From 11e1d82d9848a96a5b292c1a4a97a65167a84205 Mon Sep 17 00:00:00 2001
From: Benni Mack <benni@typo3.org>
Date: Tue, 12 Jan 2016 23:39:50 +0100
Subject: [PATCH] [!!!][TASK] Remove obsolete HtmlParser methods
Some now unused methods (after the XHTML cleanup)
within HtmlParser can be removed as well.
Resolves: #72667
Releases: master
Change-Id: I164505001d76db9b3e111d466f7b823e5cfe6b65
Reviewed-on: https://review.typo3.org/45854
Reviewed-by: Wouter Wolters <typo3@wouterwolters.nl>
Tested-by: Wouter Wolters <typo3@wouterwolters.nl>
Reviewed-by: Georg Ringer <georg.ringer@gmail.com>
Tested-by: Georg Ringer <georg.ringer@gmail.com>
---
typo3/sysext/core/Classes/Html/HtmlParser.php | 179 +-----------------
.../core/Classes/Html/RteHtmlParser.php | 18 +-
...-72667-RTEUnusedInternalMethodsRemoved.rst | 30 +++
3 files changed, 43 insertions(+), 184 deletions(-)
create mode 100644 typo3/sysext/core/Documentation/Changelog/master/Breaking-72667-RTEUnusedInternalMethodsRemoved.rst
diff --git a/typo3/sysext/core/Classes/Html/HtmlParser.php b/typo3/sysext/core/Classes/Html/HtmlParser.php
index 3fdf78324a0f..a5d049eaa5a0 100644
--- a/typo3/sysext/core/Classes/Html/HtmlParser.php
+++ b/typo3/sysext/core/Classes/Html/HtmlParser.php
@@ -183,21 +183,20 @@ class HtmlParser
/**
* Returns an array with either tag or non-tag content of the result from ->splitIntoBlock()/->splitTags()
+ * Does not include the tags in the tag parts.
*
* @param array $parts Parts generated by ->splitIntoBlock() or >splitTags()
* @param bool $tag_parts Whether to return the tag-parts (default,TRUE) or what was outside the tags.
- * @param bool $include_tag Whether to include the tags in the tag-parts (most useful for input made by ->splitIntoBlock())
* @return array Tag-parts/Non-tag-parts depending on input argument settings
* @see splitIntoBlock(), splitTags()
+ * @private Currently only use to remove content from table cells inside RteHtmlParser
*/
- public function getAllParts($parts, $tag_parts = true, $include_tag = true)
+ public function getAllParts($parts, $tag_parts = true)
{
$newParts = array();
foreach ($parts as $k => $v) {
if (($k + ($tag_parts ? 0 : 1)) % 2) {
- if (!$include_tag) {
- $v = $this->removeFirstAndLastTag($v);
- }
+ $v = $this->removeFirstAndLastTag($v);
$newParts[] = $v;
}
}
@@ -332,66 +331,6 @@ class HtmlParser
return array($value, $metaValue);
}
- /**
- * Checks whether block/solo tags are found in the correct amounts in HTML content
- * Block tags are tags which are required to have an equal amount of start and end tags, eg. "<table>...</table>"
- * Solo tags are tags which are required to have ONLY start tags (possibly with an XHTML ending like ".../>")
- * NOTICE: Correct XHTML might actually fail since "<br></br>" is allowed as well as "<br/>". However only the LATTER is accepted by this function (with "br" in the "solo-tag" list), the first example will result in a warning.
- * NOTICE: Correct XHTML might actually fail since "<p/>" is allowed as well as "<p></p>". However only the LATTER is accepted by this function (with "p" in the "block-tag" list), the first example will result in an ERROR!
- * NOTICE: Correct HTML version "something" allows eg. <p> and <li> to be NON-ended (implicitly ended by other tags). However this is NOT accepted by this function (with "p" and "li" in the block-tag list) and it will result in an ERROR!
- *
- * @param string $content HTML content to analyze
- * @param string $blockTags Tag names for block tags (eg. table or div or p) in lowercase, commalist (eg. "table,div,p")
- * @param string $soloTags Tag names for solo tags (eg. img, br or input) in lowercase, commalist ("img,br,input")
- * @return array Analyse data.
