Prepare version 4.10.0

- Add vectors similarity functions from CEP-30 into the numeric functions listed into the DatabaseMetadata. - Allow disabling DSE tests using Maven profile.
ing-bank · Sep 30, 2023 · 20a2264 · 20a2264
1 parent 7b4c5b8
commit 20a2264
Show file tree

Hide file tree

Showing 10 changed files with 123 additions and 79 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -4,10 +4,11 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) and this project adheres to 
 [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
-## [4.10.0] - Unreleased
+## [4.10.0] - 2023-09-30
 ### Added
 - Add support for new [`vector` CQL type](https://datastax-oss.atlassian.net/browse/JAVA-3060)
-  defined in [CEP-30](https://cwiki.apache.org/confluence/x/OQ40Dw).
+  defined in [CEP-30](https://cwiki.apache.org/confluence/x/OQ40Dw) 
+  Also see PR [#27](https://github.com/ing-bank/cassandra-jdbc-wrapper/pull/27).
 - Implement the method `getWarnings()` in `CassandraResultSet`.
 - Implement the following methods of `CassandraDatabaseMetaData`: 
   `getBestRowIdentifier(String, String, String, int, boolean)` and `getAttributes(String, String, String, String)`.

diff --git a/README.md b/README.md
@@ -49,6 +49,17 @@ To compile and run tests, execute the following Maven command:
 ```bash
 mvn clean package
 ```
+
+#### Some considerations about running tests
+
+If for some reason the tests using DataStax Enterprise server (`*DseContainerTest`) fail in your local environment, you 
+might disable them using the Maven profile `disableDseTests`: 
+```bash
+mvn clean package -PdisableDseTests
+```
+
+The test suite also includes integration tests with AstraDB (`DbaasAstraIntegrationTest`). These tests require an
+AstraDB token configured in the environment variable `ASTRA_DB_APPLICATION_TOKEN`, otherwise they are skipped.
 
 ### Integration in Maven projects
 
@@ -265,8 +276,11 @@ For further information about custom implementations of `SslEngineFactory`, see
 
 ### Connecting to DBaaS
 
-In order to connect to the cloud [Cassandra-based DBaaS AstraDB](https://www.datastax.com/astra) cluster, one would 
-need to specify:
+An alternative JDBC driver based on this one exists to ease the connection to the cloud 
+[Cassandra-based DBaaS AstraDB](https://www.datastax.com/astra) cluster:
+[Astra JDBC driver](https://github.com/DataStax-Examples/astra-jdbc-connector/tree/main). Do not hesitate to use it if you are in this specific situation.
+
+It's still possible to connect to AstraDB using this JDBC wrapper, so one would need to specify:
 * `secureconnectbundle`: the fully qualified path of the cloud secure connect bundle file
 * `keyspace`: the keyspace to connect to
 * `user`: the username
@@ -352,14 +366,16 @@ CREATE TABLE example_table (
     varint_col varint,
     string_set_col set<text>,
     string_list_col list<text>, 
-    string_map_col map<text, text>
+    string_map_col map<text, text>,
+    vector_col vector<float, 5>
 );
 ```
 
 To insert a record into `example_table` using a prepared statement:
 
 ```java
 import com.datastax.oss.driver.api.core.data.CqlDuration;
+import com.datastax.oss.driver.api.core.data.CqlVector;
 
