Fixed whitespace on entities and char refs in XmlParser. (#3781)

* Fixed whitespace on entities and char refs in XmlParser. * Polish. Updated MavenParserTest#malformedPom.
openrewrite · Dec 7, 2023 · 726abc6 · 726abc6
1 parent a928dc2
commit 726abc6
Show file tree

Hide file tree

Showing 3 changed files with 37 additions and 9 deletions.
diff --git a/rewrite-maven/src/test/java/org/openrewrite/maven/MavenParserTest.java b/rewrite-maven/src/test/java/org/openrewrite/maven/MavenParserTest.java
@@ -19,6 +19,7 @@
 import okhttp3.mockwebserver.MockResponse;
 import okhttp3.mockwebserver.MockWebServer;
 import okhttp3.mockwebserver.RecordedRequest;
+import org.intellij.lang.annotations.Language;
 import org.junit.jupiter.api.Disabled;
 import org.junit.jupiter.api.Test;
 import org.openrewrite.InMemoryExecutionContext;
@@ -28,6 +29,7 @@
 import org.openrewrite.maven.tree.*;
 import org.openrewrite.test.RewriteTest;
 import org.openrewrite.tree.ParseError;
+import org.openrewrite.xml.tree.Xml;
 
 import java.io.ByteArrayInputStream;
 import java.io.IOException;
@@ -2240,6 +2242,7 @@ void exclusions() {
 
     @Test
     void malformedPom() {
+        @Language("xml")
         String malformedPomXml = """
           <project>
             <groupId>com.mycompany.app</groupId>
@@ -2250,9 +2253,8 @@ void malformedPom() {
               <dependency>
                 <groupId>junit</groupId>
                 <artifactId>junit</artifactId>
-                <version>4.11</version>&gt;
+                <version>4.11</version>
               </dependency>
-            </dependencies>
           </project>
           """;
         assertThat(MavenParser.builder().build().parse(malformedPomXml))

diff --git a/rewrite-xml/src/main/java/org/openrewrite/xml/internal/XmlParserVisitor.java b/rewrite-xml/src/main/java/org/openrewrite/xml/internal/XmlParserVisitor.java
@@ -110,17 +110,19 @@ public Xml visitContent(XMLParser.ContentContext ctx) {
             return charData;
         } else if (ctx.reference() != null) {
             if (ctx.reference().EntityRef() != null) {
-                cursor += ctx.reference().EntityRef().getSymbol().getStopIndex() + 1;
+                String prefix = prefix(ctx);
+                cursor = ctx.reference().EntityRef().getSymbol().getStopIndex() + 1;
                 return new Xml.CharData(randomId(),
-                        "",
+                        prefix,
                         Markers.EMPTY,
                         false,
                         ctx.reference().EntityRef().getText(),
                         "");
             } else if (ctx.reference().CharRef() != null) {
-                cursor += ctx.reference().CharRef().getSymbol().getStopIndex() + 1;
+                String prefix = prefix(ctx);
+                cursor = ctx.reference().CharRef().getSymbol().getStopIndex() + 1;
                 return new Xml.CharData(randomId(),
-                        "",
+                        prefix,
                         Markers.EMPTY,
                         false,
                         ctx.reference().CharRef().getText(),

diff --git a/rewrite-xml/src/test/java/org/openrewrite/xml/XmlParserTest.java b/rewrite-xml/src/test/java/org/openrewrite/xml/XmlParserTest.java
@@ -297,7 +297,8 @@ void utf8BOM() {
           xml(
             """
               %s<?xml version="1.0" encoding="UTF-8"?><test></test>
-              """.formatted("\uFEFF"))
+              """.formatted("\uFEFF")
+          )
         );
     }
 
@@ -308,7 +309,8 @@ void preserveSpaceBeforeAttributeAssignment() {
           xml(
             """
               <?xml version = "1.0" encoding    =   "UTF-8" standalone = "no" ?><blah></blah>
-              """)
+              """
+          )
         );
     }
 
@@ -321,7 +323,29 @@ void linkWithQuestionMark() {
               <?xml version="1.0" encoding="ISO-8859-1"?>
               <?xml-stylesheet type="text/xsl" href="/name/other?link"?>
               <blah></blah>
-              """)
+              """
+          )
+        );
+    }
+
+    @Issue("https://github.com/openrewrite/rewrite/issues/3442")
+    @Test
+    void preserveWhitespaceOnEntities() {
+        rewriteRun(
+          xml(
+            """
+              <?xml version="1.0" encoding="UTF-8"?>
+              <message><text>&lt;?xml version='1.0' encoding='UTF-8'?&gt;&#13;
+              &lt;note&gt;&#13;
+                  &lt;to&gt;Tove&lt;/to&gt;&#13;
+                  &lt;from&gt;Jani&lt;/from&gt;&#13;
+                  &lt;heading&gt;Reminder&lt;/heading&gt;&#13;
+                  &lt;body&gt;Don't forget me this weekend!&lt;/body&gt;&#13;
+              &lt;/note&gt;&#13;
+              &#13;
+              </text></message>
+              """
+          )
         );
     }
 }