Fix malformed URL parsing bug

jlinn · Sep 25, 2017 · 78a0286 · 78a0286
1 parent 4996c25
commit 78a0286
Show file tree

Hide file tree

Showing 4 changed files with 13 additions and 4 deletions.
diff --git a/README.md b/README.md
@@ -9,6 +9,7 @@ This plugin enables URL tokenization and token filtering by URL part.
 
 | Elasticsearch Version | Plugin Version |
 |-----------------------|----------------|
+| 2.4.6 | 2.4.6.0 |
 | 2.4.5 | 2.4.5.0 |
 | 2.4.4 | 2.4.4.0 |
 | 2.4.3 | 2.4.3.0 |

diff --git a/pom.xml b/pom.xml
@@ -6,7 +6,7 @@
 
     <groupId>org.elasticsearch</groupId>
     <artifactId>elasticsearch-analysis-url</artifactId>
-    <version>2.4.5.0</version>
+    <version>2.4.6.0</version>
     <packaging>jar</packaging>
     <description>Elasticsearch URL token filter plugin</description>
 
@@ -18,7 +18,7 @@
 
     <properties>
         <project.build.sourceEncodint>UTF-8</project.build.sourceEncodint>
-        <elasticsearch.version>2.4.5</elasticsearch.version>
+        <elasticsearch.version>2.4.6</elasticsearch.version>
         <lucene.version>5.5.4</lucene.version>
         <hamcrest.version>1.3</hamcrest.version>
         <tests.output>onerror</tests.output>

diff --git a/src/main/java/org/elasticsearch/index/analysis/url/URLTokenizer.java b/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;
     }
 

diff --git a/src/test/java/org/elasticsearch/index/analysis/url/URLTokenizerIntegrationTest.java b/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");