mirrors/wallabag
1
0
Fork 0
mirror of https://github.com/wallabag/wallabag.git synced 2025-01-08 16:05:25 +00:00
Code Issues Projects Releases Wiki Activity

Run php-cs-fixer for fixing coding standard issues

Browse source
This commit is contained in:
Tobi823 2018-09-23 22:20:43 +02:00
parent d64139d812
commit 83f1c3274f
1 changed files with 53 additions and 41 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

94
src/Wallabag/CoreBundle/Helper/ContentProxy.php
View file

@ -69,47 +69,6 @@ class ContentProxy
$this->stockEntry($entry, $content);
}
/**
* Try to sanitize the title of the fetched content from wrong character encodings and invalid UTF-8 character.
* @param $title
* @param $contentType
* @return string
*/
private function sanitizeContentTitle($title, $contentType) {
if ('application/pdf' === $contentType) {
$title = $this->convertPdfEncodingToUTF8($title);
}
return $this->sanitizeUTF8Text($title);
}
/**
* If the title from the fetched content comes from a PDF, then its very possible that the character encoding is not
* UTF-8. This methods tries to identify the character encoding and translate the title to UTF-8.
* @param $title
* @return string (maybe contains invalid UTF-8 character)
*/
private function convertPdfEncodingToUTF8($title) {
// first try UTF-8 because its easier to detect its present/absence
foreach (array('UTF-8', 'UTF-16BE', 'WINDOWS-1252') as $encoding) {
if (mb_check_encoding($title, $encoding)) {
return mb_convert_encoding($title, 'UTF-8', $encoding);
}
}
return $title;
}
/**
* Remove invalid UTF-8 characters from the given string.
* @param String $rawText
* @return string
*/
private function sanitizeUTF8Text($rawText) {
if (mb_check_encoding($rawText, 'UTF-8')) {
return $rawText;
}
return iconv("UTF-8", "UTF-8//IGNORE", $rawText);
}
/**
* Use a Symfony validator to ensure the language is well formatted.
*
@ -218,6 +177,59 @@ class ContentProxy
$entry->setTitle($path);
}
/**
* Try to sanitize the title of the fetched content from wrong character encodings and invalid UTF-8 character.
*
* @param $title
* @param $contentType
*
* @return string
*/
private function sanitizeContentTitle($title, $contentType)
{
if ('application/pdf' === $contentType) {
$title = $this->convertPdfEncodingToUTF8($title);
}
return $this->sanitizeUTF8Text($title);
}
/**
* If the title from the fetched content comes from a PDF, then its very possible that the character encoding is not
* UTF-8. This methods tries to identify the character encoding and translate the title to UTF-8.
*
* @param $title
*
* @return string (maybe contains invalid UTF-8 character)
*/
private function convertPdfEncodingToUTF8($title)
{
// first try UTF-8 because its easier to detect its present/absence
foreach (['UTF-8', 'UTF-16BE', 'WINDOWS-1252'] as $encoding) {
if (mb_check_encoding($title, $encoding)) {
return mb_convert_encoding($title, 'UTF-8', $encoding);
}
}
return $title;
}
/**
* Remove invalid UTF-8 characters from the given string.
*
* @param string $rawText
*
* @return string
*/
private function sanitizeUTF8Text($rawText)
{
if (mb_check_encoding($rawText, 'UTF-8')) {
return $rawText;
}
return iconv('UTF-8', 'UTF-8//IGNORE', $rawText);
}
/**
* Stock entry with fetched or imported content.
* Will fall back to OpenGraph data if available.