Sdlxliff entities (#93)

* Sdlxliff entities * Stop adding <count-group> in sdlxliff files
matecat · Sep 5, 2024 · 80b17ae · 80b17ae
1 parent 4e66a6b
commit 80b17ae
Show file tree

Hide file tree

Showing 4 changed files with 122 additions and 1 deletion.
diff --git a/src/XliffReplacer/Xliff12.php b/src/XliffReplacer/Xliff12.php
@@ -238,7 +238,7 @@ protected function rebuildMarks( array $seg, string $translation ): string {
      *
      * @return string
      */
-    private function createTargetTag( string $translation, string $stateProp ): string {
+    protected function createTargetTag( string $translation, string $stateProp ): string {
         $targetLang = ' xml:lang="' . $this->targetLang . '"';
         $tag        = "<target $targetLang $stateProp>$translation</target>";
         $tag        .= "\n<count-group name=\"$this->currentTransUnitId\"><count count-type=\"x-matecat-raw\">" . $this->counts[ 'raw_word_count' ] . "</count><count count-type=\"x-matecat-weighted\">" . $this->counts[ 'eq_word_count' ] . '</count></count-group>';

diff --git a/src/XliffReplacer/XliffSdl.php b/src/XliffReplacer/XliffSdl.php
@@ -13,6 +13,18 @@
 
 class XliffSdl extends Xliff12 {
 
+    /**
+     * @var array
+     */
+    protected array $nodesToBuffer = [
+        'source',
+        'seg-source',
+        'note',
+        'context-group',
+        'tag-defs',
+        'cxt-defs'
+    ];
+
     /**
      * @inheritDoc
      */
@@ -100,4 +112,17 @@ protected function rebuildMarks( array $seg, string $translation ): string {
 
     }
 
+    /**
+     * This function creates a <target>
+     *
+     * @param string $translation
+     * @param string $stateProp
+     *
+     * @return string
+     */
+    protected function createTargetTag( string $translation, string $stateProp ): string {
+        $targetLang = ' xml:lang="' . $this->targetLang . '"';
+
+        return "<target $targetLang $stateProp>$translation</target>";
+    }
 }
diff --git a/tests/XliffReplacerTest.php b/tests/XliffReplacerTest.php
@@ -1209,6 +1209,41 @@ public function should_replace_12_units_with_empty_segments_with_the_correct_sta
         $this->assertEquals( $status, $output[ 'files' ][ 1 ][ 'trans-units' ][ 2 ][ 'target' ][ 'attr' ][ 'state' ] );
         $this->assertEquals( $status, $output[ 'files' ][ 1 ][ 'trans-units' ][ 3 ][ 'target' ][ 'attr' ][ 'state' ] );
     }
+
+    /**
+     * @test
+     */
+    public function should_replace_12_units_with_entities() {
+
+        $data = $this->getData( [
+            [
+                'sid' => '1',
+                'segment' => 'Hello&apos;&apos; ',
+                'internal_id' => '2973331',
+                'mrk_id' => '0',
+                'prev_tags' => '',
+                'succ_tags' => '',
+                'mrk_prev_tags' => NULL,
+                'mrk_succ_tags' => NULL,
+                'translation' => 'Ciao&apos;&apos; ',
+                'status' => 'APPROVED',
+                'error' => '',
+                'eq_word_count' => '1.34',
+                'raw_word_count' => '2.00',
+                'source_page' => NULL,
+                'r2' => NULL,
+                'data_ref_map' => NULL,
+            ],
+        ] );
+
+        $inputFile  = __DIR__ . '/../tests/files/with-entities.xliff';
+        $outputFile = __DIR__ . '/../tests/files/output/with-entities.xliff';
+
+        ( new XliffParser() )->replaceTranslation( $inputFile, $data[ 'data' ], $data[ 'transUnits' ], 'it-it', $outputFile, false );
+        $output = ( new XliffParser() )->xliffToArray( file_get_contents( $outputFile ) );
+
+        $this->assertEquals( "<mrk mid=\"0\" mtype=\"seg\">Ciao''</mrk> ", $output[ 'files' ][ 1 ][ 'trans-units' ][ 1 ][ 'target' ][ 'raw-content' ] );
+    }
 }
 
