@@ -6,26 +6,26 @@ |
||
| 6 | 6 | class HTMLPurifier_PropertyListIterator extends FilterIterator |
| 7 | 7 | { |
| 8 | 8 | |
| 9 | - protected $l; |
|
| 10 | - protected $filter; |
|
| 9 | + protected $l; |
|
| 10 | + protected $filter; |
|
| 11 | 11 | |
| 12 | - /** |
|
| 13 | - * @param $data Array of data to iterate over |
|
| 14 | - * @param $filter Optional prefix to only allow values of |
|
| 15 | - */ |
|
| 16 | - public function __construct(Iterator $iterator, $filter = null) { |
|
| 17 | - parent::__construct($iterator); |
|
| 18 | - $this->l = strlen($filter); |
|
| 19 | - $this->filter = $filter; |
|
| 20 | - } |
|
| 12 | + /** |
|
| 13 | + * @param $data Array of data to iterate over |
|
| 14 | + * @param $filter Optional prefix to only allow values of |
|
| 15 | + */ |
|
| 16 | + public function __construct(Iterator $iterator, $filter = null) { |
|
| 17 | + parent::__construct($iterator); |
|
| 18 | + $this->l = strlen($filter); |
|
| 19 | + $this->filter = $filter; |
|
| 20 | + } |
|
| 21 | 21 | |
| 22 | - public function accept() { |
|
| 23 | - $key = $this->getInnerIterator()->key(); |
|
| 24 | - if( strncmp($key, $this->filter, $this->l) !== 0 ) { |
|
| 25 | - return false; |
|
| 26 | - } |
|
| 27 | - return true; |
|
| 28 | - } |
|
| 22 | + public function accept() { |
|
| 23 | + $key = $this->getInnerIterator()->key(); |
|
| 24 | + if( strncmp($key, $this->filter, $this->l) !== 0 ) { |
|
| 25 | + return false; |
|
| 26 | + } |
|
| 27 | + return true; |
|
| 28 | + } |
|
| 29 | 29 | |
| 30 | 30 | } |
| 31 | 31 | |
@@ -21,7 +21,7 @@ |
||
| 21 | 21 | |
| 22 | 22 | public function accept() { |
| 23 | 23 | $key = $this->getInnerIterator()->key(); |
| 24 | - if( strncmp($key, $this->filter, $this->l) !== 0 ) { |
|
| 24 | + if (strncmp($key, $this->filter, $this->l) !== 0) { |
|
| 25 | 25 | return false; |
| 26 | 26 | } |
| 27 | 27 | return true; |
@@ -12,14 +12,14 @@ |
||
| 12 | 12 | abstract class HTMLPurifier_Strategy |
| 13 | 13 | { |
| 14 | 14 | |
| 15 | - /** |
|
| 16 | - * Executes the strategy on the tokens. |
|
| 17 | - * |
|
| 18 | - * @param $tokens Array of HTMLPurifier_Token objects to be operated on. |
|
| 19 | - * @param $config Configuration options |
|
| 20 | - * @returns Processed array of token objects. |
|
| 21 | - */ |
|
| 22 | - abstract public function execute($tokens, $config, $context); |
|
| 15 | + /** |
|
| 16 | + * Executes the strategy on the tokens. |
|
| 17 | + * |
|
| 18 | + * @param $tokens Array of HTMLPurifier_Token objects to be operated on. |
|
| 19 | + * @param $config Configuration options |
|
| 20 | + * @returns Processed array of token objects. |
|
| 21 | + */ |
|
| 22 | + abstract public function execute($tokens, $config, $context); |
|
| 23 | 23 | |
| 24 | 24 | } |
| 25 | 25 | |
@@ -6,17 +6,17 @@ |
||
| 6 | 6 | abstract class HTMLPurifier_Strategy_Composite extends HTMLPurifier_Strategy |
| 7 | 7 | { |
| 8 | 8 | |
| 9 | - /** |
|
| 10 | - * List of strategies to run tokens through. |
|
| 11 | - */ |
|
| 12 | - protected $strategies = array(); |
|
| 9 | + /** |
|
| 10 | + * List of strategies to run tokens through. |
|
| 11 | + */ |
|
| 12 | + protected $strategies = array(); |
|
| 13 | 13 | |
| 14 | - public function execute($tokens, $config, $context) { |
|
| 15 | - foreach ($this->strategies as $strategy) { |
|
| 16 | - $tokens = $strategy->execute($tokens, $config, $context); |
|
| 17 | - } |
|
| 18 | - return $tokens; |
|
| 19 | - } |
|
| 14 | + public function execute($tokens, $config, $context) { |
|
| 15 | + foreach ($this->strategies as $strategy) { |
|
| 16 | + $tokens = $strategy->execute($tokens, $config, $context); |
|
| 17 | + } |
|
| 18 | + return $tokens; |
|
| 19 | + } |
|
| 20 | 20 | |
| 21 | 21 | } |
| 22 | 22 | |
@@ -6,12 +6,12 @@ |
||
| 6 | 6 | class HTMLPurifier_Strategy_Core extends HTMLPurifier_Strategy_Composite |
| 7 | 7 | { |
| 8 | 8 | |
| 9 | - public function __construct() { |
|
| 10 | - $this->strategies[] = new HTMLPurifier_Strategy_RemoveForeignElements(); |
|
| 11 | - $this->strategies[] = new HTMLPurifier_Strategy_MakeWellFormed(); |
|
| 12 | - $this->strategies[] = new HTMLPurifier_Strategy_FixNesting(); |
|
| 13 | - $this->strategies[] = new HTMLPurifier_Strategy_ValidateAttributes(); |
|
| 14 | - } |
|
| 9 | + public function __construct() { |
|
| 10 | + $this->strategies[] = new HTMLPurifier_Strategy_RemoveForeignElements(); |
|
| 11 | + $this->strategies[] = new HTMLPurifier_Strategy_MakeWellFormed(); |
|
| 12 | + $this->strategies[] = new HTMLPurifier_Strategy_FixNesting(); |
|
| 13 | + $this->strategies[] = new HTMLPurifier_Strategy_ValidateAttributes(); |
|
| 14 | + } |
|
| 15 | 15 | |
| 16 | 16 | } |
| 17 | 17 | |
@@ -31,297 +31,297 @@ |
||
| 31 | 31 | class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy |
| 32 | 32 | { |
| 33 | 33 | |
| 34 | - public function execute($tokens, $config, $context) { |
|
| 35 | - //####################################################################// |
|
| 36 | - // Pre-processing |
|
| 37 | - |
|
| 38 | - // get a copy of the HTML definition |
|
| 39 | - $definition = $config->getHTMLDefinition(); |
|
| 40 | - |
|
| 41 | - // insert implicit "parent" node, will be removed at end. |
|
| 42 | - // DEFINITION CALL |
|
| 43 | - $parent_name = $definition->info_parent; |
|
| 44 | - array_unshift($tokens, new HTMLPurifier_Token_Start($parent_name)); |
|
| 45 | - $tokens[] = new HTMLPurifier_Token_End($parent_name); |
|
| 46 | - |
|
| 47 | - // setup the context variable 'IsInline', for chameleon processing |
|
| 48 | - // is 'false' when we are not inline, 'true' when it must always |
|
| 49 | - // be inline, and an integer when it is inline for a certain |
|
| 50 | - // branch of the document tree |
|
| 51 | - $is_inline = $definition->info_parent_def->descendants_are_inline; |
|
| 52 | - $context->register('IsInline', $is_inline); |
|
| 53 | - |
|
| 54 | - // setup error collector |
|
| 55 | - $e =& $context->get('ErrorCollector', true); |
|
| 56 | - |
|
| 57 | - //####################################################################// |
|
| 58 | - // Loop initialization |
|
| 59 | - |
|
| 60 | - // stack that contains the indexes of all parents, |
|
| 61 | - // $stack[count($stack)-1] being the current parent |
|
| 62 | - $stack = array(); |
|
| 63 | - |
|
| 64 | - // stack that contains all elements that are excluded |
|
| 65 | - // it is organized by parent elements, similar to $stack, |
|
| 66 | - // but it is only populated when an element with exclusions is |
|
| 67 | - // processed, i.e. there won't be empty exclusions. |
|
| 68 | - $exclude_stack = array(); |
|
| 69 | - |
|
| 70 | - // variable that contains the start token while we are processing |
|
| 71 | - // nodes. This enables error reporting to do its job |
|
| 72 | - $start_token = false; |
|
| 73 | - $context->register('CurrentToken', $start_token); |
|
| 74 | - |
|
| 75 | - //####################################################################// |
|
| 76 | - // Loop |
|
| 77 | - |
|
| 78 | - // iterate through all start nodes. Determining the start node |
|
| 79 | - // is complicated so it has been omitted from the loop construct |
|
| 80 | - for ($i = 0, $size = count($tokens) ; $i < $size; ) { |
|
| 81 | - |
|
| 82 | - //################################################################// |
|
| 83 | - // Gather information on children |
|
| 84 | - |
|
| 85 | - // child token accumulator |
|
| 86 | - $child_tokens = array(); |
|
| 87 | - |
|
| 88 | - // scroll to the end of this node, report number, and collect |
|
| 89 | - // all children |
|
| 90 | - for ($j = $i, $depth = 0; ; $j++) { |
|
| 91 | - if ($tokens[$j] instanceof HTMLPurifier_Token_Start) { |
|
| 92 | - $depth++; |
|
| 93 | - // skip token assignment on first iteration, this is the |
|
| 94 | - // token we currently are on |
|
| 95 | - if ($depth == 1) continue; |
|
| 96 | - } elseif ($tokens[$j] instanceof HTMLPurifier_Token_End) { |
|
| 97 | - $depth--; |
|
| 98 | - // skip token assignment on last iteration, this is the |
|
| 99 | - // end token of the token we're currently on |
|
| 100 | - if ($depth == 0) break; |
|
| 101 | - } |
|
| 102 | - $child_tokens[] = $tokens[$j]; |
|
| 103 | - } |
|
| 104 | - |
|
| 105 | - // $i is index of start token |
|
| 106 | - // $j is index of end token |
|
| 107 | - |
|
| 108 | - $start_token = $tokens[$i]; // to make token available via CurrentToken |
|
| 109 | - |
|
| 110 | - //################################################################// |
|
| 111 | - // Gather information on parent |
|
| 112 | - |
|
| 113 | - // calculate parent information |
|
| 114 | - if ($count = count($stack)) { |
|
| 115 | - $parent_index = $stack[$count-1]; |
|
| 116 | - $parent_name = $tokens[$parent_index]->name; |
|
| 117 | - if ($parent_index == 0) { |
|
| 118 | - $parent_def = $definition->info_parent_def; |
|
| 119 | - } else { |
|
| 120 | - $parent_def = $definition->info[$parent_name]; |
|
| 121 | - } |
|
| 122 | - } else { |
|
| 123 | - // processing as if the parent were the "root" node |
|
| 124 | - // unknown info, it won't be used anyway, in the future, |
|
| 125 | - // we may want to enforce one element only (this is |
|
| 126 | - // necessary for HTML Purifier to clean entire documents |
|
| 127 | - $parent_index = $parent_name = $parent_def = null; |
|
| 128 | - } |
|
| 129 | - |
|
| 130 | - // calculate context |
|
| 131 | - if ($is_inline === false) { |
|
| 132 | - // check if conditions make it inline |
|
| 133 | - if (!empty($parent_def) && $parent_def->descendants_are_inline) { |
|
| 134 | - $is_inline = $count - 1; |
|
| 135 | - } |
|
| 136 | - } else { |
|
| 137 | - // check if we're out of inline |
|
| 138 | - if ($count === $is_inline) { |
|
| 139 | - $is_inline = false; |
|
| 140 | - } |
|
| 141 | - } |
|
| 142 | - |
|
| 143 | - //################################################################// |
|
| 144 | - // Determine whether element is explicitly excluded SGML-style |
|
| 145 | - |
|
| 146 | - // determine whether or not element is excluded by checking all |
|
| 147 | - // parent exclusions. The array should not be very large, two |
|
| 148 | - // elements at most. |
|
| 149 | - $excluded = false; |
|
| 150 | - if (!empty($exclude_stack)) { |
|
| 151 | - foreach ($exclude_stack as $lookup) { |
|
| 152 | - if (isset($lookup[$tokens[$i]->name])) { |
|
| 153 | - $excluded = true; |
|
| 154 | - // no need to continue processing |
|
| 155 | - break; |
|
| 156 | - } |
|
| 157 | - } |
|
| 158 | - } |
|
| 159 | - |
|
| 160 | - //################################################################// |
|
| 161 | - // Perform child validation |
|
| 162 | - |
|
| 163 | - if ($excluded) { |
|
| 164 | - // there is an exclusion, remove the entire node |
|
| 165 | - $result = false; |
|
| 166 | - $excludes = array(); // not used, but good to initialize anyway |
|
| 167 | - } else { |
|
| 168 | - // DEFINITION CALL |
|
| 169 | - if ($i === 0) { |
|
| 170 | - // special processing for the first node |
|
| 171 | - $def = $definition->info_parent_def; |
|
| 172 | - } else { |
|
| 173 | - $def = $definition->info[$tokens[$i]->name]; |
|
| 174 | - |
|
| 175 | - } |
|
| 176 | - |
|
| 177 | - if (!empty($def->child)) { |
|
| 178 | - // have DTD child def validate children |
|
| 179 | - $result = $def->child->validateChildren( |
|
| 180 | - $child_tokens, $config, $context); |
|
| 181 | - } else { |
|
| 182 | - // weird, no child definition, get rid of everything |
|
| 183 | - $result = false; |
|
| 184 | - } |
|
| 185 | - |
|
| 186 | - // determine whether or not this element has any exclusions |
|
| 187 | - $excludes = $def->excludes; |
|
| 188 | - } |
|
| 189 | - |
|
| 190 | - // $result is now a bool or array |
|
| 191 | - |
|
| 192 | - //################################################################// |
|
| 193 | - // Process result by interpreting $result |
|
| 194 | - |
|
| 195 | - if ($result === true || $child_tokens === $result) { |
|
| 196 | - // leave the node as is |
|
| 197 | - |
|
| 198 | - // register start token as a parental node start |
|
| 199 | - $stack[] = $i; |
|
| 200 | - |
|
| 201 | - // register exclusions if there are any |
|
| 202 | - if (!empty($excludes)) $exclude_stack[] = $excludes; |
|
| 203 | - |
|
| 204 | - // move cursor to next possible start node |
|
| 205 | - $i++; |
|
| 206 | - |
|
| 207 | - } elseif($result === false) { |
|
| 208 | - // remove entire node |
|
| 209 | - |
|
| 210 | - if ($e) { |
|
| 211 | - if ($excluded) { |
|
| 212 | - $e->send(E_ERROR, 'Strategy_FixNesting: Node excluded'); |
|
| 213 | - } else { |
|
| 214 | - $e->send(E_ERROR, 'Strategy_FixNesting: Node removed'); |
|
| 215 | - } |
|
| 216 | - } |
|
| 217 | - |
|
| 218 | - // calculate length of inner tokens and current tokens |
|
| 219 | - $length = $j - $i + 1; |
|
| 220 | - |
|
| 221 | - // perform removal |
|
| 222 | - array_splice($tokens, $i, $length); |
|
| 223 | - |
|
| 224 | - // update size |
|
| 225 | - $size -= $length; |
|
| 226 | - |
|
| 227 | - // there is no start token to register, |
|
| 228 | - // current node is now the next possible start node |
|
| 229 | - // unless it turns out that we need to do a double-check |
|
| 230 | - |
|
| 231 | - // this is a rought heuristic that covers 100% of HTML's |
|
| 232 | - // cases and 99% of all other cases. A child definition |
|
| 233 | - // that would be tricked by this would be something like: |
|
| 234 | - // ( | a b c) where it's all or nothing. Fortunately, |
|
| 235 | - // our current implementation claims that that case would |
|
| 236 | - // not allow empty, even if it did |
|
| 237 | - if (!$parent_def->child->allow_empty) { |
|
| 238 | - // we need to do a double-check |
|
| 239 | - $i = $parent_index; |
|
| 240 | - array_pop($stack); |
|
