MensBeam
/
HTML-Parser
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
Browse Source

Cleaning up

ns
Dustin Wilson 3 years ago
parent
a8d146fad4
commit
9177ed6074
6 changed files with 61 additions and 99 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 105
      lib/DOM/Document.php
  2. 11
      lib/DOM/ElementMap.php
  3. 22
      lib/DOM/traits/Moonwalk.php
  4. 14
      lib/DOM/traits/MoonwalkShallow.php
  5. 4
      lib/DOM/traits/Walk.php
  6. 4
      lib/DOM/traits/WalkShallow.php

105
lib/DOM/Document.php
View File

@ -20,7 +20,7 @@ class Document extends AbstractDocument {
protected $_body = null;
// List of elements that are treated as block elements for the purposes of
// output formatting when serializing
protected const BLOCK_ELEMENTS = [ 'address', 'article', 'aside', 'blockquote', 'base', 'body', 'details', 'dialog', 'dd', 'div', 'dl', 'dt', 'fieldset', 'figcaption', 'figure', 'footer', 'form', 'frame', 'frameset', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'head', 'header', 'hr', 'html', 'isindex', 'li', 'link', 'main', 'meta', 'nav', 'ol', 'p', 'pre', 'section', 'script', 'source', 'style', 'table', 'template', 'td', 'tfoot', 'th', 'thead', 'title', 'tr', 'ul' ];
protected const BLOCK_ELEMENTS = [ 'address', 'article', 'aside', 'blockquote', 'base', 'body', 'details', 'dialog', 'dd', 'div', 'dl', 'dt', 'fieldset', 'figcaption', 'figure', 'footer', 'form', 'frame', 'frameset', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'head', 'header', 'hr', 'html', 'isindex', 'li', 'link', 'main', 'meta', 'nav', 'ol', 'p', 'picture', 'pre', 'section', 'script', 'source', 'style', 'table', 'template', 'td', 'tfoot', 'th', 'thead', 'title', 'tr', 'ul' ];
// List of preformatted elements where content is ignored for the purposes of
// output formatting when serializing
protected const PREFORMATTED_ELEMENTS = [ 'iframe', 'listing', 'noembed', 'noframes', 'noscript', 'plaintext', 'pre', 'style', 'script', 'textarea', 'title', 'xmp' ];
@ -317,12 +317,6 @@ class Document extends AbstractDocument {
}
};
$blockElementOrForeignFilter = function($n) use ($currentNode) {
if (!$n->isSameNode($currentNode) && $n instanceof Element && ($n->namespaceURI !== null || in_array($n->nodeName, self::BLOCK_ELEMENTS))) {
return true;
}
};
$blockElementOrForeignSiblings = false;
$modify = false;
}
@ -370,36 +364,6 @@ class Document extends AbstractDocument {
}
}
/*if ($foreignElement === null && $foreign) {
if ($hasChildNodes) {
$foreignElement = $currentNode;
$children = $currentNode->parentNode->walkShallow($blockElementOrForeignFilter);
$blockChild = false;
foreach ($children as $child) {
if ($child->namespaceURI === null) {
$blockChild = true;
break;
}
}
if ($blockChild) {
$blockElementOrForeignSiblings = true;
$foreignElementWithBlockElementSiblings = true;
$modify = true;
}
}
}
elseif ($foreignElement !== null && $foreign && $foreignElementWithBlockElementSiblings) {
$modify = true;
}
elseif ($currentNode->parentNode !== null && in_array($tagName, self::BLOCK_ELEMENTS) && (($currentNode->previousElementSibling === null && $currentNode->nextElementSibling === null) || $currentNode->parentNode->walkShallow($blockElementOrForeignFilter)->current() !== null)) {
$blockElementOrForeignSiblings = true;
$modify = true;
}*/
if ($modify) {
$s .= "\n" . str_repeat(' ', $indent);
}
@ -514,7 +478,7 @@ class Document extends AbstractDocument {
// If a foreign element with a foreign element ancestor with block element
// siblings and has at least one element child or any element with a block
// element descendant...
if (($foreign && $foreignElementWithBlockElementSiblings && $currentNode->firstElementChild !== null) || ($hasChildNodes && $currentNode->walk($blockElementFilter)->current() !== null)) {
if (($foreign && $foreignElementWithBlockElementSiblings && $currentNode->firstElementChild !== null) || ($currentNode->walk($blockElementFilter)->current() !== null)) {
$s .= "\n" . str_repeat(' ', $indent);
}
}
@ -538,7 +502,7 @@ class Document extends AbstractDocument {
