@@ -1,181 +1,181 @@ |
||
1 | 1 | <?php |
2 | 2 | class WhitelistGenerator extends Object implements Flushable { |
3 | 3 | |
4 | - public static function generateWhitelist(){ |
|
5 | - $whitelist = self::generateWhitelistRules(); |
|
6 | - self::syncCacheFilesystem($whitelist); |
|
7 | - } |
|
8 | - |
|
9 | - public static function generateWhitelistRules(){ |
|
10 | - //get all URL rules |
|
11 | - $rules = Config::inst()->get('Director', 'rules'); |
|
12 | - |
|
13 | - $allTopLevelRules = array(); |
|
14 | - foreach ($rules as $pattern => $controllerOptions) { |
|
15 | - //allow for route rules starting with double-slash (//) |
|
16 | - $pattern = ltrim($pattern, '/!'); |
|
17 | - |
|
18 | - //match first portion of the URL, either delimited by slash or colon or end-of-line |
|
19 | - if (preg_match('/^(.*?)(\/|:|$)/', $pattern, $matches)){ |
|
20 | - if (!empty($matches[1])){ |
|
21 | - array_push($allTopLevelRules, $matches[1]); |
|
22 | - } |
|
23 | - } |
|
24 | - } |
|
25 | - $filteredRules = array('home'); //add 'home' url, as default |
|
4 | + public static function generateWhitelist(){ |
|
5 | + $whitelist = self::generateWhitelistRules(); |
|
6 | + self::syncCacheFilesystem($whitelist); |
|
7 | + } |
|
8 | + |
|
9 | + public static function generateWhitelistRules(){ |
|
10 | + //get all URL rules |
|
11 | + $rules = Config::inst()->get('Director', 'rules'); |
|
12 | + |
|
13 | + $allTopLevelRules = array(); |
|
14 | + foreach ($rules as $pattern => $controllerOptions) { |
|
15 | + //allow for route rules starting with double-slash (//) |
|
16 | + $pattern = ltrim($pattern, '/!'); |
|
17 | + |
|
18 | + //match first portion of the URL, either delimited by slash or colon or end-of-line |
|
19 | + if (preg_match('/^(.*?)(\/|:|$)/', $pattern, $matches)){ |
|
20 | + if (!empty($matches[1])){ |
|
21 | + array_push($allTopLevelRules, $matches[1]); |
|
22 | + } |
|
23 | + } |
|
24 | + } |
|
25 | + $filteredRules = array('home'); //add 'home' url, as default |
|
26 | 26 | |
27 | - $addToWhitelist = Config::inst()->get('WhitelistGenerator', 'addToWhitelist'); |
|
28 | - if ($addToWhitelist && is_array($addToWhitelist)) { |
|
29 | - $filteredRules = array_merge($filteredRules, $addToWhitelist); |
|
30 | - } |
|
27 | + $addToWhitelist = Config::inst()->get('WhitelistGenerator', 'addToWhitelist'); |
|
28 | + if ($addToWhitelist && is_array($addToWhitelist)) { |
|
29 | + $filteredRules = array_merge($filteredRules, $addToWhitelist); |
|
30 | + } |
|
31 | 31 | |
32 | - foreach($allTopLevelRules as $rule) { |
|
33 | - if (strpos($rule, '$') !== false) { |
|
34 | - if ($rule === '$Controller') { |
|
35 | - //special case for Controllers, add all possible controllers |
|
36 | - $subControllers = ClassInfo::subclassesFor(new Controller()); |
|
37 | - |
|
38 | - foreach ($subControllers as $controller){ |
|
39 | - array_push($filteredRules, $controller); //add the controller name as a link |
|
40 | - } |
|
41 | - |
|
42 | - } elseif ($rule === '$URLSegment') { |
|
43 | - $topLevelPagesArray = array(); //temporary array to store top-level pages |
|
44 | - |
|
45 | - //special case for SiteTree, add all possible top Level Pages |
|
46 | - $topLevelPages = SiteTree::get()->filter('ParentID', 0); |
|
47 | - |
|
48 | - foreach ($topLevelPages as $page) { |
|
49 | - $link = $page->RelativeLink(); |
|
50 | - array_push($topLevelPagesArray, trim($link, '\/ ')); //remove trailing or leading slashes from links |
|
51 | - } |
|
52 | - |
|
53 | - //fetch old top-level pages URLs from the SiteTree_versions table |
|
54 | - if (Config::inst()->get('WhitelistGenerator', 'includeSiteTreeVersions')) { |
|
55 | - $oldTopLevelPagesArray = self::find_old_top_level_pages($topLevelPagesArray); |
|
56 | - } |
|
57 | - |
|
