#134: Allowed specifying charset for MySQL tables

doc/changes/changes_3.6.0.md

@@ -6,7 +6,7 @@ Code name: Fix CVE-2024-7254 in test dependency `com.google.protobuf:protobuf-ja
 
 This release fixes CVE-2024-7254 in test dependency `com.google.protobuf:protobuf-java:3.25.1`.
 
-The release also speeds up inserting rows into a table by using batch insert.
+The release also speeds up inserting rows into a table by using batch insert and allows specifying a charset when creating MySQL tables, see the [user guide](../user_guide/user_guide.md#mysql-specific-database-objects) for details.
 
 ## Security
 
@@ -15,6 +15,7 @@ The release also speeds up inserting rows into a table by using batch insert.
 ## Features
 
 * #137: Updated `AbstractImmediateDatabaseObjectWriter#write()` to use batching for inserting rows
+* #134: Allowed specifying charset for MySQL tables
 
 ## Dependency Updates
 

doc/user_guide/user_guide.md

@@ -62,7 +62,7 @@ final Table table = schema.createTable("DAYS","DAY_NAME","VARCHAR(9), "SHORT_NAM
 In case you want to create more complex tables, you can also use a builder.
 
 ```java
-final Table table=schema.createTableBuilder("DAYS")
+final Table table = schema.createTableBuilder("DAYS")
         .column("DAY_NAME","VARCHAR(9)"
         .column("SHORT_NAME","VARCHAR(3)"
         .column("DAY_IN_WEEK","DECIMAL(1,0)"
@@ -390,6 +390,17 @@ Given that a script of that name exists, you can then [execute the script](#exec
 
 ## MySQL-Specific Database Objects
 
-So far there are no MySQL Specific Database Objects that are not described in [Dialect-Agnostic Database Objects](#dialect-agnostic-database-objects) section.
+In addition to [Dialect-Agnostic Database Objects](#dialect-agnostic-database-objects), MySQL allows specifying a charset when creating a new table using the table builder of a `MySqlSchema`. When no charset is specified, MySql uses UTF8 as default.
+
+```java
+final MySqlSchema schema = (MySqlSchema) factory.createSchema("TEST"));
+final MySqlTable table = schema.createTableBuilder("ASCII_DAYS")
+        .charset("ASCII")
+        .column("DAY_NAME","VARCHAR(9)"
+        .column("SHORT_NAME","VARCHAR(3)"
+        .column("DAY_IN_WEEK","DECIMAL(1,0)"
+        // ...
+        .build()
+```
 
 Please keep in mind that Schema object represents a database in MySQL as a schema is a [synonym](https://dev.mysql.com/doc/refman/8.0/en/create-database.html) for a database in MySQL syntax.

src/main/java/com/exasol/dbbuilder/dialects/AbstractSchema.java

@@ -63,7 +63,7 @@ public Table.Builder createTableBuilder(final String name) {
     public Table createTable(final String name, final List<String> columnNames, final List<String> columnTypes) {
         verifyNotDeleted();
         if (columnNames.size() == columnTypes.size()) {
-            final Table.Builder builder = Table.builder(getWriter(), this, getIdentifier(name));
+            final Table.Builder builder = createTableBuilder(name);
             passColumnsToTableBuilder(columnNames, columnTypes, builder);
             final Table table = builder.build();
             this.tables.add(table);

src/main/java/com/exasol/dbbuilder/dialects/mysql/MySqlImmediateDatabaseObjectWriter.java

@@ -46,9 +46,31 @@ public void write(final User user, final DatabaseObject object, final ObjectPriv
         }
     }
 
+    @Override
+    public void write(final Table table) {
+        final MySqlTable mySqlTable = (MySqlTable) table;
+        final StringBuilder builder = new StringBuilder("CREATE TABLE ");
+        builder.append(mySqlTable.getFullyQualifiedName()).append(" (");
+        int i = 0;
+        for (final Column column : mySqlTable.getColumns()) {
+            if (i++ > 0) {
+                builder.append(", ");
+            }
+            builder.append(getQuotedColumnName(column.getName())) //
+                    .append(" ") //
+                    .append(column.getType());
+        }
+        builder.append(")");
