Merge branch 'ing-bank:release/next' into add-test

CHANGELOG.md

@@ -38,7 +38,11 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
 ### Fixed
 - Fix `NullPointerException` issue [#38](https://github.com/ing-bank/cassandra-jdbc-wrapper/issues/38) when a null 
   type name pattern is specified in a call to `CassandraDatabaseMetaData.getUDTs(String, String, String, int[])`.
-- Add null safety on some methods of `CassandraMetadataResultSet`.
+- Fix issue [#39](https://github.com/ing-bank/cassandra-jdbc-wrapper/issues/39): return `false` when the method 
+  `isSearchable(int)` is called on the metadata of a result set without table or schema name (typically on 
+  `CassandraMetadataResultSet`s).
+- Fix issue preventing to retrieve the metadata of an empty `CassandraMetadataResultSet`.
+- Add null safety on some methods of `CassandraResultSet` and `CassandraMetadataResultSet`.
 
 ## [4.10.2] - 2023-11-01
 ### Fixed

CONTRIBUTING.md

@@ -52,6 +52,12 @@ To run the tests, execute the following command:
 ```
 mvn test
 ```
+Regarding the tests, you need **Docker** installed on your machine.
+Indeed, most of the tests in this project are based on 
+[Testcontainers for Cassandra](https://java.testcontainers.org/modules/databases/cassandra/), because testing a JDBC API
+implementation requires ensuring that the driver is able to connect to a database and execute queries correctly. 
+For example, very common JDBC drivers like those for [PostgreSQL](https://github.com/pgjdbc/pgjdbc) or
+[MS SQL Server](https://github.com/Microsoft/mssql-jdbc/) are also tested against a real database.
 
 ### Submit a pull request
 
@@ -63,12 +69,20 @@ Once your changes and tests are ready for review, submit them:
    to verify it or simply run `mvn clean install` and check the logs).
 
 3. Rebase your changes: update your local repository with the most recent code from the original repository, and rebase
-your branch on top of the latest `release/next` branch. It is better that your initial changes are squashed into a
-single commit. If more changes are required to validate the pull request, we invite you to add them as separate commits.
+   your branch on top of the latest `release/next` branch. It is better that your initial changes are squashed into a
+   single commit. If more changes are required to validate the pull request, we invite you to add them as separate 
+   commits.
 
 4. Finally, push your local changes to your forked repository and submit a pull request into the branch `release/next`
-with a title which sums up the changes that you have made (try to not exceed 50 characters), and provide more details in
-the body. If necessary, also mention the number of the issue solved by your changes, e.g. "Closes #123".
+   with a title which sums up the changes that you have made (try to not exceed 50 characters), and provide more details
+   in the body. If necessary, also mention the number of the issue solved by your changes, e.g. "Closes #123".
+
+### About dependencies
+
+If your changes require to add a new dependency or update an existing one, be sure to check these points first of all:
+* the dependency is the latest stable version of the library compatible with JDK 8
+* the dependency does not introduce vulnerabilities
+* the version of the dependency is specified in a property `<artifactId>.version` in `pom.xml`.
 