58 | - $filteredRules = array_merge($filteredRules, $topLevelPagesArray, $oldTopLevelPagesArray); |
|
59 | - } else { |
|
60 | - user_error('Unknown wildcard URL match found: '.$rule, E_WARNING); |
|
61 | - } |
|
62 | - } else { |
|
63 | - //add the rule to a new list of rules |
|
64 | - array_push($filteredRules, $rule); |
|
65 | - } |
|
66 | - |
|
67 | - } |
|
68 | - |
|
69 | - //filter duplicates (order doesn't matter here, as we are only interested in the first level of the rules) |
|
70 | - $filteredRules = array_unique($filteredRules); |
|
71 | - |
|
72 | - $removeFromWhitelist = Config::inst()->get('WhitelistGenerator', 'removeFromWhitelist'); |
|
73 | - if ($removeFromWhitelist && is_array($removeFromWhitelist)) { |
|
74 | - $filteredRules = array_merge(array_diff($filteredRules, $removeFromWhitelist)); |
|
75 | - } |
|
76 | - |
|
77 | - return $filteredRules; |
|
78 | - } |
|
79 | - |
|
80 | - protected static function array_delete($array, $element) { |
|
32 | + foreach($allTopLevelRules as $rule) { |
|
33 | + if (strpos($rule, '$') !== false) { |
|
34 | + if ($rule === '$Controller') { |
|
35 | + //special case for Controllers, add all possible controllers |
|
36 | + $subControllers = ClassInfo::subclassesFor(new Controller()); |
|
37 | + |
|
38 | + foreach ($subControllers as $controller){ |
|
39 | + array_push($filteredRules, $controller); //add the controller name as a link |
|
40 | + } |
|
41 | + |
|
42 | + } elseif ($rule === '$URLSegment') { |
|
43 | + $topLevelPagesArray = array(); //temporary array to store top-level pages |
|
44 | + |
|
45 | + //special case for SiteTree, add all possible top Level Pages |
|
46 | + $topLevelPages = SiteTree::get()->filter('ParentID', 0); |
|
47 | + |
|
48 | + foreach ($topLevelPages as $page) { |
|
49 | + $link = $page->RelativeLink(); |
|
50 | + array_push($topLevelPagesArray, trim($link, '\/ ')); //remove trailing or leading slashes from links |
|
51 | + } |
|
52 | + |
|
53 | + //fetch old top-level pages URLs from the SiteTree_versions table |
|
54 | + if (Config::inst()->get('WhitelistGenerator', 'includeSiteTreeVersions')) { |
|
55 | + $oldTopLevelPagesArray = self::find_old_top_level_pages($topLevelPagesArray); |
|
56 | + } |
|
57 | + |
|
58 | + $filteredRules = array_merge($filteredRules, $topLevelPagesArray, $oldTopLevelPagesArray); |
|
59 | + } else { |
|
60 | + user_error('Unknown wildcard URL match found: '.$rule, E_WARNING); |
|
61 | + } |
|
62 | + } else { |
|
63 | + //add the rule to a new list of rules |
|
64 | + array_push($filteredRules, $rule); |
|
65 | + } |
|
66 | + |
|
67 | + } |
|
68 | + |
|
69 | + //filter duplicates (order doesn't matter here, as we are only interested in the first level of the rules) |
|
70 | + $filteredRules = array_unique($filteredRules); |
|
71 | + |
|
72 | + $removeFromWhitelist = Config::inst()->get('WhitelistGenerator', 'removeFromWhitelist'); |
|
73 | + if ($removeFromWhitelist && is_array($removeFromWhitelist)) { |
|
74 | + $filteredRules = array_merge(array_diff($filteredRules, $removeFromWhitelist)); |
|
75 | + } |
|
76 | + |
|
77 | + return $filteredRules; |
|
78 | + } |
|
79 | + |
|
80 | + protected static function array_delete($array, $element) { |
|
81 | 81 | $elementArray = array($element); |
82 | - return array_diff($array, $elementArray); |
|
83 | - } |
|
84 | - |
|
85 | - /** |
|
86 | - * Sync the list of all top-level routes with the file system whitelist cache |
|
87 | - */ |
|
88 | - protected static function syncCacheFilesystem($whitelist) { |
|
89 | - $dir = BASE_PATH . DIRECTORY_SEPARATOR . Config::inst()->get('WhitelistGenerator', 'dir'); |
|
90 | - |
|
91 | - $whitelistFolderContents = scandir($dir); |
|
92 | - |
|
93 | - //create list of files to create |
|
94 | - $toCreate = array(); |
|
95 | - foreach ($whitelist as $listItem){ |