- */
- public function checkTagTypeCounts($content, $blockTags = 'a,b,blockquote,body,div,em,font,form,h1,h2,h3,h4,h5,h6,i,li,map,ol,option,p,pre,select,span,strong,table,td,textarea,tr,u,ul', $soloTags = 'br,hr,img,input,area')
- {
- $content = strtolower($content);
- $analyzedOutput = array();
- // Counts appearances of start-tags
- $analyzedOutput['counts'] = array();
- // Lists ERRORS
- $analyzedOutput['errors'] = array();
- // Lists warnings.
- $analyzedOutput['warnings'] = array();
- // Lists stats for block-tags
- $analyzedOutput['blocks'] = array();
- // Lists stats for solo-tags
- $analyzedOutput['solo'] = array();
- // Block tags, must have endings...
- $blockTags = explode(',', $blockTags);
- foreach ($blockTags as $tagName) {
- $countBegin = count(preg_split(('/\\<' . preg_quote($tagName, '/') . '(\\s|\\>)/s'), $content)) - 1;
- $countEnd = count(preg_split(('/\\<\\/' . preg_quote($tagName, '/') . '(\\s|\\>)/s'), $content)) - 1;
- $analyzedOutput['blocks'][$tagName] = array($countBegin, $countEnd, $countBegin - $countEnd);
- if ($countBegin) {
- $analyzedOutput['counts'][$tagName] = $countBegin;
- }
- if ($countBegin - $countEnd) {
- if ($countBegin - $countEnd > 0) {
- $analyzedOutput['errors'][$tagName] = 'There were more start-tags (' . $countBegin . ') than end-tags (' . $countEnd . ') for the element "' . $tagName . '". There should be an equal amount!';
- } else {
- $analyzedOutput['warnings'][$tagName] = 'There were more end-tags (' . $countEnd . ') than start-tags (' . $countBegin . ') for the element "' . $tagName . '". There should be an equal amount! However the problem is not fatal.';
- }
- }
- }
- // Solo tags, must NOT have endings...
- $soloTags = explode(',', $soloTags);
- foreach ($soloTags as $tagName) {
- $countBegin = count(preg_split(('/\\<' . preg_quote($tagName, '/') . '(\\s|\\>)/s'), $content)) - 1;
- $countEnd = count(preg_split(('/\\<\\/' . preg_quote($tagName, '/') . '(\\s|\\>)/s'), $content)) - 1;
- $analyzedOutput['solo'][$tagName] = array($countBegin, $countEnd);
- if ($countBegin) {
- $analyzedOutput['counts'][$tagName] = $countBegin;
- }
- if ($countEnd) {
- $analyzedOutput['warnings'][$tagName] = 'There were end-tags found (' . $countEnd . ') for the element "' . $tagName . '". This was not expected (although XHTML technically allows it).';
- }
- }
- return $analyzedOutput;
- }
-
/*********************************
*
* Clean HTML code
@@ -869,45 +808,6 @@ class HtmlParser
return $srcVal;
}
- /**
- * Cleans up the input $value for fonttags.
- * If keepFace,-Size and -Color is set then font-tags with an allowed property is kept. Else deleted.
- *
- * @param string HTML content with font-tags inside to clean up.
- * @param bool If set, keep "face" attribute
- * @param bool If set, keep "size" attribute
- * @param bool If set, keep "color" attribute
- * @return string Processed HTML content
- */
- public function cleanFontTags($value, $keepFace = 0, $keepSize = 0, $keepColor = 0)
- {
- // ,1 ?? - could probably be more stable if splitTags() was used since this depends on end-tags being properly set!
- $fontSplit = $this->splitIntoBlock('font', $value);
- foreach ($fontSplit as $k => $v) {
- // Font
- if ($k % 2) {
- $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v));
- $newAttribs = array();
- if ($keepFace && $attribArray['face']) {
- $newAttribs[] = 'face="' . $attribArray['face'] . '"';
- }
- if ($keepSize && $attribArray['size']) {
- $newAttribs[] = 'size="' . $attribArray['size'] . '"';
- }
- if ($keepColor && $attribArray['color']) {
- $newAttribs[] = 'color="' . $attribArray['color'] . '"';
- }
- $innerContent = $this->cleanFontTags($this->removeFirstAndLastTag($v), $keepFace, $keepSize, $keepColor);
- if (!empty($newAttribs)) {
- $fontSplit[$k] = '<font ' . implode(' ', $newAttribs) . '>' . $innerContent . '</font>';
- } else {
- $fontSplit[$k] = $innerContent;
- }
- }
- }
- return implode('', $fontSplit);
- }
-
/**
* This is used to map certain tag-names into other names.