 ### License headers
 

README.md

@@ -50,6 +50,11 @@ To compile and run tests, execute the following Maven command:
 mvn clean package
 ```
 
+To build a bundled version of the JDBC wrapper, run the following command:
+```bash
+mvn clean package -Pbundle
+```
+
 #### Some considerations about running tests
 
 If for some reason the tests using DataStax Enterprise server (`*DseContainerTest`) fail in your local environment, you 
@@ -90,6 +95,17 @@ JetBrains DataGrip, you can have a look to the following links:
   this example uses Astra JDBC driver (based on this project), so refer to the "Usage" section below to adapt driver
   class and JDBC URL values.
 
+This JDBC wrapper for Apache Cassandra® is also used to run 
+[Liquibase for Cassandra databases](https://github.com/liquibase/liquibase-cassandra) (from Liquibase 4.25.0). To execute Liquibase scripts on
+your Cassandra database, specify the following properties in your Liquibase properties file:
+```
+driver: com.ing.data.cassandra.jdbc.CassandraDriver
+url: jdbc:cassandra://<host>:<port>/<keyspaceName>?compliancemode=Liquibase
+```
+See the "Usage" section below for further details about the allowed parameters in the JDBC URL.
+For further details about Liquibase usage, please check the 
+[official documentation](https://contribute.liquibase.com/extensions-integrations/directory/database-tutorials/cassandra/).
+
 ## Usage
 
 Connect to a Cassandra cluster using the following arguments:

pom.xml

@@ -119,7 +119,7 @@
         <lombok.version>1.18.30</lombok.version>
         <mockito.version>3.12.4</mockito.version>
         <slf4j.version>1.7.36</slf4j.version>
-        <testcontainers.version>1.19.1</testcontainers.version>
+        <testcontainers.version>1.19.2</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>

src/main/java/com/ing/data/cassandra/jdbc/AbstractResultSet.java

@@ -55,8 +55,9 @@ abstract class AbstractResultSet implements Wrapper {
      * @param columnIndex The column index (first column is 1).
      * @param type        The data type to check.
      * @return {@code true} if the column CQL data type is the given one, {@code false} otherwise.
+     * @throws SQLException when the CQL type cannot be determined for the given column.
      */
-    boolean isCqlType(final int columnIndex, @Nonnull final DataTypeEnum type) {
+    boolean isCqlType(final int columnIndex, @Nonnull final DataTypeEnum type) throws SQLException {
         final String columnType = StringUtils.substringBefore(DataTypeEnum.cqlName(getCqlDataType(columnIndex)), "<");
         return type.cqlType.equalsIgnoreCase(columnType);
     }
@@ -67,8 +68,9 @@ boolean isCqlType(final int columnIndex, @Nonnull final DataTypeEnum type) {
      * @param columnLabel The column name.
      * @param type        The data type to check.
      * @return {@code true} if the column CQL data type is the given one, {@code false} otherwise.
+     * @throws SQLException when the CQL type cannot be determined for the given column.
      */
-    boolean isCqlType(final String columnLabel, @Nonnull final DataTypeEnum type) {
+    boolean isCqlType(final String columnLabel, @Nonnull final DataTypeEnum type) throws SQLException {
         final String columnType = StringUtils.substringBefore(DataTypeEnum.cqlName(getCqlDataType(columnLabel)), "<");
         return type.cqlType.equalsIgnoreCase(columnType);
     }
@@ -78,16 +80,18 @@ boolean isCqlType(final String columnLabel, @Nonnull final DataTypeEnum type) {
      *
      * @param columnIndex The column index (first column is 1).
      * @return The CQL data type of the column.
+     * @throws SQLException when the CQL type cannot be determined for the given column.
      */
-    abstract DataType getCqlDataType(int columnIndex);
+    abstract DataType getCqlDataType(int columnIndex) throws SQLException;
 
     /**
      * Gets the CQL type of the column with the given name.
      *
      * @param columnLabel The column name.
      * @return The CQL data type of the column.
+     * @throws SQLException when the CQL type cannot be determined for the given column.
      */
-    abstract DataType getCqlDataType(String columnLabel);
+    abstract DataType getCqlDataType(String columnLabel) throws SQLException;
 
     public boolean absolute(final int row) throws SQLException {
         throw new SQLFeatureNotSupportedException(NOT_SUPPORTED);

src/main/java/com/ing/data/cassandra/jdbc/CassandraMetadataResultSet.java

@@ -196,13 +196,25 @@ private void populateColumns() {
     }
 
     @Override
-    DataType getCqlDataType(final int columnIndex) {
-        return this.currentRow.getColumnDefinitions().getType(columnIndex - 1);
+    DataType getCqlDataType(final int columnIndex) throws SQLException {
+        if (this.currentRow != null && this.currentRow.getColumnDefinitions() != null) {
+            return this.currentRow.getColumnDefinitions().getType(columnIndex - 1);
+        }
+        if (this.driverResultSet != null && this.driverResultSet.getColumnDefinitions() != null) {
+            return this.driverResultSet.getColumnDefinitions().getType(columnIndex - 1);
+        }
+        throw new SQLException(UNABLE_TO_RETRIEVE_METADATA);
     }
 
     @Override
-    DataType getCqlDataType(final String columnLabel) {
-        return this.currentRow.getColumnDefinitions().getType(columnLabel);
