fix: add 'CONTAIN' support in query builder

This fix to address the issue #114
linkedin · Jul 27, 2021 · 200ebfe · 200ebfe
1 parent 77df5d8
commit 200ebfe
Show file tree

Hide file tree

Showing 6 changed files with 147 additions and 3 deletions.
diff --git a/dao-impl/elasticsearch-dao-7/src/main/java/com/linkedin/metadata/dao/utils/SearchUtils.java b/dao-impl/elasticsearch-dao-7/src/main/java/com/linkedin/metadata/dao/utils/SearchUtils.java
@@ -60,6 +60,10 @@ public static Map<String, String> getRequestMap(@Nullable Filter requestParams)
    * <p>If the condition between a field and value is not the same as EQUAL, a Range query is constructed. This
    * condition does not support multiple values for the same field.
    *
+   * <p>When CONTAIN, START_WITH and END_WITH conditions are used, the underlying logic is using wildcard query which is
+   * not performant according to ES. For details, please refer to:
+   * https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-wildcard-query.html#wildcard-query-field-params
+   *
    * @param criterion {@link Criterion} single criterion which contains field, value and a comparison operator
    */
   @Nonnull
@@ -78,6 +82,15 @@ public static QueryBuilder getQueryBuilderFromCriterion(@Nonnull Criterion crite
       return QueryBuilders.rangeQuery(criterion.getField()).lt(criterion.getValue().trim());
     } else if (condition == Condition.LESS_THAN_OR_EQUAL_TO) {
       return QueryBuilders.rangeQuery(criterion.getField()).lte(criterion.getValue().trim());
+    }  else if (condition == Condition.CONTAIN) {
+      return QueryBuilders.wildcardQuery(criterion.getField(),
+          "*" + ESUtils.escapeReservedCharacters(criterion.getValue().trim()) + "*");
+    } else if (condition == Condition.START_WITH) {
+      return QueryBuilders.wildcardQuery(criterion.getField(),
+          ESUtils.escapeReservedCharacters(criterion.getValue().trim()) + "*");
+    } else if (condition == Condition.END_WITH) {
+      return QueryBuilders.wildcardQuery(criterion.getField(),
+          "*" + ESUtils.escapeReservedCharacters(criterion.getValue().trim()));
     }
 
     throw new UnsupportedOperationException("Unsupported condition: " + condition);

diff --git a/...l/elasticsearch-dao-7/src/test/java/com/linkedin/metadata/dao/search/ESSearchDAOTest.java b/...l/elasticsearch-dao-7/src/test/java/com/linkedin/metadata/dao/search/ESSearchDAOTest.java
@@ -248,7 +248,7 @@ public void testFilteredQueryUnsupportedCondition() {
     int from = 0;
     int size = 10;
     final Filter filter2 = new Filter().setCriteria(new CriterionArray(Arrays.asList(
-        new Criterion().setField("field_contain").setValue("value_contain").setCondition(Condition.CONTAIN))));
+        new Criterion().setField("field_contain").setValue("value_contain").setCondition(Condition.IN))));
     SortCriterion sortCriterion = new SortCriterion().setOrder(SortOrder.ASCENDING).setField("urn");
     assertThrows(UnsupportedOperationException.class,
         () -> _searchDAO.getFilteredSearchQuery(filter2, sortCriterion, from, size));

diff --git a/...pl/elasticsearch-dao-7/src/test/java/com/linkedin/metadata/dao/utils/SearchUtilsTest.java b/...pl/elasticsearch-dao-7/src/test/java/com/linkedin/metadata/dao/utils/SearchUtilsTest.java
@@ -7,6 +7,8 @@
 import java.util.Collections;
 import java.util.HashMap;
 import java.util.Map;
+import org.elasticsearch.index.query.QueryBuilder;
+import org.elasticsearch.index.query.WildcardQueryBuilder;
 import org.testng.annotations.Test;
 
