mirrors/wallabag
1
0
Fork 0
mirror of https://github.com/wallabag/wallabag.git synced 2024-11-27 03:21:01 +00:00
Code Issues Projects Releases Wiki Activity

fix of issue #718: Error parsing file imported from Pocket #718

Browse source
This commit is contained in:
Maryana Rozhankivska 2014-06-25 19:34:14 +03:00
parent c9563378ea
commit aa126ba458
1 changed files with 65 additions and 40 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

105
inc/3rdparty/simple_html_dom.php vendored Normal file → Executable file
View file

@ -34,7 +34,7 @@
* @author S.C. Chen <me578022@gmail.com> * @author S.C. Chen <me578022@gmail.com>
* @author John Schlick * @author John Schlick
* @author Rus Carroll * @author Rus Carroll
* @version 1.5 ($Rev: 202 $) * @version 1.5 ($Rev: 210 $)
* @package PlaceLocalInclude * @package PlaceLocalInclude
* @subpackage simple_html_dom * @subpackage simple_html_dom
*/ */
@ -269,7 +269,10 @@ class simple_html_dom_node
{ {
return $this->children; return $this->children;
} }
if (isset($this->children[$idx])) return $this->children[$idx]; if (isset($this->children[$idx]))
{
return $this->children[$idx];
}
return null; return null;
} }
@ -330,14 +333,14 @@ class simple_html_dom_node
function find_ancestor_tag($tag) function find_ancestor_tag($tag)
{ {
global $debug_object; global $debug_object;
if (is_object($debug_object)) { $debug_object->debugLogEntry(1); } if (is_object($debug_object)) { $debug_object->debug_log_entry(1); }
// Start by including ourselves in the comparison. // Start by including ourselves in the comparison.
$returnDom = $this; $returnDom = $this;
while (!is_null($returnDom)) while (!is_null($returnDom))
{ {
if (is_object($debug_object)) { $debug_object->debugLog(2, "Current tag is: " . $returnDom->tag); } if (is_object($debug_object)) { $debug_object->debug_log(2, "Current tag is: " . $returnDom->tag); }
if ($returnDom->tag == $tag) if ($returnDom->tag == $tag)
{ {
@ -374,7 +377,7 @@ class simple_html_dom_node
$text = " with text: " . $this->text; $text = " with text: " . $this->text;
} }
} }
$debug_object->debugLog(1, 'Innertext of tag: ' . $this->tag . $text); $debug_object->debug_log(1, 'Innertext of tag: ' . $this->tag . $text);
} }
if ($this->tag==='root') return $this->innertext(); if ($this->tag==='root') return $this->innertext();
@ -532,7 +535,9 @@ class simple_html_dom_node
foreach ($head as $k=>$v) foreach ($head as $k=>$v)
{ {
if (!isset($found_keys[$k])) if (!isset($found_keys[$k]))
{
$found_keys[$k] = 1; $found_keys[$k] = 1;
}
} }
} }
@ -554,7 +559,7 @@ class simple_html_dom_node
protected function seek($selector, &$ret, $lowercase=false) protected function seek($selector, &$ret, $lowercase=false)
{ {
global $debug_object; global $debug_object;
if (is_object($debug_object)) { $debug_object->debugLogEntry(1); } if (is_object($debug_object)) { $debug_object->debug_log_entry(1); }
list($tag, $key, $val, $exp, $no_key) = $selector; list($tag, $key, $val, $exp, $no_key) = $selector;
@ -615,7 +620,7 @@ class simple_html_dom_node
// this is a normal search, we want the value of that attribute of the tag. // this is a normal search, we want the value of that attribute of the tag.
