improve handling of unknown non-UTF8 EAC logs

src/Parser/EAC/Translator.php

@@ -54,6 +54,7 @@ class Translator
             file_get_contents(__DIR__ . DIRECTORY_SEPARATOR . 'languages' . DIRECTORY_SEPARATOR . 'master.json'),
             true
         );
+
         foreach ($languages as $code => $lang) {
             foreach ($lang['eac_strings'] as $eac_string) {
                 if (preg_match('/' . preg_quote($eac_string, "/") . '/ui', $log) === 1) {

src/Util.php

@@ -41,6 +41,12 @@ class Util
             if ($results['charset'] !== 'utf-8' && $results['confidence'] > 0.7) {
                 // $log = mb_convert_encoding($log, 'UTF-8', $results['charset']);
                 $log = iconv($results['charset'], 'UTF-8', $log);
+            } elseif ($results['charset'] !== 'utf-8' && $results['confidence'] > 0.3) {
+                // If we've got a poor confidence on our decoding, we just use a generic
+                // ISO-8859-1 as that covers a decent range of things that people would
+                // inadvertently re-encode a log into. I seriously cannot express how
+                // much I hate how EAC does not use always UTF-8.
+                $log = iconv('ISO-8859-1', 'UTF-8', $log);
             }
         }
         return $log;

tests/Parser/EAC/TranslatorTest.php

@@ -6,6 +6,7 @@ namespace OrpheusNET\Logchecker\Parser\EAC;
 
 use FilesystemIterator;
 use OrpheusNET\Logchecker\Exception\UnknownLanguageException;
+use OrpheusNET\Logchecker\Util;
 use PHPUnit\Framework\TestCase;
 
 class TranslatorTest extends TestCase
@@ -51,4 +52,33 @@ class TranslatorTest extends TestCase
             )
         );
     }
+
+    public function englishLogProvider()
+    {
+        return array_map(
+            function ($file) {
+                return [$file];
+            },
+            array_filter(
+                scandir(implode(DIRECTORY_SEPARATOR, [__DIR__, '..', '..', 'logs', 'eac', 'originals'])),
+                function ($file) {
+                    return substr($file, 0, 2) === 'en';
+                }
+            )
+        );
+    }
+
+    /**
+     * @dataProvider englishLogProvider
+     */
+    public function testEnglishLanguage(string $file)
+    {
+        $logPath = implode(DIRECTORY_SEPARATOR, [__DIR__, '..', '..', 'logs', 'eac', 'originals', $file]);
+        $log = file_get_contents($logPath);
+        $log = Util::decodeEncoding($log, $logPath);
+        $langDetails = Translator::getLanguage($log);
+        $this->assertSame('en', $langDetails['code']);
+        $this->assertSame('English', $langDetails['name']);
+        $this->assertSame('English', $langDetails['name_english']);
+    }
 }