+    DataType getCqlDataType(final String columnLabel) throws SQLException {
+        if (this.currentRow != null && this.currentRow.getColumnDefinitions() != null) {
+            return this.currentRow.getColumnDefinitions().getType(columnLabel);
+        }
+        if (this.driverResultSet != null && this.driverResultSet.getColumnDefinitions() != null) {
+            return this.driverResultSet.getColumnDefinitions().getType(columnLabel);
+        }
+        throw new SQLException(UNABLE_TO_RETRIEVE_METADATA);
     }
 
     @Override
@@ -1071,7 +1083,7 @@ public String getColumnLabel(final int column) throws SQLException {
 
         @Override
         public String getColumnName(final int column) throws SQLException {
-            if (currentRow != null) {
+            if (currentRow != null && currentRow.getColumnDefinitions() != null) {
                 return currentRow.getColumnDefinitions().getName(column - 1);
             }
             if (driverResultSet != null && driverResultSet.getColumnDefinitions() != null) {
@@ -1085,7 +1097,7 @@ public int getColumnDisplaySize(final int column) {
             try {
                 final AbstractJdbcType<?> jdbcEquivalentType;
                 final ColumnDefinitions.Definition columnDefinition;
-                if (currentRow != null) {
+                if (currentRow != null && currentRow.getColumnDefinitions() != null) {
                     columnDefinition = currentRow.getColumnDefinitions().asList().get(column - 1);
                 } else if (driverResultSet != null && driverResultSet.getColumnDefinitions() != null) {
                     columnDefinition = driverResultSet.getColumnDefinitions().asList().get(column - 1);
@@ -1105,7 +1117,7 @@ public int getColumnDisplaySize(final int column) {
         }
 
         @Override
-        public int getColumnType(final int column) {
+        public int getColumnType(final int column) throws SQLException {
             final DataType type;
             if (currentRow != null) {
                 type = getCqlDataType(column);
@@ -1118,7 +1130,7 @@ public int getColumnType(final int column) {
         }
 
         @Override
-        public String getColumnTypeName(final int column) {
+        public String getColumnTypeName(final int column) throws SQLException {
             // Specification says "database specific type name"; for Cassandra this means the AbstractType.
             final DataType type;
             if (currentRow != null) {
@@ -1141,7 +1153,7 @@ public int getScale(final int column) {
             try {
                 final AbstractJdbcType<?> jdbcEquivalentType;
                 final ColumnDefinitions.Definition columnDefinition;
-                if (currentRow != null) {
+                if (currentRow != null && currentRow.getColumnDefinitions() != null) {
                     columnDefinition = currentRow.getColumnDefinitions().asList().get(column - 1);
                 } else if (driverResultSet != null && driverResultSet.getColumnDefinitions() != null) {
                     columnDefinition = driverResultSet.getColumnDefinitions().asList().get(column - 1);
@@ -1171,7 +1183,7 @@ public String getSchemaName(final int column) throws SQLException {
         @Override
         public String getTableName(final int column) {
             final String tableName;
-            if (currentRow != null) {
+            if (currentRow != null && currentRow.getColumnDefinitions() != null) {
                 tableName = currentRow.getColumnDefinitions().getTable(column - 1);
             } else if (driverResultSet != null && driverResultSet.getColumnDefinitions() != null) {
                 tableName = driverResultSet.getColumnDefinitions().getTable(column - 1);
@@ -1231,9 +1243,16 @@ public boolean isSearchable(final int column) throws SQLException {
                 return false;
             }
             final String columnName = getColumnName(column);
+            final String schemaName = getSchemaName(column);
+            final String tableName = getTableName(column);
+            // If the schema or table name is not defined, always returns false since we cannot determine if the column
+            // is searchable in this context.
+            if (StringUtils.isEmpty(schemaName) || StringUtils.isEmpty(tableName)) {
+                return false;
+            }
             final AtomicBoolean searchable = new AtomicBoolean(false);
-            statement.connection.getSession().getMetadata().getKeyspace(getSchemaName(column))
-                .flatMap(metadata -> metadata.getTable(getTableName(column)))
+            statement.connection.getSession().getMetadata().getKeyspace(schemaName)
+                .flatMap(metadata -> metadata.getTable(tableName))
                 .ifPresent(tableMetadata -> {
                     boolean result;
                     // Check first if the column is a clustering column or in a partitioning key.
@@ -1250,7 +1269,7 @@ public boolean isSearchable(final int column) throws SQLException {
         }
 
         @Override
-        public boolean isSigned(final int column) {
+        public boolean isSigned(final int column) throws SQLException {
             final DataType type;
             if (currentRow != null) {
                 type = getCqlDataType(column);

src/main/java/com/ing/data/cassandra/jdbc/CassandraResultSet.java

@@ -100,6 +100,7 @@