 import static com.linkedin.metadata.dao.utils.SearchUtils.*;
@@ -37,4 +39,62 @@ public void testGetRequestMap() {
     ));
     assertThrows(UnsupportedOperationException.class, () -> getRequestMap(filter3));
   }
+
+
+  @Test
+  public void testGetQueryBuilderFromContainCriterion() {
+
+    // Given: a 'contain' criterion
+    Criterion containCriterion = new Criterion();
+    containCriterion.setValue("match * text");
+    containCriterion.setCondition(Condition.CONTAIN);
+    containCriterion.setField("text");
+
+    // Expect 'contain' criterion creates a MatchQueryBuilder
+    QueryBuilder queryBuilder = SearchUtils.getQueryBuilderFromCriterion(containCriterion);
+    assertNotNull(queryBuilder);
+    assertTrue(queryBuilder instanceof WildcardQueryBuilder);
+
+    // Expect 'field name' and search terms
+    assertEquals(((WildcardQueryBuilder) queryBuilder).fieldName(), "text");
+    assertEquals(((WildcardQueryBuilder) queryBuilder).value(), "*match \\* text*");
+  }
+
+  @Test
+  public void testGetQueryBuilderFromStartWithCriterion() {
+
+    // Given: a 'start_with' criterion
+    Criterion containCriterion = new Criterion();
+    containCriterion.setValue("match * text");
+    containCriterion.setCondition(Condition.START_WITH);
+    containCriterion.setField("text");
+
+    // Expect 'start_with' criterion creates a WildcardQueryBuilder
+    QueryBuilder queryBuilder = SearchUtils.getQueryBuilderFromCriterion(containCriterion);
+    assertNotNull(queryBuilder);
+    assertTrue(queryBuilder instanceof WildcardQueryBuilder);
+
+    // Expect 'field name' and search terms
+    assertEquals(((WildcardQueryBuilder) queryBuilder).fieldName(), "text");
+    assertEquals(((WildcardQueryBuilder) queryBuilder).value(), "match \\* text*");
+  }
+
+  @Test
+  public void testGetQueryBuilderFromEndWithCriterion() {
+
+    // Given: a 'end_with' criterion
+    Criterion containCriterion = new Criterion();
+    containCriterion.setValue("match * text");
+    containCriterion.setCondition(Condition.END_WITH);
+    containCriterion.setField("text");
+
+    // Expect 'end_with' criterion creates a MatchQueryBuilder
+    QueryBuilder queryBuilder = SearchUtils.getQueryBuilderFromCriterion(containCriterion);
+    assertNotNull(queryBuilder);
+    assertTrue(queryBuilder instanceof WildcardQueryBuilder);
+
+    // Expect 'field name' and search terms
+    assertEquals(((WildcardQueryBuilder) queryBuilder).fieldName(), "text");
+    assertEquals(((WildcardQueryBuilder) queryBuilder).value(), "*match \\* text");
+  }
 }
diff --git a/dao-impl/elasticsearch-dao/src/main/java/com/linkedin/metadata/dao/utils/SearchUtils.java b/dao-impl/elasticsearch-dao/src/main/java/com/linkedin/metadata/dao/utils/SearchUtils.java
@@ -68,6 +68,10 @@ static boolean isUrn(@Nonnull String value) {
    * <p>If the condition between a field and value is not the same as EQUAL, a Range query is constructed. This
    * condition does not support multiple values for the same field.
    *
+   * <p>When CONTAIN, START_WITH and END_WITH conditions are used, the underlying logic is using wildcard query which is
+   * not performant according to ES. For details, please refer to:
+   * https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-wildcard-query.html#wildcard-query-field-params
+   *
    * @param criterion {@link Criterion} single criterion which contains field, value and a comparison operator
    */
   @Nonnull
@@ -88,8 +92,16 @@ public static QueryBuilder getQueryBuilderFromCriterion(@Nonnull Criterion crite
       return QueryBuilders.rangeQuery(criterion.getField()).lt(criterion.getValue().trim());
     } else if (condition == Condition.LESS_THAN_OR_EQUAL_TO) {
       return QueryBuilders.rangeQuery(criterion.getField()).lte(criterion.getValue().trim());
+    } else if (condition == Condition.CONTAIN) {
+      return QueryBuilders.wildcardQuery(criterion.getField(),
+          "*" + ESUtils.escapeReservedCharacters(criterion.getValue().trim()) + "*");
+    } else if (condition == Condition.START_WITH) {
+      return QueryBuilders.wildcardQuery(criterion.getField(),
+          ESUtils.escapeReservedCharacters(criterion.getValue().trim()) + "*");
+    } else if (condition == Condition.END_WITH) {
+      return QueryBuilders.wildcardQuery(criterion.getField(),
+          "*" + ESUtils.escapeReservedCharacters(criterion.getValue().trim()));
     }
-
     throw new UnsupportedOperationException("Unsupported condition: " + condition);
   }
 