 import java.io.ByteArrayInputStream;
 import java.sql.Date;
@@ -370,8 +386,8 @@ public class HelloCassandra {
     final String insertCql = "INSERT INTO example_table (bigint_col, ascii_col, blob_col, boolean_col, decimal_col, "
       + "double_col, float_col, inet_col, int_col, smallint_col, text_col, timestamp_col, time_col, date_col, "
       + "tinyint_col, duration_col, uuid_col, timeuuid_col, varchar_col, varint_col, string_set_col, "
-      + "string_list_col, string_map_col) "
-      + "VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, now(), ?, ?, ?, ?, ?);";
+      + "string_list_col, string_map_col, vector_col) "
+      + "VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, now(), ?, ?, ?, ?, ?, ?);";
     final PreparedStatement preparedStatement = connection.prepareStatement(insertCql);
     preparedStatement.setObject(1, 1L);                             // bigint
     preparedStatement.setObject(2, "test");                         // ascii
@@ -401,14 +417,16 @@ public class HelloCassandra {
     sampleSet.add("test1");
     sampleSet.add("test2");
     preparedStatement.setObject(20, sampleSet);                     // set
-    ArrayList<String> sampleList = new ArrayList<String>();
+    final ArrayList<String> sampleList = new ArrayList<String>();
     sampleList.add("test1");
     sampleList.add("test2");
     preparedStatement.setObject(21, sampleList);                    // list
-    HashMap<String, String> sampleMap = new HashMap<String, String>();
+    final HashMap<String, String> sampleMap = new HashMap<String, String>();
     sampleMap.put("1", "test1");
     sampleMap.put("2", "test2");
     preparedStatement.setObject(22, sampleMap);                     // map
+    final CqlVector<Float> sampleVector = CqlVector.newInstance(1.0f, 0.0f, 1.0f, 0.5f, 0.2f);
+    preparedStatement.setObject(23, sampleVector);                  // vector
     // Execute the prepare statement.
     preparedStatement.execute();
   }
@@ -696,6 +714,7 @@ We use [SemVer](http://semver.org/) for versioning.
 * Madhavan Sridharan - **[@msmygit](https://github.com/msmygit)**
 * Marius Jokubauskas - **[@mjok](https://github.com/mjok)**
 * Sualeh Fatehi - **[@sualeh](https://github.com/sualeh)**
+* Cedrick Lunven - **[@clun](https://github.com/clun)**
 
 And special thanks to the developer of the original project on which is based this one:
 * Alexander Dejanovski - **[@adejanovski](https://github.com/adejanovski)**

diff --git a/pom.xml b/pom.xml
@@ -5,7 +5,7 @@
 
 	<groupId>com.ing.data</groupId>
 	<artifactId>cassandra-jdbc-wrapper</artifactId>
-    <version>4.9.1</version>
+    <version>4.10.0</version>
 	<packaging>jar</packaging>
 
     <name>Cassandra JDBC Wrapper</name>
@@ -109,10 +109,10 @@
         <hamcrest.version>2.2</hamcrest.version>
         <junit5.version>5.10.0</junit5.version>
         <junit-platform.version>1.10.0</junit-platform.version>
-        <lombok.version>1.18.28</lombok.version>
+        <lombok.version>1.18.30</lombok.version>
         <mockito.version>3.12.4</mockito.version>
         <slf4j.version>1.7.36</slf4j.version>
-        <testcontainers.version>1.18.3</testcontainers.version>
+        <testcontainers.version>1.19.0</testcontainers.version>
         <astra-sdk.version>0.6.11</astra-sdk.version>
         <!-- Versions for plugins -->
         <maven-checkstyle-plugin.version>3.3.0</maven-checkstyle-plugin.version>
@@ -241,14 +241,14 @@
             <version>${testcontainers.version}</version>
             <scope>test</scope>
         </dependency>
-        <!-- Astra Test instances for integration tests -->
+        <!-- Astra SDK for integration tests with AstraDB -->
         <dependency>
             <groupId>com.datastax.astra</groupId>
             <artifactId>astra-sdk-devops</artifactId>
             <version>${astra-sdk.version}</version>
             <scope>test</scope>
         </dependency>
-        <!-- handy to build the Cql Queries -->
+        <!-- Handy to build the CQL queries -->
         <dependency>
             <groupId>com.datastax.oss</groupId>
             <artifactId>java-driver-query-builder</artifactId>
@@ -461,5 +461,24 @@
                 </plugins>
             </build>
         </profile>
+
+        <!-- 'disableDseTests' excludes integration tests using DataStax Enterprise server. This can be handy if
+         these specific tests failed for some reason in a local environment. -->
+        <profile>
+            <id>disableDseTests</id>
+            <build>
+                <plugins>
+                    <plugin>
+                        <artifactId>maven-surefire-plugin</artifactId>
+                        <version>${maven-surefire-plugin.version}</version>
+                        <configuration>
+                            <excludes>
+                                <exclude>*DseContainerTest.java</exclude>
+                            </excludes>
+                        </configuration>
+                    </plugin>
+                </plugins>
+            </build>
+        </profile>
     </profiles>
 </project>
diff --git a/src/main/java/com/ing/data/cassandra/jdbc/CassandraDatabaseMetaData.java b/src/main/java/com/ing/data/cassandra/jdbc/CassandraDatabaseMetaData.java
@@ -511,8 +511,9 @@ public int getMaxUserNameLength() {
     @Override
     public String getNumericFunctions() throws SQLException {
         checkStatementClosed();
-        // Cassandra does not implement natively numeric functions.
-        return StringUtils.EMPTY;
+        // We consider here the vectors similarity functions introduced by CEP-30 as numeric functions (see
+        // https://issues.apache.org/jira/browse/CASSANDRA-18640).
+        return "similarity_cosine,similarity_euclidean,similarity_dot_product";
     }
 