# noframes, or plaintext element, or if the parent of current node is a noscript
# element and scripting is enabled for the node, then append the value of
# current node’s data IDL attribute literally.
if ($currentNode->parentNode->namespaceURI === null && in_array($currentNode->parentNode->nodeName, [ 'style', 'script', 'xmp', 'iframe', 'noembed', 'noframes', 'noscript', 'plaintext' ])) {
if ($currentNode->parentNode->namespaceURI === null && in_array($currentNode->parentNode->nodeName, [ 'style', 'script', 'xmp', 'iframe', 'noembed', 'noframes', 'plaintext' ])) {
$s .= $text;
}
# Otherwise, append the value of current node’s data IDL attribute, escaped as
@ -546,24 +510,8 @@ class Document extends AbstractDocument {
else {
if ($formatOutput) {
if ($preformattedElement === null) {
if ($foreignElement !== null) {
$modify = true;
} elseif ($currentNode->parentNode instanceof Element && in_array($currentNode->parentNode->nodeName, self::BLOCK_ELEMENTS)) {
$children = $currentNode->parentNode->walkShallow($blockElementOrForeignFilter);
$blockChild = false;
foreach ($children as $child) {
if ($child->namespaceURI === null) {
$blockChild = true;
break;
}
}
if ($blockChild) {
$modify = true;
}
}
if ($modify) {
$text = preg_replace('/ +/', ' ', str_replace("\t", ' ', $text));
if ($foreignElementWithBlockElementSiblings || $currentNode->parentNode->walk($blockElementFilter)->current() !== null) {
// If the text node's data is made up of only whitespace characters continue
// onto the next node
if (strspn($text, Data::WHITESPACE) === strlen($text)) {
@ -574,7 +522,7 @@ class Document extends AbstractDocument {
// the next node.
// Normalization here means that newlines are removed and simple spaces and tabs
// are condensed a single space.
$text = preg_replace([ '/[\n\x0C\x0D]+/', '/[ \t]+/' ], [ '', ' ' ], $text);
$text = preg_replace('/[\n\x0C\x0D]+/', '', $text);
if ($text === '') {
continue;
}
@ -587,39 +535,22 @@ class Document extends AbstractDocument {
}
# If current node is a Comment
elseif ($currentNode instanceof Comment) {
if ($formatOutput) {
$children = $currentNode->parentNode->walkShallow($blockElementOrForeignFilter);
$blockChild = false;
foreach ($children as $child) {
if ($child->namespaceURI === null) {
$blockChild = true;
break;
}
}
$text = trim($currentNode->data);
if ($blockChild) {
$indention = str_repeat(' ', $indent);
$s .= "\n$indention<!--";
if (strpos($text, "\n") === false) {
$s .= " $text -->";
} else {
$s .= "\n$indention " . str_replace("\n", "\n$indention ", $text) . "\n$indention-->";
}
} else {
$s .= "<!-- $text -->";
}
} else {
# Append the literal string "<!--" (U+003C LESS-THAN SIGN, U+0021 EXCLAMATION
# MARK, U+002D HYPHEN-MINUS, U+002D HYPHEN-MINUS), followed by the value of
# current node’s data IDL attribute, followed by the literal string "-->"
# (U+002D HYPHEN-MINUS, U+002D HYPHEN-MINUS, U+003E GREATER-THAN SIGN).
$s .= "<!--{$currentNode->data}-->";
if ($formatOutput && $preformattedElement === null && $foreignElementWithBlockElementSiblings || $currentNode->parentNode->walk($blockElementFilter)->current() !== null) {
$s .= "\n" . str_repeat(' ', $indent);
}
# Append the literal string "<!--" (U+003C LESS-THAN SIGN, U+0021 EXCLAMATION
# MARK, U+002D HYPHEN-MINUS, U+002D HYPHEN-MINUS), followed by the value of
# current node’s data IDL attribute, followed by the literal string "-->"
# (U+002D HYPHEN-MINUS, U+002D HYPHEN-MINUS, U+003E GREATER-THAN SIGN).
$s .= "<!--{$currentNode->data}-->";
}
# If current node is a ProcessingInstruction
elseif ($currentNode instanceof ProcessingInstruction) {
if ($formatOutput && $preformattedElement === null && $foreignElementWithBlockElementSiblings || $currentNode->parentNode->walk($blockElementFilter)->current() !== null) {
$s .= "\n" . str_repeat(' ', $indent);
}
# Append the literal string "<?" (U+003C LESS-THAN SIGN, U+003F QUESTION MARK),
# followed by the value of current node’s target IDL attribute, followed by a
# single U+0020 SPACE character, followed by the value of current node’s data