+        if (mySqlTable.getCharset() != null) {
+            builder.append(" CHARACTER SET ") //
+                    .append(mySqlTable.getCharset());
+        }
+        writeToObject(mySqlTable, builder.toString());
+    }
+
     @Override
     // [impl->dsn~dropping-schemas~2]
     public void drop(final Schema schema) {
         writeToObject(schema, "DROP SCHEMA " + schema.getFullyQualifiedName());
     }
-}
+}

src/main/java/com/exasol/dbbuilder/dialects/mysql/MySqlSchema.java

@@ -31,4 +31,9 @@ public DatabaseObjectWriter getWriter() {
     protected Identifier getIdentifier(final String name) {
         return MySQLIdentifier.of(name);
     }
+
+    @Override
+    public MySqlTable.MySqlTableBuilder createTableBuilder(final String name) {
+        return MySqlTable.builder(getWriter(), this, getIdentifier(name));
+    }
 }

src/main/java/com/exasol/dbbuilder/dialects/mysql/MySqlTable.java

@@ -0,0 +1,51 @@
+package com.exasol.dbbuilder.dialects.mysql;
+
+import com.exasol.db.Identifier;
+import com.exasol.dbbuilder.dialects.*;
+
+public class MySqlTable extends Table {
+
+    private final String charset;
+
+    protected MySqlTable(final MySqlTableBuilder builder) {
+        super(builder);
+        this.charset = builder.charset;
+    }
+
+    public String getCharset() {
+        return charset;
+    }
+
+    public static MySqlTableBuilder builder(final DatabaseObjectWriter writer, final Schema parentSchema,
+            final Identifier tableName) {
+        return new MySqlTableBuilder(writer, parentSchema, tableName);
+    }
+
+    public static class MySqlTableBuilder extends Table.Builder {
+
+        private String charset;
+
+        private MySqlTableBuilder(final DatabaseObjectWriter writer, final Schema parentSchema,
+                final Identifier tableName) {
+            super(writer, parentSchema, tableName);
+        }
+
+        @Override
+        public MySqlTableBuilder column(final String columnName, final String columnType) {
+            super.column(columnName, columnType);
+            return this;
+        }
+
+        public MySqlTableBuilder charset(final String charset) {
+            this.charset = charset;
+            return this;
+        }
+
+        @Override
+        public MySqlTable build() {
+            final MySqlTable table = new MySqlTable(this);
+            this.writer.write(table);
+            return table;
+        }
+    }
+}

src/test/java/com/exasol/dbbuilder/dialects/AbstractObjectFactoryTest.java

@@ -7,9 +7,9 @@
 
 public abstract class AbstractObjectFactoryTest {
 
-    abstract protected AbstractImmediateDatabaseObjectWriter getWriterMock();
+    protected abstract AbstractImmediateDatabaseObjectWriter getWriterMock();
 
-    abstract protected DatabaseObjectFactory testee();
+    protected abstract DatabaseObjectFactory testee();
 
     @Test
     void createSchemaWritesObject() {

src/test/java/com/exasol/dbbuilder/dialects/mysql/MySQLDatabaseObjectCreationAndDeletionIT.java

@@ -9,12 +9,14 @@
 import static org.hamcrest.Matchers.containsString;
 import static org.hamcrest.Matchers.equalTo;
 import static org.junit.jupiter.api.Assertions.assertAll;
+import static org.junit.jupiter.api.Assertions.assertTrue;
 
 import java.sql.*;
 
 import org.hamcrest.Matcher;
 import org.junit.jupiter.api.Tag;
 import org.junit.jupiter.api.Test;
+import org.testcontainers.containers.JdbcDatabaseContainer.NoDriverFoundException;
 import org.testcontainers.containers.MySQLContainer;
 import org.testcontainers.junit.jupiter.Container;
 import org.testcontainers.junit.jupiter.Testcontainers;
@@ -81,19 +83,21 @@ private void assertUserHasSchemaPrivilege(final String username, final String ob
 
     @Test
     void testGrantSchemaPrivilegeToUser() {
-        final Schema schema = this.factory.createSchema("OBJPRIVSCHEMA");
-        final User user = this.factory.createUser("OBJPRIVUSER").grant(schema, SELECT, DELETE);
