From dae48d9fe8b0babc2be87547ded64dbc08d335ba Mon Sep 17 00:00:00 2001
From: Alex Ksikes <alex.ksikes@gmail.com>
Date: Tue, 6 May 2014 16:19:23 +0200
Subject: [PATCH] Added the ability to include the queried document for More
 Like This API.

By default More Like This API excludes the queried document from the response.
However, when debugging or when comparing scores across different queries, it
could be useful to have the best possible matched hit. So this option lets users
explicitly specify the desired behavior.

Closes #6067
---
 docs/reference/search/more-like-this.asciidoc |  3 ++
 .../action/mlt/MoreLikeThisRequest.java       | 26 ++++++++++++++
 .../mlt/MoreLikeThisRequestBuilder.java       |  8 +++++
 .../mlt/TransportMoreLikeThisAction.java      |  8 +++--
 .../action/mlt/RestMoreLikeThisAction.java    |  1 +
 .../mlt/MoreLikeThisActionTests.java          | 36 +++++++++++++++++++
 6 files changed, 79 insertions(+), 3 deletions(-)

diff --git a/docs/reference/search/more-like-this.asciidoc b/docs/reference/search/more-like-this.asciidoc
index 25e8158c461..ffdad3a4d3d 100644
--- a/docs/reference/search/more-like-this.asciidoc
+++ b/docs/reference/search/more-like-this.asciidoc
@@ -25,5 +25,8 @@ Rest parameters relating to search are also allowed, including
 When no `mlt_fields` are specified, all the fields of the document will
 be used in the `more_like_this` query generated.
 
+By default, the queried document is excluded from the response (`include`
+set to false).
+
 Note: In order to use the `mlt` feature a `mlt_field` needs to be either
 be `stored`, store `term_vector` or `source` needs to be enabled.
diff --git a/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequest.java b/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequest.java
index 3da63a051fe..5fa6bba4c55 100644
--- a/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequest.java
+++ b/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequest.java
@@ -21,6 +21,7 @@ package org.elasticsearch.action.mlt;
 
 import org.elasticsearch.ElasticsearchGenerationException;
 import org.elasticsearch.ElasticsearchIllegalArgumentException;
+import org.elasticsearch.Version;
 import org.elasticsearch.action.ActionRequest;
 import org.elasticsearch.action.ActionRequestValidationException;
 import org.elasticsearch.action.ValidateActions;
@@ -72,6 +73,7 @@ public class MoreLikeThisRequest extends ActionRequest<MoreLikeThisRequest> {
     private int minWordLength = -1;
     private int maxWordLength = -1;
     private float boostTerms = -1;
+    private boolean include = false;
 
     private SearchType searchType = SearchType.DEFAULT;
     private int searchSize = 0;
@@ -313,6 +315,21 @@ public class MoreLikeThisRequest extends ActionRequest<MoreLikeThisRequest> {
         return this.boostTerms;
     }
 
+    /**
+     * Whether to include the queried document. Defaults to <tt>false</tt>.
+     */
+    public MoreLikeThisRequest include(boolean include) {
+        this.include = include;
+        return this;
+    }
+
+    /**
+     * Whether to include the queried document. Defaults to <tt>false</tt>.
+     */
+    public boolean include() {
+        return this.include;
+    }
+
     void beforeLocalFork() {
         if (searchSourceUnsafe) {
             searchSource = searchSource.copyBytesArray();
@@ -553,6 +570,12 @@ public class MoreLikeThisRequest extends ActionRequest<MoreLikeThisRequest> {
         minWordLength = in.readVInt();
         maxWordLength = in.readVInt();
         boostTerms = in.readFloat();
+        if (in.getVersion().onOrAfter(Version.V_1_2_0)) {
+            include = in.readBoolean();
+        } else {
+            include = false; // hard-coded behavior until Elasticsearch 1.2
+        }
+
         searchType = SearchType.fromId(in.readByte());
         if (in.readBoolean()) {
             searchQueryHint = in.readString();
@@ -622,6 +645,9 @@ public class MoreLikeThisRequest extends ActionRequest<MoreLikeThisRequest> {
         out.writeVInt(minWordLength);
         out.writeVInt(maxWordLength);
         out.writeFloat(boostTerms);
+        if (out.getVersion().onOrAfter(Version.V_1_2_0)) {
+            out.writeBoolean(include);
+        }
 
         out.writeByte(searchType.id());
         if (searchQueryHint == null) {
diff --git a/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequestBuilder.java b/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequestBuilder.java
index cbeeb386d97..32300f41666 100644
--- a/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequestBuilder.java
+++ b/src/main/java/org/elasticsearch/action/mlt/MoreLikeThisRequestBuilder.java
@@ -140,6 +140,14 @@ public class MoreLikeThisRequestBuilder extends ActionRequestBuilder<MoreLikeThi
         return this;
     }
 
