Merge pull request #36 from matecat/new-sprintf-regex

Sprintf regex refactoring
matecat · Aug 10, 2023 · 903773a · 903773a
2 parents 0d451eb + bd46e92
commit 903773a
Show file tree

Hide file tree

Showing 8 changed files with 17 additions and 49 deletions.
diff --git a/src/Filters/Sprintf/language/de-AT/not_allowed.php b/src/Filters/Sprintf/language/de-AT/not_allowed.php
diff --git a/src/Filters/Sprintf/language/de-CH/not_allowed.php b/src/Filters/Sprintf/language/de-CH/not_allowed.php
diff --git a/src/Filters/Sprintf/language/de-DE/not_allowed.php b/src/Filters/Sprintf/language/de-DE/not_allowed.php
diff --git a/src/Filters/SprintfToPH.php b/src/Filters/SprintfToPH.php
@@ -33,6 +33,12 @@ public function __construct() {
      * 20%-zar - ignored
      *</code>
      *
+     * @see
+     * - https://developer.apple.com/library/archive/documentation/Cocoa/Conceptual/Strings/Articles/formatSpecifiers.html#//apple_ref/doc/uid/TP40004265-SW1
+     * - https://en.cppreference.com/w/c/io/fprintf
+     * - https://www.php.net/manual/en/function.sprintf.php
+     * - https://www.w3resource.com/c-programming/stdio/c_library_method_sprintf.php
+     *
      * @param $segment
      *
      * @return string
@@ -41,12 +47,14 @@ public function transform( $segment ) {
 
         $sprintfLocker = new SprintfLocker( $this->pipeline->getSource(), $this->pipeline->getTarget() );
 
-        //placeholding
+        // placeholding
         $segment = $sprintfLocker->lock( $segment );
 
         // Octal parsing is disabled due to Hungarian percentages 20%-os
-        // preg_match_all( '/(?:\x25\x25)|(\x25(?:(?:[1-9]\d*)\$|\((?:[^\)]+)\))?(?:\+)?(?:0|\'[^$])?(?:-)?(?:\d+)?(?:\.(?:\d+))?(?:[b-fiosuxX]))/', $segment, $vars, PREG_SET_ORDER );
-        preg_match_all( '/(?:\x25\x25)|(\x25(?:(?:[1-9]\d*)\$|\((?:[^\)]+)\))?(?:\+)?(?:-)?(?:0|\'[^$])?(?:\d+)?(?:\.(?:\d+))?(?:[b-fiosuxX]))/', $segment, $vars, PREG_SET_ORDER );
+        $regex = '/(?:\x25\x25)|(\x25(?:(?:[1-9]\d*)\$|\((?:[^\)]+)\))?(?:\+)?(?:0|[+-]?\'[^$])?(?:-)?(?:\d+)?(?:\.(?:\d+))?((?:[hjlqtzL]{0,2}[ac-giopsuxAC-GOSUX]{1})(?![\d\w])|(?:#@[\w]+@)|(?:@)))/';
+
+
+        preg_match_all( $regex, $segment, $vars, PREG_SET_ORDER );
         foreach ( $vars as $pos => $variable ) {
 
             //replace subsequent elements excluding already encoded

diff --git a/src/MateCatFilter.php b/src/MateCatFilter.php
@@ -40,6 +40,7 @@
 use Matecat\SubFiltering\Filters\StandardPHToMateCatCustomPH;
 use Matecat\SubFiltering\Filters\SubFilteredPhToHtml;
 use Matecat\SubFiltering\Filters\TwigToPh;
+use Matecat\SubFiltering\Filters\Variables;
 
 /**
  * Class Filter
@@ -160,12 +161,11 @@ public function fromLayer0ToLayer1( $segment ) {
         $channel->addLast( new LtGtDecode() );
         $channel->addLast( new HtmlToPh() );
         $channel->addLast( new TwigToPh() );
+        $channel->addLast( new Variables() );
         $channel->addLast( new RubyOnRailsI18n() );
         $channel->addLast( new Snails() );
         $channel->addLast( new DoubleSquareBrackets() );
-        //$channel->addLast( new DoubleUnderscore() );
         $channel->addLast( new DollarCurlyBrackets() );
-        $channel->addLast( new PercentSnail() );
         $channel->addLast( new PercentNumberSnail() );
         $channel->addLast( new Percentages() );
         $channel->addLast( new SquareSprintf() );

diff --git a/src/MyMemoryFilter.php b/src/MyMemoryFilter.php
@@ -64,19 +64,17 @@ public function fromLayer0ToLayer1( $segment, $cid = null ) {
         $channel->addLast( new PlaceHoldXliffTags() );
         $channel->addLast( new LtGtDecode() );
         $channel->addLast( new HtmlToPh() );
+        $channel->addLast( new TwigToPh() );
+        $channel->addLast( new Variables() );
 
         if ( $cid == 'airbnb' ) {
-            $channel->addLast( new Variables() );
             $channel->addLast( new SmartCounts() );
         }
 
