adding the possibility for extracting www-based URLs as well

Author: MTDdk

.gitignore

@@ -9,3 +9,4 @@
 
 # mvn
 target/
+/bin/

src/main/java/org/nibor/autolink/LinkExtractor.java

@@ -3,6 +3,7 @@
 import org.nibor.autolink.internal.EmailScanner;
 import org.nibor.autolink.internal.Scanner;
 import org.nibor.autolink.internal.UrlScanner;
+import org.nibor.autolink.internal.WwwUrlScanner;
 
 import java.util.*;
 
@@ -14,10 +15,12 @@
 public class LinkExtractor {
 
     private final Scanner urlScanner;
+    private final Scanner wwwScanner;
     private final Scanner emailScanner;
 
-    private LinkExtractor(UrlScanner urlScanner, EmailScanner emailScanner) {
+    private LinkExtractor(UrlScanner urlScanner, WwwUrlScanner wwwScanner, EmailScanner emailScanner) {
         this.urlScanner = urlScanner;
+        this.wwwScanner = wwwScanner;
         this.emailScanner = emailScanner;
     }
 
@@ -46,6 +49,8 @@ private Scanner trigger(char c) {
                 return urlScanner;
             case '@':
                 return emailScanner;
+            case 'w':
+                return wwwScanner;
         }
         return null;
     }
@@ -88,8 +93,9 @@ public Builder emailDomainMustHaveDot(boolean emailDomainMustHaveDot) {
          */
         public LinkExtractor build() {
             UrlScanner urlScanner = linkTypes.contains(LinkType.URL) ? new UrlScanner() : null;
+            WwwUrlScanner wwwScanner = linkTypes.contains(LinkType.URL) ? new WwwUrlScanner() : null;
             EmailScanner emailScanner = linkTypes.contains(LinkType.EMAIL) ? new EmailScanner(emailDomainMustHaveDot) : null;
-            return new LinkExtractor(urlScanner, emailScanner);
+            return new LinkExtractor(urlScanner, wwwScanner, emailScanner);
         }
     }
 

src/main/java/org/nibor/autolink/LinkSpan.java

@@ -21,4 +21,8 @@ public interface LinkSpan {
      */
     int getEndIndex();
 
+    /**
+     * @return the found sequence
+     */
+    CharSequence sequence();
 }

src/main/java/org/nibor/autolink/internal/EmailScanner.java

@@ -25,12 +25,12 @@ public LinkSpan scan(CharSequence input, int triggerIndex, int rewindIndex) {
         }
 
         int afterAt = triggerIndex + 1;
-        int last = findLast(input, afterAt);
-        if (last == -1) {
+        int last = findLast(input, afterAt) + 1;
+        if (last == 0) {
             return null;
         }
 
-        return new LinkSpanImpl(LinkType.EMAIL, first, last + 1);
+        return new LinkSpanImpl(LinkType.EMAIL, first, last, input.subSequence(first, last));
     }
 
     // See "Local-part" in RFC 5321, plus extensions in RFC 6531

src/main/java/org/nibor/autolink/internal/LinkSpanImpl.java

@@ -8,11 +8,13 @@ public class LinkSpanImpl implements LinkSpan {
     private final LinkType linkType;
     private final int beginIndex;
     private final int endIndex;
+    private final CharSequence sequence;
 
-    public LinkSpanImpl(LinkType linkType, int beginIndex, int endIndex) {
+    public LinkSpanImpl(LinkType linkType, int beginIndex, int endIndex, CharSequence sequence) {
         this.linkType = linkType;
         this.beginIndex = beginIndex;
         this.endIndex = endIndex;
+        this.sequence = sequence;
     }
 
     @Override
@@ -29,6 +31,11 @@ public int getBeginIndex() {
     public int getEndIndex() {
         return endIndex;
     }
+    
+    @Override
+    public CharSequence sequence() {
+        return sequence;
+    }
 
     @Override
     public String toString() {

src/main/java/org/nibor/autolink/internal/UrlScanner.java

@@ -23,13 +23,13 @@ public LinkSpan scan(CharSequence input, int triggerIndex, int rewindIndex) {
             return null;
         }
 
-        int last = findLast(input, afterSlashSlash);
+        int last = findLast(input, afterSlashSlash) + 1;
 
-        return new LinkSpanImpl(LinkType.URL, first, last + 1);
+        return new LinkSpanImpl(LinkType.URL, first, last, input.subSequence(first, last));
     }
 