| 241 | - } |
|
| 242 | - |
|
| 243 | - // PROJECTED OPTIMIZATION: Process all children elements before |
|
| 244 | - // reprocessing parent node. |
|
| 245 | - |
|
| 246 | - } else { |
|
| 247 | - // replace node with $result |
|
| 248 | - |
|
| 249 | - // calculate length of inner tokens |
|
| 250 | - $length = $j - $i - 1; |
|
| 251 | - |
|
| 252 | - if ($e) { |
|
| 253 | - if (empty($result) && $length) { |
|
| 254 | - $e->send(E_ERROR, 'Strategy_FixNesting: Node contents removed'); |
|
| 255 | - } else { |
|
| 256 | - $e->send(E_WARNING, 'Strategy_FixNesting: Node reorganized'); |
|
| 257 | - } |
|
| 258 | - } |
|
| 259 | - |
|
| 260 | - // perform replacement |
|
| 261 | - array_splice($tokens, $i + 1, $length, $result); |
|
| 262 | - |
|
| 263 | - // update size |
|
| 264 | - $size -= $length; |
|
| 265 | - $size += count($result); |
|
| 266 | - |
|
| 267 | - // register start token as a parental node start |
|
| 268 | - $stack[] = $i; |
|
| 269 | - |
|
| 270 | - // register exclusions if there are any |
|
| 271 | - if (!empty($excludes)) $exclude_stack[] = $excludes; |
|
| 272 | - |
|
| 273 | - // move cursor to next possible start node |
|
| 274 | - $i++; |
|
| 275 | - |
|
| 276 | - } |
|
| 277 | - |
|
| 278 | - //################################################################// |
|
| 279 | - // Scroll to next start node |
|
| 280 | - |
|
| 281 | - // We assume, at this point, that $i is the index of the token |
|
| 282 | - // that is the first possible new start point for a node. |
|
| 283 | - |
|
| 284 | - // Test if the token indeed is a start tag, if not, move forward |
|
| 285 | - // and test again. |
|
| 286 | - $size = count($tokens); |
|
| 287 | - while ($i < $size and !$tokens[$i] instanceof HTMLPurifier_Token_Start) { |
|
| 288 | - if ($tokens[$i] instanceof HTMLPurifier_Token_End) { |
|
| 289 | - // pop a token index off the stack if we ended a node |
|
| 290 | - array_pop($stack); |
|
| 291 | - // pop an exclusion lookup off exclusion stack if |
|
| 292 | - // we ended node and that node had exclusions |
|
| 293 | - if ($i == 0 || $i == $size - 1) { |
|
| 294 | - // use specialized var if it's the super-parent |
|
| 295 | - $s_excludes = $definition->info_parent_def->excludes; |
|
| 296 | - } else { |
|
| 297 | - $s_excludes = $definition->info[$tokens[$i]->name]->excludes; |
|
| 298 | - } |
|
| 299 | - if ($s_excludes) { |
|
| 300 | - array_pop($exclude_stack); |
|
| 301 | - } |
|
| 302 | - } |
|
| 303 | - $i++; |
|
| 304 | - } |
|
| 305 | - |
|
| 306 | - } |
|
| 307 | - |
|
| 308 | - //####################################################################// |
|
| 309 | - // Post-processing |
|
| 310 | - |
|
| 311 | - // remove implicit parent tokens at the beginning and end |
|
| 312 | - array_shift($tokens); |
|
| 313 | - array_pop($tokens); |
|
| 314 | - |
|
| 315 | - // remove context variables |
|
| 316 | - $context->destroy('IsInline'); |
|
| 317 | - $context->destroy('CurrentToken'); |
|
| 318 | - |
|
| 319 | - //####################################################################// |
|
| 320 | - // Return |
|
| 321 | - |
|
| 322 | - return $tokens; |
|
| 323 | - |
|
| 324 | - } |
|
| 34 | + public function execute($tokens, $config, $context) { |
|
| 35 | + //####################################################################// |
|
| 36 | + // Pre-processing |
|
| 37 | + |
|
| 38 | + // get a copy of the HTML definition |
|
| 39 | + $definition = $config->getHTMLDefinition(); |
|
| 40 | + |
|
| 41 | + // insert implicit "parent" node, will be removed at end. |
|
| 42 | + // DEFINITION CALL |
|
| 43 | + $parent_name = $definition->info_parent; |
|
| 44 | + array_unshift($tokens, new HTMLPurifier_Token_Start($parent_name)); |
|
| 45 | + $tokens[] = new HTMLPurifier_Token_End($parent_name); |
|
| 46 | + |
|
| 47 | + // setup the context variable 'IsInline', for chameleon processing |
|
| 48 | + // is 'false' when we are not inline, 'true' when it must always |
|
| 49 | + // be inline, and an integer when it is inline for a certain |
|
| 50 | + // branch of the document tree |
|
| 51 | + $is_inline = $definition->info_parent_def->descendants_are_inline; |
|
| 52 | + $context->register('IsInline', $is_inline); |
|
| 53 | + |
|
| 54 | + // setup error collector |
|
| 55 | + $e =& $context->get('ErrorCollector', true); |
|
| 56 | + |
|
| 57 | + //####################################################################// |
|
| 58 | + // Loop initialization |
|
| 59 | + |
|
| 60 | + // stack that contains the indexes of all parents, |
|
| 61 | + // $stack[count($stack)-1] being the current parent |
|
| 62 | + $stack = array(); |
|
| 63 | + |
|
| 64 | + // stack that contains all elements that are excluded |
|
| 65 | + // it is organized by parent elements, similar to $stack, |
|
| 66 | + // but it is only populated when an element with exclusions is |
|
| 67 | + // processed, i.e. there won't be empty exclusions. |
|
| 68 | + $exclude_stack = array(); |
|
| 69 | + |
|
| 70 | + // variable that contains the start token while we are processing |
|
| 71 | + // nodes. This enables error reporting to do its job |
|
| 72 | + $start_token = false; |
|
| 73 | + $context->register('CurrentToken', $start_token); |
|
| 74 | + |
|
| 75 | + //####################################################################// |
|
| 76 | + // Loop |
|
| 77 | + |
|
| 78 | + // iterate through all start nodes. Determining the start node |
|
| 79 | + // is complicated so it has been omitted from the loop construct |
|
| 80 | + for ($i = 0, $size = count($tokens) ; $i < $size; ) { |
|
| 81 | + |
|
| 82 | + //################################################################// |
|
| 83 | + // Gather information on children |
|
| 84 | + |
|
| 85 | + // child token accumulator |
|
| 86 | + $child_tokens = array(); |
|
| 87 | + |
|
| 88 | + // scroll to the end of this node, report number, and collect |
|
| 89 | + // all children |
|
| 90 | + for ($j = $i, $depth = 0; ; $j++) { |
|
| 91 | + if ($tokens[$j] instanceof HTMLPurifier_Token_Start) { |
|
| 92 | + $depth++; |
|
| 93 | + // skip token assignment on first iteration, this is the |
|
| 94 | + // token we currently are on |
|
| 95 | + if ($depth == 1) continue; |
|
| 96 | + } elseif ($tokens[$j] instanceof HTMLPurifier_Token_End) { |
|
| 97 | + $depth--; |
|
| 98 | + // skip token assignment on last iteration, this is the |
|
| 99 | + // end token of the token we're currently on |
|
| 100 | + if ($depth == 0) break; |
|
| 101 | + } |
|
| 102 | + $child_tokens[] = $tokens[$j]; |
|
| 103 | + } |
|
| 104 | + |
|
| 105 | + // $i is index of start token |
|
| 106 | + // $j is index of end token |
|
| 107 | + |
|
| 108 | + $start_token = $tokens[$i]; // to make token available via CurrentToken |
|
| 109 | + |
|
| 110 | + //################################################################// |
|
| 111 | + // Gather information on parent |
|
| 112 | + |
|
| 113 | + // calculate parent information |
|
| 114 | + if ($count = count($stack)) { |
|
| 115 | + $parent_index = $stack[$count-1]; |
|
| 116 | + $parent_name = $tokens[$parent_index]->name; |
|
| 117 | + if ($parent_index == 0) { |
|
| 118 | + $parent_def = $definition->info_parent_def; |
|
| 119 | + } else { |
|
| 120 | + $parent_def = $definition->info[$parent_name]; |
|
| 121 | + } |
|
| 122 | + } else { |
|
| 123 | + // processing as if the parent were the "root" node |
|
| 124 | + // unknown info, it won't be used anyway, in the future, |
|
| 125 | + // we may want to enforce one element only (this is |
|
| 126 | + // necessary for HTML Purifier to clean entire documents |
|
| 127 | + $parent_index = $parent_name = $parent_def = null; |
|
| 128 | + } |
|
| 129 | + |
|
| 130 | + // calculate context |
|
| 131 | + if ($is_inline === false) { |
|
| 132 | + // check if conditions make it inline |
|
| 133 | + if (!empty($parent_def) && $parent_def->descendants_are_inline) { |
|
| 134 | + $is_inline = $count - 1; |
|
| 135 | + } |
|
| 136 | + } else { |
|
| 137 | + // check if we're out of inline |
|
| 138 | + if ($count === $is_inline) { |
|
| 139 | + $is_inline = false; |
|
| 140 | + } |
|
| 141 | + } |
|
| 142 | + |
|
| 143 | + //################################################################// |
|
| 144 | + // Determine whether element is explicitly excluded SGML-style |
|
| 145 | + |
|
| 146 | + // determine whether or not element is excluded by checking all |
|
| 147 | + // parent exclusions. The array should not be very large, two |
|
| 148 | + // elements at most. |
|
| 149 | + $excluded = false; |
|
| 150 | + if (!empty($exclude_stack)) { |
|
| 151 | + foreach ($exclude_stack as $lookup) { |
|
| 152 | + if (isset($lookup[$tokens[$i]->name])) { |
|
| 153 | + $excluded = true; |
|
| 154 | + // no need to continue processing |
|
| 155 | + break; |
|
| 156 | + } |
|
| 157 | + } |
|
| 158 | + } |
|
| 159 | + |
|
| 160 | + //################################################################// |
|
| 161 | + // Perform child validation |
|
| 162 | + |
|
| 163 | + if ($excluded) { |
|
| 164 | + // there is an exclusion, remove the entire node |
|
| 165 | + $result = false; |
|
| 166 | + $excludes = array(); // not used, but good to initialize anyway |
|
| 167 | + } else { |
|
| 168 | + // DEFINITION CALL |
|
| 169 | + if ($i === 0) { |
|
| 170 | + // special processing for the first node |
|
| 171 | + $def = $definition->info_parent_def; |
|
| 172 | + } else { |
|
| 173 | + $def = $definition->info[$tokens[$i]->name]; |
|
| 174 | + |
|
| 175 | + } |
|
| 176 | + |
|
| 177 | + if (!empty($def->child)) { |
|
| 178 | + // have DTD child def validate children |
|
| 179 | + $result = $def->child->validateChildren( |
|
| 180 | + $child_tokens, $config, $context); |
|
| 181 | + } else { |
|
| 182 | + // weird, no child definition, get rid of everything |
|
| 183 | + $result = false; |
|
| 184 | + } |
|
| 185 | + |
|
| 186 | + // determine whether or not this element has any exclusions |
|
| 187 | + $excludes = $def->excludes; |
|
| 188 | + } |
|
| 189 | + |
|
| 190 | + // $result is now a bool or array |
|
| 191 | + |
|
| 192 | + //################################################################// |
|
| 193 | + // Process result by interpreting $result |
|
| 194 | + |
|
| 195 | + if ($result === true || $child_tokens === $result) { |
|
| 196 | + // leave the node as is |
|
| 197 | + |
|
| 198 | + // register start token as a parental node start |
|
| 199 | + $stack[] = $i; |
|
| 200 | + |
|
| 201 | + // register exclusions if there are any |
|
| 202 | + if (!empty($excludes)) $exclude_stack[] = $excludes; |
|
| 203 | + |
|
| 204 | + // move cursor to next possible start node |
|
| 205 | + $i++; |
|
| 206 | + |
|
| 207 | + } elseif($result === false) { |
|
| 208 | + // remove entire node |
|
| 209 | + |
|
| 210 | + if ($e) { |
|
| 211 | + if ($excluded) { |
|
| 212 | + $e->send(E_ERROR, 'Strategy_FixNesting: Node excluded'); |
|
| 213 | + } else { |
|
| 214 | + $e->send(E_ERROR, 'Strategy_FixNesting: Node removed'); |
|
| 215 | + } |
|
| 216 | + } |
|
| 217 | + |
|
| 218 | + // calculate length of inner tokens and current tokens |
|
| 219 | + $length = $j - $i + 1; |
|
| 220 | + |
|
| 221 | + // perform removal |
|
| 222 | + array_splice($tokens, $i, $length); |
|
| 223 | + |
|
| 224 | + // update size |
|
| 225 | + $size -= $length; |
|
| 226 | + |
|
| 227 | + // there is no start token to register, |
|
| 228 | + // current node is now the next possible start node |
|
| 229 | + // unless it turns out that we need to do a double-check |
|
| 230 | + |
|
| 231 | + // this is a rought heuristic that covers 100% of HTML's |
|
| 232 | + // cases and 99% of all other cases. A child definition |
|
| 233 | + // that would be tricked by this would be something like: |
|
| 234 | + // ( | a b c) where it's all or nothing. Fortunately, |
|
| 235 | + // our current implementation claims that that case would |
|
| 236 | + // not allow empty, even if it did |
|
| 237 | + if (!$parent_def->child->allow_empty) { |
|
| 238 | + // we need to do a double-check |
|
| 239 | + $i = $parent_index; |
|
| 240 | + array_pop($stack); |
|
| 241 | + } |
|
| 242 | + |
|
| 243 | + // PROJECTED OPTIMIZATION: Process all children elements before |
|
| 244 | + // reprocessing parent node. |
|
| 245 | + |
|
| 246 | + } else { |
|
| 247 | + // replace node with $result |
|
| 248 | + |
|
| 249 | + // calculate length of inner tokens |
|
| 250 | + $length = $j - $i - 1; |
|
| 251 | + |
|
| 252 | + if ($e) { |
|
| 253 | + if (empty($result) && $length) { |
|
| 254 | + $e->send(E_ERROR, 'Strategy_FixNesting: Node contents removed'); |
|
| 255 | + } else { |
|
| 256 | + $e->send(E_WARNING, 'Strategy_FixNesting: Node reorganized'); |
|
| 257 | + } |
|
| 258 | + } |
|
| 259 | + |
|
| 260 | + // perform replacement |
|
| 261 | + array_splice($tokens, $i + 1, $length, $result); |
|
| 262 | + |
|
| 263 | + // update size |
|
| 264 | + $size -= $length; |
|
| 265 | + $size += count($result); |
|
| 266 | + |
|
| 267 | + // register start token as a parental node start |
|
| 268 | + $stack[] = $i; |
|
| 269 | + |
|
| 270 | + // register exclusions if there are any |
|
| 271 | + if (!empty($excludes)) $exclude_stack[] = $excludes; |
|
| 272 | + |
|
| 273 | + // move cursor to next possible start node |
|
| 274 | + $i++; |
|
| 275 | + |
|
| 276 | + } |
|
| 277 | + |
|
| 278 | + //################################################################// |
|
| 279 | + // Scroll to next start node |
|
| 280 | + |
|
| 281 | + // We assume, at this point, that $i is the index of the token |
|
| 282 | + // that is the first possible new start point for a node. |
|
| 283 | + |
|
| 284 | + // Test if the token indeed is a start tag, if not, move forward |
|
| 285 | + // and test again. |
|
| 286 | + $size = count($tokens); |