tests/logs/eac/originals/en_8.log

@@ -0,0 +1,194 @@
+EAC extraction logfile from 20. November 2006, 20:20 for CD
+Banco del Mutuo Soccorso / Seguendo le Tracce
+
+Used drive  : _NEC    DVD+-RW ND-6450A   Adapter: 1  ID: 0
+Read mode   : Secure with NO C2, accurate stream, disable cache
+Read offset correction : 594
+Overread into Lead-In and Lead-Out : No
+
+Used output format : Internal WAV Routines
+                     44.100 Hz; 16 Bit; Stereo
+
+Other options      : 
+    Fill up missing offset samples with silence : Yes
+    Delete leading and trailing silent blocks : No
+    Installed external ASPI interface
+
+
+Track  1
+     Filename C:\Documents and Settings\user\Documenti\file flac\01R.I.P. (english version).wav
+
+     Pre-gap length  0:00:02.00
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC EC43BAD3
+     Copy CRC EC43BAD3
+     Copy OK
+
+Track  2
+     Filename C:\Documents and Settings\user\Documenti\file flac\02L'albero del pane.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC 70584B1D
+     Copy CRC 70584B1D
+     Copy OK
+
+Track  3
+     Filename C:\Documents and Settings\user\Documenti\file flac\03La danza dei grandi rettili.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC 4E17BAEF
+     Copy CRC 4E17BAEF
+     Copy OK
+
+Track  4
+     Filename C:\Documents and Settings\user\Documenti\file flac\04Passaggio.wav
+
+     Peak level 44.4 %
+     Track quality 100.0 %
+     Test CRC D76BE99A
+     Copy CRC D76BE99A
+     Copy OK
+
+Track  5
+     Filename C:\Documents and Settings\user\Documenti\file flac\05Non mi rompete.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC 826A9F81
+     Copy CRC 826A9F81
+     Copy OK
+
+Track  6
+     Filename C:\Documents and Settings\user\Documenti\file flac\06Dopo... niente più è lo stesso.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC CFB6C6BA
+     Copy CRC CFB6C6BA
+     Copy OK
+
+Track  7
+     Filename C:\Documents and Settings\user\Documenti\file flac\07Traccia II.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC D7BAF9A2
+     Copy CRC D7BAF9A2
+     Copy OK
+
+Track  8
+     Filename C:\Documents and Settings\user\Documenti\file flac\08Metamorfosi.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC BBB2254A
+     Copy CRC BBB2254A
+     Copy OK
+
+No errors occured
+
+
+End of status report
+
+------------------------------------------------------------
+
+EAC extraction logfile from 20. November 2006, 21:08 for CD
+Banco del Mutuo Soccorso / Seguendo le Tracce
+
+Used drive  : _NEC    DVD+-RW ND-6450A   Adapter: 1  ID: 0
+Read mode   : Secure with NO C2, accurate stream, disable cache
+Read offset correction : 594
+Overread into Lead-In and Lead-Out : No
+
+Used output format : C:\Programmi\FLAC\flac.exe   (User Defined Encoder)
+                     320 kBit/s
+                     Additional command line options : -V --replay-gain -8 -T "artist=%a" -T "title=%t" -T "album=%g" -T "date=%y" -T "tracknumber=%n" -T "genre=%m" %s 
+
+Other options      : 
+    Fill up missing offset samples with silence : Yes
+    Delete leading and trailing silent blocks : No
+    Installed external ASPI interface
+
+
+Track  1
+     Filename C:\Documents and Settings\user\Documenti\file flac\01R.I.P. (english version).wav
+
+     Pre-gap length  0:00:02.00
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC EC43BAD3
+     Copy CRC EC43BAD3
+     Copy OK
+
+Track  2
+     Filename C:\Documents and Settings\user\Documenti\file flac\02L'albero del pane.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC 70584B1D
+     Copy CRC 70584B1D
+     Copy OK
+
+Track  3
+     Filename C:\Documents and Settings\user\Documenti\file flac\03La danza dei grandi rettili.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC 4E17BAEF
+     Copy CRC 4E17BAEF
+     Copy OK
+
+Track  4
+     Filename C:\Documents and Settings\user\Documenti\file flac\04Passaggio.wav
+
+     Peak level 44.4 %
+     Track quality 100.0 %
+     Test CRC D76BE99A
+     Copy CRC D76BE99A
+     Copy OK
+
+Track  5
+     Filename C:\Documents and Settings\user\Documenti\file flac\05Non mi rompete.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC 826A9F81
+     Copy CRC 826A9F81
+     Copy OK
+
+Track  6
+     Filename C:\Documents and Settings\user\Documenti\file flac\06Dopo... niente più è lo stesso.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC CFB6C6BA
+     Copy CRC CFB6C6BA
+     Copy OK
+
+Track  7
+     Filename C:\Documents and Settings\user\Documenti\file flac\07Traccia II.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC D7BAF9A2
+     Copy CRC D7BAF9A2
+     Copy OK
+
+Track  8
+     Filename C:\Documents and Settings\user\Documenti\file flac\08Metamorfosi.wav
+
+     Peak level 97.7 %
+     Track quality 100.0 %
+     Test CRC BBB2254A
+     Copy CRC BBB2254A
+     Copy OK
+
+No errors occured
+
+
+End of status report