 import static com.ing.data.cassandra.jdbc.utils.ErrorConstants.NOT_SUPPORTED;
 import static com.ing.data.cassandra.jdbc.utils.ErrorConstants.NO_INTERFACE;
 import static com.ing.data.cassandra.jdbc.utils.ErrorConstants.UNABLE_TO_READ_VALUE;
+import static com.ing.data.cassandra.jdbc.utils.ErrorConstants.UNABLE_TO_RETRIEVE_METADATA;
 import static com.ing.data.cassandra.jdbc.utils.ErrorConstants.UNSUPPORTED_JSON_TYPE_CONVERSION;
 import static com.ing.data.cassandra.jdbc.utils.ErrorConstants.UNSUPPORTED_TYPE_CONVERSION;
 import static com.ing.data.cassandra.jdbc.utils.ErrorConstants.VALID_LABELS;
@@ -270,12 +271,18 @@ private void populateColumns() {
 
     @Override
     DataType getCqlDataType(final int columnIndex) {
-        return this.currentRow.getColumnDefinitions().get(columnIndex - 1).getType();
+        if (this.currentRow != null) {
+            return this.currentRow.getColumnDefinitions().get(columnIndex - 1).getType();
+        }
+        return this.driverResultSet.getColumnDefinitions().get(columnIndex - 1).getType();
     }
 
     @Override
     DataType getCqlDataType(final String columnLabel) {
-        return this.currentRow.getColumnDefinitions().get(columnLabel).getType();
+        if (this.currentRow != null) {
+            return this.currentRow.getColumnDefinitions().get(columnLabel).getType();
+        }
+        return this.driverResultSet.getColumnDefinitions().get(columnLabel).getType();
     }
 
     @Override
@@ -1780,9 +1787,16 @@ public boolean isSearchable(final int column) throws SQLException {
                 return false;
             }
             final String columnName = getColumnName(column);
+            final String schemaName = getSchemaName(column);
+            final String tableName = getTableName(column);
+            // If the schema or table name is not defined (this should not happen here, but better to be careful),
+            // always returns false since we cannot determine if the column is searchable in this context.
+            if (StringUtils.isEmpty(schemaName) || StringUtils.isEmpty(tableName)) {
+                return false;
+            }
             final AtomicBoolean searchable = new AtomicBoolean(false);
-            statement.connection.getSession().getMetadata().getKeyspace(getSchemaName(column))
-                .flatMap(metadata -> metadata.getTable(getTableName(column)))
+            statement.connection.getSession().getMetadata().getKeyspace(schemaName)
+                .flatMap(metadata -> metadata.getTable(tableName))
                 .ifPresent(tableMetadata -> {
                     boolean result;
                     // Check first if the column is a clustering column or in a partitioning key.

src/main/java/com/ing/data/cassandra/jdbc/ColumnDefinitions.java

@@ -16,6 +16,7 @@
 package com.ing.data.cassandra.jdbc;
 
 import com.datastax.oss.driver.api.core.type.DataType;
+import org.apache.commons.lang3.StringUtils;
 
 import javax.annotation.Nonnull;
 import java.util.Arrays;
@@ -310,6 +311,17 @@ public Definition(final String keyspace, final String table, final String name,
             this.type = type;
         }
 
+        /**
+         * Builds a column definition in an anonymous table (useful for metadata result sets built programmatically).
+         *
+         * @param name The column name.
+         * @param type The column type.
+         * @return A new column definition instance.
+         */
+        public static Definition buildDefinitionInAnonymousTable(final String name, final DataType type) {
+            return new Definition(StringUtils.EMPTY, StringUtils.EMPTY, name, type);
+        }
+
         /**
          * Gets the name of the keyspace this column is part of.
          *

src/main/java/com/ing/data/cassandra/jdbc/metadata/CatalogMetadataResultSetBuilder.java

@@ -15,13 +15,16 @@
 
 package com.ing.data.cassandra.jdbc.metadata;
 
+import com.datastax.oss.driver.api.core.type.DataTypes;
 import com.ing.data.cassandra.jdbc.CassandraMetadataResultSet;
 import com.ing.data.cassandra.jdbc.CassandraStatement;
 
 import java.sql.DatabaseMetaData;
 import java.sql.SQLException;
 import java.util.ArrayList;
 
+import static com.ing.data.cassandra.jdbc.ColumnDefinitions.Definition.buildDefinitionInAnonymousTable;
+
 /**
  * Utility class building metadata result sets ({@link CassandraMetadataResultSet} objects) related to catalogs.
  */
@@ -52,10 +55,14 @@ public CatalogMetadataResultSetBuilder(final CassandraStatement statement) throw
      */
     public CassandraMetadataResultSet buildCatalogs() throws SQLException {
         final ArrayList<MetadataRow> catalogs = new ArrayList<>();
-        final MetadataRow row = new MetadataRow().addEntry(TABLE_CATALOG_SHORTNAME,
-            this.statement.getConnection().getCatalog());
-        catalogs.add(row);
-        return CassandraMetadataResultSet.buildFrom(this.statement, new MetadataResultSet().setRows(catalogs));
+        final MetadataRow.MetadataRowTemplate rowTemplate = new MetadataRow.MetadataRowTemplate(
+            buildDefinitionInAnonymousTable(TABLE_CATALOG_SHORTNAME, DataTypes.TEXT)
+        );
+
+        catalogs.add(new MetadataRow().withTemplate(rowTemplate, this.statement.getConnection().getCatalog()));
+
+        return CassandraMetadataResultSet.buildFrom(this.statement,
+            new MetadataResultSet(rowTemplate).setRows(catalogs));
     }
 
 }