|
| 287 | + while ($i < $size and !$tokens[$i] instanceof HTMLPurifier_Token_Start) { |
|
| 288 | + if ($tokens[$i] instanceof HTMLPurifier_Token_End) { |
|
| 289 | + // pop a token index off the stack if we ended a node |
|
| 290 | + array_pop($stack); |
|
| 291 | + // pop an exclusion lookup off exclusion stack if |
|
| 292 | + // we ended node and that node had exclusions |
|
| 293 | + if ($i == 0 || $i == $size - 1) { |
|
| 294 | + // use specialized var if it's the super-parent |
|
| 295 | + $s_excludes = $definition->info_parent_def->excludes; |
|
| 296 | + } else { |
|
| 297 | + $s_excludes = $definition->info[$tokens[$i]->name]->excludes; |
|
| 298 | + } |
|
| 299 | + if ($s_excludes) { |
|
| 300 | + array_pop($exclude_stack); |
|
| 301 | + } |
|
| 302 | + } |
|
| 303 | + $i++; |
|
| 304 | + } |
|
| 305 | + |
|
| 306 | + } |
|
| 307 | + |
|
| 308 | + //####################################################################// |
|
| 309 | + // Post-processing |
|
| 310 | + |
|
| 311 | + // remove implicit parent tokens at the beginning and end |
|
| 312 | + array_shift($tokens); |
|
| 313 | + array_pop($tokens); |
|
| 314 | + |
|
| 315 | + // remove context variables |
|
| 316 | + $context->destroy('IsInline'); |
|
| 317 | + $context->destroy('CurrentToken'); |
|
| 318 | + |
|
| 319 | + //####################################################################// |
|
| 320 | + // Return |
|
| 321 | + |
|
| 322 | + return $tokens; |
|
| 323 | + |
|
| 324 | + } |
|
| 325 | 325 | |
| 326 | 326 | } |
| 327 | 327 | |
@@ -52,7 +52,7 @@ discard block |
||
| 52 | 52 | $context->register('IsInline', $is_inline); |
| 53 | 53 | |
| 54 | 54 | // setup error collector |
| 55 | - $e =& $context->get('ErrorCollector', true); |
|
| 55 | + $e = & $context->get('ErrorCollector', true); |
|
| 56 | 56 | |
| 57 | 57 | //####################################################################// |
| 58 | 58 | // Loop initialization |
@@ -77,7 +77,7 @@ discard block |
||
| 77 | 77 | |
| 78 | 78 | // iterate through all start nodes. Determining the start node |
| 79 | 79 | // is complicated so it has been omitted from the loop construct |
| 80 | - for ($i = 0, $size = count($tokens) ; $i < $size; ) { |
|
| 80 | + for ($i = 0, $size = count($tokens); $i < $size;) { |
|
| 81 | 81 | |
| 82 | 82 | //################################################################// |
| 83 | 83 | // Gather information on children |
@@ -112,7 +112,7 @@ discard block |
||
| 112 | 112 | |
| 113 | 113 | // calculate parent information |
| 114 | 114 | if ($count = count($stack)) { |
| 115 | - $parent_index = $stack[$count-1]; |
|
| 115 | + $parent_index = $stack[$count - 1]; |
|
| 116 | 116 | $parent_name = $tokens[$parent_index]->name; |
| 117 | 117 | if ($parent_index == 0) { |
| 118 | 118 | $parent_def = $definition->info_parent_def; |
@@ -204,7 +204,7 @@ discard block |
||
| 204 | 204 | // move cursor to next possible start node |
| 205 | 205 | $i++; |
| 206 | 206 | |
| 207 | - } elseif($result === false) { |
|
| 207 | + } elseif ($result === false) { |
|
| 208 | 208 | // remove entire node |
| 209 | 209 | |
| 210 | 210 | if ($e) { |
@@ -92,12 +92,16 @@ discard block |
||
| 92 | 92 | $depth++; |
| 93 | 93 | // skip token assignment on first iteration, this is the |
| 94 | 94 | // token we currently are on |
| 95 | - if ($depth == 1) continue; |
|
| 95 | + if ($depth == 1) { |
|
| 96 | + continue; |
|
| 97 | + } |
|
| 96 | 98 | } elseif ($tokens[$j] instanceof HTMLPurifier_Token_End) { |
| 97 | 99 | $depth--; |
| 98 | 100 | // skip token assignment on last iteration, this is the |
| 99 | 101 | // end token of the token we're currently on |
| 100 | - if ($depth == 0) break; |
|
| 102 | + if ($depth == 0) { |
|
| 103 | + break; |
|
| 104 | + } |
|
| 101 | 105 | } |
| 102 | 106 | $child_tokens[] = $tokens[$j]; |
| 103 | 107 | } |
@@ -199,7 +203,9 @@ discard block |
||
| 199 | 203 | $stack[] = $i; |
| 200 | 204 | |
| 201 | 205 | // register exclusions if there are any |
| 202 | - if (!empty($excludes)) $exclude_stack[] = $excludes; |
|
| 206 | + if (!empty($excludes)) { |
|
| 207 | + $exclude_stack[] = $excludes; |
|
| 208 | + } |
|
| 203 | 209 | |
| 204 | 210 | // move cursor to next possible start node |
| 205 | 211 | $i++; |
@@ -268,7 +274,9 @@ discard block |
||
| 268 | 274 | $stack[] = $i; |
| 269 | 275 | |
| 270 | 276 | // register exclusions if there are any |
| 271 | - if (!empty($excludes)) $exclude_stack[] = $excludes; |
|
| 277 | + if (!empty($excludes)) { |
|
| 278 | + $exclude_stack[] = $excludes; |
|
| 279 | + } |
|
| 272 | 280 | |
| 273 | 281 | // move cursor to next possible start node |
| 274 | 282 | $i++; |
@@ -14,518 +14,518 @@ |
||
| 14 | 14 | class HTMLPurifier_Strategy_MakeWellFormed extends HTMLPurifier_Strategy |
| 15 | 15 | { |
| 16 | 16 | |
| 17 | - /** |
|
| 18 | - * Array stream of tokens being processed. |
|
| 19 | - */ |
|
| 20 | - protected $tokens; |
|
| 21 | - |
|
| 22 | - /** |
|
| 23 | - * Current index in $tokens. |
|
| 24 | - */ |
|
| 25 | - protected $t; |
|
| 26 | - |
|
| 27 | - /** |
|
| 28 | - * Current nesting of elements. |
|
| 29 | - */ |
|
| 30 | - protected $stack; |
|
| 31 | - |
|
| 32 | - /** |
|
| 33 | - * Injectors active in this stream processing. |
|
| 34 | - */ |
|
| 35 | - protected $injectors; |
|
| 36 | - |
|
| 37 | - /** |
|
| 38 | - * Current instance of HTMLPurifier_Config. |
|
| 39 | - */ |
|
| 40 | - protected $config; |
|
| 41 | - |
|
| 42 | - /** |
|
| 43 | - * Current instance of HTMLPurifier_Context. |
|
| 44 | - */ |
|
| 45 | - protected $context; |
|
| 46 | - |
|
| 47 | - public function execute($tokens, $config, $context) { |
|
| 48 | - |
|
| 49 | - $definition = $config->getHTMLDefinition(); |
|
| 50 | - |
|
| 51 | - // local variables |
|
| 52 | - $generator = new HTMLPurifier_Generator($config, $context); |
|
| 53 | - $escape_invalid_tags = $config->get('Core.EscapeInvalidTags'); |
|
| 54 | - // used for autoclose early abortion |
|
| 55 | - $global_parent_allowed_elements = array(); |
|
| 56 | - if (isset($definition->info[$definition->info_parent])) { |
|
| 57 | - // may be unset under testing circumstances |
|
| 58 | - $global_parent_allowed_elements = $definition->info[$definition->info_parent]->child->getAllowedElements($config); |
|
| 59 | - } |
|
| 60 | - $e = $context->get('ErrorCollector', true); |
|
| 61 | - $t = false; // token index |
|
| 62 | - $i = false; // injector index |
|
| 63 | - $token = false; // the current token |
|
| 64 | - $reprocess = false; // whether or not to reprocess the same token |
|
| 65 | - $stack = array(); |
|
| 66 | - |
|
| 67 | - // member variables |
|
| 68 | - $this->stack =& $stack; |
|
| 69 | - $this->t =& $t; |
|
| 70 | - $this->tokens =& $tokens; |
|
| 71 | - $this->config = $config; |
|
| 72 | - $this->context = $context; |
|
| 73 | - |
|
| 74 | - // context variables |
|
| 75 | - $context->register('CurrentNesting', $stack); |
|
| 76 | - $context->register('InputIndex', $t); |
|
| 77 | - $context->register('InputTokens', $tokens); |
|
| 78 | - $context->register('CurrentToken', $token); |
|
| 79 | - |
|
| 80 | - // -- begin INJECTOR -- |
|
| 81 | - |
|
| 82 | - $this->injectors = array(); |
|
| 83 | - |
|
| 84 | - $injectors = $config->getBatch('AutoFormat'); |
|
| 85 | - $def_injectors = $definition->info_injector; |
|
| 86 | - $custom_injectors = $injectors['Custom']; |
|
| 87 | - unset($injectors['Custom']); // special case |
|
| 88 | - foreach ($injectors as $injector => $b) { |
|
| 89 | - // XXX: Fix with a legitimate lookup table of enabled filters |
|
| 90 | - if (strpos($injector, '.') !== false) continue; |
|
| 91 | - $injector = "HTMLPurifier_Injector_$injector"; |
|
| 92 | - if (!$b) continue; |
|
| 93 | - $this->injectors[] = new $injector; |
|
| 94 | - } |
|
| 95 | - foreach ($def_injectors as $injector) { |
|
| 96 | - // assumed to be objects |
|
| 97 | - $this->injectors[] = $injector; |
|
| 98 | - } |
|
| 99 | - foreach ($custom_injectors as $injector) { |
|
| 100 | - if (!$injector) continue; |
|
| 101 | - if (is_string($injector)) { |
|
| 102 | - $injector = "HTMLPurifier_Injector_$injector"; |
|
| 103 | - $injector = new $injector; |
|
| 104 | - } |
|
| 105 | - $this->injectors[] = $injector; |
|
| 106 | - } |
|
| 107 | - |
|
| 108 | - // give the injectors references to the definition and context |
|
| 109 | - // variables for performance reasons |
|
| 110 | - foreach ($this->injectors as $ix => $injector) { |
|
| 111 | - $error = $injector->prepare($config, $context); |
|
| 112 | - if (!$error) continue; |
|
| 113 | - array_splice($this->injectors, $ix, 1); // rm the injector |
|
| 114 | - trigger_error("Cannot enable {$injector->name} injector because $error is not allowed", E_USER_WARNING); |
|
| 115 | - } |
|
| 116 | - |
|
| 117 | - // -- end INJECTOR -- |
|
| 118 | - |
|
| 119 | - // a note on reprocessing: |
|
| 120 | - // In order to reduce code duplication, whenever some code needs |
|
| 121 | - // to make HTML changes in order to make things "correct", the |
|
| 122 | - // new HTML gets sent through the purifier, regardless of its |
|
| 123 | - // status. This means that if we add a start token, because it |
|
| 124 | - // was totally necessary, we don't have to update nesting; we just |
|
| 125 | - // punt ($reprocess = true; continue;) and it does that for us. |
|
| 126 | - |
|
| 127 | - // isset is in loop because $tokens size changes during loop exec |
|
| 128 | - for ( |
|
| 129 | - $t = 0; |
|
| 130 | - $t == 0 || isset($tokens[$t - 1]); |
|
| 131 | - // only increment if we don't need to reprocess |
|
| 132 | - $reprocess ? $reprocess = false : $t++ |
|
| 133 | - ) { |
|
| 134 | - |
|
| 135 | - // check for a rewind |
|
| 136 | - if (is_int($i) && $i >= 0) { |
|
| 137 | - // possibility: disable rewinding if the current token has a |
|
| 138 | - // rewind set on it already. This would offer protection from |
|
| 139 | - // infinite loop, but might hinder some advanced rewinding. |
|
| 140 | - $rewind_to = $this->injectors[$i]->getRewind(); |
|
| 141 | - if (is_int($rewind_to) && $rewind_to < $t) { |
|
| 142 | - if ($rewind_to < 0) $rewind_to = 0; |
|
| 143 | - while ($t > $rewind_to) { |
|
| 144 | - $t--; |
|
| 145 | - $prev = $tokens[$t]; |
|
| 146 | - // indicate that other injectors should not process this token, |
|
| 147 | - // but we need to reprocess it |
|
| 148 | - unset($prev->skip[$i]); |
|
| 149 | - $prev->rewind = $i; |
|
| 150 | - if ($prev instanceof HTMLPurifier_Token_Start) array_pop($this->stack); |
|
| 151 | - elseif ($prev instanceof HTMLPurifier_Token_End) $this->stack[] = $prev->start; |
|
| 152 | - } |
|
| 153 | - } |
|
| 154 | - $i = false; |
|
| 155 | - } |
|
| 156 | - |
|
| 157 | - // handle case of document end |
|
| 158 | - if (!isset($tokens[$t])) { |
|
| 159 | - // kill processing if stack is empty |
|
| 160 | - if (empty($this->stack)) break; |
|
| 161 | - |
|
| 162 | - // peek |
|
| 163 | - $top_nesting = array_pop($this->stack); |
|
| 164 | - $this->stack[] = $top_nesting; |
|
| 165 | - |
|
| 166 | - // send error [TagClosedSuppress] |
|
| 167 | - if ($e && !isset($top_nesting->armor['MakeWellFormed_TagClosedError'])) { |
|
| 168 | - $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag closed by document end', $top_nesting); |
|
| 169 | - } |
|
| 170 | - |
|
| 171 | - // append, don't splice, since this is the end |
|
| 172 | - $tokens[] = new HTMLPurifier_Token_End($top_nesting->name); |
|
| 173 | - |
|
| 174 | - // punt! |
|
| 175 | - $reprocess = true; |
|
| 176 | - continue; |
|
| 177 | - } |
|
| 178 | - |
|
| 179 | - $token = $tokens[$t]; |
|
| 180 | - |
|
| 181 | - //echo '<br>'; printTokens($tokens, $t); printTokens($this->stack); |
|
| 182 | - //flush(); |
|
| 183 | - |
|
| 184 | - // quick-check: if it's not a tag, no need to process |
|
| 185 | - if (empty($token->is_tag)) { |
|
| 186 | - if ($token instanceof HTMLPurifier_Token_Text) { |
|
| 187 | - foreach ($this->injectors as $i => $injector) { |
|
| 188 | - if (isset($token->skip[$i])) continue; |
|
| 189 | - if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 190 | - $injector->handleText($token); |
|
| 191 | - $this->processToken($token, $i); |
|
| 192 | - $reprocess = true; |
|
| 193 | - break; |
|
| 194 | - } |
|
| 195 | - } |
|
| 196 | - // another possibility is a comment |
|
| 197 | - continue; |
|
| 198 | - } |
|
| 199 | - |
|
| 200 | - if (isset($definition->info[$token->name])) { |
|
| 201 | - $type = $definition->info[$token->name]->child->type; |
|
| 202 | - } else { |
|
| 203 | - $type = false; // Type is unknown, treat accordingly |
|
| 204 | - } |
|
| 205 | - |
|
| 206 | - // quick tag checks: anything that's *not* an end tag |
|
| 207 | - $ok = false; |
|
| 208 | - if ($type === 'empty' && $token instanceof HTMLPurifier_Token_Start) { |
|
| 209 | - // claims to be a start tag but is empty |
|
| 210 | - $token = new HTMLPurifier_Token_Empty($token->name, $token->attr, $token->line, $token->col, $token->armor); |
|
| 211 | - $ok = true; |
|
| 212 | - } elseif ($type && $type !== 'empty' && $token instanceof HTMLPurifier_Token_Empty) { |
|
| 213 | - // claims to be empty but really is a start tag |
|
| 214 | - $this->swap(new HTMLPurifier_Token_End($token->name)); |
|
| 215 | - $this->insertBefore(new HTMLPurifier_Token_Start($token->name, $token->attr, $token->line, $token->col, $token->armor)); |
|
| 216 | - // punt (since we had to modify the input stream in a non-trivial way) |
|
| 217 | - $reprocess = true; |
|
| 218 | - continue; |
|
| 219 | - } elseif ($token instanceof HTMLPurifier_Token_Empty) { |
|
| 220 | - // real empty token |
|
| 221 | - $ok = true; |
|
| 222 | - } elseif ($token instanceof HTMLPurifier_Token_Start) { |
|
| 223 | - // start tag |
|
| 224 | - |
|
