feat(cubejs-databricks-jdbc-driver): read-only mode (#8440)

* Add .editorconfig for DX convenience

* feat(cubejs-databricks-jdbc-driver): set readonly mode if no exportBicket options provided

* chore(cubejs-databricks-jdbc-driver): polish driver interfaces/implementation

* Update docs

* docs(jdbc-driver): update driver installation on macos

---------

Co-authored-by: Igor Lukanin <[email protected]>

.editorconfig

@@ -0,0 +1,10 @@
+# editorconfig.org
+root = true
+
+[*]
+indent_style = space
+indent_size = 2
+end_of_line = lf
+charset = utf-8
+trim_trailing_whitespace = true
+insert_final_newline = true

docs/pages/product/configuration/data-sources/databricks-jdbc.mdx

@@ -84,7 +84,7 @@ here][ref-caching-using-preaggs-build-strats].
 
 | Feature       | Works with read-only mode? | Is default? |
 | ------------- | :------------------------: | :---------: |
-| Simple        |             ❌             |     ✅      |
+| Simple        |             ✅             |     ✅      |
 | Export Bucket |             ❌             |     ❌      |
 
 By default, Databricks JDBC uses a [simple][self-preaggs-simple] strategy to

packages/cubejs-base-driver/src/driver.interface.ts

@@ -230,6 +230,7 @@ export interface DriverInterface {
   queryColumnTypes: (sql: string, params: unknown[]) => Promise<{ name: any; type: string; }[]>;
   //
   getSchemas: () => Promise<QuerySchemasResult[]>;
+  tablesSchema: () => Promise<any>;
   getTablesForSpecificSchemas: (schemas: QuerySchemasResult[]) => Promise<QueryTablesResult[]>;
   getColumnsForSpecificTables: (tables: QueryTablesResult[]) => Promise<QueryColumnsResult[]>;
   // eslint-disable-next-line camelcase
@@ -246,7 +247,7 @@ export interface DriverInterface {
    * queried fields types.
    */
   stream?: (table: string, values: unknown[], options: StreamOptions) => Promise<StreamTableData>;
-  
+
   /**
    * Returns to the Cubestore an object with links to unloaded to an
    * export bucket data.
@@ -258,7 +259,7 @@ export interface DriverInterface {
    * Determines whether export bucket feature is configured or not.
    */
   isUnloadSupported?: (options: UnloadOptions) => Promise<boolean>;
-  
+
   // Current timestamp, defaults to new Date().getTime()
   nowTimestamp(): number;
   // Shutdown the driver

packages/cubejs-databricks-jdbc-driver/src/DatabricksDriver.ts

@@ -71,7 +71,7 @@ export type DatabricksDriverConfiguration = JDBCDriverConfiguration &
      * Export bucket AWS account region.
      */
     awsRegion?: string,
-    
+
     /**
      * Export bucket Azure account key.
      */
@@ -119,7 +119,7 @@ export class DatabricksDriver extends JDBCDriver {
   private showSparkProtocolWarn: boolean;
 
   /**
-   * Read-only mode flag.
+   * Driver Configuration.
    */
   protected readonly config: DatabricksDriverConfiguration;
 
@@ -221,6 +221,11 @@ export class DatabricksDriver extends JDBCDriver {
       exportBucketCsvEscapeSymbol:
         getEnv('dbExportBucketCsvEscapeSymbol', { dataSource }),
     };
+    if (config.readOnly === undefined) {
+      // we can set readonly to true if there is no bucket config provided
+      config.readOnly = !config.exportBucket;
+    }
+
     super(config);
     this.config = config;
     this.showSparkProtocolWarn = showSparkProtocolWarn;
@@ -289,10 +294,11 @@ export class DatabricksDriver extends JDBCDriver {
     values: unknown[],
   ): Promise<R[]> {
     if (this.config.catalog) {
+      const preAggSchemaName = this.getPreAggrSchemaName();
       return super.query(
         query.replace(
-          new RegExp(`(?<=\\s)${this.getPreaggsSchemaName()}\\.(?=[^\\s]+)`, 'g'),
-          `${this.config.catalog}.${this.getPreaggsSchemaName()}.`
+          new RegExp(`(?<=\\s)${preAggSchemaName}\\.(?=[^\\s]+)`, 'g'),
+          `${this.config.catalog}.${preAggSchemaName}.`
         ),
         values,
       );
@@ -304,7 +310,7 @@ export class DatabricksDriver extends JDBCDriver {
   /**
    * Returns pre-aggregation schema name.
    */
-  public getPreaggsSchemaName(): string {
+  protected getPreAggrSchemaName(): string {
     const schema = getEnv('preAggregationsSchema');
     if (schema) {
       return schema;
@@ -327,7 +333,7 @@ export class DatabricksDriver extends JDBCDriver {
     return super.dropTable(tableFullName, options);
   }
 
