mirrors/htmlpurifier
0
0
Fork 0
mirror of https://github.com/ezyang/htmlpurifier.git synced 2025-03-23 14:27:02 +00:00
Code Releases Activity

Profusely comment FixNesting.

Browse Source 
git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/trunk@183 48356398-32a2-884e-a903-53898d9a118a
This commit is contained in:
Edward Z. Yang 2006-08-07 20:28:12 +00:00
parent 7e2693fdea
commit 9cec089f97
1 changed files with 69 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

83
library/HTMLPurifier/Strategy/FixNesting.php
View File

@ -42,31 +42,53 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
function execute($tokens) { function execute($tokens) {
// insert implicit "parent" node, will be removed at end //####################################################################//
// Pre-processing
// insert implicit "parent" node, will be removed at end.
// ! we might want to move this to configuration
// DEFINITION CALL
$parent_name = $this->definition->info_parent; $parent_name = $this->definition->info_parent;
array_unshift($tokens, new HTMLPurifier_Token_Start($parent_name)); array_unshift($tokens, new HTMLPurifier_Token_Start($parent_name));
$tokens[] = new HTMLPurifier_Token_End($parent_name); $tokens[] = new HTMLPurifier_Token_End($parent_name);
//####################################################################//
// Loop initialization
// stack that contains the indexes of all parents, // stack that contains the indexes of all parents,
// $stack[count($stack)-1] being the current parent // $stack[count($stack)-1] being the current parent
$stack = array(); $stack = array();
// stack that contains all elements that are excluded // stack that contains all elements that are excluded
// same structure as $stack, but it is only populated when an element
// with exclusions is processed, i.e. there won't be empty exclusions.
$exclude_stack = array(); $exclude_stack = array();
//####################################################################//
// Loop
// iterate through all start nodes. Determining the start node
// is complicated so it has been omitted from the loop construct
for ($i = 0, $size = count($tokens) ; $i < $size; ) { for ($i = 0, $size = count($tokens) ; $i < $size; ) {
//################################################################//
// Gather information on children
// child token accumulator
$child_tokens = array(); $child_tokens = array();
// scroll to the end of this node, and report number // scroll to the end of this node, report number, and collect
// all children
for ($j = $i, $depth = 0; ; $j++) { for ($j = $i, $depth = 0; ; $j++) {
if ($tokens[$j]->type == 'start') { if ($tokens[$j]->type == 'start') {
$depth++; $depth++;
// skip token assignment on first iteration // skip token assignment on first iteration, this is the
// token we currently are on
if ($depth == 1) continue; if ($depth == 1) continue;
} elseif ($tokens[$j]->type == 'end') { } elseif ($tokens[$j]->type == 'end') {
$depth--; $depth--;
// skip token assignment on last iteration // skip token assignment on last iteration, this is the
// end token of the token we're currently on
if ($depth == 0) break; if ($depth == 0) break;
} }
$child_tokens[] = $tokens[$j]; $child_tokens[] = $tokens[$j];
@ -75,12 +97,16 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
// $i is index of start token // $i is index of start token
// $j is index of end token // $j is index of end token
//################################################################//
// Gather information on parent
// calculate parent information // calculate parent information
if ($count = count($stack)) { if ($count = count($stack)) {
$parent_index = $stack[$count-1]; $parent_index = $stack[$count-1];
$parent_name = $tokens[$parent_index]->name; $parent_name = $tokens[$parent_index]->name;
$parent_def = $this->definition->info[$parent_name]; $parent_def = $this->definition->info[$parent_name];
} else { } else {
// unknown info, it won't be used anyway
$parent_index = $parent_name = $parent_def = null; $parent_index = $parent_name = $parent_def = null;
} }
@ -88,26 +114,36 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
if (isset($parent_def)) { if (isset($parent_def)) {
$context = $parent_def->type; $context = $parent_def->type;
} else { } else {
// generally found in specialized elements like UL
$context = 'unknown'; $context = 'unknown';
} }
// determine whether or not element is excluded //################################################################//
// Determine whether element is explicitly excluded SGML-style
// determine whether or not element is excluded by checking all
// parent exclusions. The array should not be very large, two
// elements at most.
$excluded = false; $excluded = false;
if (!empty($exclude_stack)) { if (!empty($exclude_stack)) {
foreach ($exclude_stack as $lookup) { foreach ($exclude_stack as $lookup) {
if (isset($lookup[$tokens[$i]->name])) { if (isset($lookup[$tokens[$i]->name])) {
$excluded = true; $excluded = true;
// no need to continue processing
break; break;
} }
} }
} }
//################################################################//
// Perform child validation
if ($excluded) { if ($excluded) {
// there is an exclusion, remove the entire node
$result = false; $result = false;
} else { } else {
// DEFINITION CALL // DEFINITION CALL
$def = $this->definition->info[$tokens[$i]->name]; $def = $this->definition->info[$tokens[$i]->name];
$child_def = $def->child; $child_def = $def->child;
// have DTD child def validate children // have DTD child def validate children
@ -117,9 +153,12 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
$excludes = $def->excludes; $excludes = $def->excludes;
} }
// process result // $result is now a bool or array
//################################################################//
// Process result by interpreting $result
if ($result === true) { if ($result === true) {
// leave the node as is // leave the node as is
// register start token as a parental node start // register start token as a parental node start
@ -132,13 +171,15 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
$i++; $i++;
} elseif($result === false) { } elseif($result === false) {
// remove entire node
// calculate length of inner tokens and current tokens
$length = $j - $i + 1; $length = $j - $i + 1;
// remove entire node // perform removal
array_splice($tokens, $i, $length); array_splice($tokens, $i, $length);
// change size // update size
$size -= $length; $size -= $length;
// there is no start token to register, // there is no start token to register,
@ -150,14 +191,19 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
$i = $parent_index; $i = $parent_index;
} }
} else { // PROJECTED OPTIMIZATION: Process all children elements before
// reprocessing parent node.
} else {
// replace node with $result
// calculate length of inner tokens
$length = $j - $i - 1; $length = $j - $i - 1;
// replace node with $result // perform replacement
array_splice($tokens, $i + 1, $length, $result); array_splice($tokens, $i + 1, $length, $result);
// change size // update size
$size -= $length; $size -= $length;
$size += count($result); $size += count($result);
@ -172,6 +218,9 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
} }
//################################################################//
// Scroll to next start node
// We assume, at this point, that $i is the index of the token // We assume, at this point, that $i is the index of the token
// that is the first possible new start point for a node. // that is the first possible new start point for a node.
@ -192,10 +241,16 @@ class HTMLPurifier_Strategy_FixNesting extends HTMLPurifier_Strategy
} }
// remove implicit divs //####################################################################//
// Post-processing
// remove implicit parent tokens at the beginning and end
array_shift($tokens); array_shift($tokens);
array_pop($tokens); array_pop($tokens);
//####################################################################//
// Return
return $tokens; return $tokens;
} }