normalize of URI array +// NOTE: Give me the uri_pickup() result with chunks +function uri_array_normalize(& $pickups, $preserve = TRUE) +{ + if (! is_array($pickups)) return $pickups; + + foreach (array_keys($pickups) as $key) { + $_key = & $pickups[$key]; + $_key['path'] = isset($_key['path']) ? strtolower($_key['path']) : ''; + $_key['file'] = isset($_key['file']) ? file_normalize($_key['file']) : ''; + $_key['query'] = isset($_key['query']) ? query_normalize(strtolower($_key['query']), TRUE) : ''; + $_key['fragment'] = (isset($_key['fragment']) && $preserve) ? + strtolower($_key['fragment']) : ''; // Just ignore + } + + return $pickups; +} + +// An URI array => An URI (See uri_pickup()) +function uri_array_implode($uri = array()) +{ + if (empty($uri) || ! is_array($uri)) return NULL; + + $tmp = array(); + if (isset($uri['scheme']) && $uri['scheme'] !== '') { + $tmp[] = & $uri['scheme']; + $tmp[] = '://'; + } + if (isset($uri['userinfo']) && $uri['userinfo'] !== '') { + $tmp[] = & $uri['userinfo']; + $tmp[] = '@'; + } + if (isset($uri['host']) && $uri['host'] !== '') { + $tmp[] = & $uri['host']; + } + if (isset($uri['port']) && $uri['port'] !== '') { + $tmp[] = ':'; + $tmp[] = & $uri['port']; + } + if (isset($uri['path']) && $uri['path'] !== '') { + $tmp[] = & $uri['path']; + } + if (isset($uri['file']) && $uri['file'] !== '') { + $tmp[] = & $uri['file']; + } + if (isset($uri['query']) && $uri['query'] !== '') { + $tmp[] = '?'; + $tmp[] = & $uri['query']; + } + if (isset($uri['fragment']) && $uri['fragment'] !== '') { + $tmp[] = '#'; + $tmp[] = & $uri['fragment']; + } + + return implode('', $tmp); +} + +// $array['something'] => $array['wanted'] +function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '') +{ + if (! is_array($array) || ! is_array($keys)) return FALSE; + + // Nondestructive test + if (! $force) + foreach(array_keys($keys) as $from) + if (! isset($array[$from])) + return FALSE; + + foreach($keys as $from => $to) { + if ($from === $to) continue; + if (! $force || isset($array[$from])) { + $array[$to] = & $array[$from]; + unset($array[$from]); + } else { + $array[$to] = $default; + } + } + + return TRUE; +} + +// --------------------- +// Area pickup + +// Pickup all of markup areas +function area_pickup($string = '', $method = array()) +{ + $area = array(); + if (empty($method)) return $area; + + // Anchor tag pair by preg_match and preg_match_all() + // [OK] + // [OK] Good site! + // [OK] test + // [OK] visit http://nasty.example.com/ + // [OK] discount foobar + // [NG] visit http://ng.example.com _not_ended_ + $regex = '#]*\bhref\b[^>]*>.*?]*(>)#i'; + if (isset($method['area_anchor'])) { + $areas = array(); + $count = isset($method['asap']) ? + preg_match($regex, $string) : + preg_match_all($regex, $string, $areas); + if (! empty($count)) $area['area_anchor'] = $count; + } + if (isset($method['uri_anchor'])) { + $areas = array(); + preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE); + foreach(array_keys($areas) as $_area) { + $areas[$_area] = array( + $areas[$_area][0][1], // Area start () + $areas[$_area][1][1], // Area end () + ); + } + if (! empty($areas)) $area['uri_anchor'] = $areas; + } + + // phpBB's "BBCode" pair by preg_match and preg_match_all() + // [OK] [url][/url] + // [OK] [url]http://nasty.example.com/[/url] + // [OK] [link]http://nasty.example.com/[/link] + // [OK] [url=http://nasty.example.com]visit http://nasty.example.com/[/url] + // [OK] [link http://nasty.example.com/]buy something[/link] + $regex = '#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#i'; + if (isset($method['area_bbcode'])) { + $areas = array(); + $count = isset($method['asap']) ? + preg_match($regex, $string) : + preg_match_all($regex, $string, $areas, PREG_SET_ORDER); + if (! empty($count)) $area['area_bbcode'] = $count; + } + if (isset($method['uri_bbcode'])) { + $areas = array(); + preg_match_all($regex, $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE); + foreach(array_keys($areas) as $_area) { + $areas[$_area] = array( + $areas[$_area][0][1], // Area start ([url]) + $areas[$_area][2][1], // Area end ([/url]) + ); + } + if (! empty($areas)) $area['uri_bbcode'] = $areas; + } + + // Various Wiki syntax + // [text_or_uri>text_or_uri] + // [text_or_uri:text_or_uri] + // [text_or_uri|text_or_uri] + // [text_or_uri->text_or_uri] + // [text_or_uri