| 225 | - // ...unless they also have to close their parent |
|
| 226 | - if (!empty($this->stack)) { |
|
| 227 | - |
|
| 228 | - // Performance note: you might think that it's rather |
|
| 229 | - // inefficient, recalculating the autoclose information |
|
| 230 | - // for every tag that a token closes (since when we |
|
| 231 | - // do an autoclose, we push a new token into the |
|
| 232 | - // stream and then /process/ that, before |
|
| 233 | - // re-processing this token.) But this is |
|
| 234 | - // necessary, because an injector can make an |
|
| 235 | - // arbitrary transformations to the autoclosing |
|
| 236 | - // tokens we introduce, so things may have changed |
|
| 237 | - // in the meantime. Also, doing the inefficient thing is |
|
| 238 | - // "easy" to reason about (for certain perverse definitions |
|
| 239 | - // of "easy") |
|
| 240 | - |
|
| 241 | - $parent = array_pop($this->stack); |
|
| 242 | - $this->stack[] = $parent; |
|
| 243 | - |
|
| 244 | - if (isset($definition->info[$parent->name])) { |
|
| 245 | - $elements = $definition->info[$parent->name]->child->getAllowedElements($config); |
|
| 246 | - $autoclose = !isset($elements[$token->name]); |
|
| 247 | - } else { |
|
| 248 | - $autoclose = false; |
|
| 249 | - } |
|
| 250 | - |
|
| 251 | - if ($autoclose && $definition->info[$token->name]->wrap) { |
|
| 252 | - // Check if an element can be wrapped by another |
|
| 253 | - // element to make it valid in a context (for |
|
| 254 | - // example, <ul><ul> needs a <li> in between) |
|
| 255 | - $wrapname = $definition->info[$token->name]->wrap; |
|
| 256 | - $wrapdef = $definition->info[$wrapname]; |
|
| 257 | - $elements = $wrapdef->child->getAllowedElements($config); |
|
| 258 | - $parent_elements = $definition->info[$parent->name]->child->getAllowedElements($config); |
|
| 259 | - if (isset($elements[$token->name]) && isset($parent_elements[$wrapname])) { |
|
| 260 | - $newtoken = new HTMLPurifier_Token_Start($wrapname); |
|
| 261 | - $this->insertBefore($newtoken); |
|
| 262 | - $reprocess = true; |
|
| 263 | - continue; |
|
| 264 | - } |
|
| 265 | - } |
|
| 266 | - |
|
| 267 | - $carryover = false; |
|
| 268 | - if ($autoclose && $definition->info[$parent->name]->formatting) { |
|
| 269 | - $carryover = true; |
|
| 270 | - } |
|
| 271 | - |
|
| 272 | - if ($autoclose) { |
|
| 273 | - // check if this autoclose is doomed to fail |
|
| 274 | - // (this rechecks $parent, which his harmless) |
|
| 275 | - $autoclose_ok = isset($global_parent_allowed_elements[$token->name]); |
|
| 276 | - if (!$autoclose_ok) { |
|
| 277 | - foreach ($this->stack as $ancestor) { |
|
| 278 | - $elements = $definition->info[$ancestor->name]->child->getAllowedElements($config); |
|
| 279 | - if (isset($elements[$token->name])) { |
|
| 280 | - $autoclose_ok = true; |
|
| 281 | - break; |
|
| 282 | - } |
|
| 283 | - if ($definition->info[$token->name]->wrap) { |
|
| 284 | - $wrapname = $definition->info[$token->name]->wrap; |
|
| 285 | - $wrapdef = $definition->info[$wrapname]; |
|
| 286 | - $wrap_elements = $wrapdef->child->getAllowedElements($config); |
|
| 287 | - if (isset($wrap_elements[$token->name]) && isset($elements[$wrapname])) { |
|
| 288 | - $autoclose_ok = true; |
|
| 289 | - break; |
|
| 290 | - } |
|
| 291 | - } |
|
| 292 | - } |
|
| 293 | - } |
|
| 294 | - if ($autoclose_ok) { |
|
| 295 | - // errors need to be updated |
|
| 296 | - $new_token = new HTMLPurifier_Token_End($parent->name); |
|
| 297 | - $new_token->start = $parent; |
|
| 298 | - if ($carryover) { |
|
| 299 | - $element = clone $parent; |
|
| 300 | - // [TagClosedAuto] |
|
| 301 | - $element->armor['MakeWellFormed_TagClosedError'] = true; |
|
| 302 | - $element->carryover = true; |
|
| 303 | - $this->processToken(array($new_token, $token, $element)); |
|
| 304 | - } else { |
|
| 305 | - $this->insertBefore($new_token); |
|
| 306 | - } |
|
| 307 | - // [TagClosedSuppress] |
|
| 308 | - if ($e && !isset($parent->armor['MakeWellFormed_TagClosedError'])) { |
|
| 309 | - if (!$carryover) { |
|
| 310 | - $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag auto closed', $parent); |
|
| 311 | - } else { |
|
| 312 | - $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag carryover', $parent); |
|
| 313 | - } |
|
| 314 | - } |
|
| 315 | - } else { |
|
| 316 | - $this->remove(); |
|
| 317 | - } |
|
| 318 | - $reprocess = true; |
|
| 319 | - continue; |
|
| 320 | - } |
|
| 321 | - |
|
| 322 | - } |
|
| 323 | - $ok = true; |
|
| 324 | - } |
|
| 325 | - |
|
| 326 | - if ($ok) { |
|
| 327 | - foreach ($this->injectors as $i => $injector) { |
|
| 328 | - if (isset($token->skip[$i])) continue; |
|
| 329 | - if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 330 | - $injector->handleElement($token); |
|
| 331 | - $this->processToken($token, $i); |
|
| 332 | - $reprocess = true; |
|
| 333 | - break; |
|
| 334 | - } |
|
| 335 | - if (!$reprocess) { |
|
| 336 | - // ah, nothing interesting happened; do normal processing |
|
| 337 | - $this->swap($token); |
|
| 338 | - if ($token instanceof HTMLPurifier_Token_Start) { |
|
| 339 | - $this->stack[] = $token; |
|
| 340 | - } elseif ($token instanceof HTMLPurifier_Token_End) { |
|
| 341 | - throw new HTMLPurifier_Exception('Improper handling of end tag in start code; possible error in MakeWellFormed'); |
|
| 342 | - } |
|
| 343 | - } |
|
| 344 | - continue; |
|
| 345 | - } |
|
| 346 | - |
|
| 347 | - // sanity check: we should be dealing with a closing tag |
|
| 348 | - if (!$token instanceof HTMLPurifier_Token_End) { |
|
| 349 | - throw new HTMLPurifier_Exception('Unaccounted for tag token in input stream, bug in HTML Purifier'); |
|
| 350 | - } |
|
| 351 | - |
|
| 352 | - // make sure that we have something open |
|
| 353 | - if (empty($this->stack)) { |
|
| 354 | - if ($escape_invalid_tags) { |
|
| 355 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag to text'); |
|
| 356 | - $this->swap(new HTMLPurifier_Token_Text( |
|
| 357 | - $generator->generateFromToken($token) |
|
| 358 | - )); |
|
| 359 | - } else { |
|
| 360 | - $this->remove(); |
|
| 361 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag removed'); |
|
| 362 | - } |
|
| 363 | - $reprocess = true; |
|
| 364 | - continue; |
|
| 365 | - } |
|
| 366 | - |
|
| 367 | - // first, check for the simplest case: everything closes neatly. |
|
| 368 | - // Eventually, everything passes through here; if there are problems |
|
| 369 | - // we modify the input stream accordingly and then punt, so that |
|
| 370 | - // the tokens get processed again. |
|
| 371 | - $current_parent = array_pop($this->stack); |
|
| 372 | - if ($current_parent->name == $token->name) { |
|
| 373 | - $token->start = $current_parent; |
|
| 374 | - foreach ($this->injectors as $i => $injector) { |
|
| 375 | - if (isset($token->skip[$i])) continue; |
|
| 376 | - if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 377 | - $injector->handleEnd($token); |
|
| 378 | - $this->processToken($token, $i); |
|
| 379 | - $this->stack[] = $current_parent; |
|
| 380 | - $reprocess = true; |
|
| 381 | - break; |
|
| 382 | - } |
|
| 383 | - continue; |
|
| 384 | - } |
|
| 385 | - |
|
| 386 | - // okay, so we're trying to close the wrong tag |
|
| 387 | - |
|
| 388 | - // undo the pop previous pop |
|
| 389 | - $this->stack[] = $current_parent; |
|
| 390 | - |
|
| 391 | - // scroll back the entire nest, trying to find our tag. |
|
| 392 | - // (feature could be to specify how far you'd like to go) |
|
| 393 | - $size = count($this->stack); |
|
| 394 | - // -2 because -1 is the last element, but we already checked that |
|
| 395 | - $skipped_tags = false; |
|
| 396 | - for ($j = $size - 2; $j >= 0; $j--) { |
|
| 397 | - if ($this->stack[$j]->name == $token->name) { |
|
| 398 | - $skipped_tags = array_slice($this->stack, $j); |
|
| 399 | - break; |
|
| 400 | - } |
|
| 401 | - } |
|
| 402 | - |
|
| 403 | - // we didn't find the tag, so remove |
|
| 404 | - if ($skipped_tags === false) { |
|
| 405 | - if ($escape_invalid_tags) { |
|
| 406 | - $this->swap(new HTMLPurifier_Token_Text( |
|
| 407 | - $generator->generateFromToken($token) |
|
| 408 | - )); |
|
| 409 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag to text'); |
|
| 410 | - } else { |
|
| 411 | - $this->remove(); |
|
| 412 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag removed'); |
|
| 413 | - } |
|
| 414 | - $reprocess = true; |
|
| 415 | - continue; |
|
| 416 | - } |
|
| 417 | - |
|
| 418 | - // do errors, in REVERSE $j order: a,b,c with </a></b></c> |
|
| 419 | - $c = count($skipped_tags); |
|
| 420 | - if ($e) { |
|
| 421 | - for ($j = $c - 1; $j > 0; $j--) { |
|
| 422 | - // notice we exclude $j == 0, i.e. the current ending tag, from |
|
| 423 | - // the errors... [TagClosedSuppress] |
|
| 424 | - if (!isset($skipped_tags[$j]->armor['MakeWellFormed_TagClosedError'])) { |
|
| 425 | - $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag closed by element end', $skipped_tags[$j]); |
|
| 426 | - } |
|
| 427 | - } |
|
| 428 | - } |
|
| 429 | - |
|
| 430 | - // insert tags, in FORWARD $j order: c,b,a with </a></b></c> |
|
| 431 | - $replace = array($token); |
|
| 432 | - for ($j = 1; $j < $c; $j++) { |
|
| 433 | - // ...as well as from the insertions |
|
| 434 | - $new_token = new HTMLPurifier_Token_End($skipped_tags[$j]->name); |
|
| 435 | - $new_token->start = $skipped_tags[$j]; |
|
| 436 | - array_unshift($replace, $new_token); |
|
| 437 | - if (isset($definition->info[$new_token->name]) && $definition->info[$new_token->name]->formatting) { |
|
| 438 | - // [TagClosedAuto] |
|
| 439 | - $element = clone $skipped_tags[$j]; |
|
| 440 | - $element->carryover = true; |
|
| 441 | - $element->armor['MakeWellFormed_TagClosedError'] = true; |
|
| 442 | - $replace[] = $element; |
|
| 443 | - } |
|
| 444 | - } |
|
| 445 | - $this->processToken($replace); |
|
| 446 | - $reprocess = true; |
|
| 447 | - continue; |
|
| 448 | - } |
|
| 449 | - |
|
| 450 | - $context->destroy('CurrentNesting'); |
|
| 451 | - $context->destroy('InputTokens'); |
|
| 452 | - $context->destroy('InputIndex'); |
|
| 453 | - $context->destroy('CurrentToken'); |
|
| 454 | - |
|
| 455 | - unset($this->injectors, $this->stack, $this->tokens, $this->t); |
|
| 456 | - return $tokens; |
|
| 457 | - } |
|
| 458 | - |
|
| 459 | - /** |
|
| 460 | - * Processes arbitrary token values for complicated substitution patterns. |
|
| 461 | - * In general: |
|
| 462 | - * |
|
| 463 | - * If $token is an array, it is a list of tokens to substitute for the |
|
| 464 | - * current token. These tokens then get individually processed. If there |
|
| 465 | - * is a leading integer in the list, that integer determines how many |
|
| 466 | - * tokens from the stream should be removed. |
|
| 467 | - * |
|
| 468 | - * If $token is a regular token, it is swapped with the current token. |
|
| 469 | - * |
|
| 470 | - * If $token is false, the current token is deleted. |
|
| 471 | - * |
|
| 472 | - * If $token is an integer, that number of tokens (with the first token |
|
| 473 | - * being the current one) will be deleted. |
|
| 474 | - * |
|
| 475 | - * @param $token Token substitution value |
|
| 476 | - * @param $injector Injector that performed the substitution; default is if |
|
| 477 | - * this is not an injector related operation. |
|
| 478 | - */ |
|
| 479 | - protected function processToken($token, $injector = -1) { |
|
| 480 | - |
|
| 481 | - // normalize forms of token |
|
| 482 | - if (is_object($token)) $token = array(1, $token); |
|
| 483 | - if (is_int($token)) $token = array($token); |
|
| 484 | - if ($token === false) $token = array(1); |
|
| 485 | - if (!is_array($token)) throw new HTMLPurifier_Exception('Invalid token type from injector'); |
|
| 486 | - if (!is_int($token[0])) array_unshift($token, 1); |
|
| 487 | - if ($token[0] === 0) throw new HTMLPurifier_Exception('Deleting zero tokens is not valid'); |
|
| 488 | - |
|
| 489 | - // $token is now an array with the following form: |
|
| 490 | - // array(number nodes to delete, new node 1, new node 2, ...) |
|
| 491 | - |
|
| 492 | - $delete = array_shift($token); |
|
| 493 | - $old = array_splice($this->tokens, $this->t, $delete, $token); |
|
| 494 | - |
|
| 495 | - if ($injector > -1) { |
|
| 496 | - // determine appropriate skips |
|
| 497 | - $oldskip = isset($old[0]) ? $old[0]->skip : array(); |
|
| 498 | - foreach ($token as $object) { |
|
| 499 | - $object->skip = $oldskip; |
|
| 500 | - $object->skip[$injector] = true; |
|
| 501 | - } |
|
| 502 | - } |
|
| 503 | - |
|
| 504 | - } |
|
| 505 | - |
|
| 506 | - /** |
|
| 507 | - * Inserts a token before the current token. Cursor now points to |
|
| 508 | - * this token. You must reprocess after this. |
|
| 509 | - */ |
|
| 510 | - private function insertBefore($token) { |
|
| 511 | - array_splice($this->tokens, $this->t, 0, array($token)); |
|
| 512 | - } |
|
| 513 | - |
|
| 514 | - /** |
|
| 515 | - * Removes current token. Cursor now points to new token occupying previously |
|
| 516 | - * occupied space. You must reprocess after this. |
|
| 517 | - */ |
|
| 518 | - private function remove() { |
|
| 519 | - array_splice($this->tokens, $this->t, 1); |
|
| 520 | - } |
|
| 521 | - |
|
| 522 | - /** |
|
| 523 | - * Swap current token with new token. Cursor points to new token (no |
|
| 524 | - * change). You must reprocess after this. |
|
| 525 | - */ |
|
| 526 | - private function swap($token) { |
|
| 527 | - $this->tokens[$this->t] = $token; |
|
| 528 | - } |
|
| 17 | + /** |
|
| 18 | + * Array stream of tokens being processed. |
|
| 19 | + */ |
|
| 20 | + protected $tokens; |
|
| 21 | + |
|
| 22 | + /** |
|
| 23 | + * Current index in $tokens. |
|
| 24 | + */ |
|
| 25 | + protected $t; |
|
| 26 | + |
|
| 27 | + /** |
|
| 28 | + * Current nesting of elements. |
