diff --git a/src/Wallabag/CoreBundle/Helper/ContentProxy.php b/src/Wallabag/CoreBundle/Helper/ContentProxy.php index 3fe31c2c7..d4ea608f3 100644 --- a/src/Wallabag/CoreBundle/Helper/ContentProxy.php +++ b/src/Wallabag/CoreBundle/Helper/ContentProxy.php @@ -53,6 +53,7 @@ class ContentProxy if ((empty($content) || false === $this->validateContent($content)) && false === $disableContentUpdate) { $fetchedContent = $this->graby->fetchContent($url); + $fetchedContent['title'] = $this->sanitizeContentTitle($fetchedContent['title'], $fetchedContent['content_type']); // when content is imported, we have information in $content // in case fetching content goes bad, we'll keep the imported information instead of overriding them @@ -176,6 +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. diff --git a/tests/Wallabag/CoreBundle/Helper/ContentProxyTest.php b/tests/Wallabag/CoreBundle/Helper/ContentProxyTest.php index 51df8de1b..3f3c60d0f 100644 --- a/tests/Wallabag/CoreBundle/Helper/ContentProxyTest.php +++ b/tests/Wallabag/CoreBundle/Helper/ContentProxyTest.php @@ -531,6 +531,250 @@ class ContentProxyTest extends TestCase $this->assertSame('1.1.1.1', $entry->getDomainName()); } + public function testWebsiteWithValidUTF8Title_doNothing() + { + // You can use https://www.online-toolz.com/tools/text-hex-convertor.php to convert UTF-8 text <=> hex + // See http://graphemica.com for more info about the characters + // '😻ℤz' (U+1F63B or F09F98BB; U+2124 or E284A4; U+007A or 7A) in hexadecimal and UTF-8 + $actualTitle = $this->hexToStr('F09F98BB' . 'E284A4' . '7A'); + + $tagger = $this->getTaggerMock(); + $tagger->expects($this->once()) + ->method('tag'); + + $graby = $this->getMockBuilder('Graby\Graby') + ->setMethods(['fetchContent']) + ->disableOriginalConstructor() + ->getMock(); + + $graby->expects($this->any()) + ->method('fetchContent') + ->willReturn([ + 'html' => false, + 'title' => $actualTitle, + 'url' => '', + 'content_type' => 'text/html', + 'language' => '', + ]); + + $proxy = new ContentProxy($graby, $tagger, $this->getValidator(), $this->getLogger(), $this->fetchingErrorMessage); + $entry = new Entry(new User()); + $proxy->updateEntry($entry, 'http://0.0.0.0'); + + // '😻ℤz' (U+1F63B or F09F98BB; U+2124 or E284A4; U+007A or 7A) in hexadecimal and UTF-8 + $expectedTitle = 'F09F98BB' . 'E284A4' . '7A'; + $this->assertSame($expectedTitle, $this->strToHex($entry->getTitle())); + } + + public function testWebsiteWithInvalidUTF8Title_removeInvalidCharacter() + { + // See http://graphemica.com for more info about the characters + // 'a€b' (61;80;62) in hexadecimal and WINDOWS-1252 - but 80 is a invalid UTF-8 character. + // The correct UTF-8 € character (U+20AC) is E282AC + $actualTitle = $this->hexToStr('61' . '80' . '62'); + + $tagger = $this->getTaggerMock(); + $tagger->expects($this->once()) + ->method('tag'); + + $graby = $this->getMockBuilder('Graby\Graby') + ->setMethods(['fetchContent']) + ->disableOriginalConstructor() + ->getMock(); + + $graby->expects($this->any()) + ->method('fetchContent') + ->willReturn([ + 'html' => false, + 'title' => $actualTitle, + 'url' => '', + 'content_type' => 'text/html', + 'language' => '', + ]); + + $proxy = new ContentProxy($graby, $tagger, $this->getValidator(), $this->getLogger(), $this->fetchingErrorMessage); + $entry = new Entry(new User()); + $proxy->updateEntry($entry, 'http://0.0.0.0'); + + // 'ab' (61;62) because all invalid UTF-8 character (like 80) are removed + $expectedTitle = '61' . '62'; + $this->assertSame($expectedTitle, $this->strToHex($entry->getTitle())); + } + + public function testPdfWithUTF16BETitle_convertToUTF8() + { + // See http://graphemica.com for more info about the characters + // '😻' (U+1F63B;D83DDE3B) in hexadecimal and as UTF16BE + $actualTitle = $this->hexToStr('D83DDE3B'); + + $tagger = $this->getTaggerMock(); + $tagger->expects($this->once()) + ->method('tag'); + + $graby = $this->getMockBuilder('Graby\Graby') + ->setMethods(['fetchContent']) + ->disableOriginalConstructor() + ->getMock(); + + $graby->expects($this->any()) + ->method('fetchContent') + ->willReturn([ + 'html' => false, + 'title' => $actualTitle, + 'url' => '', + 'content_type' => 'application/pdf', + 'language' => '', + ]); + + $proxy = new ContentProxy($graby, $tagger, $this->getValidator(), $this->getLogger(), $this->fetchingErrorMessage); + $entry = new Entry(new User()); + $proxy->updateEntry($entry, 'http://0.0.0.0'); + + // '😻' (U+1F63B or F09F98BB) in hexadecimal and UTF-8 + $expectedTitle = 'F09F98BB'; + $this->assertSame($expectedTitle, $this->strToHex($entry->getTitle())); + } + + public function testPdfWithUTF8Title_doNothing() + { + // See http://graphemica.com for more info about the characters + // '😻' (U+1F63B;D83DDE3B) in hexadecimal and as UTF8 + $actualTitle = $this->hexToStr('F09F98BB'); + + $tagger = $this->getTaggerMock(); + $tagger->expects($this->once()) + ->method('tag'); + + $graby = $this->getMockBuilder('Graby\Graby') + ->setMethods(['fetchContent']) + ->disableOriginalConstructor() + ->getMock(); + + $graby->expects($this->any()) + ->method('fetchContent') + ->willReturn([ + 'html' => false, + 'title' => $actualTitle, + 'url' => '', + 'content_type' => 'application/pdf', + 'language' => '', + ]); + + $proxy = new ContentProxy($graby, $tagger, $this->getValidator(), $this->getLogger(), $this->fetchingErrorMessage); + $entry = new Entry(new User()); + $proxy->updateEntry($entry, 'http://0.0.0.0'); + + // '😻' (U+1F63B or F09F98BB) in hexadecimal and UTF-8 + $expectedTitle = 'F09F98BB'; + $this->assertSame($expectedTitle, $this->strToHex($entry->getTitle())); + } + + public function testPdfWithWINDOWS1252Title_convertToUTF8() + { + // See http://graphemica.com for more info about the characters + // '€' (80) in hexadecimal and WINDOWS-1252 + $actualTitle = $this->hexToStr('80'); + + $tagger = $this->getTaggerMock(); + $tagger->expects($this->once()) + ->method('tag'); + + $graby = $this->getMockBuilder('Graby\Graby') + ->setMethods(['fetchContent']) + ->disableOriginalConstructor() + ->getMock(); + + $graby->expects($this->any()) + ->method('fetchContent') + ->willReturn([ + 'html' => false, + 'title' => $actualTitle, + 'url' => '', + 'content_type' => 'application/pdf', + 'language' => '', + ]); + + $proxy = new ContentProxy($graby, $tagger, $this->getValidator(), $this->getLogger(), $this->fetchingErrorMessage); + $entry = new Entry(new User()); + $proxy->updateEntry($entry, 'http://0.0.0.0'); + + // '€' (U+20AC or E282AC) in hexadecimal and UTF-8 + $expectedTitle = 'E282AC'; + $this->assertSame($expectedTitle, $this->strToHex($entry->getTitle())); + } + + public function testPdfWithInvalidCharacterInTitle_removeInvalidCharacter() + { + // See http://graphemica.com for more info about the characters + // '😻ℤ�z' (U+1F63B or F09F98BB; U+2124 or E284A4; invalid character 81; U+007A or 7A) in hexadecimal and UTF-8 + // 0x81 is not a valid character for UTF16, UTF8 and WINDOWS-1252 + $actualTitle = $this->hexToStr('F09F98BB' . 'E284A4' . '81' . '7A'); + + $tagger = $this->getTaggerMock(); + $tagger->expects($this->once()) + ->method('tag'); + + $graby = $this->getMockBuilder('Graby\Graby') + ->setMethods(['fetchContent']) + ->disableOriginalConstructor() + ->getMock(); + + $graby->expects($this->any()) + ->method('fetchContent') + ->willReturn([ + 'html' => false, + 'title' => $actualTitle, + 'url' => '', + 'content_type' => 'application/pdf', + 'language' => '', + ]); + + $proxy = new ContentProxy($graby, $tagger, $this->getValidator(), $this->getLogger(), $this->fetchingErrorMessage); + $entry = new Entry(new User()); + $proxy->updateEntry($entry, 'http://0.0.0.0'); + + // '😻ℤz' (U+1F63B or F09F98BB; U+2124 or E284A4; U+007A or 7A) in hexadecimal and UTF-8 + // the 0x81 (represented by �) is invalid for UTF16, UTF8 and WINDOWS-1252 and is removed + $expectedTitle = 'F09F98BB' . 'E284A4' . '7A'; + $this->assertSame($expectedTitle, $this->strToHex($entry->getTitle())); + } + + /** + * https://stackoverflow.com/a/18506801. + * + * @param $string + * + * @return string + */ + private function strToHex($string) + { + $hex = ''; + for ($i = 0; $i < \strlen($string); ++$i) { + $ord = \ord($string[$i]); + $hexCode = dechex($ord); + $hex .= substr('0' . $hexCode, -2); + } + + return strtoupper($hex); + } + + /** + * https://stackoverflow.com/a/18506801. + * + * @param $hex + * + * @return string + */ + private function hexToStr($hex) + { + $string = ''; + for ($i = 0; $i < \strlen($hex) - 1; $i += 2) { + $string .= \chr(hexdec($hex[$i] . $hex[$i + 1])); + } + + return $string; + } + private function getTaggerMock() { return $this->getMockBuilder(RuleBasedTagger::class)