text_or_uri] // MediaWiki + // MediaWiki: [http://nasty.example.com/ visit http://nasty.example.com/] + + return $area; +} + +// If in doubt, it's a little doubtful +// if (Area => inside <= Area) $brief += -1 +function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset') +{ + if (! is_array($areas) || ! is_array($array)) return; + + $areas_keys = array_keys($areas); + foreach(array_keys($array) as $u_index) { + $offset = isset($array[$u_index][$o_key]) ? + intval($array[$u_index][$o_key]) : 0; + foreach($areas_keys as $a_index) { + if (isset($array[$u_index][$a_key])) { + $offset_s = intval($areas[$a_index][0]); + $offset_e = intval($areas[$a_index][1]); + // [Area => inside <= Area] + if ($offset_s < $offset && $offset < $offset_e) { + $array[$u_index][$a_key] += $belief; + } + } + } + } +} + +// --------------------- +// Spam-uri pickup + // Domain exposure callback (See spam_uri_pickup_preprocess()) // http://victim.example.org/?foo+site:nasty.example.com+bar // => http://nasty.example.com/?refer=victim.example.org @@ -106,19 +323,23 @@ function _preg_replace_callback_domain_exposure($matches = array()) } // Flipped URI - $result = - $matches[1] . '://' . // scheme - $matches[4] . // nasty.example.com - '/?refer=' . strtolower($matches[2]) . // victim.example.org - ' ' . $result; + if (isset($matches[4])) { + $result = + $matches[1] . '://' . // scheme + $matches[4] . // nasty.example.com + '/?refer=' . strtolower($matches[2]) . // victim.example.org + ' ' . $result; + } return $result; } -// Preprocess: rawurldecode() and adding space(s) to detect/count some URIs _if possible_ +// Preprocess: rawurldecode() and adding space(s) and something +// to detect/count some URIs _if possible_ // NOTE: It's maybe danger to var_dump(result). [e.g. 'javascript:'] // [OK] http://victim.example.org/go?http%3A%2F%2Fnasty.example.org // [OK] http://victim.example.org/http://nasty.example.org +// TODO: link.toolbot.com, urlx.org function spam_uri_pickup_preprocess($string = '') { if (! is_string($string)) return ''; @@ -128,11 +349,19 @@ function spam_uri_pickup_preprocess($string = '') // Domain exposure (See _preg_replace_callback_domain_exposure()) $string = preg_replace_callback( array( - // Something Google: http://www.google.com/supported_domains - '#(http)://([a-z0-9.]+\.google\.[a-z]{2,3}(?:\.[a-z]{2})?)/' . - '([a-z0-9?=&.%_+-]+)' . // ?query=foo+ - '\bsite:([a-z0-9.%_-]+)' . // site:nasty.example.com - '()' . // Preserve? + '#(http)://' . + '(' . + // Something Google: http://www.google.com/supported_domains + '(?:[a-z0-9.]+\.)?google\.[a-z]{2,3}(?:\.[a-z]{2})?' . + '|' . + // AltaVista + '(?:[a-z0-9.]+\.)?altavista.com' . + + ')' . + '/' . + '([a-z0-9?=&.%_/\'\\\+-]+)' . // path/?query=foo+bar+ + '\bsite:([a-z0-9.%_-]+\.[a-z0-9.%_-]+)' . // site:nasty.example.com + //'()' . // Preserve or remove? '#i', ), '_preg_replace_callback_domain_exposure', @@ -142,7 +371,7 @@ function spam_uri_pickup_preprocess($string = '') // URI exposure (uriuri => uri uri) $string = preg_replace( array( - '#(?visit http://nasty.example.com/ - // [OK] discount foobar - // [NG] visit http://ng.example.com _not_ended_ - // [NG] Good site! test - $areas = array(); - preg_match_all('#]*href[^>]*>.*?]*(>)#i', - $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE); - //var_dump(recursive_map('htmlspecialchars', $areas)); - foreach(array_keys($areas) as $area) { - $areas[$area] = array( - $areas[$area][0][1], // Area start () - $areas[$area][1][1], // Area end () - ); - } - area_measure($areas, $array); - - // phpBB's "BBCode" by preg_match_all() - // [url]http://nasty.example.com/[/url] - // [link]http://nasty.example.com/[/link] - // [url=http://nasty.example.com]visit http://nasty.example.com/[/url] - // [link http://nasty.example.com/]buy something[/link] - // ?? [url=][/url] - $areas = array(); - preg_match_all('#\[(url|link)\b[^\]]*\].*?