infong
/
php-readability
mirror of https://github.com/j0k3r/php-readability
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #30 from j0k3r/avoid-php-warning

Browse Source 
Avoid php warning
pull/32/head 1.1.8
Jérémy Benoist 9 years ago committed by GitHub
parent 6def743902 85fb92a042
commit 5ac82f3c87
2 changed files with 49 additions and 42 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Unified View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 89
      src/JSLikeHTMLElement.php
  2. 2
      tests/ReadabilityTest.php

89
src/JSLikeHTMLElement.php
Unescape View File

@ -45,52 +45,59 @@ class JSLikeHTMLElement extends \DOMElement
*/ */
public function __set($name, $value) public function __set($name, $value)
{ {
if ($name === 'innerHTML') { if ($name !== 'innerHTML') {
// first, empty the element $trace = debug_backtrace();
for ($x = $this->childNodes->length - 1; $x >= 0; --$x) { trigger_error('Undefined property via __set(): ' . $name . ' in ' . $trace[0]['file'] . ' on line ' . $trace[0]['line'], E_USER_NOTICE);
$this->removeChild($this->childNodes->item($x));
return;
}
// first, empty the element
for ($x = $this->childNodes->length - 1; $x >= 0; --$x) {
$this->removeChild($this->childNodes->item($x));
}
// $value holds our new inner HTML
$value = trim($value);
if (empty($value)) {
return;
}
// ensure bad entity won't generate warning
$previousError = libxml_use_internal_errors(true);
$f = $this->ownerDocument->createDocumentFragment();
// appendXML() expects well-formed markup (XHTML)
$result = $f->appendXML($value);
if ($result) {
if ($f->hasChildNodes()) {
$this->appendChild($f);
} }
} else {
// $value is probably ill-formed
$f = new \DOMDocument();
$value = mb_convert_encoding($value, 'HTML-ENTITIES', 'UTF-8');
// Using <htmlfragment> will generate a warning, but so will bad HTML
// (and by this point, bad HTML is what we've got).
// We use it (and suppress the warning) because an HTML fragment will
// be wrapped around <html><body> tags which we don't really want to keep.
// Note: despite the warning, if loadHTML succeeds it will return true.
$result = $f->loadHTML('<htmlfragment>' . $value . '</htmlfragment>');
if ($result) {
$import = $f->getElementsByTagName('htmlfragment')->item(0);
// $value holds our new inner HTML foreach ($import->childNodes as $child) {
if ($value !== '') { $importedNode = $this->ownerDocument->importNode($child, true);
$f = $this->ownerDocument->createDocumentFragment(); $this->appendChild($importedNode);
// appendXML() expects well-formed markup (XHTML)
// @ to suppress PHP warnings
$result = @$f->appendXML($value);
if ($result) {
if ($f->hasChildNodes()) {
$this->appendChild($f);
}
} else {
// $value is probably ill-formed
$f = new \DOMDocument();
$value = mb_convert_encoding($value, 'HTML-ENTITIES', 'UTF-8');
// Using <htmlfragment> will generate a warning, but so will bad HTML
// (and by this point, bad HTML is what we've got).
// We use it (and suppress the warning) because an HTML fragment will
// be wrapped around <html><body> tags which we don't really want to keep.
// Note: despite the warning, if loadHTML succeeds it will return true.
$result = @$f->loadHTML('<htmlfragment>' . $value . '</htmlfragment>');
if ($result) {
$import = $f->getElementsByTagName('htmlfragment')->item(0);
foreach ($import->childNodes as $child) {
$importedNode = $this->ownerDocument->importNode($child, true);
$this->appendChild($importedNode);
}
} else {
// oh well, we tried, we really did. :(
// this element is now empty
}
} }
} }
} else {
$trace = debug_backtrace();
trigger_error('Undefined property via __set(): ' . $name . ' in ' . $trace[0]['file'] . ' on line ' . $trace[0]['line'], E_USER_NOTICE);
} }
libxml_clear_errors();
libxml_use_internal_errors($previousError);
} }
/** /**

2
tests/ReadabilityTest.php
Unescape View File

@ -443,7 +443,7 @@ class ReadabilityTest extends \PHPUnit_Framework_TestCase
public function testPostFilters() public function testPostFilters()
{ {
$readability = $this->getReadability('<div>' . str_repeat('<p>This <b>is</b> the awesome content :)</p>', 7) . '</div>', 'http://0.0.0.0'); $readability = $this->getReadability('<div>' . str_repeat('<p>This <b>is</b> the awesome content :)</p>', 10) . '</div>', 'http://0.0.0.0');
$readability->addPostFilter('!<strong[^>]*>(.*?)</strong>!is', ''); $readability->addPostFilter('!<strong[^>]*>(.*?)</strong>!is', '');
$res = $readability->init(); $res = $readability->init();

Write Preview
Loading…
Cancel
Save