Correct generation of <target> nodes with translate NO attribute (#95)

* fix not translatable <target> * wip * Tests all green * Xliff 2.0 * Refactoring
matecat · Oct 3, 2024 · a562ef4 · a562ef4
1 parent f74067c
commit a562ef4
Show file tree

Hide file tree

Showing 5 changed files with 104 additions and 2 deletions.
diff --git a/src/XliffReplacer/AbstractXliffReplacer.php b/src/XliffReplacer/AbstractXliffReplacer.php
@@ -398,6 +398,13 @@ protected function setInBuffer( string $name ) {
         if ( in_array( $name, $this->nodesToBuffer ) ) {
             $this->bufferIsActive = true;
         }
+
+        // We need bufferIsActive for <target> nodes with currentTransUnitIsTranslatable = 'NO'
+        // because in the other case, the target can be chunked into pieces by xml_set_character_data_handler()
+        // and this can potentially lead to a wrong string rebuild by postProcAndFlush function if the internal placeholders are split
+        if($name === 'target' and $this->currentTransUnitIsTranslatable === 'no'){
+            $this->bufferIsActive = true;
+        }
     }
 
     /**

diff --git a/src/XliffReplacer/Xliff12.php b/src/XliffReplacer/Xliff12.php
@@ -108,7 +108,9 @@ protected function tagClose( $parser, string $name ) {
          */
         if ( !$this->isEmpty ) {
 
-            if ( !$this->inTarget ) {
+            // write closing tag if is not a target
+            // EXCLUDE the target nodes with currentTransUnitIsTranslatable = 'NO'
+            if ( !$this->inTarget and $this->currentTransUnitIsTranslatable !== 'no' ) {
                 $tag = "</$name>";
             }
 
@@ -120,6 +122,12 @@ protected function tagClose( $parser, string $name ) {
                     // actually there may be more than one segment to that ID if there are two mrk of the same source segment
                     $tag = $this->rebuildTarget();
 
+                } elseif( !empty($this->CDATABuffer) and $this->currentTransUnitIsTranslatable === 'no' ) {
+
+                    // These are target nodes with currentTransUnitIsTranslatable = 'NO'
+                    $this->bufferIsActive = false;
+                    $tag                  = $this->CDATABuffer . "</$name>";
+                    $this->CDATABuffer    = "";
                 }
 
                 $this->targetWasWritten = true;

diff --git a/src/XliffReplacer/Xliff20.php b/src/XliffReplacer/Xliff20.php
@@ -167,7 +167,9 @@ protected function tagClose( $parser, string $name ) {
          */
         if ( !$this->isEmpty ) {
 
-            if ( !$this->inTarget ) {
+            // write closing tag if is not a target
+            // EXCLUDE the target nodes with currentTransUnitIsTranslatable = 'NO'
+            if ( !$this->inTarget and $this->currentTransUnitIsTranslatable !== 'no' ) {
                 $tag = "</$name>";
             }
 
@@ -195,6 +197,12 @@ protected function tagClose( $parser, string $name ) {
                     //append translation
                     $tag = "<target>$translation</target>";
 
+                } elseif( !empty($this->CDATABuffer) and $this->currentTransUnitIsTranslatable === 'no' ) {
+
+                    // These are target nodes with currentTransUnitIsTranslatable = 'NO'
+                    $this->bufferIsActive = false;
+                    $tag                  = $this->CDATABuffer . "</$name>";
+                    $this->CDATABuffer    = "";
                 }
 
                 // signal we are leaving a target

diff --git a/tests/XliffReplacerTest.php b/tests/XliffReplacerTest.php
@@ -1244,6 +1244,49 @@ public function should_replace_12_units_with_entities() {
 
         $this->assertEquals( "<mrk mid=\"0\" mtype=\"seg\">Ciao''</mrk> ", $output[ 'files' ][ 1 ][ 'trans-units' ][ 1 ][ 'target' ][ 'raw-content' ] );
     }
+
+    /**
+     * @test
+     */
+    public function other_tests_replacing_12_units_with_entities() {
+
+        $data = $this->getData( [
+            [
+                'sid' => '1',
+                'segment' => 'Hello&apos;&apos; ',
+                'internal_id' => '3142672',
+                'mrk_id' => '0',
+                'prev_tags' => '',
+                'succ_tags' => '',
+                'mrk_prev_tags' => NULL,
+                'mrk_succ_tags' => NULL,
+                'translation' => 'با دقت به کد زیر نگاه کنید. با کلیک روی «اجرا»، این برنامه کدام طراحی را انجام می‌دهد؟',
+                '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/entities.xliff';
+        $outputFile = __DIR__ . '/../tests/files/output/entities.xliff';
+
+        ( new XliffParser() )->replaceTranslation( $inputFile, $data[ 'data' ], $data[ 'transUnits' ], 'it-it', $outputFile, false );
+        $output = ( new XliffParser() )->xliffToArray( file_get_contents( $outputFile ) );
+
+        $expected = '&lt;table&gt;
+&lt;tr&gt;&lt;td&gt;A&lt;/td&gt;&lt;td&gt;&lt;img src="https://images.code.org/cfc3f8206438a60afe3be9afe7fc0a22-image-1489118742610.10.15.png" width="100px" style="mix-blend-mode: multiply;"/&gt;&lt;/td&gt;&lt;td&gt;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&lt;/td&gt;&lt;td&gt;B&lt;/td&gt;&lt;td&gt;&lt;img src="https://images.code.org/975b027684d2f5411b960bf82987663e-image-1489119999013.11.13.png" width="100px" style="mix-blend-mode: multiply;"/&gt;&lt;/td&gt;&lt;td&gt;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&lt;/td&gt;&lt;td&gt;C&lt;/td&gt;&lt;td&gt;&lt;img src="https://images.code.org/635ac54ed7cb2e2d24eb341b3ec4eecb-image-1489120024059.12.00.png" width="80px" style="mix-blend-mode: multiply; clip: rect(0px,0px,0px,40px);"/&gt;&lt;/td&gt;&lt;/tr&gt;
+&lt;/table&gt;
+
+&lt;br/&gt;&lt;br/&gt;
+
+';
+
+        $this->assertEquals( $expected, $output[ 'files' ][ 1 ][ 'trans-units' ][ 1 ][ 'target' ][ 'raw-content' ] );
+    }
 }
 