|
| 29 | + */ |
|
| 30 | + protected $stack; |
|
| 31 | + |
|
| 32 | + /** |
|
| 33 | + * Injectors active in this stream processing. |
|
| 34 | + */ |
|
| 35 | + protected $injectors; |
|
| 36 | + |
|
| 37 | + /** |
|
| 38 | + * Current instance of HTMLPurifier_Config. |
|
| 39 | + */ |
|
| 40 | + protected $config; |
|
| 41 | + |
|
| 42 | + /** |
|
| 43 | + * Current instance of HTMLPurifier_Context. |
|
| 44 | + */ |
|
| 45 | + protected $context; |
|
| 46 | + |
|
| 47 | + public function execute($tokens, $config, $context) { |
|
| 48 | + |
|
| 49 | + $definition = $config->getHTMLDefinition(); |
|
| 50 | + |
|
| 51 | + // local variables |
|
| 52 | + $generator = new HTMLPurifier_Generator($config, $context); |
|
| 53 | + $escape_invalid_tags = $config->get('Core.EscapeInvalidTags'); |
|
| 54 | + // used for autoclose early abortion |
|
| 55 | + $global_parent_allowed_elements = array(); |
|
| 56 | + if (isset($definition->info[$definition->info_parent])) { |
|
| 57 | + // may be unset under testing circumstances |
|
| 58 | + $global_parent_allowed_elements = $definition->info[$definition->info_parent]->child->getAllowedElements($config); |
|
| 59 | + } |
|
| 60 | + $e = $context->get('ErrorCollector', true); |
|
| 61 | + $t = false; // token index |
|
| 62 | + $i = false; // injector index |
|
| 63 | + $token = false; // the current token |
|
| 64 | + $reprocess = false; // whether or not to reprocess the same token |
|
| 65 | + $stack = array(); |
|
| 66 | + |
|
| 67 | + // member variables |
|
| 68 | + $this->stack =& $stack; |
|
| 69 | + $this->t =& $t; |
|
| 70 | + $this->tokens =& $tokens; |
|
| 71 | + $this->config = $config; |
|
| 72 | + $this->context = $context; |
|
| 73 | + |
|
| 74 | + // context variables |
|
| 75 | + $context->register('CurrentNesting', $stack); |
|
| 76 | + $context->register('InputIndex', $t); |
|
| 77 | + $context->register('InputTokens', $tokens); |
|
| 78 | + $context->register('CurrentToken', $token); |
|
| 79 | + |
|
| 80 | + // -- begin INJECTOR -- |
|
| 81 | + |
|
| 82 | + $this->injectors = array(); |
|
| 83 | + |
|
| 84 | + $injectors = $config->getBatch('AutoFormat'); |
|
| 85 | + $def_injectors = $definition->info_injector; |
|
| 86 | + $custom_injectors = $injectors['Custom']; |
|
| 87 | + unset($injectors['Custom']); // special case |
|
| 88 | + foreach ($injectors as $injector => $b) { |
|
| 89 | + // XXX: Fix with a legitimate lookup table of enabled filters |
|
| 90 | + if (strpos($injector, '.') !== false) continue; |
|
| 91 | + $injector = "HTMLPurifier_Injector_$injector"; |
|
| 92 | + if (!$b) continue; |
|
| 93 | + $this->injectors[] = new $injector; |
|
| 94 | + } |
|
| 95 | + foreach ($def_injectors as $injector) { |
|
| 96 | + // assumed to be objects |
|
| 97 | + $this->injectors[] = $injector; |
|
| 98 | + } |
|
| 99 | + foreach ($custom_injectors as $injector) { |
|
| 100 | + if (!$injector) continue; |
|
| 101 | + if (is_string($injector)) { |
|
| 102 | + $injector = "HTMLPurifier_Injector_$injector"; |
|
| 103 | + $injector = new $injector; |
|
| 104 | + } |
|
| 105 | + $this->injectors[] = $injector; |
|
| 106 | + } |
|
| 107 | + |
|
| 108 | + // give the injectors references to the definition and context |
|
| 109 | + // variables for performance reasons |
|
| 110 | + foreach ($this->injectors as $ix => $injector) { |
|
| 111 | + $error = $injector->prepare($config, $context); |
|
| 112 | + if (!$error) continue; |
|
| 113 | + array_splice($this->injectors, $ix, 1); // rm the injector |
|
| 114 | + trigger_error("Cannot enable {$injector->name} injector because $error is not allowed", E_USER_WARNING); |
|
| 115 | + } |
|
| 116 | + |
|
| 117 | + // -- end INJECTOR -- |
|
| 118 | + |
|
| 119 | + // a note on reprocessing: |
|
| 120 | + // In order to reduce code duplication, whenever some code needs |
|
| 121 | + // to make HTML changes in order to make things "correct", the |
|
| 122 | + // new HTML gets sent through the purifier, regardless of its |
|
| 123 | + // status. This means that if we add a start token, because it |
|
| 124 | + // was totally necessary, we don't have to update nesting; we just |
|
| 125 | + // punt ($reprocess = true; continue;) and it does that for us. |
|
| 126 | + |
|
| 127 | + // isset is in loop because $tokens size changes during loop exec |
|
| 128 | + for ( |
|
| 129 | + $t = 0; |
|
| 130 | + $t == 0 || isset($tokens[$t - 1]); |
|
| 131 | + // only increment if we don't need to reprocess |
|
| 132 | + $reprocess ? $reprocess = false : $t++ |
|
| 133 | + ) { |
|
| 134 | + |
|
| 135 | + // check for a rewind |
|
| 136 | + if (is_int($i) && $i >= 0) { |
|
| 137 | + // possibility: disable rewinding if the current token has a |
|
| 138 | + // rewind set on it already. This would offer protection from |
|
| 139 | + // infinite loop, but might hinder some advanced rewinding. |
|
| 140 | + $rewind_to = $this->injectors[$i]->getRewind(); |
|
| 141 | + if (is_int($rewind_to) && $rewind_to < $t) { |
|
| 142 | + if ($rewind_to < 0) $rewind_to = 0; |
|
| 143 | + while ($t > $rewind_to) { |
|
| 144 | + $t--; |
|
| 145 | + $prev = $tokens[$t]; |
|
| 146 | + // indicate that other injectors should not process this token, |
|
| 147 | + // but we need to reprocess it |
|
| 148 | + unset($prev->skip[$i]); |
|
| 149 | + $prev->rewind = $i; |
|
| 150 | + if ($prev instanceof HTMLPurifier_Token_Start) array_pop($this->stack); |
|
| 151 | + elseif ($prev instanceof HTMLPurifier_Token_End) $this->stack[] = $prev->start; |
|
| 152 | + } |
|
| 153 | + } |
|
| 154 | + $i = false; |
|
| 155 | + } |
|
| 156 | + |
|
| 157 | + // handle case of document end |
|
| 158 | + if (!isset($tokens[$t])) { |
|
| 159 | + // kill processing if stack is empty |
|
| 160 | + if (empty($this->stack)) break; |
|
| 161 | + |
|
| 162 | + // peek |
|
| 163 | + $top_nesting = array_pop($this->stack); |
|
| 164 | + $this->stack[] = $top_nesting; |
|
| 165 | + |
|
| 166 | + // send error [TagClosedSuppress] |
|
| 167 | + if ($e && !isset($top_nesting->armor['MakeWellFormed_TagClosedError'])) { |
|
| 168 | + $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag closed by document end', $top_nesting); |
|
| 169 | + } |
|
| 170 | + |
|
| 171 | + // append, don't splice, since this is the end |
|
| 172 | + $tokens[] = new HTMLPurifier_Token_End($top_nesting->name); |
|
| 173 | + |
|
| 174 | + // punt! |
|
| 175 | + $reprocess = true; |
|
| 176 | + continue; |
|
| 177 | + } |
|
| 178 | + |
|
| 179 | + $token = $tokens[$t]; |
|
| 180 | + |
|
| 181 | + //echo '<br>'; printTokens($tokens, $t); printTokens($this->stack); |
|
| 182 | + //flush(); |
|
| 183 | + |
|
| 184 | + // quick-check: if it's not a tag, no need to process |
|
| 185 | + if (empty($token->is_tag)) { |
|
| 186 | + if ($token instanceof HTMLPurifier_Token_Text) { |
|
| 187 | + foreach ($this->injectors as $i => $injector) { |
|
| 188 | + if (isset($token->skip[$i])) continue; |
|
| 189 | + if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 190 | + $injector->handleText($token); |
|
| 191 | + $this->processToken($token, $i); |
|
| 192 | + $reprocess = true; |
|
| 193 | + break; |
|
| 194 | + } |
|
| 195 | + } |
|
| 196 | + // another possibility is a comment |
|
| 197 | + continue; |
|
| 198 | + } |
|
| 199 | + |
|
| 200 | + if (isset($definition->info[$token->name])) { |
|
| 201 | + $type = $definition->info[$token->name]->child->type; |
|
| 202 | + } else { |
|
| 203 | + $type = false; // Type is unknown, treat accordingly |
|
| 204 | + } |
|
| 205 | + |
|
| 206 | + // quick tag checks: anything that's *not* an end tag |
|
| 207 | + $ok = false; |
|
| 208 | + if ($type === 'empty' && $token instanceof HTMLPurifier_Token_Start) { |
|
| 209 | + // claims to be a start tag but is empty |
|
| 210 | + $token = new HTMLPurifier_Token_Empty($token->name, $token->attr, $token->line, $token->col, $token->armor); |
|
| 211 | + $ok = true; |
|
| 212 | + } elseif ($type && $type !== 'empty' && $token instanceof HTMLPurifier_Token_Empty) { |
|
| 213 | + // claims to be empty but really is a start tag |
|
| 214 | + $this->swap(new HTMLPurifier_Token_End($token->name)); |
|
| 215 | + $this->insertBefore(new HTMLPurifier_Token_Start($token->name, $token->attr, $token->line, $token->col, $token->armor)); |
|
| 216 | + // punt (since we had to modify the input stream in a non-trivial way) |
|
| 217 | + $reprocess = true; |
|
| 218 | + continue; |
|
| 219 | + } elseif ($token instanceof HTMLPurifier_Token_Empty) { |
|
| 220 | + // real empty token |
|
| 221 | + $ok = true; |
|
| 222 | + } elseif ($token instanceof HTMLPurifier_Token_Start) { |
|
| 223 | + // start tag |
|
| 224 | + |
|
| 225 | + // ...unless they also have to close their parent |
|
| 226 | + if (!empty($this->stack)) { |
|
| 227 | + |
|
| 228 | + // Performance note: you might think that it's rather |
|
| 229 | + // inefficient, recalculating the autoclose information |
|
| 230 | + // for every tag that a token closes (since when we |
|
| 231 | + // do an autoclose, we push a new token into the |
|
| 232 | + // stream and then /process/ that, before |
|
| 233 | + // re-processing this token.) But this is |
|
| 234 | + // necessary, because an injector can make an |
|
| 235 | + // arbitrary transformations to the autoclosing |
|
| 236 | + // tokens we introduce, so things may have changed |
|
| 237 | + // in the meantime. Also, doing the inefficient thing is |
|
| 238 | + // "easy" to reason about (for certain perverse definitions |
|
| 239 | + // of "easy") |
|
| 240 | + |
|
| 241 | + $parent = array_pop($this->stack); |
|
| 242 | + $this->stack[] = $parent; |
|
| 243 | + |
|
| 244 | + if (isset($definition->info[$parent->name])) { |
|
| 245 | + $elements = $definition->info[$parent->name]->child->getAllowedElements($config); |
|
| 246 | + $autoclose = !isset($elements[$token->name]); |
|
| 247 | + } else { |
|
| 248 | + $autoclose = false; |
|
| 249 | + } |
|
| 250 | + |
|
| 251 | + if ($autoclose && $definition->info[$token->name]->wrap) { |
|
| 252 | + // Check if an element can be wrapped by another |
|
| 253 | + // element to make it valid in a context (for |
|
| 254 | + // example, <ul><ul> needs a <li> in between) |
|
| 255 | + $wrapname = $definition->info[$token->name]->wrap; |
|
| 256 | + $wrapdef = $definition->info[$wrapname]; |
|
| 257 | + $elements = $wrapdef->child->getAllowedElements($config); |
|
| 258 | + $parent_elements = $definition->info[$parent->name]->child->getAllowedElements($config); |
|
| 259 | + if (isset($elements[$token->name]) && isset($parent_elements[$wrapname])) { |
|
| 260 | + $newtoken = new HTMLPurifier_Token_Start($wrapname); |
|
| 261 | + $this->insertBefore($newtoken); |
|
| 262 | + $reprocess = true; |
|
| 263 | + continue; |
|
| 264 | + } |
|
| 265 | + } |
|
| 266 | + |
|
| 267 | + $carryover = false; |
|
| 268 | + if ($autoclose && $definition->info[$parent->name]->formatting) { |
|
| 269 | + $carryover = true; |
|
| 270 | + } |
|
| 271 | + |
|
| 272 | + if ($autoclose) { |
|
| 273 | + // check if this autoclose is doomed to fail |
|
| 274 | + // (this rechecks $parent, which his harmless) |
|
| 275 | + $autoclose_ok = isset($global_parent_allowed_elements[$token->name]); |
|
| 276 | + if (!$autoclose_ok) { |
|
| 277 | + foreach ($this->stack as $ancestor) { |
|
| 278 | + $elements = $definition->info[$ancestor->name]->child->getAllowedElements($config); |
|
| 279 | + if (isset($elements[$token->name])) { |
|
| 280 | + $autoclose_ok = true; |
|
| 281 | + break; |
|
| 282 | + } |
|
| 283 | + if ($definition->info[$token->name]->wrap) { |
|
| 284 | + $wrapname = $definition->info[$token->name]->wrap; |
|
| 285 | + $wrapdef = $definition->info[$wrapname]; |
|
| 286 | + $wrap_elements = $wrapdef->child->getAllowedElements($config); |
|
| 287 | + if (isset($wrap_elements[$token->name]) && isset($elements[$wrapname])) { |
|
| 288 | + $autoclose_ok = true; |
|
| 289 | + break; |
|
| 290 | + } |
|
| 291 | + } |
|
| 292 | + } |
|
| 293 | + } |
|
| 294 | + if ($autoclose_ok) { |
|
| 295 | + // errors need to be updated |
|
| 296 | + $new_token = new HTMLPurifier_Token_End($parent->name); |
|
| 297 | + $new_token->start = $parent; |
|
| 298 | + if ($carryover) { |
|
| 299 | + $element = clone $parent; |
|
| 300 | + // [TagClosedAuto] |
|
| 301 | + $element->armor['MakeWellFormed_TagClosedError'] = true; |
|
| 302 | + $element->carryover = true; |
|
| 303 | + $this->processToken(array($new_token, $token, $element)); |
|
| 304 | + } else { |
|
| 305 | + $this->insertBefore($new_token); |
|
| 306 | + } |
|
| 307 | + // [TagClosedSuppress] |
|
| 308 | + if ($e && !isset($parent->armor['MakeWellFormed_TagClosedError'])) { |
|
| 309 | + if (!$carryover) { |
|
| 310 | + $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag auto closed', $parent); |
|
| 311 | + } else { |
|
| 312 | + $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag carryover', $parent); |
|
| 313 | + } |
|
| 314 | + } |
|
| 315 | + } else { |
|
| 316 | + $this->remove(); |
|
| 317 | + } |
|
| 318 | + $reprocess = true; |
|
| 319 | + continue; |
|
| 320 | + } |
|
| 321 | + |
|
| 322 | + } |
|
| 323 | + $ok = true; |
|
| 324 | + } |
|
| 325 | + |
|
| 326 | + if ($ok) { |
|
| 327 | + foreach ($this->injectors as $i => $injector) { |
|
| 328 | + if (isset($token->skip[$i])) continue; |
|
| 329 | + if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 330 | + $injector->handleElement($token); |
|
| 331 | + $this->processToken($token, $i); |
|
| 332 | + $reprocess = true; |
|
| 333 | + break; |
|
| 334 | + } |
|
| 335 | + if (!$reprocess) { |
|
| 336 | + // ah, nothing interesting happened; do normal processing |
|
| 337 | + $this->swap($token); |
|
| 338 | + if ($token instanceof HTMLPurifier_Token_Start) { |
|
| 339 | + $this->stack[] = $token; |