     @Override
@@ -779,7 +780,7 @@ public String getTimeDateFunctions() throws SQLException {
         checkStatementClosed();
         // See: https://cassandra.apache.org/doc/latest/cassandra/cql/functions.html
         return "dateOf,now,minTimeuuid,maxTimeuuid,unixTimestampOf,toDate,toTimestamp,toUnixTimestamp,currentTimestamp,"
-            + "currentDate,currentTime,currentTimeUUID,";
+            + "currentDate,currentTime,currentTimeUUID";
     }
 
     @Override

diff --git a/src/main/java/com/ing/data/cassandra/jdbc/metadata/TableMetadataResultSetBuilder.java b/src/main/java/com/ing/data/cassandra/jdbc/metadata/TableMetadataResultSetBuilder.java
@@ -330,8 +330,8 @@ public CassandraMetadataResultSet buildPrimaryKeys(final String schema, final St
      * </p>
      *
      * @param schema   A schema name pattern. It must match the schema name as it is stored in the database; {@code ""}
-     *                 retrieves those without a schema and {@code null} means that the schema name should not be used to
-     *                 narrow the search. Using {@code ""} as the same effect as {@code null} because here the schema
+     *                 retrieves those without a schema and {@code null} means that the schema name should not be used
+     *                 to narrow the search. Using {@code ""} as the same effect as {@code null} because here the schema
      *                 corresponds to the keyspace and Cassandra tables cannot be defined outside a keyspace.
      * @param table    A table name. It must match the table name as it is stored in the database.
      * @param scope    The scope of interest, using the same values as {@code SCOPE} in the result set.

diff --git a/src/main/java/com/ing/data/cassandra/jdbc/types/DataTypeEnum.java b/src/main/java/com/ing/data/cassandra/jdbc/types/DataTypeEnum.java
@@ -172,6 +172,8 @@ public enum DataTypeEnum {
      */
     VECTOR(DataType.LIST, CqlVector.class, "Vector");
 
+    static final String VECTOR_CLASSNAME = "org.apache.cassandra.db.marshal.VectorType";
+
     private static final Map<String, DataTypeEnum> CQL_DATATYPE_TO_DATATYPE;
 
     /**
@@ -185,8 +187,6 @@ public enum DataTypeEnum {
 
     final int protocolId;
 
-    static final String VECTOR_CLASSNAME = "org.apache.cassandra.db.marshal.VectorType";
-
     static {
         CQL_DATATYPE_TO_DATATYPE = new HashMap<>();
         for (final DataTypeEnum dataType : DataTypeEnum.values()) {
@@ -220,6 +220,7 @@ public static DataTypeEnum fromCqlTypeName(final String cqlTypeName) {
         if (cqlTypeName.startsWith(UDT.cqlType)) {
             return UDT;
         }
+        // Manage vector type
         if (cqlTypeName.contains(VECTOR_CLASSNAME)) {
             return VECTOR;
         }
@@ -337,7 +338,10 @@ public String toString() {
      */
     public static String cqlName(@Nonnull final com.datastax.oss.driver.api.core.type.DataType dataType) {
         final String rawCql = dataType.asCql(false, false);
-        return rawCql.contains(VECTOR_CLASSNAME) ? VECTOR.cqlType : rawCql;
+        if (rawCql.contains(VECTOR_CLASSNAME)) {
+            return VECTOR.cqlType;
+        }
+        return rawCql;
     }
 }