 class DummyXliffReplacerCallbackWhichReturnFalse implements XliffReplacerCallbackInterface {

diff --git a/tests/files/entities.xliff b/tests/files/entities.xliff
@@ -0,0 +1,36 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<xliff version="1.2" xmlns="urn:oasis:names:tc:xliff:document:1.2">
+  <file id="179174" original="/course_content/2022/coursee-2022.json" source-language="en-US" target-language="fa" datatype="plaintext" project-id="26074" tool-id="crowdin">
+    <header>
+      <tool tool-id="crowdin" tool-name="Crowdin" tool-version="1.1"/>
+    </header>
+    <body>
+      <trans-unit id="3142670" resname="&quot;https://studio.code.org/s/coursee-2022/lessons/7/levels/7&quot;.sublevels.courseD_artist_nestedLoops9_predict12022.contained levels.0.dsls.markdown" translate="no" approved="yes">
+        <source>&lt;table&gt;
+&lt;tr&gt;&lt;td&gt;A&lt;/td&gt;&lt;td&gt;&lt;img src=&quot;https://images.code.org/cfc3f8206438a60afe3be9afe7fc0a22-image-1489118742610.10.15.png&quot; width=&quot;100px&quot; style=&quot;mix-blend-mode: multiply;&quot;/&gt;&lt;/td&gt;&lt;td&gt;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&lt;/td&gt;&lt;td&gt;B&lt;/td&gt;&lt;td&gt;&lt;img src=&quot;https://images.code.org/975b027684d2f5411b960bf82987663e-image-1489119999013.11.13.png&quot; width=&quot;100px&quot; style=&quot;mix-blend-mode: multiply;&quot;/&gt;&lt;/td&gt;&lt;td&gt;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&lt;/td&gt;&lt;td&gt;C&lt;/td&gt;&lt;td&gt;&lt;img src=&quot;https://images.code.org/635ac54ed7cb2e2d24eb341b3ec4eecb-image-1489120024059.12.00.png&quot; width=&quot;80px&quot; style=&quot;mix-blend-mode: multiply; clip: rect(0px,0px,0px,40px);&quot;/&gt;&lt;/td&gt;&lt;/tr&gt;
+&lt;/table&gt;
+
+&lt;br/&gt;&lt;br/&gt;
+
+</source>
+        <target state="final">&lt;table&gt;
+&lt;tr&gt;&lt;td&gt;A&lt;/td&gt;&lt;td&gt;&lt;img src=&quot;https://images.code.org/cfc3f8206438a60afe3be9afe7fc0a22-image-1489118742610.10.15.png&quot; width=&quot;100px&quot; style=&quot;mix-blend-mode: multiply;&quot;/&gt;&lt;/td&gt;&lt;td&gt;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&lt;/td&gt;&lt;td&gt;B&lt;/td&gt;&lt;td&gt;&lt;img src=&quot;https://images.code.org/975b027684d2f5411b960bf82987663e-image-1489119999013.11.13.png&quot; width=&quot;100px&quot; style=&quot;mix-blend-mode: multiply;&quot;/&gt;&lt;/td&gt;&lt;td&gt;&amp;nbsp;&amp;nbsp;&amp;nbsp;&amp;nbsp;&lt;/td&gt;&lt;td&gt;C&lt;/td&gt;&lt;td&gt;&lt;img src=&quot;https://images.code.org/635ac54ed7cb2e2d24eb341b3ec4eecb-image-1489120024059.12.00.png&quot; width=&quot;80px&quot; style=&quot;mix-blend-mode: multiply; clip: rect(0px,0px,0px,40px);&quot;/&gt;&lt;/td&gt;&lt;/tr&gt;
+&lt;/table&gt;
+
+&lt;br/&gt;&lt;br/&gt;
+
+</target>
+        <context-group purpose="information">
+          <context context-type="source">https://studio.code.org/s/coursee-2022/lessons/7/levels/7 -&gt; sublevels -&gt; courseD_artist_nestedLoops9_predict12022 -&gt; contained levels -&gt; 0 -&gt; dsls -&gt; markdown</context>
+        </context-group>
+      </trans-unit>
+      <trans-unit id="3142672" resname="&quot;https://studio.code.org/s/coursee-2022/lessons/7/levels/7&quot;.sublevels.courseD_artist_nestedLoops9_predict12022.contained levels.0.dsls.questions.0.text" approved="yes">
+        <source>Take a good look at the code below. Which drawing will this program make when you click &quot;Run&quot;?</source>
+        <target state="final">با دقت به کد زیر نگاه کنید. با کلیک روی «اجرا»، این برنامه کدام طراحی را انجام می‌دهد؟</target>
+        <context-group purpose="information">
+          <context context-type="source">https://studio.code.org/s/coursee-2022/lessons/7/levels/7 -&gt; sublevels -&gt; courseD_artist_nestedLoops9_predict12022 -&gt; contained levels -&gt; 0 -&gt; dsls -&gt; questions -&gt; 0 -&gt; text</context>
+        </context-group>
+      </trans-unit>
+    </body>
+  </file>
+</xliff>