|
96 | - if (!in_array($listItem, $whitelistFolderContents)) { |
|
97 | - if (!empty($listItem)) { //don't include empty files, such as the file for / |
|
98 | - array_push($toCreate, $listItem); |
|
99 | - } |
|
100 | - } |
|
101 | - } |
|
102 | - |
|
103 | - //create list of files to delete |
|
104 | - $toDelete = array(); |
|
105 | - foreach ($whitelistFolderContents as $file){ |
|
106 | - if (!in_array($file, array('','..','.','.htaccess'))) { //exclude things that should stay in the folder |
|
107 | - if (!in_array($file, $whitelist)) { |
|
108 | - array_push($toDelete, $file); |
|
109 | - } |
|
110 | - } |
|
111 | - } |
|
112 | - |
|
113 | - //delete files which are no longer necessary |
|
114 | - foreach ($toDelete as $delete) { |
|
115 | - unlink($dir . DIRECTORY_SEPARATOR . $delete); |
|
116 | - } |
|
117 | - |
|
118 | - //create new whitelist items as files |
|
119 | - foreach ($toCreate as $create) { |
|
120 | - touch($dir . DIRECTORY_SEPARATOR . $create); |
|
121 | - } |
|
122 | - } |
|
123 | - |
|
124 | - /** |
|
125 | - * Does a database query searching through past URLs of top-level pages, returning any URLs previously used for |
|
126 | - * pages in the SiteTree. This is to ensure that OldPageRedirector rules still apply correctly. That is, to ensure |
|
127 | - * that pages that have been renamed continue to redirect to their current versions, we add the pages' old URLs |
|
128 | - * to the whitelist. |
|
129 | - * @param $currentTopLevelPages |
|
130 | - * @return array URLs of past top-level pages |
|
131 | - */ |
|
132 | - protected static function find_old_top_level_pages($currentTopLevelPages){ |
|
133 | - $oldPageURLs = array(); |
|
134 | - |
|
135 | - $queryClass = 'SQLSelect'; |
|
136 | - if (!class_exists($queryClass) && class_exists('SQLQuery')){ |
|
137 | - $queryClass = 'SQLQuery'; |
|
138 | - } |
|
82 | + return array_diff($array, $elementArray); |
|
83 | + } |
|
84 | + |
|
85 | + /** |
|
86 | + * Sync the list of all top-level routes with the file system whitelist cache |
|
87 | + */ |
|
88 | + protected static function syncCacheFilesystem($whitelist) { |
|
89 | + $dir = BASE_PATH . DIRECTORY_SEPARATOR . Config::inst()->get('WhitelistGenerator', 'dir'); |
|
90 | + |
|
91 | + $whitelistFolderContents = scandir($dir); |
|
92 | + |
|
93 | + //create list of files to create |
|
94 | + $toCreate = array(); |
|
95 | + foreach ($whitelist as $listItem){ |
|
96 | + if (!in_array($listItem, $whitelistFolderContents)) { |
|
97 | + if (!empty($listItem)) { //don't include empty files, such as the file for / |
|
98 | + array_push($toCreate, $listItem); |
|
99 | + } |
|
100 | + } |
|
101 | + } |
|
102 | + |
|
103 | + //create list of files to delete |
|
104 | + $toDelete = array(); |
|
105 | + foreach ($whitelistFolderContents as $file){ |
|
106 | + if (!in_array($file, array('','..','.','.htaccess'))) { //exclude things that should stay in the folder |
|
107 | + if (!in_array($file, $whitelist)) { |
|
108 | + array_push($toDelete, $file); |
|
109 | + } |
|
110 | + } |
|
111 | + } |
|
112 | + |
|
113 | + //delete files which are no longer necessary |
|
114 | + foreach ($toDelete as $delete) { |
|
115 | + unlink($dir . DIRECTORY_SEPARATOR . $delete); |
|
116 | + } |
|
117 | + |
|
118 | + //create new whitelist items as files |
|
119 | + foreach ($toCreate as $create) { |
|
120 | + touch($dir . DIRECTORY_SEPARATOR . $create); |
|
121 | + } |
|
122 | + } |
|
123 | + |
|
124 | + /** |
|
125 | + * Does a database query searching through past URLs of top-level pages, returning any URLs previously used for |
|
126 | + * pages in the SiteTree. This is to ensure that OldPageRedirector rules still apply correctly. That is, to ensure |
|
