Fix malformed url bug

README.md

@@ -9,7 +9,7 @@ This plugin enables URL tokenization and token filtering by URL part.
 
 | Elasticsearch Version | Plugin Version |
 |-----------------------|----------------|
-| 2.3.4 | 2.3.4.3 |
+| 2.3.4 | 2.3.4.4 |
 | 2.3.3 | 2.3.3.5 |
 | 2.3.2 | 2.3.2.1 |
 | 2.3.1 | 2.3.1.1 |

pom.xml

@@ -6,7 +6,7 @@
 
     <groupId>org.elasticsearch</groupId>
     <artifactId>elasticsearch-analysis-url</artifactId>
-    <version>2.3.4.3</version>
+    <version>2.3.4.4</version>
     <packaging>jar</packaging>
     <description>Elasticsearch URL token filter plugin</description>
 

src/main/java/org/elasticsearch/index/analysis/url/URLTokenizer.java

@@ -504,8 +504,10 @@ private List<Token> tokenizeSpecial(URL url) {
         // protocol://host
         token = getPart(url, URLPart.PROTOCOL) + "://" + getPart(url, URLPart.HOST);
         start = getStartIndex(url, token);
-        end = getEndIndex(start, token);
-        tokens.add(new Token(token, URLPart.WHOLE, start, end));
+        if (start != -1) {
+            end = getEndIndex(start, token);
+            tokens.add(new Token(token, URLPart.WHOLE, start, end));
+        }
         return tokens;
     }
 

src/test/java/org/elasticsearch/index/analysis/url/URLTokenizerIntegrationTest.java

@@ -33,6 +33,12 @@ public void testAnalyze() {
     }
 
 
+    @Test
+    public void testAnalyzePartial() throws Exception {
+        assertTokensContain("http://", "tokenizer_url_all", ":80", "http:", "http", "80");
+    }
+
+
     @Test
     public void testAnalyzeWhole() throws Exception {
         List<AnalyzeResponse.AnalyzeToken> tokens = analyzeURL("http://foo.bar.com", "tokenizer_url_all_malformed");