\[/\1\b[^\]]*(\])#i', - $string, $areas, PREG_SET_ORDER | PREG_OFFSET_CAPTURE); - //var_dump(recursive_map('htmlspecialchars', $areas)); - foreach(array_keys($areas) as $area) { - $areas[$area] = array( - $areas[$area][0][1], // Area start ([url]) - $areas[$area][2][1], // Area end ([/url]) - ); - } - area_measure($areas, $array); - - // Various Wiki syntax - // [text_or_uri>text_or_uri] - // [text_or_uri:text_or_uri] - // [text_or_uri|text_or_uri] - // [text_or_uri->text_or_uri] - // [text_or_uri text_or_uri] // MediaWiki - // MediaWiki: [http://nasty.example.com/ visit http://nasty.example.com/] - - // Remove 'offset's for area_measure() - //foreach(array_keys($array) as $key) - // unset($array[$key]['offset']); - } - - return $array; -} - -// $array['something'] => $array['wanted'] -function array_rename_keys(& $array, $keys = array('from' => 'to'), $force = FALSE, $default = '') -{ - if (! is_array($array) || ! is_array($keys)) - return FALSE; - - // Nondestructive test - if (! $force) - foreach(array_keys($keys) as $from) - if (! isset($array[$from])) - return FALSE; - - foreach($keys as $from => $to) { - if ($from === $to) continue; - if (! $force || isset($array[$from])) { - $array[$to] = & $array[$from]; - unset($array[$from]); - } else { - $array[$to] = $default; - } - } - - return TRUE; -} - -// If in doubt, it's a little doubtful -function area_measure($areas, & $array, $belief = -1, $a_key = 'area', $o_key = 'offset') -{ - if (! is_array($areas) || ! is_array($array)) return; - - $areas_keys = array_keys($areas); - foreach(array_keys($array) as $u_index) { - $offset = isset($array[$u_index][$o_key]) ? - intval($array[$u_index][$o_key]) : 0; - foreach($areas_keys as $a_index) { - if (isset($array[$u_index][$a_key])) { - $offset_s = intval($areas[$a_index][0]); - $offset_e = intval($areas[$a_index][1]); - // [Area => inside <= Area] - if ($offset_s < $offset && $offset < $offset_e) { - $array[$u_index][$a_key] += $belief; + $_method = array(); + if (isset($method['uri_anchor'])) $_method['uri_anchor'] = & $method['uri_anchor']; + if (isset($method['uri_bbcode'])) $_method['uri_bbcode'] = & $method['uri_bbcode']; + $areas = area_pickup($string, $_method, TRUE); + if (! empty($areas)) { + $area_shadow = array(); + foreach (array_keys($array) as $key) { + $area_shadow[$key] = & $array[$key]['area']; + foreach (array_keys($_method) as $_key) { + $area_shadow[$key][$_key] = 0; + } + } + foreach (array_keys($_method) as $_key) { + if (isset($areas[$_key])) { + area_measure($areas[$_key], $area_shadow, 1, $_key); } } } } + + // Remove 'offset's for area_measure() + foreach(array_keys($array) as $key) + unset($array[$key]['area']['offset']); + + return $array; } // --------------------- -// Part Two +// Normalization // Scheme normalization: Renaming the schemes // snntp://example.org => nntps://example.org @@ -300,6 +461,25 @@ function scheme_normalize($scheme = '', $considerd_harmfull = TRUE) return $scheme; } +// Hostname normlization +// www.foo => www.foo ('foo' seems TLD) +// www.foo.bar => foo.bar +// www.10.20 => www.10.20 (Invalid hostname) +// NOTE: +// 'www' is mostly used as traditional hostname of WWW server. +// 'www.foo.bar' may be identical with 'foo.bar'. +function host_normalize($host = '') +{ + $host = strtolower($host); + + $matches = array(); + if (preg_match('/^www\.(.+\.[a-z]+)$/', $host, $matches)) { + return $matches[1]; + } else { + return $host; + } +} + // Port normalization: Suppress the (redundant) default port // HTTP://example.org:80/ => http://example.org/ // HTTP://example.org:8080/ => http://example.org:8080/ @@ -353,99 +533,140 @@ function port_normalize($port, $scheme, $scheme_normalize = TRUE) // http://example.org/path/../../a/../back => http://example.org/back function path_normalize($path = '', $divider = '/', $addroot = TRUE) { - if (! is_string($path) || $path == '') { - $path = $addroot ? $divider : ''; - } else { - $path = trim($path); - $last = ($path[strlen($path) - 1] == $divider) ? $divider : ''; - $array = explode($divider, $path); - - // Remove paddings - foreach(array_keys($array) as $key) { - if ($array[$key] == '' || $array[$key] == '.') - unset($array[$key]); - } - // Back-track - $tmp = array(); - foreach($array as $value) { - if ($value == '..') { - array_pop($tmp); - } else { - array_push($tmp, $value); - } - } - $array = & $tmp; + if (! is_string($path) || $path == '') + return $addroot ? $divider : ''; + + $path = trim($path); + $last = ($path[strlen($path) - 1] == $divider) ? $divider : ''; + $array = explode($divider, $path); - $path = $addroot ? $divider : ''; - if (! empty($array)) $path .