127 | + * that pages that have been renamed continue to redirect to their current versions, we add the pages' old URLs |
|
128 | + * to the whitelist. |
|
129 | + * @param $currentTopLevelPages |
|
130 | + * @return array URLs of past top-level pages |
|
131 | + */ |
|
132 | + protected static function find_old_top_level_pages($currentTopLevelPages){ |
|
133 | + $oldPageURLs = array(); |
|
134 | + |
|
135 | + $queryClass = 'SQLSelect'; |
|
136 | + if (!class_exists($queryClass) && class_exists('SQLQuery')){ |
|
137 | + $queryClass = 'SQLQuery'; |
|
138 | + } |
|
139 | 139 | |
140 | - $query = new $queryClass( |
|
141 | - 'DISTINCT (URLSegment)', |
|
142 | - 'SiteTree_versions', |
|
143 | - array( |
|
144 | - 'ParentID = 0', |
|
145 | - 'WasPublished = 1', |
|
146 | - 'URLSegment NOT IN (\''.implode("','",array_filter($currentTopLevelPages)).'\')' |
|
147 | - ) |
|
148 | - ); |
|
149 | - |
|
150 | - $records = $query->execute(); |
|
151 | - if ($records) { |
|
152 | - foreach($records as $record) { |
|
153 | - array_push($oldPageURLs, $record['URLSegment']); |
|
154 | - } |
|
155 | - } |
|
156 | - |
|
157 | - return $oldPageURLs; |
|
158 | - } |
|
140 | + $query = new $queryClass( |
|
141 | + 'DISTINCT (URLSegment)', |
|
142 | + 'SiteTree_versions', |
|
143 | + array( |
|
144 | + 'ParentID = 0', |
|
145 | + 'WasPublished = 1', |
|
146 | + 'URLSegment NOT IN (\''.implode("','",array_filter($currentTopLevelPages)).'\')' |
|
147 | + ) |
|
148 | + ); |
|
149 | + |
|
150 | + $records = $query->execute(); |
|
151 | + if ($records) { |
|
152 | + foreach($records as $record) { |
|
153 | + array_push($oldPageURLs, $record['URLSegment']); |
|
154 | + } |
|
155 | + } |
|
156 | + |
|
157 | + return $oldPageURLs; |
|
158 | + } |
|
159 | 159 | |
160 | 160 | public static function ensureWhitelistFolderExists(){ |
161 | - $dir = BASE_PATH . DIRECTORY_SEPARATOR . Config::inst()->get('WhitelistGenerator', 'dir'); |
|
162 | - if (!file_exists($dir)) { |
|
163 | - mkdir($dir); //create a new whitelist dir |
|
164 | - chmod($dir,0777); //make sure it is readable by the web-server user |
|
165 | - //create a htaccess file to ensure that the whitelist cache directory is not web-accessible |
|
166 | - file_put_contents($dir.DIRECTORY_SEPARATOR.'.htaccess', "Deny from all\n"); |
|
167 | - } |
|
168 | - } |
|
169 | - |
|
170 | - public static function clearWhitelist(){ |
|
171 | - $dir = BASE_PATH . DIRECTORY_SEPARATOR . Config::inst()->get('WhitelistGenerator', 'dir'); |
|
172 | - if (file_exists($dir)) { |
|
173 | - array_map('unlink', glob($dir."/*")); |
|
174 | - } |
|
175 | - } |
|
176 | - |
|
177 | - public static function flush() { |
|
178 | - self::ensureWhitelistFolderExists(); //only create folder on flush, not on sitetree changes |
|
179 | - self::generateWhitelist(); |
|
180 | - } |
|
161 | + $dir = BASE_PATH . DIRECTORY_SEPARATOR . Config::inst()->get('WhitelistGenerator', 'dir'); |
|
162 | + if (!file_exists($dir)) { |
|
163 | + mkdir($dir); //create a new whitelist dir |
|
164 | + chmod($dir,0777); //make sure it is readable by the web-server user |
|
165 | + //create a htaccess file to ensure that the whitelist cache directory is not web-accessible |
|
166 | + file_put_contents($dir.DIRECTORY_SEPARATOR.'.htaccess', "Deny from all\n"); |
|
167 | + } |
|
168 | + } |
|
169 | + |
|
170 | + public static function clearWhitelist(){ |
|
171 | + $dir = BASE_PATH . DIRECTORY_SEPARATOR . Config::inst()->get('WhitelistGenerator', 'dir'); |
|
172 | + if (file_exists($dir)) { |
|
173 | + array_map('unlink', glob($dir."/*")); |
|
174 | + } |
|
175 | + } |
|
176 | + |
|
177 | + public static function flush() { |
|
178 | + self::ensureWhitelistFolderExists(); //only create folder on flush, not on sitetree changes |
|
179 | + self::generateWhitelist(); |
|
180 | + } |
|
181 | 181 | } |