Issue #6717: add support to RegexpMultiline to match across lines

Author: rnveach

src/main/java/com/puppycrawl/tools/checkstyle/checks/regexp/RegexpMultilineCheck.java

@@ -43,6 +43,8 @@ public class RegexpMultilineCheck extends AbstractFileSetCheck {
     private int maximum;
     /** Whether to ignore case when matching. */
     private boolean ignoreCase;
+    /** Whether to match across multiple lines. */
+    private boolean matchAcrossLines;
 
     /** The detector to use. */
     private MultilineDetector detector;
@@ -51,7 +53,7 @@ public class RegexpMultilineCheck extends AbstractFileSetCheck {
     public void beginProcessing(String charset) {
         final DetectorOptions options = DetectorOptions.newBuilder()
             .reporter(this)
-            .compileFlags(Pattern.MULTILINE)
+            .compileFlags(getRegexCompileFlags())
             .format(format)
             .message(message)
             .minimum(minimum)
@@ -66,6 +68,24 @@ protected void processFiltered(File file, FileText fileText) {
         detector.processLines(fileText);
     }
 
+    /**
+     * Retrieves the compile flags for the regular expression being built based
+     * on {@code matchAcrossLines}.
+     * @return The compile flags.
+     */
+    private int getRegexCompileFlags() {
+        final int result;
+
+        if (matchAcrossLines) {
+            result = Pattern.DOTALL;
+        }
+        else {
+            result = Pattern.MULTILINE;
+        }
+
+        return result;
+    }
+
     /**
      * Sets the format of the regular expression to match.
      * @param format the format of the regular expression to match.
@@ -106,4 +126,12 @@ public void setIgnoreCase(boolean ignoreCase) {
         this.ignoreCase = ignoreCase;
     }
 
+    /**
+     * Sets whether to match across multiple lines.
+     * @param matchAcrossLines whether to match across multiple lines.
+     */
+    public void setMatchAcrossLines(boolean matchAcrossLines) {
+        this.matchAcrossLines = matchAcrossLines;
+    }
+
 }

src/test/java/com/puppycrawl/tools/checkstyle/checks/regexp/RegexpMultilineCheckTest.java

@@ -255,4 +255,27 @@ public void testGoodLimit() throws Exception {
         verify(checkConfig, getPath("InputRegexpMultilineSemantic.java"), expected);
     }
 
+    @Test
+    public void testMultilineSupport() throws Exception {
+        final DefaultConfiguration checkConfig = createModuleConfig(RegexpMultilineCheck.class);
+        checkConfig.addAttribute("format", "abc.*def");
+        checkConfig.addAttribute("matchAcrossLines", "true");
+        final String[] expected = {
+            "9: " + getCheckMessage(MSG_REGEXP_EXCEEDED, "abc.*def"),
+        };
+        verify(checkConfig, getPath("InputRegexpMultilineMultilineSupport.java"), expected);
+    }
+
+    @Test
+    public void testMultilineSupportNotGreedy() throws Exception {
+        final DefaultConfiguration checkConfig = createModuleConfig(RegexpMultilineCheck.class);
+        checkConfig.addAttribute("format", "abc.*?def");
+        checkConfig.addAttribute("matchAcrossLines", "true");
+        final String[] expected = {
+            "9: " + getCheckMessage(MSG_REGEXP_EXCEEDED, "abc.*?def"),
+            "11: " + getCheckMessage(MSG_REGEXP_EXCEEDED, "abc.*?def"),
+        };
+        verify(checkConfig, getPath("InputRegexpMultilineMultilineSupport2.java"), expected);
+    }
+
 }

src/test/resources/com/puppycrawl/tools/checkstyle/checks/regexp/regexpmultiline/InputRegexpMultilineMultilineSupport.java

@@ -0,0 +1,18 @@
+package com.puppycrawl.tools.checkstyle.checks.regexp.regexpmultiline;
+
+/**
+ * Config format = 'ABC.*DEF'(lowercased)
+ * matchAcrossLines = true
+ */
+public class InputRegexpMultilineMultilineSupport {
+    void method() {
+// abc - violation
+// def
+// abc
+    }
+
+    void method2() {
+// def
+// abc
+    }
+}

src/test/resources/com/puppycrawl/tools/checkstyle/checks/regexp/regexpmultiline/InputRegexpMultilineMultilineSupport2.java

@@ -0,0 +1,18 @@
+package com.puppycrawl.tools.checkstyle.checks.regexp.regexpmultiline;
+
+/**
+ * Config format = 'ABC.*?DEF'(lowercased)
+ * matchAcrossLines = true
+ */
+public class InputRegexpMultilineMultilineSupport2 {
+    void method() {
+// abc - violation
+// def
+// abc - violation
+    }
+
+    void method2() {
+// def
+// abc
+    }
+}

src/xdocs/config_regexp.xml

@@ -550,6 +550,13 @@ public static final int A_SETPOINT = 1;
             <td><code>0</code></td>
             <td>5.0</td>
           </tr>
+          <tr>
+            <td>matchAcrossLines</td>
+            <td>Controls whether to match expressions across multiple lines.</td>
+            <td><a href="property_types.html#boolean">Boolean</a></td>
+            <td><code>false</code></td>
+            <td>8.25</td>
+          </tr>
           <tr>
             <td>fileExtensions</td>
             <td>file type extension of files to process</td>
@@ -570,6 +577,36 @@ public static final int A_SETPOINT = 1;
     value=&quot;System\.(out)|(err)\.print(ln)?\(&quot;/&gt;
 &lt;/module&gt;
         </source>
+
+        <p>
+          To configure the check to match text that spans multiple lines,
+          like normal code in a Java file:
+        </p>
+        <source>
+&lt;module name=&quot;RegexpMultiline&quot;&gt;
+  &lt;property name=&quot;matchAcrossLines&quot; value=&quot;true&quot;/&gt;
+  &lt;property name=&quot;format&quot; value=&quot;System\.out.*print\(&quot;/&gt;
+&lt;/module&gt;
+        </source>
+        <p>
+          Example of violation from the above config:
+        </p>
+        <source>
+void method() {
+  System.out. // violation
+  print("Example");
+  System.out.
+  print("Example");
+}
+        </source>
+        <p>
+          Note: Beware of the greedy regular expression used in the above example.
+          <code>.*</code> will match as much as possible and not produce multiple violations in
+          the file if multiple groups of lines could match the expression. To prevent an
+          expression being too greedy, avoid overusing matching all text or allow it to be
+          optional, like <code>.*?</code>. Changing the example expression to not be greedy
+          will allow multiple violations in the example to be found in the same file.
+        </p>
       </subsection>
 
       <subsection name="Example of Usage" id="RegexpMultiline_Example_of_Usage">