11
lib/DOM/ElementMap.php
View File

@ -1,7 +1,8 @@
<?php
/** @license MIT
* Copyright 2017 , Dustin Wilson, J. King et al.
* See LICENSE and AUTHORS files for details */
* Copyright 2017 Dustin Wilson, J. King et al.
* See LICENSE and AUTHORS files for details
*/
declare(strict_types=1);
namespace MensBeam\HTML;
@ -42,6 +43,12 @@ class ElementMap {
return false;
}
public static function getIterator(): \Traversable {
foreach (self::$_storage as $v) {
yield $v;
}
}
public static function has(Element $element) {
foreach (self::$_storage as $v) {
if ($v->isSameNode($element)) {

22
lib/DOM/traits/Moonwalk.php
View File

@ -14,8 +14,26 @@ trait Moonwalk {
private function moonwalkGenerator(\DOMNode $node, ?\Closure $filter = null) {
do {
if ($filter === null || $filter($node)) {
yield $node;
while (true) {
if ($filter === null || $filter($node)) {
yield $node;
}
// If node is an instance of DocumentFragment then it might be the content
// fragment of a template element, so iterate through all template elements
// stored in the element map and see if node is the fragment of one of the
// templates; if it is change node to the template element and reprocess. Magic!
// Can walk backwards THROUGH templates!
if ($node instanceof DocumentFragment) {
foreach (ElementMap::getIterator() as $element) {
if ($element->ownerDocument->isSameNode($node->ownerDocument) && $element instanceof TemplateElement && $element->content->isSameNode($node)) {
$node = $element;
continue;
}
}
}
break;
}
} while ($node = $node->parentNode);
}

14
lib/DOM/traits/MoonwalkShallow.php
View File

@ -14,13 +14,13 @@ trait MoonwalkShallow {
* @param ?\Closure $filter - An optional closure to use to filter
*/
public function moonwalkShallow(?\Closure $filter = null): \Generator {
if ($this->hasChildNodes()) {
$childNodesLength = $this->childNodes->length;
for ($childNodesLength = $this->childNodes->length, $i = $childNodesLength - 1; $i >= 0; $i--) {
$child = $this->childNodes[$i];
if ($filter === null || $filter($child)) {
yield $child;
}
$node = (!$this instanceof TemplateElement) ? $this : $this->content;
$childNodesLength = $node->childNodes->length;
for ($childNodesLength = $node->childNodes->length, $i = $childNodesLength - 1; $i >= 0; $i--) {
$child = $node->childNodes[$i];
if ($filter === null || $filter($child)) {
yield $child;
}
}
}

4
lib/DOM/traits/Walk.php
View File

@ -17,6 +17,10 @@ trait Walk {
yield $node;
}
if ($node instanceof TemplateElement) {
$node = $node->content;
}
if ($node->hasChildNodes()) {
$children = $node->childNodes;
foreach ($children as $c) {

4
lib/DOM/traits/WalkShallow.php
View File

@ -13,7 +13,9 @@ trait WalkShallow {
* @param ?\Closure $filter - An optional closure to use to filter
*/
public function walkShallow(?\Closure $filter = null): \Generator {
foreach ($this->childNodes as $child) {
$node = (!$this instanceof TemplateElement) ? $this : $this->content;
foreach ($node->childNodes as $child) {
if ($filter === null || $filter($child)) {
yield $child;
}

Write Preview
Loading…
Cancel
Save