*
@@ -925,44 +825,6 @@ class HtmlParser
return $value;
}
- /**
- * This converts htmlspecialchar()'ed tags (from $tagList) back to real tags. Eg. '<strong>' would be converted back to '<strong>' if found in $tagList
- *
- * @param string $content HTML content
- * @param string $tagList Tag list, separated by comma. Lowercase!
- * @return string Processed HTML content
- */
- public function unprotectTags($content, $tagList = '')
- {
- $tagsArray = GeneralUtility::trimExplode(',', $tagList, true);
- $contentParts = explode('<', $content);
- // bypass the first
- $contentPartsSliced = array_slice($contentParts, 1, null, true);
- foreach ($contentPartsSliced as $k => $tok) {
- $firstChar = $tok[0];
- if (trim($firstChar) !== '') {
- $subparts = explode('>', $tok, 2);
- $tagEnd = strlen($subparts[0]);
- if (strlen($tok) != $tagEnd) {
- $endTag = $firstChar == '/' ? 1 : 0;
- $tagContent = substr($tok, $endTag, $tagEnd - $endTag);
- $tagParts = preg_split('/\\s+/s', $tagContent, 2);
- $tagName = strtolower($tagParts[0]);
- if ((string)$tagList === '' || in_array($tagName, $tagsArray)) {
- $contentParts[$k] = '<' . $subparts[0] . '>' . $subparts[1];
- } else {
- $contentParts[$k] = '<' . $tok;
- }
- } else {
- $contentParts[$k] = '<' . $tok;
- }
- } else {
- $contentParts[$k] = '<' . $tok;
- }
- }
- return implode('', $contentParts);
- }
-
/**
* Internal function for case shifting of a string or whole array
*
@@ -1017,39 +879,6 @@ class HtmlParser
return implode(' ', $accu);
}
- /**
- * Get tag attributes, the classic version (which had some limitations?)
- *
- * @param string $tag The tag
- * @param bool $deHSC De-htmlspecialchar flag.
- * @return array
- * @access private
- */
- public function get_tag_attributes_classic($tag, $deHSC = 0)
- {
- $attr = $this->get_tag_attributes($tag, $deHSC);
- return is_array($attr[0]) ? $attr[0] : array();
- }
-
- /**
- * Indents input content with $number instances of $indentChar
- *
- * @param string $content Content string, multiple lines.
- * @param int $number Number of indents
- * @param string $indentChar Indent character/string
- * @return string Indented code (typ. HTML)
- */
- public function indentLines($content, $number = 1, $indentChar = TAB)
- {
- $preTab = str_pad('', $number * strlen($indentChar), $indentChar);
- $lines = explode(LF, str_replace(CR, '', $content));
- foreach ($lines as &$line) {
- $line = $preTab . $line;
- }
- unset($line);
- return implode(LF, $lines);
- }
-
/**
* Converts TSconfig into an array for the HTMLcleaner function.
*
diff --git a/typo3/sysext/core/Classes/Html/RteHtmlParser.php b/typo3/sysext/core/Classes/Html/RteHtmlParser.php
index 2680257e6cea..f63c5fbf7ef3 100644
--- a/typo3/sysext/core/Classes/Html/RteHtmlParser.php
+++ b/typo3/sysext/core/Classes/Html/RteHtmlParser.php
@@ -330,7 +330,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
// Image found, do processing:
if ($k % 2) {
// Get attributes
- $attribArray = $this->get_tag_attributes_classic($v, 1);
+ list($attribArray) = $this->get_tag_attributes($v, true);
// It's always an absolute URL coming from the RTE into the Database.