-        assertAll(() -> assertUserHasSchemaPrivilege(user.getName(), schema.getName(), "Select_priv"),
-                () -> assertUserHasSchemaPrivilege(user.getName(), schema.getName(), "Delete_priv"));
+        try (final Schema schema = this.factory.createSchema("OBJPRIVSCHEMA")) {
+            final User user = this.factory.createUser("OBJPRIVUSER").grant(schema, SELECT, DELETE);
+            assertAll(() -> assertUserHasSchemaPrivilege(user.getName(), schema.getName(), "Select_priv"),
+                    () -> assertUserHasSchemaPrivilege(user.getName(), schema.getName(), "Delete_priv"));
+        }
     }
 
     @Test
     void testGrantTablePrivilegeToUser() {
-        final Schema schema = this.factory.createSchema("TABPRIVSCHEMA");
-        final Table table = schema.createTable("TABPRIVTABLE", "COL1", "DATE", "COL2", "INT");
-        final User user = this.factory.createUser("TABPRIVUSER").grant(table, SELECT, DELETE);
-        assertAll(() -> assertUserHasTablePrivilege(user.getName(), table.getName(), "Select"),
-                () -> assertUserHasTablePrivilege(user.getName(), table.getName(), "Delete"));
+        try (final Schema schema = this.factory.createSchema("TABPRIVSCHEMA")) {
+            final Table table = schema.createTable("TABPRIVTABLE", "COL1", "DATE", "COL2", "INT");
+            final User user = this.factory.createUser("TABPRIVUSER").grant(table, SELECT, DELETE);
+            assertAll(() -> assertUserHasTablePrivilege(user.getName(), table.getName(), "Select"),
+                    () -> assertUserHasTablePrivilege(user.getName(), table.getName(), "Delete"));
+        }
     }
 
     private void assertUserHasTablePrivilege(final String username, final String objectName,
@@ -114,17 +118,76 @@ private void assertUserHasTablePrivilege(final String username, final String obj
 
     @Test
     void testInsertIntoTable() {
-        final Schema schema = this.factory.createSchema("INSERTSCHEMA");
-        final Table table = schema.createTable("INSERTTABLE", "ID", "INT", "NAME", "VARCHAR(10)");
-        table.insert(1, "FOO").insert(2, "BAR");
-        try {
-            final ResultSet result = this.adminConnection.createStatement()
-                    .executeQuery("SELECT ID, NAME FROM " + table.getFullyQualifiedName() + "ORDER BY ID ASC");
-            assertThat(result, table().row(1, "FOO").row(2, "BAR").matches());
-        } catch (final SQLException exception) {
-            throw new AssertionError(ExaError.messageBuilder("E-TDBJ-25")
-                    .message("Unable to validate contents of table {{table}}", table.getFullyQualifiedName())
-                    .toString(), exception);
+        try (final Schema schema = this.factory.createSchema("INSERTSCHEMA")) {
+            final Table table = schema.createTable("INSERTTABLE", "ID", "INT", "NAME", "VARCHAR(10)");
+            table.insert(1, "FOO").insert(2, "BAR");
+            try {
+                final ResultSet result = this.adminConnection.createStatement()
+                        .executeQuery("SELECT ID, NAME FROM " + table.getFullyQualifiedName() + "ORDER BY ID ASC");
+                assertThat(result, table().row(1, "FOO").row(2, "BAR").matches());
+            } catch (final SQLException exception) {
+                throw new AssertionError(ExaError.messageBuilder("E-TDBJ-25")
+                        .message("Unable to validate contents of table {{table}}", table.getFullyQualifiedName())
+                        .toString(), exception);
+            }
+        }
+    }
+
+    @Test
+    void testCreateTableWithDefaultCharsetUsesUtf8() {
+        try (final MySqlSchema schema = (MySqlSchema) this.factory.createSchema("CHARSET_SCHEMA_DEFAULT")) {
+            final MySqlTable table = schema.createTableBuilder("TABLE_WITH_CHARSET").column("ID", "INT")
+                    .column("NAME", "VARCHAR(10)").build();
+            assertAll(
+                    () -> assertThat("column charset",