= implode($divider, $array) . $last; + // Remove paddings + foreach(array_keys($array) as $key) { + if ($array[$key] == '' || $array[$key] == '.') + unset($array[$key]); + } + // Back-track + $tmp = array(); + foreach($array as $value) { + if ($value == '..') { + array_pop($tmp); + } else { + array_push($tmp, $value); + } } + $array = & $tmp; + + $path = $addroot ? $divider : ''; + if (! empty($array)) $path .= implode($divider, $array) . $last; return $path; } -// An URI array => An URI (See uri_pickup()) -function uri_array_implode($uri = array()) +// DirectoryIndex normalize (Destructive and rough) +function file_normalize($string = 'index.html.en') { - if (empty($uri) || ! is_array($uri)) return NULL; - - $tmp = array(); - if (isset($uri['scheme']) && $uri['scheme'] !== '') { - $tmp[] = & $uri['scheme']; - $tmp[] = '://'; - } - if (isset($uri['userinfo']) && $uri['userinfo'] !== '') { - $tmp[] = & $uri['userinfo']; - $tmp[] = '@'; - } - if (isset($uri['host']) && $uri['host'] !== '') { - $tmp[] = & $uri['host']; - } - if (isset($uri['port']) && $uri['port'] !== '') { - $tmp[] = ':'; - $tmp[] = & $uri['port']; + static $array = array( + 'index' => TRUE, // Some system can omit the suffix + 'index.htm' => TRUE, + 'index.html' => TRUE, + 'index.shtml' => TRUE, + 'index.jsp' => TRUE, + 'index.php' => TRUE, + 'index.php3' => TRUE, + 'index.php4' => TRUE, + //'index.pl' => TRUE, + //'index.py' => TRUE, + //'index.rb' => TRUE, + 'index.cgi' => TRUE, + 'default.htm' => TRUE, + 'default.html' => TRUE, + 'default.asp' => TRUE, + 'default.aspx' => TRUE, + ); + + // Content-negothiation filter: + // Roughly removing ISO 639 -like + // 2-letter suffixes (See RFC3066) + $matches = array(); + if (preg_match('/(.*)\.[a-z][a-z](?:-[a-z][a-z])?$/i', $string, $matches)) { + $_string = $matches[1]; + } else { + $_string = & $string; } - if (isset($uri['path']) && $uri['path'] !== '') { - $tmp[] = & $uri['path']; + + if (isset($array[strtolower($_string)])) { + return ''; + } else { + return $string; } - if (isset($uri['file']) && $uri['file'] !== '') { - $tmp[] = & $uri['file']; +} + +// Sort query-strings if possible (Destructive and rough) +// [OK] &&&&f=d&b&d&c&a=0dd => a=0dd&b&c&d&f=d +// [OK] nothing==&eg=dummy&eg=padding&eg=foobar => eg=foobar +function query_normalize($string = '', $equal = FALSE, $equal_cutempty = TRUE) +{ + $array = explode('&', $string); + + // Remove '&' paddings + foreach(array_keys($array) as $key) { + if ($array[$key] == '') { + unset($array[$key]); + } } - if (isset($uri['fragment']) && $uri['fragment'] !== '') { - $tmp[] = '#'; - $tmp[] = & $uri['fragment']; + + // Consider '='-sepalated input and paddings + if ($equal) { + $equals = $not_equals = array(); + foreach ($array as $part) { + if (strpos($part, '=') === FALSE) { + $not_equals[] = $part; + } else { + list($key, $value) = explode('=', $part, 2); + $value = ltrim($value, '='); + if (! $equal_cutempty || $value != '') { + $equals[$key] = $value; + } + } + } + + $array = & $not_equals; + foreach ($equals as $key => $value) { + $array[] = $key . '=' . $value; + } + unset($equals); } - return implode('', $tmp); + natsort($array); + return implode('&', $array); } // --------------------- // Part One : Checker +// Rough implementation of globbing +// +// USAGE: $regex = '/^' . generate_glob_regex('*.txt', '/') . '$/i'; +// function generate_glob_regex($string = '', $divider = '/') { static $from = array( - 0 => '*', - 1 => '?', - 2 => '\[', - 3 => '\]', - 4 => '[', - 5 => ']', + 1 => '*', + 11 => '?', + // 22 => '[', // Maybe cause regex compilation error (e.g. '[]') + // 23 => ']', // ); static $mid = array( - 0 => '_AST_', - 1 => '_QUE_', - 2 => '_eRBR_', - 3 => '_eLBR_', - 4 => '_RBR_', - 5 => '_LBR_', + 1 => '_AST_', + 11 => '_QUE_', + // 22 => '_RBR_', + // 23 => '_LBR_', ); static $to = array( - 0 => '.*', - 1 => '.', - 2 => '\[', - 3 => '\]', - 4 => '[', - 5 => ']', + 1 => '.*', + 11 => '.', + // 22 => '[', + // 23 => ']', ); $string = str_replace($from, $mid, $string); // Hide @@ -455,204 +676,512 @@ function generate_glob_regex($string = '', $divider = '/') return $string; } -// TODO: Ignore list -// TODO: require_or_include_once(another file) -function is_badhost($hosts = '', $asap = TRUE) +// Rough hostname checker +// [OK] 192.168. +// TODO: Strict digit, 0x, CIDR, IPv6 +function is_ip($string = '') { - static $blocklist_regex; - - if (! isset($blocklist_regex)) { - $blocklist_regex = array(); - $blocklist = array( - // Deny all uri - //'*', - - // IP address or ... - //'10.20.