From 7915f08bc95538e717ad6051336c0673c5d715ad Mon Sep 17 00:00:00 2001
From: Jan van Nunen <jan.van.nunen@tiobe.com>
Date: Thu, 18 Feb 2016 13:40:32 +0100
Subject: [PATCH] Extended Objective-C grammar to accept UTF-8 escapes (\uXXXX)
 in string literals.

---
 pmd-objectivec/etc/grammar/ObjC2.0.jj         |  4 +-
 ...capesInStringLiteralObjCTokenizerTest.java | 37 +++++++++++++++++++
 .../cpd/FileWithUTF8EscapeInStringLiteral.m   |  8 ++++
 3 files changed, 47 insertions(+), 2 deletions(-)
 create mode 100644 pmd-objectivec/src/test/java/net/sourceforge/pmd/cpd/UTF8EscapesInStringLiteralObjCTokenizerTest.java
 create mode 100644 pmd-objectivec/src/test/resources/net/sourceforge/pmd/cpd/FileWithUTF8EscapeInStringLiteral.m

diff --git a/pmd-objectivec/etc/grammar/ObjC2.0.jj b/pmd-objectivec/etc/grammar/ObjC2.0.jj
index ba888ee9ed..f25936bdfa 100644
--- a/pmd-objectivec/etc/grammar/ObjC2.0.jj
+++ b/pmd-objectivec/etc/grammar/ObjC2.0.jj
@@ -375,8 +375,8 @@ TOKEN : {
 |  <FLOATING_POINT_LITERAL: (["0"-"9"])+ "." (["0"-"9"])* (<EXPONENT>)? (["f","F","d","D"])? | "." (["0"-"9"])+ (<EXPONENT>)? (["f","F","d","D"])? | (["0"-"9"])+ <EXPONENT> (["f","F","d","D"])? | (["0"-"9"])+ (<EXPONENT>)? ["f","F","d","D"]>
 |  <#EXPONENT: ["e","E"] (["+","-"])? (["0"-"9"])+>
 |  <CHARACTER_LITERAL: "\'" (~["\'","\\","\n","\r"] | "\\" (["n","t","b","r","f","a","v","\\","\'","\""] | ["0"-"7"] (["0"-"7"])? | ["0"-"3"] ["0"-"7"] ["0"-"7"])) "\'">
-|  <STRING_LITERAL: [ "@" ] "\"" ( ~["\"","\\","\n","\r"] | "\\" ["x","X"] (["0"-"9","a"-"f","A"-"F"]) | "\\" ( ["n","t","b","r","f","a","v","\\","\'","\""] | ["0"-"7"] (["0"-"7"])?  | ["0"-"3"] ["0"-"7"] ["0"-"7"] | ( ["\n","\r"] | "\r\n")))* "\"">
-|  <CSTRING_LITERAL: "\"" ( ~["\"","\\","\n","\r"] | "\\" ["x","X"] (["0"-"9","a"-"f","A"-"F"]) | "\\" ( ["n","t","b","r","f","a","v","\\","\'","\""] | ["0"-"7"] (["0"-"7"])?  | ["0"-"3"] ["0"-"7"] ["0"-"7"] | ( ["\n","\r"] | "\r\n")))* "\"">
+|  <STRING_LITERAL: [ "@" ] "\"" ( ~["\"","\\","\n","\r"] | "\\" ["x","X","u","U"] (["0"-"9","a"-"f","A"-"F"]) | "\\" ( ["n","t","b","r","f","a","v","\\","\'","\""] | ["0"-"7"] (["0"-"7"])?  | ["0"-"3"] ["0"-"7"] ["0"-"7"] | ( ["\n","\r"] | "\r\n")))* "\"">
+|  <CSTRING_LITERAL: "\"" ( ~["\"","\\","\n","\r"] | "\\" ["x","X","u","U"] (["0"-"9","a"-"f","A"-"F"]) | "\\" ( ["n","t","b","r","f","a","v","\\","\'","\""] | ["0"-"7"] (["0"-"7"])?  | ["0"-"3"] ["0"-"7"] ["0"-"7"] | ( ["\n","\r"] | "\r\n")))* "\"">
 }
 
 TOKEN :
diff --git a/pmd-objectivec/src/test/java/net/sourceforge/pmd/cpd/UTF8EscapesInStringLiteralObjCTokenizerTest.java b/pmd-objectivec/src/test/java/net/sourceforge/pmd/cpd/UTF8EscapesInStringLiteralObjCTokenizerTest.java
new file mode 100644
index 0000000000..bd3d87d930
--- /dev/null
+++ b/pmd-objectivec/src/test/java/net/sourceforge/pmd/cpd/UTF8EscapesInStringLiteralObjCTokenizerTest.java
@@ -0,0 +1,37 @@
+/**
+ * BSD-style license; for more info see http://pmd.sourceforge.net/license.html
+ */
+package net.sourceforge.pmd.cpd;
+
+import java.io.IOException;
+
+import net.sourceforge.pmd.testframework.AbstractTokenizerTest;
+
+import org.apache.commons.io.IOUtils;
+import org.junit.Before;
+import org.junit.Test;
+
+
+//Tests if the ObjectiveC tokenizer supports UTF-8 escapes in string literals
+public class UTF8EscapesInStringLiteralObjCTokenizerTest extends AbstractTokenizerTest {
+
+    private static final String FILENAME = "FileWithUTF8EscapeInStringLiteral.m";
+
+    @Before
+    @Override
+    public void buildTokenizer() throws IOException {
+        this.tokenizer = new ObjectiveCTokenizer();
+        this.sourceCode = new SourceCode(new SourceCode.StringCodeLoader(this.getSampleCode(), FILENAME));
+    }
+
+    @Override
+    public String getSampleCode() throws IOException {
+        return IOUtils.toString(ObjectiveCTokenizer.class.getResourceAsStream(FILENAME), "UTF-8");
+    }
+
+    @Test
+    public void tokenizeTest() throws IOException {
+        this.expectedTokenCount = 45;
+        super.tokenizeTest();
+    }
+}
diff --git a/pmd-objectivec/src/test/resources/net/sourceforge/pmd/cpd/FileWithUTF8EscapeInStringLiteral.m b/pmd-objectivec/src/test/resources/net/sourceforge/pmd/cpd/FileWithUTF8EscapeInStringLiteral.m
new file mode 100644
index 0000000000..dafd9e3c88
--- /dev/null
+++ b/pmd-objectivec/src/test/resources/net/sourceforge/pmd/cpd/FileWithUTF8EscapeInStringLiteral.m
@@ -0,0 +1,8 @@
+@property (nonatomic, strong) UIButton *copyrightsButton;
+
+- (void)setupCopyrightsButton {
+    self.copyrightsButton = [[UIButton alloc] initWithFrame:CGRectZero];
+    [self.copyrightsButton setTitle:@"\u00a9" forState:UIControlStateNormal];
+}
+
+@end