$nodeKeyValue = $node->attr[$key]; $nodeKeyValue = $node->attr[$key];
} }
if (is_object($debug_object)) {$debug_object->debugLog(2, "testing node: " . $node->tag . " for attribute: " . $key . $exp . $val . " where nodes value is: " . $nodeKeyValue);} if (is_object($debug_object)) {$debug_object->debug_log(2, "testing node: " . $node->tag . " for attribute: " . $key . $exp . $val . " where nodes value is: " . $nodeKeyValue);}
//PaperG - If lowercase is set, do a case insensitive test of the value of the selector. //PaperG - If lowercase is set, do a case insensitive test of the value of the selector.
if ($lowercase) { if ($lowercase) {
@ -623,7 +628,7 @@ class simple_html_dom_node
} else { } else {
$check = $this->match($exp, $val, $nodeKeyValue); $check = $this->match($exp, $val, $nodeKeyValue);
} }
if (is_object($debug_object)) {$debug_object->debugLog(2, "after match: " . ($check ? "true" : "false"));} if (is_object($debug_object)) {$debug_object->debug_log(2, "after match: " . ($check ? "true" : "false"));}
// handle multiple class // handle multiple class
if (!$check && strcasecmp($key, 'class')===0) { if (!$check && strcasecmp($key, 'class')===0) {
@ -645,12 +650,12 @@ class simple_html_dom_node
unset($node); unset($node);
} }
// It's passed by reference so this is actually what this function returns. // It's passed by reference so this is actually what this function returns.
if (is_object($debug_object)) {$debug_object->debugLog(1, "EXIT - ret: ", $ret);} if (is_object($debug_object)) {$debug_object->debug_log(1, "EXIT - ret: ", $ret);}
} }
protected function match($exp, $pattern, $value) { protected function match($exp, $pattern, $value) {
global $debug_object; global $debug_object;
if (is_object($debug_object)) {$debug_object->debugLogEntry(1);} if (is_object($debug_object)) {$debug_object->debug_log_entry(1);}
switch ($exp) { switch ($exp) {
case '=': case '=':
@ -672,7 +677,7 @@ class simple_html_dom_node
protected function parse_selector($selector_string) { protected function parse_selector($selector_string) {
global $debug_object; global $debug_object;
if (is_object($debug_object)) {$debug_object->debugLogEntry(1);} if (is_object($debug_object)) {$debug_object->debug_log_entry(1);}
// pattern of CSS selectors, modified from mootools // pattern of CSS selectors, modified from mootools
// Paperg: Add the colon to the attrbute, so that it properly finds <tag attr:ibute="something" > like google does. // Paperg: Add the colon to the attrbute, so that it properly finds <tag attr:ibute="something" > like google does.
@ -683,7 +688,7 @@ class simple_html_dom_node
// $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is"; // $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is";
$pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-:]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is"; $pattern = "/([\w-:\*]*)(?:\#([\w-]+)|\.([\w-]+))?(?:\[@?(!?[\w-:]+)(?:([!*^$]?=)[\"']?(.*?)[\"']?)?\])?([\/, ]+)/is";
preg_match_all($pattern, trim($selector_string).' ', $matches, PREG_SET_ORDER); preg_match_all($pattern, trim($selector_string).' ', $matches, PREG_SET_ORDER);
if (is_object($debug_object)) {$debug_object->debugLog(2, "Matches Array: ", $matches);} if (is_object($debug_object)) {$debug_object->debug_log(2, "Matches Array: ", $matches);}
$selectors = array(); $selectors = array();
$result = array(); $result = array();
@ -718,12 +723,14 @@ class simple_html_dom_node
return $selectors; return $selectors;
} }
function __get($name) { function __get($name)
{
if (isset($this->attr[$name])) if (isset($this->attr[$name]))