 class DummyXliffReplacerCallbackWhichReturnFalse implements XliffReplacerCallbackInterface {

diff --git a/tests/files/with-entities.xliff b/tests/files/with-entities.xliff
@@ -0,0 +1,61 @@
+<?xml version="1.0" encoding="utf-8"?>
+<xliff xmlns:sdl="http://sdl.com/FileTypes/SdlXliff/1.0" xmlns="urn:oasis:names:tc:xliff:document:1.2" version="1.2" sdl:version="1.0">
+  <file datatype="x-textunit_dto" original="unknown" source-language="en" target-language="fr-FR">
+  <header>
+  			<file-info
+  				xmlns="http://sdl.com/FileTypes/SdlXliff/1.0">
+  				<value key="SDL:FileId">2bc03c8e-0d02-49e7-8788-f3e94f511c00</value>
+  				<value key="SDL:CreationDate">06/04/2024 15:25:58</value>
+  				<value key="SDL:OriginalFilePath">C:\Users\StefanieVeldhuis\AppData\Local\Temp\SDL_Wizard_Project_cb33f86e-d120-4cb5-af9f-9ca3532362fd\en-GB\M4_output.json</value>
+  				<value key="SDL:OriginalEncoding">utf-8</value>
+  				<value key="EscapeUnicode">False</value>
+  				<value key="SetPrettyPrinting">True</value>
+  				<value key="Utf8Bom">DontAddIfNotPresent</value>
+  				<value key="SDL_SUBCONTENT_STREAM">true</value>
+  				<value key="ParagraphTextDirections"></value>
+  				<sniff-info>
+  					<detected-encoding detection-level="Likely" encoding="utf-8"/>
+  					<detected-source-lang detection-level="Guess" lang="en-GB"/>
+  				</sniff-info>
+  			</file-info>
+  			<sdl:filetype-info>
+  				<sdl:filetype-id>JSON v 1.0.0.0</sdl:filetype-id>
+  			</sdl:filetype-info>
+  			<cxt-defs
+  				xmlns="http://sdl.com/FileTypes/SdlXliff/1.0">
+  				<cxt-def id="1" type="StructureParagraphUnit">
+  					<props>
+  						<value key="StructureParagraphPath">[239].Text</value>
+  					</props>
+  				</cxt-def>
+  			</cxt-defs>
+  			<tag-defs
+  				xmlns="http://sdl.com/FileTypes/SdlXliff/1.0">
+  				<tag id="4407">
+  					<st>#SDL_SUBCONTENT_BOUNDARY#</st>
+  				</tag>
+  				<tag id="4409">
+  					<st>#SDL_SUBCONTENT_BOUNDARY#</st>
+  				</tag>
+  				<tag id="4410">
+  					<st>#SDL-SUBCONTENT-MARKER#</st>
+  				</tag>
+  				<tag id="4411">
+  					<st>&apos;&apos;I cannot compensate you, as it is Colgate-Palmolive policy not to compensate Customers for margin loss. Our Customers are free to sell Colgate-Palmolive products at any price they choose. Our Customers are responsible for their own profitability.&apos;&apos; </st>
+  				</tag>
+  			</tag-defs>
+  		</header>
+    <body>
+      <trans-unit id="2973331" resname="creators.PAGE_TITLE" xml:space="preserve">
+        <source xml:lang="en">Hello&apos;&apos; </source>
+        <seg-source>
+          <mrk mid="0" mtype="seg">Hello&apos;&apos; </mrk>
+        </seg-source>
+        <target state="new" xml:lang="it-IT">
+          <mrk mid="0" mtype="seg">Ciao&apos;&apos; </mrk>
+        </target>
+        <note>Silly note</note>
+      </trans-unit>
+    </body>
+  </file>
+</xliff>