|
| 340 | + } elseif ($token instanceof HTMLPurifier_Token_End) { |
|
| 341 | + throw new HTMLPurifier_Exception('Improper handling of end tag in start code; possible error in MakeWellFormed'); |
|
| 342 | + } |
|
| 343 | + } |
|
| 344 | + continue; |
|
| 345 | + } |
|
| 346 | + |
|
| 347 | + // sanity check: we should be dealing with a closing tag |
|
| 348 | + if (!$token instanceof HTMLPurifier_Token_End) { |
|
| 349 | + throw new HTMLPurifier_Exception('Unaccounted for tag token in input stream, bug in HTML Purifier'); |
|
| 350 | + } |
|
| 351 | + |
|
| 352 | + // make sure that we have something open |
|
| 353 | + if (empty($this->stack)) { |
|
| 354 | + if ($escape_invalid_tags) { |
|
| 355 | + if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag to text'); |
|
| 356 | + $this->swap(new HTMLPurifier_Token_Text( |
|
| 357 | + $generator->generateFromToken($token) |
|
| 358 | + )); |
|
| 359 | + } else { |
|
| 360 | + $this->remove(); |
|
| 361 | + if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag removed'); |
|
| 362 | + } |
|
| 363 | + $reprocess = true; |
|
| 364 | + continue; |
|
| 365 | + } |
|
| 366 | + |
|
| 367 | + // first, check for the simplest case: everything closes neatly. |
|
| 368 | + // Eventually, everything passes through here; if there are problems |
|
| 369 | + // we modify the input stream accordingly and then punt, so that |
|
| 370 | + // the tokens get processed again. |
|
| 371 | + $current_parent = array_pop($this->stack); |
|
| 372 | + if ($current_parent->name == $token->name) { |
|
| 373 | + $token->start = $current_parent; |
|
| 374 | + foreach ($this->injectors as $i => $injector) { |
|
| 375 | + if (isset($token->skip[$i])) continue; |
|
| 376 | + if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 377 | + $injector->handleEnd($token); |
|
| 378 | + $this->processToken($token, $i); |
|
| 379 | + $this->stack[] = $current_parent; |
|
| 380 | + $reprocess = true; |
|
| 381 | + break; |
|
| 382 | + } |
|
| 383 | + continue; |
|
| 384 | + } |
|
| 385 | + |
|
| 386 | + // okay, so we're trying to close the wrong tag |
|
| 387 | + |
|
| 388 | + // undo the pop previous pop |
|
| 389 | + $this->stack[] = $current_parent; |
|
| 390 | + |
|
| 391 | + // scroll back the entire nest, trying to find our tag. |
|
| 392 | + // (feature could be to specify how far you'd like to go) |
|
| 393 | + $size = count($this->stack); |
|
| 394 | + // -2 because -1 is the last element, but we already checked that |
|
| 395 | + $skipped_tags = false; |
|
| 396 | + for ($j = $size - 2; $j >= 0; $j--) { |
|
| 397 | + if ($this->stack[$j]->name == $token->name) { |
|
| 398 | + $skipped_tags = array_slice($this->stack, $j); |
|
| 399 | + break; |
|
| 400 | + } |
|
| 401 | + } |
|
| 402 | + |
|
| 403 | + // we didn't find the tag, so remove |
|
| 404 | + if ($skipped_tags === false) { |
|
| 405 | + if ($escape_invalid_tags) { |
|
| 406 | + $this->swap(new HTMLPurifier_Token_Text( |
|
| 407 | + $generator->generateFromToken($token) |
|
| 408 | + )); |
|
| 409 | + if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag to text'); |
|
| 410 | + } else { |
|
| 411 | + $this->remove(); |
|
| 412 | + if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag removed'); |
|
| 413 | + } |
|
| 414 | + $reprocess = true; |
|
| 415 | + continue; |
|
| 416 | + } |
|
| 417 | + |
|
| 418 | + // do errors, in REVERSE $j order: a,b,c with </a></b></c> |
|
| 419 | + $c = count($skipped_tags); |
|
| 420 | + if ($e) { |
|
| 421 | + for ($j = $c - 1; $j > 0; $j--) { |
|
| 422 | + // notice we exclude $j == 0, i.e. the current ending tag, from |
|
| 423 | + // the errors... [TagClosedSuppress] |
|
| 424 | + if (!isset($skipped_tags[$j]->armor['MakeWellFormed_TagClosedError'])) { |
|
| 425 | + $e->send(E_NOTICE, 'Strategy_MakeWellFormed: Tag closed by element end', $skipped_tags[$j]); |
|
| 426 | + } |
|
| 427 | + } |
|
| 428 | + } |
|
| 429 | + |
|
| 430 | + // insert tags, in FORWARD $j order: c,b,a with </a></b></c> |
|
| 431 | + $replace = array($token); |
|
| 432 | + for ($j = 1; $j < $c; $j++) { |
|
| 433 | + // ...as well as from the insertions |
|
| 434 | + $new_token = new HTMLPurifier_Token_End($skipped_tags[$j]->name); |
|
| 435 | + $new_token->start = $skipped_tags[$j]; |
|
| 436 | + array_unshift($replace, $new_token); |
|
| 437 | + if (isset($definition->info[$new_token->name]) && $definition->info[$new_token->name]->formatting) { |
|
| 438 | + // [TagClosedAuto] |
|
| 439 | + $element = clone $skipped_tags[$j]; |
|
| 440 | + $element->carryover = true; |
|
| 441 | + $element->armor['MakeWellFormed_TagClosedError'] = true; |
|
| 442 | + $replace[] = $element; |
|
| 443 | + } |
|
| 444 | + } |
|
| 445 | + $this->processToken($replace); |
|
| 446 | + $reprocess = true; |
|
| 447 | + continue; |
|
| 448 | + } |
|
| 449 | + |
|
| 450 | + $context->destroy('CurrentNesting'); |
|
| 451 | + $context->destroy('InputTokens'); |
|
| 452 | + $context->destroy('InputIndex'); |
|
| 453 | + $context->destroy('CurrentToken'); |
|
| 454 | + |
|
| 455 | + unset($this->injectors, $this->stack, $this->tokens, $this->t); |
|
| 456 | + return $tokens; |
|
| 457 | + } |
|
| 458 | + |
|
| 459 | + /** |
|
| 460 | + * Processes arbitrary token values for complicated substitution patterns. |
|
| 461 | + * In general: |
|
| 462 | + * |
|
| 463 | + * If $token is an array, it is a list of tokens to substitute for the |
|
| 464 | + * current token. These tokens then get individually processed. If there |
|
| 465 | + * is a leading integer in the list, that integer determines how many |
|
| 466 | + * tokens from the stream should be removed. |
|
| 467 | + * |
|
| 468 | + * If $token is a regular token, it is swapped with the current token. |
|
| 469 | + * |
|
| 470 | + * If $token is false, the current token is deleted. |
|
| 471 | + * |
|
| 472 | + * If $token is an integer, that number of tokens (with the first token |
|
| 473 | + * being the current one) will be deleted. |
|
| 474 | + * |
|
| 475 | + * @param $token Token substitution value |
|
| 476 | + * @param $injector Injector that performed the substitution; default is if |
|
| 477 | + * this is not an injector related operation. |
|
| 478 | + */ |
|
| 479 | + protected function processToken($token, $injector = -1) { |
|
| 480 | + |
|
| 481 | + // normalize forms of token |
|
| 482 | + if (is_object($token)) $token = array(1, $token); |
|
| 483 | + if (is_int($token)) $token = array($token); |
|
| 484 | + if ($token === false) $token = array(1); |
|
| 485 | + if (!is_array($token)) throw new HTMLPurifier_Exception('Invalid token type from injector'); |
|
| 486 | + if (!is_int($token[0])) array_unshift($token, 1); |
|
| 487 | + if ($token[0] === 0) throw new HTMLPurifier_Exception('Deleting zero tokens is not valid'); |
|
| 488 | + |
|
| 489 | + // $token is now an array with the following form: |
|
| 490 | + // array(number nodes to delete, new node 1, new node 2, ...) |
|
| 491 | + |
|
| 492 | + $delete = array_shift($token); |
|
| 493 | + $old = array_splice($this->tokens, $this->t, $delete, $token); |
|
| 494 | + |
|
| 495 | + if ($injector > -1) { |
|
| 496 | + // determine appropriate skips |
|
| 497 | + $oldskip = isset($old[0]) ? $old[0]->skip : array(); |
|
| 498 | + foreach ($token as $object) { |
|
| 499 | + $object->skip = $oldskip; |
|
| 500 | + $object->skip[$injector] = true; |
|
| 501 | + } |
|
| 502 | + } |
|
| 503 | + |
|
| 504 | + } |
|
| 505 | + |
|
| 506 | + /** |
|
| 507 | + * Inserts a token before the current token. Cursor now points to |
|
| 508 | + * this token. You must reprocess after this. |
|
| 509 | + */ |
|
| 510 | + private function insertBefore($token) { |
|
| 511 | + array_splice($this->tokens, $this->t, 0, array($token)); |
|
| 512 | + } |
|
| 513 | + |
|
| 514 | + /** |
|
| 515 | + * Removes current token. Cursor now points to new token occupying previously |
|
| 516 | + * occupied space. You must reprocess after this. |
|
| 517 | + */ |
|
| 518 | + private function remove() { |
|
| 519 | + array_splice($this->tokens, $this->t, 1); |
|
| 520 | + } |
|
| 521 | + |
|
| 522 | + /** |
|
| 523 | + * Swap current token with new token. Cursor points to new token (no |
|
| 524 | + * change). You must reprocess after this. |
|
| 525 | + */ |
|
| 526 | + private function swap($token) { |
|
| 527 | + $this->tokens[$this->t] = $token; |
|
| 528 | + } |
|
| 529 | 529 | |
| 530 | 530 | } |
| 531 | 531 | |
@@ -65,17 +65,17 @@ |
||
| 65 | 65 | $stack = array(); |
| 66 | 66 | |
| 67 | 67 | // member variables |
| 68 | - $this->stack =& $stack; |
|
| 69 | - $this->t =& $t; |
|
| 70 | - $this->tokens =& $tokens; |
|
| 68 | + $this->stack = & $stack; |
|
| 69 | + $this->t = & $t; |
|
| 70 | + $this->tokens = & $tokens; |
|
| 71 | 71 | $this->config = $config; |
| 72 | 72 | $this->context = $context; |
| 73 | 73 | |
| 74 | 74 | // context variables |
| 75 | 75 | $context->register('CurrentNesting', $stack); |
| 76 | - $context->register('InputIndex', $t); |
|
| 77 | - $context->register('InputTokens', $tokens); |
|
| 78 | - $context->register('CurrentToken', $token); |
|
| 76 | + $context->register('InputIndex', $t); |
|
| 77 | + $context->register('InputTokens', $tokens); |
|
| 78 | + $context->register('CurrentToken', $token); |
|
| 79 | 79 | |
| 80 | 80 | // -- begin INJECTOR -- |
| 81 | 81 | |
@@ -87,9 +87,13 @@ discard block |
||
| 87 | 87 | unset($injectors['Custom']); // special case |
| 88 | 88 | foreach ($injectors as $injector => $b) { |
| 89 | 89 | // XXX: Fix with a legitimate lookup table of enabled filters |
| 90 | - if (strpos($injector, '.') !== false) continue; |
|
| 90 | + if (strpos($injector, '.') !== false) { |
|
| 91 | + continue; |
|
| 92 | + } |
|
| 91 | 93 | $injector = "HTMLPurifier_Injector_$injector"; |
| 92 | - if (!$b) continue; |
|
| 94 | + if (!$b) { |
|
| 95 | + continue; |
|
| 96 | + } |
|
| 93 | 97 | $this->injectors[] = new $injector; |
| 94 | 98 | } |
| 95 | 99 | foreach ($def_injectors as $injector) { |
@@ -97,7 +101,9 @@ discard block |
||
| 97 | 101 | $this->injectors[] = $injector; |
| 98 | 102 | } |
| 99 | 103 | foreach ($custom_injectors as $injector) { |
| 100 | - if (!$injector) continue; |
|
| 104 | + if (!$injector) { |
|
| 105 | + continue; |
|
| 106 | + } |
|
| 101 | 107 | if (is_string($injector)) { |
| 102 | 108 | $injector = "HTMLPurifier_Injector_$injector"; |
| 103 | 109 | $injector = new $injector; |
@@ -109,7 +115,9 @@ discard block |
||
| 109 | 115 | // variables for performance reasons |
| 110 | 116 | foreach ($this->injectors as $ix => $injector) { |
| 111 | 117 | $error = $injector->prepare($config, $context); |
| 112 | - if (!$error) continue; |
|
| 118 | + if (!$error) { |
|
| 119 | + continue; |
|
| 120 | + } |
|
| 113 | 121 | array_splice($this->injectors, $ix, 1); // rm the injector |
| 114 | 122 | trigger_error("Cannot enable {$injector->name} injector because $error is not allowed", E_USER_WARNING); |
| 115 | 123 | } |
@@ -139,7 +147,9 @@ discard block |
||
| 139 | 147 | // infinite loop, but might hinder some advanced rewinding. |
| 140 | 148 | $rewind_to = $this->injectors[$i]->getRewind(); |
| 141 | 149 | if (is_int($rewind_to) && $rewind_to < $t) { |
| 142 | - if ($rewind_to < 0) $rewind_to = 0; |
|
| 150 | + if ($rewind_to < 0) { |
|
| 151 | + $rewind_to = 0; |
|
| 152 | + } |
|
| 143 | 153 | while ($t > $rewind_to) { |
| 144 | 154 | $t--; |
| 145 | 155 | $prev = $tokens[$t]; |
@@ -147,8 +157,11 @@ discard block |
||
| 147 | 157 | // but we need to reprocess it |
| 148 | 158 | unset($prev->skip[$i]); |
| 149 | 159 | $prev->rewind = $i; |
| 150 | - if ($prev instanceof HTMLPurifier_Token_Start) array_pop($this->stack); |
|
| 151 | - elseif ($prev instanceof HTMLPurifier_Token_End) $this->stack[] = $prev->start; |
|
| 160 | + if ($prev instanceof HTMLPurifier_Token_Start) { |
|
| 161 | + array_pop($this->stack); |
|
| 162 | + } elseif ($prev instanceof HTMLPurifier_Token_End) { |
|
| 163 | + $this->stack[] = $prev->start; |
|
| 164 | + } |
|
| 152 | 165 | } |
| 153 | 166 | } |
| 154 | 167 | $i = false; |
@@ -157,7 +170,9 @@ discard block |
||
| 157 | 170 | // handle case of document end |
| 158 | 171 | if (!isset($tokens[$t])) { |
| 159 | 172 | // kill processing if stack is empty |
| 160 | - if (empty($this->stack)) break; |
|
| 173 | + if (empty($this->stack)) { |
|
| 174 | + break; |
|
| 175 | + } |
|
| 161 | 176 | |
| 162 | 177 | // peek |
| 163 | 178 | $top_nesting = array_pop($this->stack); |
@@ -185,8 +200,12 @@ discard block |
||
| 185 | 200 | if (empty($token->is_tag)) { |
| 186 | 201 | if ($token instanceof HTMLPurifier_Token_Text) { |
| 187 | 202 | foreach ($this->injectors as $i => $injector) { |
| 188 | - if (isset($token->skip[$i])) continue; |
|
| 189 | - if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 203 | + if (isset($token->skip[$i])) { |
|
| 204 | + continue; |
|
| 205 | + } |
|
| 206 | + if ($token->rewind !== null && $token->rewind !== $i) { |
|
| 207 | + continue; |
|
| 208 | + } |
|
| 190 | 209 | $injector->handleText($token); |
| 191 | 210 | $this->processToken($token, $i); |
| 192 | 211 | $reprocess = true; |
@@ -325,8 +344,12 @@ discard block |
||
| 325 | 344 | |
| 326 | 345 | if ($ok) { |
| 327 | 346 | foreach ($this->injectors as $i => $injector) { |
| 328 | - if (isset($token->skip[$i])) continue; |
|
| 329 | - if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 347 | + if (isset($token->skip[$i])) { |
|
| 348 | + continue; |
|
| 349 | + } |
|
| 350 | + if ($token->rewind !== null && $token->rewind !== $i) { |
|
| 351 | + continue; |
|
| 352 | + } |
|
| 330 | 353 | $injector->handleElement($token); |
| 331 | 354 | $this->processToken($token, $i); |
| 332 | 355 | $reprocess = true; |
@@ -352,13 +375,17 @@ discard block |