{ {
return $this->convert_text($this->attr[$name]); return $this->convert_text($this->attr[$name]);
} }
switch ($name) { switch ($name)
{
case 'outertext': return $this->outertext(); case 'outertext': return $this->outertext();
case 'innertext': return $this->innertext(); case 'innertext': return $this->innertext();
case 'plaintext': return $this->text(); case 'plaintext': return $this->text();
@ -732,22 +739,30 @@ class simple_html_dom_node
} }
} }
function __set($name, $value) { function __set($name, $value)
switch ($name) { {
global $debug_object;
if (is_object($debug_object)) {$debug_object->debug_log_entry(1);}
switch ($name)
{
case 'outertext': return $this->_[HDOM_INFO_OUTER] = $value; case 'outertext': return $this->_[HDOM_INFO_OUTER] = $value;
case 'innertext': case 'innertext':
if (isset($this->_[HDOM_INFO_TEXT])) return $this->_[HDOM_INFO_TEXT] = $value; if (isset($this->_[HDOM_INFO_TEXT])) return $this->_[HDOM_INFO_TEXT] = $value;
return $this->_[HDOM_INFO_INNER] = $value; return $this->_[HDOM_INFO_INNER] = $value;
} }
if (!isset($this->attr[$name])) { if (!isset($this->attr[$name]))
{
$this->_[HDOM_INFO_SPACE][] = array(' ', '', ''); $this->_[HDOM_INFO_SPACE][] = array(' ', '', '');
$this->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_DOUBLE; $this->_[HDOM_INFO_QUOTE][] = HDOM_QUOTE_DOUBLE;
} }
$this->attr[$name] = $value; $this->attr[$name] = $value;
} }
function __isset($name) { function __isset($name)
switch ($name) { {
switch ($name)
{
case 'outertext': return true; case 'outertext': return true;
case 'innertext': return true; case 'innertext': return true;
case 'plaintext': return true; case 'plaintext': return true;
@ -765,7 +780,7 @@ class simple_html_dom_node
function convert_text($text) function convert_text($text)
{ {
global $debug_object; global $debug_object;
if (is_object($debug_object)) {$debug_object->debugLogEntry(1);} if (is_object($debug_object)) {$debug_object->debug_log_entry(1);}
$converted_text = $text; $converted_text = $text;
@ -777,7 +792,7 @@ class simple_html_dom_node
$sourceCharset = strtoupper($this->dom->_charset); $sourceCharset = strtoupper($this->dom->_charset);
$targetCharset = strtoupper($this->dom->_target_charset); $targetCharset = strtoupper($this->dom->_target_charset);
} }
if (is_object($debug_object)) {$debug_object->debugLog(3, "source charset: " . $sourceCharset . " target charaset: " . $targetCharset);} if (is_object($debug_object)) {$debug_object->debug_log(3, "source charset: " . $sourceCharset . " target charaset: " . $targetCharset);}
if (!empty($sourceCharset) && !empty($targetCharset) && (strcasecmp($sourceCharset, $targetCharset) != 0)) if (!empty($sourceCharset) && !empty($targetCharset) && (strcasecmp($sourceCharset, $targetCharset) != 0))
{ {
@ -1045,10 +1060,10 @@ class simple_html_dom
// prepare // prepare
$this->prepare($str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText); $this->prepare($str, $lowercase, $stripRN, $defaultBRText, $defaultSpanText);
// strip out comments
$this->remove_noise("'<!--(.*?)-->'is");
// strip out cdata // strip out cdata
$this->remove_noise("'<!\[CDATA\[(.*?)\]\]>'is", true); $this->remove_noise("'<!\[CDATA\[(.*?)\]\]>'is", true);
// strip out comments
$this->remove_noise("'<!--(.*?)-->'is");
// Per sourceforge http://sourceforge.net/tracker/?func=detail&aid=2949097&group_id=218559&atid=1044037 // Per sourceforge http://sourceforge.net/tracker/?func=detail&aid=2949097&group_id=218559&atid=1044037
// Script tags removal now preceeds style tag removal. // Script tags removal now preceeds style tag removal.