-  public showDeprecations() {
+  private showDeprecations() {
     if (this.config.url) {
       const result = this.config.url
         .split(';')
@@ -387,7 +393,7 @@ export class DatabricksDriver extends JDBCDriver {
   /**
    * Returns tables meta data object.
    */
-  public async tablesSchema(): Promise<Record<string, Record<string, object>>> {
+  public override async tablesSchema(): Promise<Record<string, Record<string, object>>> {
     const tables = await this.getTables();
 
     const metadata: Record<string, Record<string, object>> = {};
@@ -407,7 +413,7 @@ export class DatabricksDriver extends JDBCDriver {
   /**
    * Returns list of accessible tables.
    */
-  public async getTables(): Promise<ShowTableRow[]> {
+  private async getTables(): Promise<ShowTableRow[]> {
     if (this.config.database) {
       return <any> this.query<ShowTableRow>(
         `SHOW TABLES IN ${
@@ -492,7 +498,7 @@ export class DatabricksDriver extends JDBCDriver {
   /**
    * Returns table columns types.
    */
-  public async tableColumnTypes(table: string): Promise<{ name: any; type: string; }[]> {
+  public override async tableColumnTypes(table: string): Promise<{ name: any; type: string; }[]> {
     let tableFullName = '';
     const tableArray = table.split('.');
 
@@ -567,7 +573,7 @@ export class DatabricksDriver extends JDBCDriver {
   /**
    * Returns schema full name.
    */
-  public getSchemaFullName(schema: string): string {
+  private getSchemaFullName(schema: string): string {
     if (this.config?.catalog) {
       return `${
         this.quoteIdentifier(this.config.catalog)
@@ -582,14 +588,14 @@ export class DatabricksDriver extends JDBCDriver {
   /**
    * Returns quoted string.
    */
-  public quoteIdentifier(identifier: string): string {
+  protected quoteIdentifier(identifier: string): string {
     return `\`${identifier}\``;
   }
 
   /**
    * Returns the JS type by the Databricks type.
    */
-  public toGenericType(columnType: string): string {
+  protected toGenericType(columnType: string): string {
     return DatabricksToGenericType[columnType.toLowerCase()] || super.toGenericType(columnType);
   }
 
@@ -639,7 +645,7 @@ export class DatabricksDriver extends JDBCDriver {
     const types = await this.queryColumnTypes(sql, params);
 
     await this.createExternalTableFromSql(tableFullName, sql, params, types);
-    
+
     return types;
   }
 
@@ -650,7 +656,7 @@ export class DatabricksDriver extends JDBCDriver {
     const types = await this.tableColumnTypes(tableFullName);
 
     await this.createExternalTableFromTable(tableFullName, types);
-    
+
     return types;
   }
 

packages/cubejs-jdbc-driver/README.md

@@ -11,20 +11,37 @@ JDBC driver.
 
 ## Support
 
-This package is **community supported** and should be used at your own risk. 
+This package is **community supported** and should be used at your own risk.
 
-While the Cube Dev team is happy to review and accept future community contributions, we don't have active plans for further development. This includes bug fixes unless they affect different parts of Cube.js. **We're looking for maintainers for this package.** If you'd like to become a maintainer, please contact us in Cube.js Slack. 
+While the Cube Dev team is happy to review and accept future community contributions, we don't have active plans for
+further development. This includes bug fixes unless they affect different parts of Cube.js. **We're looking for
+maintainers for this package.** If you'd like to become a maintainer, please contact us in Cube.js Slack.
 
 ## Java installation
 
 ### macOS
 
 ```sh
-brew install openjdk@8
-sudo ln -sfn /usr/local/opt/openjdk@8/libexec/openjdk.jdk /Library/Java/JavaVirtualMachines/openjdk-8.jdk
+brew install openjdk
+# At the moment of writing, openjdk 22.0.1 is the latest and proven to work on Intel/M1 Mac's
+# Follow the brew suggested advice at the end of installation:
+# For the system Java wrappers to find this JDK, symlink it with
+sudo ln -sfn /usr/local/opt/openjdk/libexec/openjdk.jdk /Library/Java/JavaVirtualMachines/openjdk.jdk
+
+# Ensure that newly installed jdk is visible
+/usr/libexec/java_home -V
+# You should see installed jdk among others, something like this:
+Matching Java Virtual Machines (3):
+    22.0.1 (x86_64) "Homebrew" - "OpenJDK 22.0.1" /usr/local/Cellar/openjdk/22.0.1/libexec/openjdk.jdk/Contents/Home
+    1.8.0_40 (x86_64) "Oracle Corporation" - "Java SE 8" /Library/Java/JavaVirtualMachines/jdk1.8.0_40.jdk/Contents/Home
+
+# Set JAVA_HOME environment variable before running yarn in the Cube repo
+export JAVA_HOME=`/usr/libexec/java_home -v 22.0.1`
 ```
 
-If this doesn't work, please run commands from `$ brew info openjdk@8`.
+**Note:** It's important to set `JAVA_HOME` prior to running `yarn/npm install` in Cube repo as Java Bridge npm package
+uses is to locate JAVA and caches it internally. In case you already run package installation you have to rebuild
+all native packages or just delete `node_modules` and run `yarn` again.
 
 ### Debian, Ubuntu, etc.
 
@@ -46,8 +63,6 @@ If you have Chocolatey packet manager:
 choco install openjdk
 ```
 
-Or download it from 
-
 ## License
 
 Cube.js JDBC Database Driver is [Apache 2.0 licensed](./LICENSE).