||
| 352 | 375 | // make sure that we have something open |
| 353 | 376 | if (empty($this->stack)) { |
| 354 | 377 | if ($escape_invalid_tags) { |
| 355 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag to text'); |
|
| 378 | + if ($e) { |
|
| 379 | + $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag to text'); |
|
| 380 | + } |
|
| 356 | 381 | $this->swap(new HTMLPurifier_Token_Text( |
| 357 | 382 | $generator->generateFromToken($token) |
| 358 | 383 | )); |
| 359 | 384 | } else { |
| 360 | 385 | $this->remove(); |
| 361 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag removed'); |
|
| 386 | + if ($e) { |
|
| 387 | + $e->send(E_WARNING, 'Strategy_MakeWellFormed: Unnecessary end tag removed'); |
|
| 388 | + } |
|
| 362 | 389 | } |
| 363 | 390 | $reprocess = true; |
| 364 | 391 | continue; |
@@ -372,8 +399,12 @@ discard block |
||
| 372 | 399 | if ($current_parent->name == $token->name) { |
| 373 | 400 | $token->start = $current_parent; |
| 374 | 401 | foreach ($this->injectors as $i => $injector) { |
| 375 | - if (isset($token->skip[$i])) continue; |
|
| 376 | - if ($token->rewind !== null && $token->rewind !== $i) continue; |
|
| 402 | + if (isset($token->skip[$i])) { |
|
| 403 | + continue; |
|
| 404 | + } |
|
| 405 | + if ($token->rewind !== null && $token->rewind !== $i) { |
|
| 406 | + continue; |
|
| 407 | + } |
|
| 377 | 408 | $injector->handleEnd($token); |
| 378 | 409 | $this->processToken($token, $i); |
| 379 | 410 | $this->stack[] = $current_parent; |
@@ -406,10 +437,14 @@ discard block |
||
| 406 | 437 | $this->swap(new HTMLPurifier_Token_Text( |
| 407 | 438 | $generator->generateFromToken($token) |
| 408 | 439 | )); |
| 409 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag to text'); |
|
| 440 | + if ($e) { |
|
| 441 | + $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag to text'); |
|
| 442 | + } |
|
| 410 | 443 | } else { |
| 411 | 444 | $this->remove(); |
| 412 | - if ($e) $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag removed'); |
|
| 445 | + if ($e) { |
|
| 446 | + $e->send(E_WARNING, 'Strategy_MakeWellFormed: Stray end tag removed'); |
|
| 447 | + } |
|
| 413 | 448 | } |
| 414 | 449 | $reprocess = true; |
| 415 | 450 | continue; |
@@ -479,12 +514,24 @@ discard block |
||
| 479 | 514 | protected function processToken($token, $injector = -1) { |
| 480 | 515 | |
| 481 | 516 | // normalize forms of token |
| 482 | - if (is_object($token)) $token = array(1, $token); |
|
| 483 | - if (is_int($token)) $token = array($token); |
|
| 484 | - if ($token === false) $token = array(1); |
|
| 485 | - if (!is_array($token)) throw new HTMLPurifier_Exception('Invalid token type from injector'); |
|
| 486 | - if (!is_int($token[0])) array_unshift($token, 1); |
|
| 487 | - if ($token[0] === 0) throw new HTMLPurifier_Exception('Deleting zero tokens is not valid'); |
|
| 517 | + if (is_object($token)) { |
|
| 518 | + $token = array(1, $token); |
|
| 519 | + } |
|
| 520 | + if (is_int($token)) { |
|
| 521 | + $token = array($token); |
|
| 522 | + } |
|
| 523 | + if ($token === false) { |
|
| 524 | + $token = array(1); |
|
| 525 | + } |
|
| 526 | + if (!is_array($token)) { |
|
| 527 | + throw new HTMLPurifier_Exception('Invalid token type from injector'); |
|
| 528 | + } |
|
| 529 | + if (!is_int($token[0])) { |
|
| 530 | + array_unshift($token, 1); |
|
| 531 | + } |
|
| 532 | + if ($token[0] === 0) { |
|
| 533 | + throw new HTMLPurifier_Exception('Deleting zero tokens is not valid'); |
|
| 534 | + } |
|
| 488 | 535 | |
| 489 | 536 | // $token is now an array with the following form: |
| 490 | 537 | // array(number nodes to delete, new node 1, new node 2, ...) |
@@ -11,177 +11,177 @@ |
||
| 11 | 11 | class HTMLPurifier_Strategy_RemoveForeignElements extends HTMLPurifier_Strategy |
| 12 | 12 | { |
| 13 | 13 | |
| 14 | - public function execute($tokens, $config, $context) { |
|
| 15 | - $definition = $config->getHTMLDefinition(); |
|
| 16 | - $generator = new HTMLPurifier_Generator($config, $context); |
|
| 17 | - $result = array(); |
|
| 18 | - |
|
| 19 | - $escape_invalid_tags = $config->get('Core.EscapeInvalidTags'); |
|
| 20 | - $remove_invalid_img = $config->get('Core.RemoveInvalidImg'); |
|
| 21 | - |
|
| 22 | - // currently only used to determine if comments should be kept |
|
| 23 | - $trusted = $config->get('HTML.Trusted'); |
|
| 24 | - $comment_lookup = $config->get('HTML.AllowedComments'); |
|
| 25 | - $comment_regexp = $config->get('HTML.AllowedCommentsRegexp'); |
|
| 26 | - $check_comments = $comment_lookup !== array() || $comment_regexp !== null; |
|
| 27 | - |
|
| 28 | - $remove_script_contents = $config->get('Core.RemoveScriptContents'); |
|
| 29 | - $hidden_elements = $config->get('Core.HiddenElements'); |
|
| 30 | - |
|
| 31 | - // remove script contents compatibility |
|
| 32 | - if ($remove_script_contents === true) { |
|
| 33 | - $hidden_elements['script'] = true; |
|
| 34 | - } elseif ($remove_script_contents === false && isset($hidden_elements['script'])) { |
|
| 35 | - unset($hidden_elements['script']); |
|
| 36 | - } |
|
| 37 | - |
|
| 38 | - $attr_validator = new HTMLPurifier_AttrValidator(); |
|
| 39 | - |
|
| 40 | - // removes tokens until it reaches a closing tag with its value |
|
| 41 | - $remove_until = false; |
|
| 42 | - |
|
| 43 | - // converts comments into text tokens when this is equal to a tag name |
|
| 44 | - $textify_comments = false; |
|
| 45 | - |
|
| 46 | - $token = false; |
|
| 47 | - $context->register('CurrentToken', $token); |
|
| 48 | - |
|
| 49 | - $e = false; |
|
| 50 | - if ($config->get('Core.CollectErrors')) { |
|
| 51 | - $e =& $context->get('ErrorCollector'); |
|
| 52 | - } |
|
| 53 | - |
|
| 54 | - foreach($tokens as $token) { |
|
| 55 | - if ($remove_until) { |
|
| 56 | - if (empty($token->is_tag) || $token->name !== $remove_until) { |
|
| 57 | - continue; |
|
| 58 | - } |
|
| 59 | - } |
|
| 60 | - if (!empty( $token->is_tag )) { |
|
| 61 | - // DEFINITION CALL |
|
| 62 | - |
|
| 63 | - // before any processing, try to transform the element |
|
| 64 | - if ( |
|
| 65 | - isset($definition->info_tag_transform[$token->name]) |
|
| 66 | - ) { |
|
| 67 | - $original_name = $token->name; |
|
| 68 | - // there is a transformation for this tag |
|
| 69 | - // DEFINITION CALL |
|
| 70 | - $token = $definition-> |
|
| 71 | - info_tag_transform[$token->name]-> |
|
| 72 | - transform($token, $config, $context); |
|
| 73 | - if ($e) $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Tag transform', $original_name); |
|
| 74 | - } |
|
| 75 | - |
|
| 76 | - if (isset($definition->info[$token->name])) { |
|
| 77 | - |
|
| 78 | - // mostly everything's good, but |
|
| 79 | - // we need to make sure required attributes are in order |
|
| 80 | - if ( |
|
| 81 | - ($token instanceof HTMLPurifier_Token_Start || $token instanceof HTMLPurifier_Token_Empty) && |
|
| 82 | - $definition->info[$token->name]->required_attr && |
|
| 83 | - ($token->name != 'img' || $remove_invalid_img) // ensure config option still works |
|
| 84 | - ) { |
|
| 85 | - $attr_validator->validateToken($token, $config, $context); |
|
| 86 | - $ok = true; |
|
| 87 | - foreach ($definition->info[$token->name]->required_attr as $name) { |
|
| 88 | - if (!isset($token->attr[$name])) { |
|
| 89 | - $ok = false; |
|
| 90 | - break; |
|
| 91 | - } |
|
| 92 | - } |
|
| 93 | - if (!$ok) { |
|
| 94 | - if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Missing required attribute', $name); |
|
| 95 | - continue; |
|
| 96 | - } |
|
| 97 | - $token->armor['ValidateAttributes'] = true; |
|
| 98 | - } |
|
| 99 | - |
|
| 100 | - if (isset($hidden_elements[$token->name]) && $token instanceof HTMLPurifier_Token_Start) { |
|
| 101 | - $textify_comments = $token->name; |
|
| 102 | - } elseif ($token->name === $textify_comments && $token instanceof HTMLPurifier_Token_End) { |
|
| 103 | - $textify_comments = false; |
|
| 104 | - } |
|
| 105 | - |
|
| 106 | - } elseif ($escape_invalid_tags) { |
|
| 107 | - // invalid tag, generate HTML representation and insert in |
|
| 108 | - if ($e) $e->send(E_WARNING, 'Strategy_RemoveForeignElements: Foreign element to text'); |
|
| 109 | - $token = new HTMLPurifier_Token_Text( |
|
| 110 | - $generator->generateFromToken($token) |
|
| 111 | - ); |
|
| 112 | - } else { |
|
| 113 | - // check if we need to destroy all of the tag's children |
|
| 114 | - // CAN BE GENERICIZED |
|
| 115 | - if (isset($hidden_elements[$token->name])) { |
|
| 116 | - if ($token instanceof HTMLPurifier_Token_Start) { |
|
| 117 | - $remove_until = $token->name; |
|
| 118 | - } elseif ($token instanceof HTMLPurifier_Token_Empty) { |
|
| 119 | - // do nothing: we're still looking |
|
| 120 | - } else { |
|
| 121 | - $remove_until = false; |
|
| 122 | - } |
|
| 123 | - if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign meta element removed'); |
|
| 124 | - } else { |
|
| 125 | - if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign element removed'); |
|
| 126 | - } |
|
| 127 | - continue; |
|
| 128 | - } |
|
| 129 | - } elseif ($token instanceof HTMLPurifier_Token_Comment) { |
|
| 130 | - // textify comments in script tags when they are allowed |
|
| 131 | - if ($textify_comments !== false) { |
|
| 132 | - $data = $token->data; |
|
| 133 | - $token = new HTMLPurifier_Token_Text($data); |
|
| 134 | - } elseif ($trusted || $check_comments) { |
|
| 135 | - // always cleanup comments |
|
| 136 | - $trailing_hyphen = false; |
|
| 137 | - if ($e) { |
|
| 138 | - // perform check whether or not there's a trailing hyphen |
|
| 139 | - if (substr($token->data, -1) == '-') { |
|
| 140 | - $trailing_hyphen = true; |
|
| 141 | - } |
|
| 142 | - } |
|
| 143 | - $token->data = rtrim($token->data, '-'); |
|
| 144 | - $found_double_hyphen = false; |
|
| 145 | - while (strpos($token->data, '--') !== false) { |
|
| 146 | - $found_double_hyphen = true; |
|
| 147 | - $token->data = str_replace('--', '-', $token->data); |
|
| 148 | - } |
|
| 149 | - if ($trusted || !empty($comment_lookup[trim($token->data)]) || ($comment_regexp !== NULL && preg_match($comment_regexp, trim($token->data)))) { |
|
| 150 | - // OK good |
|
| 151 | - if ($e) { |
|
| 152 | - if ($trailing_hyphen) { |
|
| 153 | - $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Trailing hyphen in comment removed'); |
|
| 154 | - } |
|
| 155 | - if ($found_double_hyphen) { |
|
| 156 | - $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Hyphens in comment collapsed'); |
|
| 157 | - } |
|
| 158 | - } |
|
| 159 | - } else { |
|
| 160 | - if ($e) { |
|
| 161 | - $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Comment removed'); |
|
| 162 | - } |
|
| 163 | - continue; |
|
| 164 | - } |
|
| 165 | - } else { |
|
| 166 | - // strip comments |
|
| 167 | - if ($e) $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Comment removed'); |
|
| 168 | - continue; |
|
| 169 | - } |
|
| 170 | - } elseif ($token instanceof HTMLPurifier_Token_Text) { |
|
| 171 | - } else { |
|
| 172 | - continue; |
|
| 173 | - } |
|
| 174 | - $result[] = $token; |
|
| 175 | - } |
|
| 176 | - if ($remove_until && $e) { |
|
| 177 | - // we removed tokens until the end, throw error |
|
| 178 | - $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Token removed to end', $remove_until); |
|
| 179 | - } |
|
| 180 | - |
|
| 181 | - $context->destroy('CurrentToken'); |
|
| 182 | - |
|
| 183 | - return $result; |
|
| 184 | - } |
|
| 14 | + public function execute($tokens, $config, $context) { |
|
| 15 | + $definition = $config->getHTMLDefinition(); |
|
| 16 | + $generator = new HTMLPurifier_Generator($config, $context); |
|
| 17 | + $result = array(); |
|
| 18 | + |
|
| 19 | + $escape_invalid_tags = $config->get('Core.EscapeInvalidTags'); |
|
| 20 | + $remove_invalid_img = $config->get('Core.RemoveInvalidImg'); |
|
| 21 | + |
|
| 22 | + // currently only used to determine if comments should be kept |
|
| 23 | + $trusted = $config->get('HTML.Trusted'); |
|
| 24 | + $comment_lookup = $config->get('HTML.AllowedComments'); |
|
| 25 | + $comment_regexp = $config->get('HTML.AllowedCommentsRegexp'); |
|
| 26 | + $check_comments = $comment_lookup !== array() || $comment_regexp !== null; |
|
| 27 | + |
|
| 28 | + $remove_script_contents = $config->get('Core.RemoveScriptContents'); |
|
| 29 | + $hidden_elements = $config->get('Core.HiddenElements'); |
|
| 30 | + |
|
| 31 | + // remove script contents compatibility |
|
| 32 | + if ($remove_script_contents === true) { |
|
| 33 | + $hidden_elements['script'] = true; |
|
| 34 | + } elseif ($remove_script_contents === false && isset($hidden_elements['script'])) { |
|
| 35 | + unset($hidden_elements['script']); |
|
| 36 | + } |
|
| 37 | + |
|
| 38 | + $attr_validator = new HTMLPurifier_AttrValidator(); |
|
| 39 | + |
|
| 40 | + // removes tokens until it reaches a closing tag with its value |
|
| 41 | + $remove_until = false; |
|
| 42 | + |
|
| 43 | + // converts comments into text tokens when this is equal to a tag name |
|
| 44 | + $textify_comments = false; |
|
| 45 | + |
|
| 46 | + $token = false; |
|
| 47 | + $context->register('CurrentToken', $token); |
|
| 48 | + |
|
| 49 | + $e = false; |
|
| 50 | + if ($config->get('Core.CollectErrors')) { |
|
| 51 | + $e =& $context->get('ErrorCollector'); |
|
| 52 | + } |
|
| 53 | + |
|
| 54 | + foreach($tokens as $token) { |
|
| 55 | + if ($remove_until) { |
|
| 56 | + if (empty($token->is_tag) || $token->name !== $remove_until) { |
|
| 57 | + continue; |
|
| 58 | + } |
|
| 59 | + } |
|
| 60 | + if (!empty( $token->is_tag )) { |
|
| 61 | + // DEFINITION CALL |
|
| 62 | + |
|
| 63 | + // before any processing, try to transform the element |
|
| 64 | + if ( |
|
| 65 | + isset($definition->info_tag_transform[$token->name]) |
|
| 66 | + ) { |
|
| 67 | + $original_name = $token->name; |