     // See "scheme" in RFC 3986
-    private int findFirst(CharSequence input, int beginIndex, int rewindIndex) {
+    protected int findFirst(CharSequence input, int beginIndex, int rewindIndex) {
         int first = -1;
         int digit = -1;
         for (int i = beginIndex; i >= rewindIndex; i--) {
@@ -50,7 +50,7 @@ private int findFirst(CharSequence input, int beginIndex, int rewindIndex) {
         return first;
     }
 
-    private int findLast(CharSequence input, int beginIndex) {
+    protected int findLast(CharSequence input, int beginIndex) {
         int round = 0;
         int square = 0;
         int curly = 0;

src/main/java/org/nibor/autolink/internal/WwwUrlScanner.java

@@ -0,0 +1,42 @@
+package org.nibor.autolink.internal;
+
+import org.nibor.autolink.LinkSpan;
+import org.nibor.autolink.LinkType;
+
+/**
+ * Scan for URLs starting from the trigger character "w", requires "www.".
+ * <p>
+ * Based on RFC 3986.
+ */
+public class WwwUrlScanner extends UrlScanner {
+
+    @Override
+    public LinkSpan scan(final CharSequence input, int triggerIndex, int rewindIndex) {
+        int afterDot = triggerIndex + 4;
+        if (afterDot >= input.length() || input.charAt(triggerIndex + 1) != 'w' || input.charAt(triggerIndex + 2) != 'w' || input.charAt(triggerIndex + 3) != '.') {
+            return null;
+        }
+
+        int first = triggerIndex;
+        int last = findLast(input, afterDot) + 1;
+        if (last == 0) {
+            return null;
+        }
+
+        return new LinkSpanImpl(LinkType.URL, first, last, input.subSequence(first, last));
+    }
+    
+    @Override
+    protected int findLast(CharSequence input, int beginIndex) {
+        int last = super.findLast(input, beginIndex);
+        
+        // Make sure there is at least one dot after the first dot,
+        // so www.something is not allowed, but www.something.co.uk is
+        int pointer = last;
+        while (--pointer > beginIndex) {
+            if (input.charAt(pointer) == '.' && pointer > beginIndex) return last;
+        }
+        
+        return -1;
+    }
+}

src/test/java/org/nibor/autolink/AutolinkBenchmark.java

@@ -12,6 +12,7 @@ public class AutolinkBenchmark extends AutolinkTestCase {
     private static final List<String> WORDS = Arrays.asList(
             "Lorem ", "ipsum ", "dolor ", "sit ", "amet ", "consectetur ", "adipiscing ", "elit ",
             ".", ",", ":", "@", "(", ")", "http://example.com", "https://test.com/foo_(bar)",
+            "www.something.com", "www.another.uk.co",
             "[email protected]", "[email protected]"
     );
 

src/test/java/org/nibor/autolink/AutolinkWwwUrlTest.java

@@ -0,0 +1,60 @@
+package org.nibor.autolink;
+
+import java.util.Arrays;
+import java.util.EnumSet;
+
+import org.junit.Test;
+import org.junit.runner.RunWith;
+import org.junit.runners.Parameterized;
+import org.junit.runners.Parameterized.Parameter;
+import org.junit.runners.Parameterized.Parameters;
+
+@RunWith(Parameterized.class)
+public class AutolinkWwwUrlTest extends AutolinkUrlTest {
+
+    @Parameters(name = "{1}")
+    public static Iterable<Object[]> data() {
+        return Arrays.asList(new Object[][]{
+                {LinkExtractor.builder().linkTypes(EnumSet.of(LinkType.URL)).build(), "URL"},
+                {LinkExtractor.builder().build(), "all"}
+        });
+    }
+
+    @Parameter(0)
+    public LinkExtractor linkExtractor;
+
+    @Parameter(1)
+    public String description;
+
+    @Test
+    public void notLinked() {
+        assertNotLinked("");
+        assertNotLinked("wwwsomething.com");
+        assertNotLinked("ww.foo.com");
+        assertNotLinked("w.bar.foo.co");
+        assertNotLinked("www.something");
+        assertNotLinked("www.go");
+    }
+    
+    @Test
+    public void linked() {
+        assertLinked("www.s.com","|www.s.com|");
+        assertLinked("www.fo.uk","|www.fo.uk|");
+    }
+
+    @Test
+    public void schemes() {
+        assertLinked("http://www.something.com","|http://www.something.com|");
+        assertLinked("http://something.com","|http://something.com|");
+        assertLinked("http://something.co.uk","|http://something.co.uk|");
+    }
+
+    @Override
+    protected LinkExtractor getLinkExtractor() {
+        return linkExtractor;
+    }
+
+    private void assertLinked(String input, String expected) {
+        super.assertLinked(input, expected, LinkType.URL);
+    }
+}