Merge remote-tracking branch 'fork/581'

yegor256 · Jun 3, 2015 · 08db041 · 08db041
2 parents 994ba15 + cf5d0a4
commit 08db041
Show file tree

Hide file tree

Showing 2 changed files with 71 additions and 1 deletion.
diff --git a/netbout-web/src/main/java/com/netbout/rest/Markdown.java b/netbout-web/src/main/java/com/netbout/rest/Markdown.java
@@ -33,6 +33,7 @@
 import java.io.IOException;
 import java.io.PrintWriter;
 import java.util.logging.Level;
+import java.util.regex.Pattern;
 import javax.validation.constraints.NotNull;
 import org.apache.commons.codec.CharEncoding;
 import org.apache.commons.io.IOUtils;
@@ -56,6 +57,24 @@ public final class Markdown {
      */
     private static final Tidy TIDY = Markdown.makeTidy();
 
+    /**
+     * Plain link detection pattern.
+     * The regex consists of the following parts:
+     * <ul>
+     * <li>Negative lookbehind to determine we skip properly formatted
+     * links (e.g. [foo](http://bar))</li>
+     * <li>Scheme prefix followed by arbitrary number of URL-valid
+     * characters</li>
+     * <li>Last character &ndash; URL-valid characters with exclusion of common
+     * punctuation like dot, comma, question mark, exclamation mark, semicolon,
+     * colon, and parentheses</li>
+     * </ul>
+     */
+    private static final Pattern LINK = Pattern.compile(
+        // @checkstyle LineLength (1 line)
+        "(?<!(\\]\\())(https?://[a-zA-Z0-9-._~:/\\?#\\[\\]\\(\\)@!$&'*+,;=%]+[a-zA-Z0-9-_~/#\\[\\]@$&'*+=%])"
+    );
+
     /**
      * The source text.
      */
@@ -78,7 +97,9 @@ public Markdown(@NotNull final String txt) {
     public String html() {
         synchronized (Markdown.TIDY) {
             return Markdown.clean(
-                new PegDownProcessor().markdownToHtml(this.text)
+                new PegDownProcessor().markdownToHtml(
+                    Markdown.formatLinks(this.text)
+                )
             );
         }
     }
@@ -133,4 +154,12 @@ private static Tidy makeTidy() {
         return tidy;
     }
 
+    /**
+     * Replace plain links with Markdown syntax.
+     * @param txt Text to find links in
+     * @return Text with Markdown-formatted links
+     */
+    private static String formatLinks(final String txt) {
+        return LINK.matcher(txt).replaceAll("[$2]($2)");
+    }
 }
diff --git a/netbout-web/src/test/java/com/netbout/rest/MarkdownTest.java b/netbout-web/src/test/java/com/netbout/rest/MarkdownTest.java
@@ -111,6 +111,7 @@ public void formatsTextFragmentsToHtml() throws Exception {
             MatcherAssert.assertThat(
                 new Markdown(pair[0]).html().trim(),
                 Matchers.equalTo(
+                    // @checkstyle MultipleStringLiteralsCheck (1 line)
                     pair[1].replace("\n", System.getProperty("line.separator"))
                 )
             );
@@ -155,4 +156,44 @@ public void leavesDivUntouched() throws Exception {
         );
     }
 
+    /**
+     * Markdown can detect plain text links and produce HTML
+     * with links wrapped correctly.
+     * @throws Exception If there is some problem inside
+     */
+    @Test
+    @SuppressWarnings("PMD.AvoidInstantiatingObjectsInLoops")
+    public void detectsLinks() throws Exception {
+        final String[][] texts = {
+            new String[] {
+                "http://foo.com",
+                "<p>\n  <a href=\"http://foo.com\">http://foo.com</a>\n</p>",
+            },
+            new String[] {
+                "(http://foo?com)",
+                "<p>(\n<a href=\"http://foo?com\">http://foo?com</a>)</p>",
+            },
+            new String[] {
+                "(http://foo#com)",
+                "<p>(\n<a href=\"http://foo#com\">http://foo#com</a>)</p>",
+            },
+            new String[] {
+                "(https://a?b=c)",
+                "<p>(\n<a href=\"https://a?b=c\">https://a?b=c</a>)</p>",
+            },
+            new String[] {
+                "[foo](http://foo)",
+                "<p>\n  <a href=\"http://foo\">foo</a>\n</p>",
+            },
+        };
+        for (final String[] pair : texts) {
+            MatcherAssert.assertThat(
+                new Markdown(pair[0]).html().trim(),
+                Matchers.equalTo(
+                    // @checkstyle MultipleStringLiteralsCheck (1 line)
+                    pair[1].replace("\n", System.getProperty("line.separator"))
+                )
+            );
+        }
+    }
 }