|
| 68 | + // there is a transformation for this tag |
|
| 69 | + // DEFINITION CALL |
|
| 70 | + $token = $definition-> |
|
| 71 | + info_tag_transform[$token->name]-> |
|
| 72 | + transform($token, $config, $context); |
|
| 73 | + if ($e) $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Tag transform', $original_name); |
|
| 74 | + } |
|
| 75 | + |
|
| 76 | + if (isset($definition->info[$token->name])) { |
|
| 77 | + |
|
| 78 | + // mostly everything's good, but |
|
| 79 | + // we need to make sure required attributes are in order |
|
| 80 | + if ( |
|
| 81 | + ($token instanceof HTMLPurifier_Token_Start || $token instanceof HTMLPurifier_Token_Empty) && |
|
| 82 | + $definition->info[$token->name]->required_attr && |
|
| 83 | + ($token->name != 'img' || $remove_invalid_img) // ensure config option still works |
|
| 84 | + ) { |
|
| 85 | + $attr_validator->validateToken($token, $config, $context); |
|
| 86 | + $ok = true; |
|
| 87 | + foreach ($definition->info[$token->name]->required_attr as $name) { |
|
| 88 | + if (!isset($token->attr[$name])) { |
|
| 89 | + $ok = false; |
|
| 90 | + break; |
|
| 91 | + } |
|
| 92 | + } |
|
| 93 | + if (!$ok) { |
|
| 94 | + if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Missing required attribute', $name); |
|
| 95 | + continue; |
|
| 96 | + } |
|
| 97 | + $token->armor['ValidateAttributes'] = true; |
|
| 98 | + } |
|
| 99 | + |
|
| 100 | + if (isset($hidden_elements[$token->name]) && $token instanceof HTMLPurifier_Token_Start) { |
|
| 101 | + $textify_comments = $token->name; |
|
| 102 | + } elseif ($token->name === $textify_comments && $token instanceof HTMLPurifier_Token_End) { |
|
| 103 | + $textify_comments = false; |
|
| 104 | + } |
|
| 105 | + |
|
| 106 | + } elseif ($escape_invalid_tags) { |
|
| 107 | + // invalid tag, generate HTML representation and insert in |
|
| 108 | + if ($e) $e->send(E_WARNING, 'Strategy_RemoveForeignElements: Foreign element to text'); |
|
| 109 | + $token = new HTMLPurifier_Token_Text( |
|
| 110 | + $generator->generateFromToken($token) |
|
| 111 | + ); |
|
| 112 | + } else { |
|
| 113 | + // check if we need to destroy all of the tag's children |
|
| 114 | + // CAN BE GENERICIZED |
|
| 115 | + if (isset($hidden_elements[$token->name])) { |
|
| 116 | + if ($token instanceof HTMLPurifier_Token_Start) { |
|
| 117 | + $remove_until = $token->name; |
|
| 118 | + } elseif ($token instanceof HTMLPurifier_Token_Empty) { |
|
| 119 | + // do nothing: we're still looking |
|
| 120 | + } else { |
|
| 121 | + $remove_until = false; |
|
| 122 | + } |
|
| 123 | + if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign meta element removed'); |
|
| 124 | + } else { |
|
| 125 | + if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign element removed'); |
|
| 126 | + } |
|
| 127 | + continue; |
|
| 128 | + } |
|
| 129 | + } elseif ($token instanceof HTMLPurifier_Token_Comment) { |
|
| 130 | + // textify comments in script tags when they are allowed |
|
| 131 | + if ($textify_comments !== false) { |
|
| 132 | + $data = $token->data; |
|
| 133 | + $token = new HTMLPurifier_Token_Text($data); |
|
| 134 | + } elseif ($trusted || $check_comments) { |
|
| 135 | + // always cleanup comments |
|
| 136 | + $trailing_hyphen = false; |
|
| 137 | + if ($e) { |
|
| 138 | + // perform check whether or not there's a trailing hyphen |
|
| 139 | + if (substr($token->data, -1) == '-') { |
|
| 140 | + $trailing_hyphen = true; |
|
| 141 | + } |
|
| 142 | + } |
|
| 143 | + $token->data = rtrim($token->data, '-'); |
|
| 144 | + $found_double_hyphen = false; |
|
| 145 | + while (strpos($token->data, '--') !== false) { |
|
| 146 | + $found_double_hyphen = true; |
|
| 147 | + $token->data = str_replace('--', '-', $token->data); |
|
| 148 | + } |
|
| 149 | + if ($trusted || !empty($comment_lookup[trim($token->data)]) || ($comment_regexp !== NULL && preg_match($comment_regexp, trim($token->data)))) { |
|
| 150 | + // OK good |
|
| 151 | + if ($e) { |
|
| 152 | + if ($trailing_hyphen) { |
|
| 153 | + $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Trailing hyphen in comment removed'); |
|
| 154 | + } |
|
| 155 | + if ($found_double_hyphen) { |
|
| 156 | + $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Hyphens in comment collapsed'); |
|
| 157 | + } |
|
| 158 | + } |
|
| 159 | + } else { |
|
| 160 | + if ($e) { |
|
| 161 | + $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Comment removed'); |
|
| 162 | + } |
|
| 163 | + continue; |
|
| 164 | + } |
|
| 165 | + } else { |
|
| 166 | + // strip comments |
|
| 167 | + if ($e) $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Comment removed'); |
|
| 168 | + continue; |
|
| 169 | + } |
|
| 170 | + } elseif ($token instanceof HTMLPurifier_Token_Text) { |
|
| 171 | + } else { |
|
| 172 | + continue; |
|
| 173 | + } |
|
| 174 | + $result[] = $token; |
|
| 175 | + } |
|
| 176 | + if ($remove_until && $e) { |
|
| 177 | + // we removed tokens until the end, throw error |
|
| 178 | + $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Token removed to end', $remove_until); |
|
| 179 | + } |
|
| 180 | + |
|
| 181 | + $context->destroy('CurrentToken'); |
|
| 182 | + |
|
| 183 | + return $result; |
|
| 184 | + } |
|
| 185 | 185 | |
| 186 | 186 | } |
| 187 | 187 | |
@@ -26,7 +26,7 @@ discard block |
||
| 26 | 26 | $check_comments = $comment_lookup !== array() || $comment_regexp !== null; |
| 27 | 27 | |
| 28 | 28 | $remove_script_contents = $config->get('Core.RemoveScriptContents'); |
| 29 | - $hidden_elements = $config->get('Core.HiddenElements'); |
|
| 29 | + $hidden_elements = $config->get('Core.HiddenElements'); |
|
| 30 | 30 | |
| 31 | 31 | // remove script contents compatibility |
| 32 | 32 | if ($remove_script_contents === true) { |
@@ -48,16 +48,16 @@ discard block |
||
| 48 | 48 | |
| 49 | 49 | $e = false; |
| 50 | 50 | if ($config->get('Core.CollectErrors')) { |
| 51 | - $e =& $context->get('ErrorCollector'); |
|
| 51 | + $e = & $context->get('ErrorCollector'); |
|
| 52 | 52 | } |
| 53 | 53 | |
| 54 | - foreach($tokens as $token) { |
|
| 54 | + foreach ($tokens as $token) { |
|
| 55 | 55 | if ($remove_until) { |
| 56 | 56 | if (empty($token->is_tag) || $token->name !== $remove_until) { |
| 57 | 57 | continue; |
| 58 | 58 | } |
| 59 | 59 | } |
| 60 | - if (!empty( $token->is_tag )) { |
|
| 60 | + if (!empty($token->is_tag)) { |
|
| 61 | 61 | // DEFINITION CALL |
| 62 | 62 | |
| 63 | 63 | // before any processing, try to transform the element |
@@ -70,7 +70,9 @@ discard block |
||
| 70 | 70 | $token = $definition-> |
| 71 | 71 | info_tag_transform[$token->name]-> |
| 72 | 72 | transform($token, $config, $context); |
| 73 | - if ($e) $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Tag transform', $original_name); |
|
| 73 | + if ($e) { |
|
| 74 | + $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Tag transform', $original_name); |
|
| 75 | + } |
|
| 74 | 76 | } |
| 75 | 77 | |
| 76 | 78 | if (isset($definition->info[$token->name])) { |
@@ -91,7 +93,9 @@ discard block |
||
| 91 | 93 | } |
| 92 | 94 | } |
| 93 | 95 | if (!$ok) { |
| 94 | - if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Missing required attribute', $name); |
|
| 96 | + if ($e) { |
|
| 97 | + $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Missing required attribute', $name); |
|
| 98 | + } |
|
| 95 | 99 | continue; |
| 96 | 100 | } |
| 97 | 101 | $token->armor['ValidateAttributes'] = true; |
@@ -105,7 +109,9 @@ discard block |
||
| 105 | 109 | |
| 106 | 110 | } elseif ($escape_invalid_tags) { |
| 107 | 111 | // invalid tag, generate HTML representation and insert in |
| 108 | - if ($e) $e->send(E_WARNING, 'Strategy_RemoveForeignElements: Foreign element to text'); |
|
| 112 | + if ($e) { |
|
| 113 | + $e->send(E_WARNING, 'Strategy_RemoveForeignElements: Foreign element to text'); |
|
| 114 | + } |
|
| 109 | 115 | $token = new HTMLPurifier_Token_Text( |
| 110 | 116 | $generator->generateFromToken($token) |
| 111 | 117 | ); |
@@ -120,9 +126,13 @@ discard block |
||
| 120 | 126 | } else { |
| 121 | 127 | $remove_until = false; |
| 122 | 128 | } |
| 123 | - if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign meta element removed'); |
|
| 129 | + if ($e) { |
|
| 130 | + $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign meta element removed'); |
|
| 131 | + } |
|
| 124 | 132 | } else { |
| 125 | - if ($e) $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign element removed'); |
|
| 133 | + if ($e) { |
|
| 134 | + $e->send(E_ERROR, 'Strategy_RemoveForeignElements: Foreign element removed'); |
|
| 135 | + } |
|
| 126 | 136 | } |
| 127 | 137 | continue; |
| 128 | 138 | } |
@@ -164,7 +174,9 @@ discard block |
||
| 164 | 174 | } |
| 165 | 175 | } else { |
| 166 | 176 | // strip comments |
| 167 | - if ($e) $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Comment removed'); |
|
| 177 | + if ($e) { |
|
| 178 | + $e->send(E_NOTICE, 'Strategy_RemoveForeignElements: Comment removed'); |
|
| 179 | + } |
|
| 168 | 180 | continue; |
| 169 | 181 | } |
| 170 | 182 | } elseif ($token instanceof HTMLPurifier_Token_Text) { |
@@ -7,32 +7,32 @@ |
||
| 7 | 7 | class HTMLPurifier_Strategy_ValidateAttributes extends HTMLPurifier_Strategy |
| 8 | 8 | { |
| 9 | 9 | |
| 10 | - public function execute($tokens, $config, $context) { |
|
| 10 | + public function execute($tokens, $config, $context) { |
|
| 11 | 11 | |
| 12 | - // setup validator |
|
| 13 | - $validator = new HTMLPurifier_AttrValidator(); |
|
| 12 | + // setup validator |
|
| 13 | + $validator = new HTMLPurifier_AttrValidator(); |
|
| 14 | 14 | |
| 15 | - $token = false; |
|
| 16 | - $context->register('CurrentToken', $token); |
|
| 15 | + $token = false; |
|
| 16 | + $context->register('CurrentToken', $token); |
|
| 17 | 17 | |
| 18 | - foreach ($tokens as $key => $token) { |
|
| 18 | + foreach ($tokens as $key => $token) { |
|
| 19 | 19 | |
| 20 | - // only process tokens that have attributes, |
|
| 21 | - // namely start and empty tags |
|
| 22 | - if (!$token instanceof HTMLPurifier_Token_Start && !$token instanceof HTMLPurifier_Token_Empty) continue; |
|
| 20 | + // only process tokens that have attributes, |
|
| 21 | + // namely start and empty tags |
|
| 22 | + if (!$token instanceof HTMLPurifier_Token_Start && !$token instanceof HTMLPurifier_Token_Empty) continue; |
|
| 23 | 23 | |
| 24 | - // skip tokens that are armored |
|
| 25 | - if (!empty($token->armor['ValidateAttributes'])) continue; |
|
| 24 | + // skip tokens that are armored |
|
| 25 | + if (!empty($token->armor['ValidateAttributes'])) continue; |
|
| 26 | 26 | |
| 27 | - // note that we have no facilities here for removing tokens |
|
| 28 | - $validator->validateToken($token, $config, $context); |
|
| 27 | + // note that we have no facilities here for removing tokens |
|
| 28 | + $validator->validateToken($token, $config, $context); |
|
| 29 | 29 | |
| 30 | - $tokens[$key] = $token; // for PHP 4 |
|
| 31 | - } |
|
| 32 | - $context->destroy('CurrentToken'); |
|
| 30 | + $tokens[$key] = $token; // for PHP 4 |
|
| 31 | + } |
|
| 32 | + $context->destroy('CurrentToken'); |
|
| 33 | 33 | |
| 34 | - return $tokens; |
|
| 35 | - } |
|
| 34 | + return $tokens; |
|
| 35 | + } |
|
| 36 | 36 | |
| 37 | 37 | } |
| 38 | 38 | |
@@ -19,10 +19,14 @@ |
||
| 19 | 19 | |
| 20 | 20 | // only process tokens that have attributes, |
| 21 | 21 | // namely start and empty tags |
| 22 | - if (!$token instanceof HTMLPurifier_Token_Start && !$token instanceof HTMLPurifier_Token_Empty) continue; |
|
| 22 | + if (!$token instanceof HTMLPurifier_Token_Start && !$token instanceof HTMLPurifier_Token_Empty) { |
|
| 23 | + continue; |
|
| 24 | + } |
|
| 23 | 25 | |
| 24 | 26 | // skip tokens that are armored |
| 25 | - if (!empty($token->armor['ValidateAttributes'])) continue; |
|
| 27 | + if (!empty($token->armor['ValidateAttributes'])) { |
|
| 28 | + continue; |
|
| 29 | + } |
|
| 26 | 30 | |
| 27 | 31 | // note that we have no facilities here for removing tokens |
| 28 | 32 | $validator->validateToken($token, $config, $context); |
@@ -10,30 +10,30 @@ |
||
| 10 | 10 | */ |
| 11 | 11 | class HTMLPurifier_StringHash extends ArrayObject |
| 12 | 12 | { |
| 13 | - protected $accessed = array(); |
|
| 13 | + protected $accessed = array(); |
|
| 14 | 14 | |
| 15 | - /** |
|
| 16 | - * Retrieves a value, and logs the access. |
|
| 17 | - */ |
|
| 18 | - public function offsetGet($index) { |
|
| 19 | - $this->accessed[$index] = true; |
|
| 20 | - return parent::offsetGet($index); |
|
| 21 | - } |
|
| 15 | + /** |
|
| 16 | + * Retrieves a value, and logs the access. |
|
| 17 | + */ |
|
| 18 | + public function offsetGet($index) { |
|
| 19 | + $this->accessed[$index] = true; |
|
| 20 | + return parent::offsetGet($index); |
|
| 21 | + } |
|
| 22 | 22 | |
| 23 | - /** |
|
| 24 | - * Returns a lookup array of all array indexes that have been accessed. |
|
| 25 | - * @return Array in form array($index => true). |
|
| 26 | - */ |
|
| 27 | - public function getAccessed() { |
|
| 28 | - return $this->accessed; |
|
| 29 | - } |
|
| 23 | + /** |
|
| 24 | + * Returns a lookup array of all array indexes that have been accessed. |
|
| 25 | + * @return Array in form array($index => true). |
|
| 26 | + */ |
|
| 27 | + public function getAccessed() { |
|
| 28 | + return $this->accessed; |
|
| 29 | + } |
|
| 30 | 30 | |
| 31 | - /** |
|
| 32 | - * Resets the access array. |
|
| 33 | - */ |
|
| 34 | - public function resetAccessed() { |
|
| 35 | - $this->accessed = array(); |
|
| 36 | - } |
|
| 31 | + /** |
|
| 32 | + * Resets the access array. |
|
| 33 | + */ |
|
| 34 | + public function resetAccessed() { |
|
| 35 | + $this->accessed = array(); |
|
| 36 | + } |
|
| 37 | 37 | } |
| 38 | 38 | |
| 39 | 39 | // vim: et sw=4 sts=4 |