diff --git a/...mpl/elasticsearch-dao/src/test/java/com/linkedin/metadata/dao/search/ESSearchDAOTest.java b/...mpl/elasticsearch-dao/src/test/java/com/linkedin/metadata/dao/search/ESSearchDAOTest.java
@@ -246,7 +246,7 @@ public void testFilteredQueryUnsupportedCondition() {
     int from = 0;
     int size = 10;
     final Filter filter2 = new Filter().setCriteria(new CriterionArray(Arrays.asList(
-        new Criterion().setField("field_contain").setValue("value_contain").setCondition(Condition.CONTAIN))));
+        new Criterion().setField("field_contain").setValue("value_contain").setCondition(Condition.IN))));
     SortCriterion sortCriterion = new SortCriterion().setOrder(SortOrder.ASCENDING).setField("urn");
     assertThrows(UnsupportedOperationException.class,
         () -> _searchDAO.getFilteredSearchQuery(filter2, sortCriterion, from, size));

diff --git a/...impl/elasticsearch-dao/src/test/java/com/linkedin/metadata/dao/utils/SearchUtilsTest.java b/...impl/elasticsearch-dao/src/test/java/com/linkedin/metadata/dao/utils/SearchUtilsTest.java
@@ -7,6 +7,8 @@
 import java.util.Collections;
 import java.util.HashMap;
 import java.util.Map;
+import org.elasticsearch.index.query.QueryBuilder;
+import org.elasticsearch.index.query.WildcardQueryBuilder;
 import org.testng.annotations.Test;
 
 import static com.linkedin.metadata.dao.utils.SearchUtils.*;
@@ -37,4 +39,61 @@ public void testGetRequestMap() {
     ));
     assertThrows(UnsupportedOperationException.class, () -> getRequestMap(filter3));
   }
+
+  @Test
+  public void testGetQueryBuilderFromContainCriterion() {
+
+    // Given: a 'contain' criterion
+    Criterion containCriterion = new Criterion();
+    containCriterion.setValue("match * text");
+    containCriterion.setCondition(Condition.CONTAIN);
+    containCriterion.setField("text");
+
+    // Expect 'contain' criterion creates a MatchQueryBuilder
+    QueryBuilder queryBuilder = SearchUtils.getQueryBuilderFromCriterion(containCriterion);
+    assertNotNull(queryBuilder);
+    assertTrue(queryBuilder instanceof WildcardQueryBuilder);
+
+    // Expect 'field name' and search terms
+    assertEquals(((WildcardQueryBuilder) queryBuilder).fieldName(), "text");
+    assertEquals(((WildcardQueryBuilder) queryBuilder).value(), "*match \\* text*");
+  }
+
+  @Test
+  public void testGetQueryBuilderFromStartWithCriterion() {
+
+    // Given: a 'start_with' criterion
+    Criterion containCriterion = new Criterion();
+    containCriterion.setValue("match * text");
+    containCriterion.setCondition(Condition.START_WITH);
+    containCriterion.setField("text");
+
+    // Expect 'start_with' criterion creates a WildcardQueryBuilder
+    QueryBuilder queryBuilder = SearchUtils.getQueryBuilderFromCriterion(containCriterion);
+    assertNotNull(queryBuilder);
+    assertTrue(queryBuilder instanceof WildcardQueryBuilder);
+
+    // Expect 'field name' and search terms
+    assertEquals(((WildcardQueryBuilder) queryBuilder).fieldName(), "text");
+    assertEquals(((WildcardQueryBuilder) queryBuilder).value(), "match \\* text*");
+  }
+
+  @Test
+  public void testGetQueryBuilderFromEndWithCriterion() {
+
+    // Given: a 'end_with' criterion
+    Criterion containCriterion = new Criterion();
+    containCriterion.setValue("match * text");
+    containCriterion.setCondition(Condition.END_WITH);
+    containCriterion.setField("text");
+
+    // Expect 'end_with' criterion creates a MatchQueryBuilder
+    QueryBuilder queryBuilder = SearchUtils.getQueryBuilderFromCriterion(containCriterion);
+    assertNotNull(queryBuilder);
+    assertTrue(queryBuilder instanceof WildcardQueryBuilder);
+
+    // Expect 'field name' and search terms
+    assertEquals(((WildcardQueryBuilder) queryBuilder).fieldName(), "text");
+    assertEquals(((WildcardQueryBuilder) queryBuilder).value(), "*match \\* text");
+  }
 }