// strip out <script> tags // strip out <script> tags
@ -1078,10 +1093,15 @@ class simple_html_dom
// load html from file // load html from file
function load_file() function load_file()
{ {
//external error: NOT related to dom loading
$extError=error_get_last();
$args = func_get_args(); $args = func_get_args();
$this->load(call_user_func_array('file_get_contents', $args), true); $this->load(call_user_func_array('file_get_contents', $args), true);
// Throw an error if we can't properly load the dom. // Throw an error if we can't properly load the dom.
if (($error=error_get_last())!==null) { $error=error_get_last();
if ($error!==$extError) {
$this->clear(); $this->clear();
return false; return false;
} }
@ -1198,22 +1218,22 @@ class simple_html_dom
if ($success) if ($success)
{ {
$charset = $matches[1]; $charset = $matches[1];
if (is_object($debug_object)) {$debug_object->debugLog(2, 'header content-type found charset of: ' . $charset);} if (is_object($debug_object)) {$debug_object->debug_log(2, 'header content-type found charset of: ' . $charset);}
} }
} }
if (empty($charset)) if (empty($charset))
{ {
$el = $this->root->find('meta[http-equiv=Content-Type]',0); $el = $this->root->find('meta[http-equiv=Content-Type]',0, true);
if (!empty($el)) if (!empty($el))
{ {
$fullvalue = $el->content; $fullvalue = $el->content;
if (is_object($debug_object)) {$debug_object->debugLog(2, 'meta content-type tag found' . $fullvalue);} if (is_object($debug_object)) {$debug_object->debug_log(2, 'meta content-type tag found' . $fullvalue);}
if (!empty($fullvalue)) if (!empty($fullvalue))
{ {
$success = preg_match('/charset=(.+)/', $fullvalue, $matches); $success = preg_match('/charset=(.+)/i', $fullvalue, $matches);
if ($success) if ($success)
{ {
$charset = $matches[1]; $charset = $matches[1];
@ -1221,7 +1241,7 @@ class simple_html_dom
else else
{ {
// If there is a meta tag, and they don't specify the character set, research says that it's typically ISO-8859-1 // If there is a meta tag, and they don't specify the character set, research says that it's typically ISO-8859-1
if (is_object($debug_object)) {$debug_object->debugLog(2, 'meta content-type tag couldn\'t be parsed. using iso-8859 default.');} if (is_object($debug_object)) {$debug_object->debug_log(2, 'meta content-type tag couldn\'t be parsed. using iso-8859 default.');}
$charset = 'ISO-8859-1'; $charset = 'ISO-8859-1';
} }
} }
@ -1231,14 +1251,19 @@ class simple_html_dom
// If we couldn't find a charset above, then lets try to detect one based on the text we got... // If we couldn't find a charset above, then lets try to detect one based on the text we got...
if (empty($charset)) if (empty($charset))
{ {
// Have php try to detect the encoding from the text given to us. // Use this in case mb_detect_charset isn't installed/loaded on this machine.
$charset = mb_detect_encoding($this->root->plaintext . "ascii", $encoding_list = array( "UTF-8", "CP1252" ) ); $charset = false;
if (is_object($debug_object)) {$debug_object->debugLog(2, 'mb_detect found: ' . $charset);} if (function_exists('mb_detect_encoding'))
{
// Have php try to detect the encoding from the text given to us.
$charset = mb_detect_encoding($this->root->plaintext . "ascii", $encoding_list = array( "UTF-8", "CP1252" ) );
if (is_object($debug_object)) {$debug_object->debug_log(2, 'mb_detect found: ' . $charset);}
}
// and if this doesn't work... then we need to just wrongheadedly assume it's UTF-8 so that we can move on - cause this will usually give us most of what we need... // and if this doesn't work... then we need to just wrongheadedly assume it's UTF-8 so that we can move on - cause this will usually give us most of what we need...
if ($charset === false) if ($charset === false)
{ {
if (is_object($debug_object)) {$debug_object->debugLog(2, 'since mb_detect failed - using default of utf-8');} if (is_object($debug_object)) {$debug_object->debug_log(2, 'since mb_detect failed - using default of utf-8');}
$charset = 'UTF-8'; $charset = 'UTF-8';
} }
} }
@ -1246,11 +1271,11 @@ class simple_html_dom
// Since CP1252 is a superset, if we get one of it's subsets, we want it instead. // Since CP1252 is a superset, if we get one of it's subsets, we want it instead.