-        $channel->addLast( new TwigToPh() );
         $channel->addLast( new RubyOnRailsI18n() );
         $channel->addLast( new Snails() );
         $channel->addLast( new DoubleSquareBrackets() );
-        //$channel->addLast( new DoubleUnderscore() );
         $channel->addLast( new DollarCurlyBrackets() );
-
         $channel->addLast( new PercentSnail() );
         $channel->addLast( new PercentNumberSnail() );
         $channel->addLast( new Percentages() );

diff --git a/tests/MateCatSubFilteringTest.php b/tests/MateCatSubFilteringTest.php
@@ -871,7 +871,7 @@ public function testPercentSnailSyntax() {
         $filter = $this->getFilterInstance();
 
         $db_segment      = 'This string: %@ is a IOS placeholder %@.';
-        $segment_from_UI = 'This string: <ph id="mtc_1" ctype="' . CTypeEnum::PERCENT_SNAILS . '" equiv-text="base64:JUA="/> is a IOS placeholder <ph id="mtc_2" ctype="' . CTypeEnum::PERCENT_SNAILS . '" equiv-text="base64:JUA="/>.';
+        $segment_from_UI = 'This string: <ph id="mtc_1" ctype="' . CTypeEnum::SPRINTF . '" equiv-text="base64:JUA="/> is a IOS placeholder <ph id="mtc_2" ctype="' . CTypeEnum::SPRINTF . '" equiv-text="base64:JUA="/>.';
 
         $this->assertEquals( $db_segment, $filter->fromLayer1ToLayer0( $segment_from_UI ) );
         $this->assertEquals( $segment_from_UI, $filter->fromLayer0ToLayer1( $db_segment ) );
@@ -891,7 +891,7 @@ public function testWithMixedPercentTags() {
         $filter = $this->getFilterInstance();
 
         $db_segment      = 'This string contains all these tags: %-4d %@ %12$@ %{{|discount|}} {% if count &lt; 3 %} but not this %placeholder%';
-        $segment_from_UI = 'This string contains all these tags: <ph id="mtc_1" ctype="' . CTypeEnum::SPRINTF . '" equiv-text="base64:JS00ZA=="/> <ph id="mtc_2" ctype="' . CTypeEnum::PERCENT_SNAILS . '" equiv-text="base64:JUA="/> <ph id="mtc_3" ctype="' . CTypeEnum::PERCENT_NUMBER_SNAILS . '" equiv-text="base64:JTEyJEA="/> %<ph id="mtc_4" ctype="' . CTypeEnum::TWIG . '" equiv-text="base64:e3t8ZGlzY291bnR8fX0="/> <ph id="mtc_5" ctype="' . CTypeEnum::TWIG . '" equiv-text="base64:eyUgaWYgY291bnQgJmx0OyAzICV9"/> but not this %placeholder%';
+        $segment_from_UI = 'This string contains all these tags: <ph id="mtc_1" ctype="' . CTypeEnum::SPRINTF . '" equiv-text="base64:JS00ZA=="/> <ph id="mtc_2" ctype="' . CTypeEnum::SPRINTF . '" equiv-text="base64:JUA="/> <ph id="mtc_3" ctype="' . CTypeEnum::PERCENT_NUMBER_SNAILS . '" equiv-text="base64:JTEyJEA="/> %<ph id="mtc_4" ctype="' . CTypeEnum::TWIG . '" equiv-text="base64:e3t8ZGlzY291bnR8fX0="/> <ph id="mtc_5" ctype="' . CTypeEnum::TWIG . '" equiv-text="base64:eyUgaWYgY291bnQgJmx0OyAzICV9"/> but not this %placeholder%';
 
         $this->assertEquals( $db_segment, $filter->fromLayer1ToLayer0( $segment_from_UI ) );
         $this->assertEquals( $segment_from_UI, $filter->fromLayer0ToLayer1( $db_segment ) );

diff --git a/tests/SprintfTest.php b/tests/SprintfTest.php
@@ -96,14 +96,6 @@ private function runTests($notAllowed, $languageToTest)
 
             $this->assertEquals($segment, $segmentL1);
             $this->assertEquals($segment, $segmentL2);
-
-            $filter = MateCatFilter::getInstance( new FeatureSet(), 'en-US', 'it-IT' );
-
-            $segmentL1 = $filter->fromLayer0ToLayer1( $segment );
-            $segmentL2 = $filter->fromLayer0ToLayer2( $segment );
-
-            $this->assertNotEquals($segment, $segmentL1);
-            $this->assertNotEquals($segment, $segmentL2);
         }
     }
 }