*.*', // 10.20.example.com also matches - //'\[1\]', - - // Too much malicious sub-domains - '*.blogspot.com', - - // 2006-11 dev - 'wwwtahoo.com', - - // 2006-11 dev - '*.infogami.com', - - // 2006/11/19 17:50 dev - '*.google0site.org', - '*.bigpricesearch.org', - '*.osfind.org', - '*.bablomira.biz', - ); - foreach ($blocklist as $part) { - $blocklist_regex[] = '#^' . generate_glob_regex($part, '#') . '$#i'; + if (preg_match('/^' . + '(?:[0-9]{1,3}\.){3}[0-9]{1,3}' . '|' . + '(?:[0-9]{1,3}\.){1,3}' . '$/', + $string)) { + return 4; // Seems IPv4(dot-decimal) + } else { + return 0; // Seems not IP + } +} + +// Generate host (FQDN, IPv4, ...) regex +// 'localhost' : Matches with 'localhost' only +// 'example.org' : Matches with 'example.org', and 'www.example.org' +// '.example.org' : Matches with ALL FQDN ended with '.example.org' +// '*.example.org' : Almost the same of '.example.org' except 'www.example.org' +// '10.20.30.40' : Matches with IPv4 address '10.20.30.40' only +// '192168.' : Matches with all IPv4 hosts started with '192.' +// TODO: IPv4, CIDR?, IPv6 +function generate_host_regex($string = '', $divider = '/') +{ + if (mb_strpos($string, '.') === FALSE) + return generate_glob_regex($string, $divider); + + $result = ''; + if (is_ip($string)) { + // IPv4 + return generate_glob_regex($string, $divider); + } else { + // FQDN or something + $part = explode('.', $string, 2); + if ($part[0] == '') { + $part[0] = '(?:.*\.)?'; // And all related FQDN + } else if ($part[0] == '*') { + $part[0] = '.*\.'; // All subdomains/hosts only + } else { + return generate_glob_regex($string, $divider); } + $part[1] = generate_glob_regex($part[1], $divider); + return implode('', $part); } +} - $result = 0; +function get_blocklist($list = '') +{ + static $regexs; + + if (! isset($regexs)) { + $regexs = array(); + if (file_exists(SPAM_INI_FILE)) { + $blocklist = array(); + include(SPAM_INI_FILE); + // $blocklist['badhost'] = array( + // '*.blogspot.com', // Blog services's subdomains (only) + // 'IANA-examples' => '#^(?:.*\.)?example\.(?:com|net|org)$#', + // ); + foreach(array('goodhost', 'badhost') as $_list) { + if (! isset($blocklist[$list])) continue; + foreach ($blocklist[$_list] as $key => $value) { + if (is_array($value)) { + $regexs[$_list][$key] = array(); + foreach($value as $_key => $_value) { + if (is_string($_key)) { + $regexs[$_list][$key][$_key] = $_value; // A regex + } else { + $regexs[$_list][$key][$_value] = + '/^' . generate_host_regex($_value, '/') . '$/i'; + } + } + } else { + if (is_string($key)) { + $regexs[$_list][$key] = $value; // A regex + } else { + $regexs[$_list][$value] = + '/^' . generate_host_regex($value, '/') . '$/i'; + } + } + } + } + } + } + + if ($list == '') { + return $regexs; // ALL + } else if (isset($regexs[$list])) { + return $regexs[$list]; + } else { + return array(); + } +} + +function is_badhost($hosts = array(), $asap = TRUE, & $remains) +{ + $result = array(); if (! is_array($hosts)) $hosts = array($hosts); - foreach($hosts as $host) { - if (! is_string($host)) $host = ''; - foreach ($blocklist_regex as $regex) { - if (preg_match($regex, $host)) { - ++$result; - if ($asap) { - return $result; - } else { - break; // Check next host + foreach(array_keys($hosts) as $key) { + if (! is_string($hosts[$key])) unset($hosts[$key]); + } + if (empty($hosts)) return $result; + + foreach (get_blocklist('goodhost') as $regex) { + $hosts = preg_grep_invert($regex, $hosts); + } + if (empty($hosts)) return $result; + + $tmp = array(); + foreach (get_blocklist('badhost') as $label => $regex) { + if (is_array($regex)) { + $result[$label] = array(); + foreach($regex as $_label => $_regex) { + $_group = preg_grep($_regex, $hosts); + if ($_group) { + $result[$label][$_label] = $_group; + $hosts = array_diff($hosts, $_group); + if ($asap) break; } } + if (empty($result[$label])) unset($result[$label]); + } else { + $_group = preg_grep($regex, $hosts); + if ($_group) { + $result[$label] = $_group; + $hosts = array_diff($hosts, $result[$label]); + if ($asap) break; + } } } + $remains = $hosts; + return $result; } -// TODO return TRUE or FALSE! -// Simple/fast spam check -function check_uri_spam($target = '', $method = array(), $asap = TRUE) +// Default (enabled) methods and thresholds (for content insertion) +function check_uri_spam_method($times = 1, $t_area = 0, $rule = TRUE) { - $is_spam = FALSE; - $progress = array( - 'quantity' => 0, - 'area' => 0, - 'non_uniq' => 0, - 'badhost' => 0, + $times = intval($times); + $t_area = intval($t_area); + + $positive = array( + // Thresholds + 'quantity' => 8 * $times, // Allow N URIs + 'non_uniqhost' => 3 * $times, // Allow N duped (and normalized) Hosts + //'non_uniquri'=> 3 * $times, // Allow N duped (and normalized) URIs + + // Areas + 'area_anchor' => $t_area, // Using HTML tag + 'area_bbcode' => $t_area, // Using [url] or [link] BBCode + //'uri_anchor' => $t_area, // URI inside HTML tag + //'uri_bbcode' => $t_area, // URI inside [url] or [link] BBCode + ); + if ($rule) { + $bool = array( + // Rules + //'asap' => TRUE, // Quit or return As Soon As Possible + 'uniqhost' => TRUE, // Show uniq host (at block notification mail) + 'badhost' => TRUE, // Check badhost ); + } else { + $bool = array(); + } + // Remove non-$positive values + foreach (array_keys($positive) as $key) { + if ($positive[$key] < 0) unset($positive[$key]); + } + + return $positive + $bool; +} + +// Simple/fast spam check +function check_uri_spam($target = '', $method = array()) +{ if (! is_array($method) || empty($method)) { - // Default - $method = array( - 'quantity' => 8, // Allow N URIs - 'area' => TRUE, - 'non_uniq' => 3, // Allow N times dupe - 'badhost' => TRUE, - ); + $method = check_uri_spam_method(); } + $progress = array( + 'sum' => array( + 'quantity' => 0, + 'uniqhost' => 0, + 'non_uniqhost'=> 0, + 'non_uniquri' => 0, + 'badhost' => 0, + 'area_anchor' => 0, + 'area_bbcode' => 0, + 'uri_anchor' => 0, + 'uri_bbcode' => 0, + ), + 'is_spam' => array(), + 'method' => & $method, + 'remains' => array(), + 'error' => array(), + ); + $sum = & $progress['sum']; + $is_spam = & $progress['is_spam']; + $remains = & $progress['remains']; + $error = & $progress['error']; + $asap = isset($method['asap']); + // Recurse if (is_array($target)) { foreach($target as $str) { // Recurse - list($is_spam, $_progress) = check_uri_spam($str, $method); - $progress['quantity'] += $_progress['quantity']; - $progress['non_uniq'] += $_progress['non_uniq']; - if ($asap || $is_spam) break; + $_progress = check_uri_spam($str, $method); + $_sum = & $_progress['sum']; + $_is_spam = & $_progress['is_spam']; + $_remains = & $_progress['remains']; + $_error = & $_progress['error']; + foreach (array_keys($_sum) as $key) { + $sum[$key] += $_sum[$key]; + } + foreach (array_keys($_is_spam) as $key) { + if (is_array($_is_spam[$key])) { + // Marge keys (badhost) + foreach(array_keys($_is_spam[$key]) as $_key) { + if (! isset($is_spam[$key][$_key])) { + $is_spam[$key][$_key] = $_is_spam[$key][$_key]; + } else { + $is_spam[$key][$_key] += $_is_spam[$key][$_key]; + } + } + } else { + $is_spam[$key] = TRUE; + } + } + foreach ($_remains as $key=>$value) { + foreach ($value as $_key=>$_value) { + if (is_int($_key)) { + $remains[$key][] = $_value; + } else { + $remains[$key][$_key] = $_value; + } + } + } + if (! empty($_error)) $error += $_error; + if ($asap && $is_spam) break; } - } else { - $pickups = spam_uri_pickup($target); - $progress['quantity'] += count($pickups); + return $progress; + } - if (! empty($pickups)) { + // Area: There's HTML anchor tag + if ((! $asap || ! $is_spam) && isset($method['area_anchor'])) { + $key = 'area_anchor'; + $_asap = isset($method['asap']) ? array('asap' => TRUE) : array(); + $result = area_pickup($target, array($key => TRUE) + $_asap); + if ($result) { + $sum[$key] = $result[$key]; + if (isset($method[$key]) && $sum[$key] > $method[$key]) { + $is_spam[$key] = TRUE; + } + } + } - // URI quantity - if ((! $is_spam || ! $asap) && isset($method['quantity']) && - $progress['quantity'] > $method['quantity']) { - $is_spam = TRUE; + // Area: There's 'BBCode' linking tag + if ((! $asap || ! $is_spam) && isset($method['area_bbcode'])) { + $key = 'area_bbcode'; + $_asap = isset($method['asap']) ? array('asap' => TRUE) : array(); + $result = area_pickup($target, array($key => TRUE) + $_asap); + if ($result) { + $sum[$key] = $result[$key]; + if (isset($method[$key]) && $sum[$key] > $method[$key]) { + $is_spam[$key] = TRUE; } - //var_dump($method['quantity'], $is_spam); - - // Using invalid area - if ((! $is_spam || ! $asap) && isset($method['area'])) { - foreach($pickups as $pickup) { - if ($pickup['area'] < 0) { - ++$progress['area']; - $is_spam = TRUE; - if ($asap) break; - } + } + } + + // Return if ... + if ($asap && $is_spam) return $progress; + + // URI: Pickup + $pickups = spam_uri_pickup($target, $method); + //$remains['uri_pickup'] = & $pickups; + + // Return if ... + if (empty($pickups)) return $progress; + + // URI: Check quantity + $sum['quantity'] += count($pickups); + // URI quantity + if ((! $asap || ! $is_spam) && isset($method['quantity']) && + $sum['quantity'] > $method['quantity']) { + $is_spam['quantity'] = TRUE; + } + + // URI: used inside HTML anchor tag pair + if ((! $asap || ! $is_spam) && isset($method['uri_anchor'])) { + $key = 'uri_anchor'; + foreach($pickups as $pickup) { + if (isset($pickup['area'][$key])) { + $sum[$key] += $pickup['area'][$key]; + if(isset($method[$key]) && + $sum[$key] > $method[$key]) { + $is_spam[$key] = TRUE; + if ($asap && $is_spam) break; } + if ($asap && $is_spam) break; } - //var_dump($method['area'], $is_spam); + } + } - // URI uniqueness (and removing non-uniques) - if ((! $is_spam || ! $asap) && isset($method['non_uniq'])) { - $uris = array(); - foreach ($pickups as $key => $pickup) { - $uris[$key] = uri_array_implode($pickup); - } - $count = count($uris); - $uris = array_unique($uris); - $progress['non_uniq'] += $count - count($uris); - if ($progress['non_uniq'] > $method['non_uniq']) { - $is_spam = TRUE; + // URI: used inside 'BBCode' pair + if ((! $asap || ! $is_spam) && isset($method['uri_bbcode'])) { + $key = 'uri_bbcode'; + foreach($pickups as $pickup) { + if (isset($pickup['area'][$key])) { + $sum[$key] += $pickup['area'][$key]; + if(isset($method[$key]) && + $sum[$key] > $method[$key]) { + $is_spam[$key] = TRUE; + if ($asap && $is_spam) break; } - if (! $asap || ! $is_spam) { - foreach (array_diff(array_keys($pickups), - array_keys($uris)) as $remove) { - unset($pickups[$remove]); - } + if ($asap && $is_spam) break; + } + } + } + + // URI: Uniqueness (and removing non-uniques) + if ((! $asap || ! $is_spam) && isset($method['non_uniquri'])) { + + // Destructive normalize of URIs + uri_array_normalize($pickups); + + $uris = array(); + foreach (array_keys($pickups) as $key) { + $uris[$key] = uri_array_implode($pickups[$key]); + } + $count = count($uris); + $uris = array_unique($uris); + $sum['non_uniquri'] += $count - count($uris); + if ($sum['non_uniquri'] > $method['non_uniquri']) { + $is_spam['non_uniquri'] = TRUE; + } + if (! $asap || ! $is_spam) { + foreach (array_diff(array_keys($pickups), + array_keys($uris)) as $remove) { + unset($pickups[$remove]); + } + } + unset($uris); + } + + // Return if ... + if ($asap && $is_spam) return $progress; + + // Host: Uniqueness (uniq / non-uniq) + $hosts = array(); + foreach ($pickups as $pickup) $hosts[] = & $pickup['host']; + $hosts = array_unique($hosts); + //$remains['uniqhost'] = & $hosts; + $sum['uniqhost'] += count($hosts); + if ((! $asap || ! $is_spam) && isset($method['non_uniqhost'])) { + $sum['non_uniqhost'] = $sum['quantity'] - $sum['uniqhost']; + if ($sum['non_uniqhost'] > $method['non_uniqhost']) { + $is_spam['non_uniqhost'] = TRUE; + } + } + + // Return if ... + if ($asap && $is_spam) return $progress; + + // URI: Bad host + if ((! $asap || ! $is_spam) && isset($method['badhost'])) { + $__remains = array(); + $badhost = is_badhost($hosts, $asap, $__remains); + if (! $asap) { + if ($__remains) { + $remains['badhost'] = array(); + foreach ($__remains as $value) { + $remains['badhost'][$value] = TRUE; } - unset($uris); - //var_dump($uris, $pickups); } - //var_dump($method['non_uniq'], $is_spam); + } + unset($__remains); + if (! empty($badhost)) { + $sum['badhost'] += array_count_leaves($badhost); + foreach(array_keys($badhost) as $keys) { + $is_spam['badhost'][$keys] = + array_count_leaves($badhost[$keys]); + } + unset($badhost); + } + } + + return $progress; +} + +// Count leaves +function array_count_leaves($array = array(), $count_empty_array = FALSE) +{ + if (! is_array($array) || (empty($array) && $count_empty_array)) + return 1; + + // Recurse + $result = 0; + foreach ($array as $part) { + $result += array_count_leaves($part, $count_empty_array); + } + return $result; +} - // Bad host - if ((! $is_spam || ! $asap) && isset($method['badhost'])) { - $hosts = array(); - foreach ($pickups as $pickup) { - $hosts[] = & $pickup['host']; +// --------------------- +// Reporting + +// TODO: Don't show unused $method! +// Summarize $progress (blocked only) +function summarize_spam_progress($progress = array(), $blockedonly = FALSE) +{ + if ($blockedonly) { + $tmp = array_keys($progress['is_spam']); + } else { + $tmp = array(); + $method = & $progress['method']; + if (isset($progress['sum'])) { + foreach ($progress['sum'] as $key => $value) { + if (isset($method[$key])) { + $tmp[] = $key . '(' . $value . ')'; } - $count = is_badhost(array_unique($hosts), $asap); - $progress['badhost'] += $count; - if ($count !== 0) $is_spam = TRUE; } - //var_dump($method['badhost'], $is_spam); } } - return array($is_spam, $progress); + return implode(', ', $tmp); } // --------------------- +// Exit -// Check User-Agent (not testing yet) -function is_invalid_useragent($ua_name = '' /*, $ua_vars = ''*/ ) +// Common bahavior for blocking +// NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked' +function spam_exit($mode = '', $data = array()) { - return $ua_name === ''; + switch ($mode) { + case '': echo("\n"); break; + case 'dump': + echo('
' . "\n");
+			echo htmlspecialchars(var_export($data, TRUE));
+			echo('
' . "\n"); + break; + }; + + // Force exit + exit; } + // --------------------- +// Simple filtering -// TODO: Separate check-part(s) and mail part -// TODO: Mail to administrator with more measurement data? +// TODO: Record them // Simple/fast spam filter ($target: 'a string' or an array()) -function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array()) +function pkwk_spamfilter($action, $page, $target = array('title' => ''), $method = array(), $exitmode = '') { - $is_spam = FALSE; - - //$is_spam = is_invalid_useragent('NOTYET'); - if ($is_spam) { - $action .= ' (Invalid User-Agent)'; - } else { - list($is_spam) = check_uri_spam($target, $method); - } + $progress = check_uri_spam($target, $method); - if ($is_spam) { + if (! empty($progress['is_spam'])) { // Mail to administrator(s) - global $notify, $notify_subject; - if ($notify) { - $footer['ACTION'] = $action; - $footer['PAGE'] = '[blocked] ' . $page; - $footer['URI'] = get_script_uri() . '?' . rawurlencode($page); - $footer['USER_AGENT'] = TRUE; - $footer['REMOTE_ADDR'] = TRUE; - pkwk_mail_notify($notify_subject, var_export($target, TRUE), $footer); - unset($footer); - } - } + pkwk_spamnotify($action, $page, $target, $progress, $method); - if ($is_spam) spam_exit(); + // Exit + spam_exit($exitmode, $progress); + } } // --------------------- +// PukiWiki original -// Common bahavior for blocking -// NOTE: Call this function from various blocking feature, to disgueise the reason 'why blocked' -function spam_exit() +// Mail to administrator(s) +function pkwk_spamnotify($action, $page, $target = array('title' => ''), $progress = array(), $method = array()) { - die("\n"); + global $notify, $notify_subject; + + if (! $notify) return; + + $asap = isset($method['asap']); + + $summary['ACTION'] = 'Blocked by: ' . summarize_spam_progress($progress, TRUE); + if (! $asap) { + $summary['METRICS'] = summarize_spam_progress($progress); + } + if (isset($progress['is_spam']['badhost'])) { + $badhost = array(); + foreach($progress['is_spam']['badhost'] as $glob=>$number) { + $badhost[] = $glob . '(' . $number . ')'; + } + $summary['DETAIL_BADHOST'] = implode(', ', $badhost); + } + if (! $asap && $progress['remains']['badhost']) { + $count = count($progress['remains']['badhost']); + $summary['DETAIL_NEUTRAL_HOST'] = $count . + ' (' . + preg_replace( + '/[^, a-z0-9.-]/i', '', + implode(', ', array_keys($progress['remains']['badhost'])) + ) . + ')'; + } + $summary['COMMENT'] = $action; + $summary['PAGE'] = '[blocked] ' . (is_pagename($page) ? $page : ''); + $summary['URI'] = get_script_uri() . '?' . rawurlencode($page); + $summary['USER_AGENT'] = TRUE; + $summary['REMOTE_ADDR'] = TRUE; + pkwk_mail_notify($notify_subject, var_export($target, TRUE), $summary, TRUE); } ?>