+                            getColumnCharset("def", schema.getName(), table.getName(), "NAME"), equalTo("utf8mb4")),
+                    () -> assertThat("table collation", getTableCollation("def", schema.getName(), table.getName()),
+                            equalTo("utf8mb4_0900_ai_ci")));
+        }
+    }
+
+    @Test
+    void testCreateTableWithCharset() {
+        try (final MySqlSchema schema = (MySqlSchema) this.factory.createSchema("CHARSET_SCHEMA_ASCII")) {
+            final MySqlTable table = schema.createTableBuilder("TABLE_WITH_CHARSET").charset("ASCII")
+                    .column("ID", "INT").column("NAME", "VARCHAR(10)").build();
+            assertAll(
+                    () -> assertThat("column charset",
+                            getColumnCharset("def", schema.getName(), table.getName(), "NAME"), equalTo("ascii")),
+                    () -> assertThat("table collation", getTableCollation("def", schema.getName(), table.getName()),
+                            equalTo("ascii_general_ci")));
+        }
+    }
+
+    private String getColumnCharset(final String catalog, final String schema, final String table,
+            final String column) {
+        final String query = "select CHARACTER_SET_NAME from information_schema.COLUMNS "
+                + "where TABLE_CATALOG=? AND TABLE_SCHEMA=? AND TABLE_NAME=? AND COLUMN_NAME=?";
+        try (Connection con = container.createConnection(""); PreparedStatement stmt = con.prepareStatement(query)) {
+            stmt.setString(1, catalog);
+            stmt.setString(2, schema);
+            stmt.setString(3, table);
+            stmt.setString(4, column);
+            final ResultSet rs = stmt.executeQuery();
+            assertTrue(rs.next());
+            return rs.getString("CHARACTER_SET_NAME");
+        } catch (NoDriverFoundException | SQLException exception) {
+            throw new IllegalStateException("Query '" + query + "' failed: " + exception.getMessage(), exception);
+        }
+    }
+
+    private String getTableCollation(final String catalog, final String schema, final String table) {
+        final String query = "select TABLE_COLLATION from information_schema.TABLES "
+                + "where TABLE_CATALOG=? AND TABLE_SCHEMA=? AND TABLE_NAME=?";
+        try (Connection con = container.createConnection(""); PreparedStatement stmt = con.prepareStatement(query)) {
+            stmt.setString(1, catalog);
+            stmt.setString(2, schema);
+            stmt.setString(3, table);
+            final ResultSet rs = stmt.executeQuery();
+            assertTrue(rs.next());
+            return rs.getString("TABLE_COLLATION");
+        } catch (NoDriverFoundException | SQLException exception) {
+            throw new IllegalStateException("Query '" + query + "' failed: " + exception.getMessage(), exception);
         }
     }
 

src/test/java/com/exasol/dbbuilder/dialects/mysql/MySqlTableTest.java

@@ -0,0 +1,34 @@
+package com.exasol.dbbuilder.dialects.mysql;
+
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.Matchers.*;
+
+import org.junit.jupiter.api.Test;
+import org.junit.jupiter.api.extension.ExtendWith;
+import org.mockito.Mock;
+import org.mockito.junit.jupiter.MockitoExtension;
+
+import com.exasol.dbbuilder.dialects.DatabaseObjectWriter;
+import com.exasol.dbbuilder.dialects.Schema;
+
+@ExtendWith(MockitoExtension.class)
+class MySqlTableTest {
+
+    @Mock
+    DatabaseObjectWriter writerMock;
+    @Mock
+    Schema schemaMock;
+
+    @Test
+    void createWithoutCharset() {
+        final MySqlTable table = MySqlTable.builder(writerMock, schemaMock, MySQLIdentifier.of("tableName")).build();
+        assertThat(table.getCharset(), is(nullValue()));
+    }
+
+    @Test
+    void createWithCharset() {
+        final MySqlTable table = MySqlTable.builder(writerMock, schemaMock, MySQLIdentifier.of("tableName"))
+                .charset("myCharset").build();
+        assertThat(table.getCharset(), equalTo("myCharset"));
+    }
+}