$absoluteUrl = trim($attribArray['src']);
// Make path absolute if it is relative and we have a site path which is not '/'
@@ -479,7 +479,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
// Image found
if ($k % 2) {
// Get the attributes of the img tag
- $attribArray = $this->get_tag_attributes_classic($v, 1);
+ list($attribArray) = $this->get_tag_attributes($v, true);
$absoluteUrl = trim($attribArray['src']);
// Transform the src attribute into an absolute url, if it not already
if (strtolower(substr($absoluteUrl, 0, 4)) !== 'http') {
@@ -521,7 +521,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
foreach ($blockSplit as $k => $v) {
// Block
if ($k % 2) {
- $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v), 1);
+ list($attribArray) = $this->get_tag_attributes($this->getFirstTag($v), true);
// If the url is local, remove url-prefix
if ($siteURL && substr($attribArray['href'], 0, strlen($siteURL)) == $siteURL) {
$attribArray['href'] = $this->relBackPath . substr($attribArray['href'], strlen($siteURL));
@@ -553,7 +553,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
foreach ($blockSplit as $k => $v) {
// If an A-tag was found:
if ($k % 2) {
- $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v), 1);
+ list($attribArray) = $this->get_tag_attributes($this->getFirstTag($v), true);
$info = $this->urlInfoForLinkTags($attribArray['href']);
// Check options:
$attribArray_copy = $attribArray;
@@ -799,7 +799,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
foreach ($blockSplit as $k => $v) {
// Block
if ($k % 2) {
- $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v));
+ list($attribArray) = $this->get_tag_attributes($this->getFirstTag($v));
if ($attribArray['specialtag']) {
$theTag = rawurldecode($attribArray['specialtag']);
$theTagName = $this->getFirstTagName($theTag);
@@ -913,7 +913,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
case 'h6':
if (!$css) {
- $attribArray = $this->get_tag_attributes_classic($tag);
+ list($attribArray) = $this->get_tag_attributes($tag);
// Processing inner content here:
$innerContent = $this->HTMLcleaner_db($this->removeFirstAndLastTag($blockSplit[$k]));
$blockSplit[$k] = '<' . $tagName . ($attribArray['align'] ? ' align="' . htmlspecialchars($attribArray['align']) . '"' : '') . ($attribArray['class'] ? ' class="' . htmlspecialchars($attribArray['class']) . '"' : '') . '>' . $innerContent . '</' . $tagName . '>' . $lastBR;
@@ -959,7 +959,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
foreach ($blockSplit as $k => $v) {
// If an A-tag was found
if ($k % 2) {
- $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v), 1);
+ list($attribArray) = $this->get_tag_attributes($this->getFirstTag($v), true);
// If "style" attribute is set and rteerror is not set!
if ($attribArray['style'] && !$attribArray['rteerror']) {
$attribArray_copy['style'] = $attribArray['style'];
@@ -1416,7 +1416,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
$rowSplit = $this->splitIntoBlock('tr', $v);
foreach ($rowSplit as $k2 => $v2) {
if ($k2 % 2) {
- $cellSplit = $this->getAllParts($this->splitIntoBlock('td', $v2), 1, 0);
+ $cellSplit = $this->getAllParts($this->splitIntoBlock('td', $v2));
foreach ($cellSplit as $k3 => $v3) {
$tableSplit[$k] .= $v3 . $breakChar;
}
@@ -1559,7 +1559,7 @@ class RteHtmlParser extends \TYPO3\CMS\Core\Html\HtmlParser
foreach ($blockSplit as $k => $v) {
// Block
if ($k % 2) {
- $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v), 1);
+ list($attribArray) = $this->get_tag_attributes($this->getFirstTag($v), true);
// Checking if there is a scheme, and if not, prepend the current url.
// ONLY do this if href has content - the <a> tag COULD be an anchor and if so, it should be preserved...
if ($attribArray['href'] !== '') {
diff --git a/typo3/sysext/core/Documentation/Changelog/master/Breaking-72667-RTEUnusedInternalMethodsRemoved.rst b/typo3/sysext/core/Documentation/Changelog/master/Breaking-72667-RTEUnusedInternalMethodsRemoved.rst
new file mode 100644
index 000000000000..37b0bcc1dd22
--- /dev/null
+++ b/typo3/sysext/core/Documentation/Changelog/master/Breaking-72667-RTEUnusedInternalMethodsRemoved.rst
@@ -0,0 +1,30 @@
+=======================================================
+Breaking: #72667 - RTE: Unused internal methods removed
+=======================================================
+
+Description
+===========
+
+The HTML parsing features for the Rich Text Editor feature related to the xhtml_cleaning were removed. The following now obsolete methods are
+removed as well:
+
+* ``HtmlParser->checkTagTypeCounts()``
+* ``HtmlParser->unprotectTags()``
+* ``HtmlParser->get_tag_attributes_classic()``
+* ``HtmlParser->cleanFontTags()``
+* ``HtmlParser->indentLines()``
+
+Additionally, the third parameter for the method ``HtmlParser->getAllParts()`` was removed as well, resulting that the method will always include
+the parsed tags in the result set.
+
+
+Impact
+======
+
+Calling any of the methods will result in a fatal PHP error.
+
+
+Affected Installations
+======================
+
+Any installation which uses a third-party extension that modifies the HtmlParsing via PHP.
\ No newline at end of file
--
GitLab