+    /**
+     * Whether to include the queried document. Defaults to <tt>false</tt>.
+     */
+    public MoreLikeThisRequestBuilder setInclude(boolean include) {
+        request.include(include);
+        return this;
+    }
+
     /**
      * An optional search source request allowing to control the search request for the
      * more like this documents.
diff --git a/src/main/java/org/elasticsearch/action/mlt/TransportMoreLikeThisAction.java b/src/main/java/org/elasticsearch/action/mlt/TransportMoreLikeThisAction.java
index 77782f10124..705e00c23c6 100644
--- a/src/main/java/org/elasticsearch/action/mlt/TransportMoreLikeThisAction.java
+++ b/src/main/java/org/elasticsearch/action/mlt/TransportMoreLikeThisAction.java
@@ -178,9 +178,11 @@ public class TransportMoreLikeThisAction extends TransportAction<MoreLikeThisReq
                     }
 
                     // exclude myself
-                    Term uidTerm = docMapper.uidMapper().term(request.type(), request.id());
-                    boolBuilder.mustNot(termQuery(uidTerm.field(), uidTerm.text()));
-                    boolBuilder.adjustPureNegative(false);
+                    if (!request.include()) {
+                        Term uidTerm = docMapper.uidMapper().term(request.type(), request.id());
+                        boolBuilder.mustNot(termQuery(uidTerm.field(), uidTerm.text()));
+                        boolBuilder.adjustPureNegative(false);
+                    }
                 } catch (Throwable e) {
                     listener.onFailure(e);
                     return;
diff --git a/src/main/java/org/elasticsearch/rest/action/mlt/RestMoreLikeThisAction.java b/src/main/java/org/elasticsearch/rest/action/mlt/RestMoreLikeThisAction.java
index be6cc09315b..cd940bbc853 100644
--- a/src/main/java/org/elasticsearch/rest/action/mlt/RestMoreLikeThisAction.java
+++ b/src/main/java/org/elasticsearch/rest/action/mlt/RestMoreLikeThisAction.java
@@ -70,6 +70,7 @@ public class RestMoreLikeThisAction extends BaseRestHandler {
         mltRequest.minWordLength(request.paramAsInt("min_word_len", request.paramAsInt("min_word_length", -1)));
         mltRequest.maxWordLength(request.paramAsInt("max_word_len", request.paramAsInt("max_word_length", -1)));
         mltRequest.boostTerms(request.paramAsFloat("boost_terms", -1));
+        mltRequest.include(request.paramAsBoolean("include", false));
 
         mltRequest.searchType(SearchType.fromString(request.param("search_type")));
         mltRequest.searchIndices(request.paramAsStringArray("search_indices", null));
diff --git a/src/test/java/org/elasticsearch/mlt/MoreLikeThisActionTests.java b/src/test/java/org/elasticsearch/mlt/MoreLikeThisActionTests.java
index f515b0ce972..880574eba63 100644
--- a/src/test/java/org/elasticsearch/mlt/MoreLikeThisActionTests.java
+++ b/src/test/java/org/elasticsearch/mlt/MoreLikeThisActionTests.java
@@ -252,4 +252,40 @@ public class MoreLikeThisActionTests extends ElasticsearchIntegrationTest {
         assertHitCount(searchResponse, 0l);
     }
 
+    @Test
+    public void testSimpleMoreLikeInclude() throws Exception {
+        logger.info("Creating index test");
+        assertAcked(prepareCreate("test").addMapping("type1",
+                jsonBuilder().startObject().startObject("type1").startObject("properties")
+                        .startObject("text").field("type", "string").endObject()
+                        .endObject().endObject().endObject()));
+
+        logger.info("Running Cluster Health");
+        assertThat(ensureGreen(), equalTo(ClusterHealthStatus.GREEN));
+
+        logger.info("Indexing...");
+        client().index(indexRequest("test").type("type1").id("1").source(
+                jsonBuilder().startObject()
+                        .field("text", "Apache Lucene is a free/open source information retrieval software library").endObject()))
+                .actionGet();
+        client().index(indexRequest("test").type("type1").id("2").source(
+                jsonBuilder().startObject()
+                        .field("text", "Lucene has been ported to other programming languages").endObject()))
+                .actionGet();
+        client().admin().indices().refresh(refreshRequest()).actionGet();
+
+        logger.info("Running More Like This with include true");
+        SearchResponse mltResponse = client().moreLikeThis(
+                moreLikeThisRequest("test").type("type1").id("1").minTermFreq(1).minDocFreq(1).include(true)).actionGet();
+        assertOrderedSearchHits(mltResponse, "1", "2");
+
+        mltResponse = client().moreLikeThis(
+                moreLikeThisRequest("test").type("type1").id("2").minTermFreq(1).minDocFreq(1).include(true)).actionGet();
+        assertOrderedSearchHits(mltResponse, "2", "1");
+
+        logger.info("Running More Like This with include false");
+        mltResponse = client().moreLikeThis(moreLikeThisRequest("test").type("type1").id("1").minTermFreq(1).minDocFreq(1)).actionGet();
+        assertSearchHits(mltResponse, "2");
+    }
+
 }