if ((strtolower($charset) == strtolower('ISO-8859-1')) || (strtolower($charset) == strtolower('Latin1')) || (strtolower($charset) == strtolower('Latin-1'))) if ((strtolower($charset) == strtolower('ISO-8859-1')) || (strtolower($charset) == strtolower('Latin1')) || (strtolower($charset) == strtolower('Latin-1')))
{ {
if (is_object($debug_object)) {$debug_object->debugLog(2, 'replacing ' . $charset . ' with CP1252 as its a superset');} if (is_object($debug_object)) {$debug_object->debug_log(2, 'replacing ' . $charset . ' with CP1252 as its a superset');}
$charset = 'CP1252'; $charset = 'CP1252';
} }
if (is_object($debug_object)) {$debug_object->debugLog(1, 'EXIT - ' . $charset);} if (is_object($debug_object)) {$debug_object->debug_log(1, 'EXIT - ' . $charset);}
return $this->_charset = $charset; return $this->_charset = $charset;
} }
@ -1616,14 +1641,14 @@ class simple_html_dom
protected function remove_noise($pattern, $remove_tag=false) protected function remove_noise($pattern, $remove_tag=false)
{ {
global $debug_object; global $debug_object;
if (is_object($debug_object)) { $debug_object->debugLogEntry(1); } if (is_object($debug_object)) { $debug_object->debug_log_entry(1); }
$count = preg_match_all($pattern, $this->doc, $matches, PREG_SET_ORDER|PREG_OFFSET_CAPTURE); $count = preg_match_all($pattern, $this->doc, $matches, PREG_SET_ORDER|PREG_OFFSET_CAPTURE);
for ($i=$count-1; $i>-1; --$i) for ($i=$count-1; $i>-1; --$i)
{ {
$key = '___noise___'.sprintf('% 5d', count($this->noise)+1000); $key = '___noise___'.sprintf('% 5d', count($this->noise)+1000);
if (is_object($debug_object)) { $debug_object->debugLog(2, 'key is: ' . $key); } if (is_object($debug_object)) { $debug_object->debug_log(2, 'key is: ' . $key); }
$idx = ($remove_tag) ? 0 : 1; $idx = ($remove_tag) ? 0 : 1;
$this->noise[$key] = $matches[$i][$idx][0]; $this->noise[$key] = $matches[$i][$idx][0];
$this->doc = substr_replace($this->doc, $key, $matches[$i][$idx][1], strlen($matches[$i][$idx][0])); $this->doc = substr_replace($this->doc, $key, $matches[$i][$idx][1], strlen($matches[$i][$idx][0]));
@ -1641,7 +1666,7 @@ class simple_html_dom
function restore_noise($text) function restore_noise($text)
{ {
global $debug_object; global $debug_object;
if (is_object($debug_object)) { $debug_object->debugLogEntry(1); } if (is_object($debug_object)) { $debug_object->debug_log_entry(1); }
while (($pos=strpos($text, '___noise___'))!==false) while (($pos=strpos($text, '___noise___'))!==false)
{ {
@ -1649,7 +1674,7 @@ class simple_html_dom
if (strlen($text) > $pos+15) if (strlen($text) > $pos+15)
{ {
$key = '___noise___'.$text[$pos+11].$text[$pos+12].$text[$pos+13].$text[$pos+14].$text[$pos+15]; $key = '___noise___'.$text[$pos+11].$text[$pos+12].$text[$pos+13].$text[$pos+14].$text[$pos+15];
if (is_object($debug_object)) { $debug_object->debugLog(2, 'located key of: ' . $key); } if (is_object($debug_object)) { $debug_object->debug_log(2, 'located key of: ' . $key); }
if (isset($this->noise[$key])) if (isset($this->noise[$key]))
{ {
@ -1674,7 +1699,7 @@ class simple_html_dom
function search_noise($text) function search_noise($text)
{ {
global $debug_object; global $debug_object;
if (is_object($debug_object)) { $debug_object->debugLogEntry(1); } if (is_object($debug_object)) { $debug_object->debug_log_entry(1); }
foreach($this->noise as $noiseElement) foreach($this->noise as $noiseElement)
{ {