schemacrawler-kylin

English | 简体中文

SchemaCrawler plug-in for Apache Kylin. It registers the kylin database system with SchemaCrawler so that jdbc:kylin:* connections can be inspected through the standard SchemaCrawler tool chain.

Table of Contents

• 1. Project Overview
• 2. Features & Status
• 3. Requirements & Compatibility
• 4. Architecture & Modules
• 5. Installation
• 6. Quick Start
• 7. Configuration
• 8. Core Usage / API
• 9. Testing & Build
• 10. Versioning & Branches
• 11. Contributing & License

1. Project Overview

What it is

schemacrawler-kylin is a SchemaCrawler database plug-in for Apache Kylin. It plugs into SchemaCrawler's DatabaseConnectorRegistry through the JDK ServiceLoader mechanism, exposing the Kylin server type (kylin) so that SchemaCrawler CLI and library users can connect to Kylin via jdbc:kylin:* URLs and inspect its schemas.

What it is not

• It is not a Kylin JDBC driver — the actual driver is provided by org.apache.kylin:kylin-jdbc (a dependency of this module).
• It is not a full database-metadata emulator. Kylin's JDBC driver exposes a limited DatabaseMetaData implementation, so the quality of metadata (databases, tables, columns) depends on what the Kylin JDBC layer itself supports.

Typical scenarios

Scenario	Description
Schema inspection of Kylin OLAP cubes	Run SchemaCrawler with --server=kylin to list databases, tables and columns of a Kylin instance.
Embedding in SchemaCrawler-based tooling	Any tool that resolves database connectors from DatabaseConnectorRegistry automatically picks up the Kylin plug-in once this jar is on the classpath.
Database documentation pipelines	Combine with SchemaCrawler commands (e.g. list, schema) to generate schema reports for Kylin.

2. Features & Status

Capability	Status	Notes
Registers server type kylin (Apache Kylin)	Available	DatabaseServerType("kylin", "Apache Kylin")
SPI auto-discovery via META-INF/services	Available	schemacrawler.tools.databaseconnector.DatabaseConnector service entry
URL matching for jdbc:kylin:*	Available	Implemented by supportsUrlPredicate()
Loads org.apache.kylin.jdbc.Driver	Available	Class.forName on connector construction
CLI help integration (Connections.kylin.txt)	Available	--host, --port, --database, --user, --password options
Information-schema views resource	Limited	kylin.information_schema resource ships with the plug-in; its content is currently minimal (empty baseline).

Status is reported as of 3.0.x.x.20260630-SNAPSHOT on the feature/3.0.x branch.

3. Requirements & Compatibility

Item	Version
JDK	21+
Maven	3.0+ (Maven Wrapper 3.5.0 bundled)
SchemaCrawler	16.7.2
Kylin JDBC driver	3.0.0 (org.apache.kylin:kylin-jdbc)

Version lines

Branch	JDK baseline	Version pattern
feature/1.0.x	JDK 8	1.0.x.*
feature/2.0.x	JDK 17	2.0.x.*
feature/3.0.x	JDK 21	3.0.x.*

4. Architecture & Modules

 SchemaCrawler CLI / library
        |
        v
 DatabaseConnectorRegistry (ServiceLoader)
        |
        v
 KylinDatabaseConnector  <--(SPI: META-INF/services)-->  plug-in jar
        |
        |  url = jdbc:kylin://host:port/database
        v
 org.apache.kylin.jdbc.Driver  (kylin-jdbc dependency)
        |
        v
 Apache Kylin (OLAP engine)

This is a single-module project (packaging jar):

Module / artifact	Role
schemacrawler-kylin	The SchemaCrawler plug-in: connector registration, URL matching, driver loading, CLI help, information-schema resource.

5. Installation

The artifact is not yet published to Maven Central. Resolve it from the project's configured artifact repository (Aliyun Packages) or install it locally from source; the snapshot version currently used on the feature/3.0.x branch is 3.0.x.x.20260630-SNAPSHOT.

Maven

<dependency>
    <groupId>io.github.easy4j</groupId>
    <artifactId>schemacrawler-kylin</artifactId>
    <version>3.0.x.x.20260630-SNAPSHOT</version>
</dependency>

Gradle

implementation 'io.github.easy4j:schemacrawler-kylin:3.0.x.x.20260630-SNAPSHOT'

6. Quick Start

Via SchemaCrawler CLI (the plug-in jar plus kylin-jdbc must be on the SchemaCrawler classpath):

schemacrawler --server=kylin \
  --host=localhost --port=7070 \
  --database=default --user=<user> --password=<password> \
  --info-level=standard --command=list

Expected result: SchemaCrawler recognizes --server=kylin, loads the Kylin JDBC driver and lists the databases/tables visible through the Kylin JDBC metadata layer (exact output depends on the Kylin version and on what the Kylin JDBC driver exposes).

Via the library API (plug-in discovery is automatic once the jar is on the classpath):

import schemacrawler.tools.databaseconnector.DatabaseConnectorRegistry;

DatabaseConnectorRegistry registry =
        DatabaseConnectorRegistry.getDatabaseConnectorRegistry();

boolean hasKylin = registry.hasDatabaseSystemIdentifier("kylin"); // true

7. Configuration

The plug-in ships schemacrawler-kylin.config.properties with the following defaults:

Property	Default	Meaning
host	localhost	Kylin host name
port	10000	Port used in the connection URL template
database	default	Database (project) name

Assumption: the shipped default port=10000 appears to have been carried over from a Hive-oriented template. Apache Kylin instances commonly listen on port 7070; verify the port against your deployment and override it explicitly when connecting.

The connection options exposed through SchemaCrawler help are --host, --port, --database, --user, --password.

8. Core Usage / API

The public API surface is intentionally small — one connector class:

Class	Package	Role
KylinDatabaseConnector	schemacrawler.server.kylin	DatabaseConnector implementation; registers kylin, matches jdbc:kylin:* URLs, loads the Kylin JDBC driver.

Constructing the connector directly:

import schemacrawler.server.kylin.KylinDatabaseConnector;

// Throws IOException if the bundled config/information-schema
// resources cannot be loaded; throws RuntimeException if the
// Kylin JDBC driver class is not on the classpath.
KylinDatabaseConnector connector = new KylinDatabaseConnector();

9. Testing & Build

# Full build with tests and JaCoCo coverage report/check
./mvnw clean verify

# Run tests only
./mvnw test

# Install into the local repository
./mvnw install

Test & gate facts (as configured in the pom):

• JUnit 4 (junit:junit:4.13.2, test scope).
• One integration-style test class, schemacrawler.integration.test.TestBundledDistributions, which verifies that the bundled distribution registers the Kylin database system identifier through DatabaseConnectorRegistry.
• JaCoCo is bound to prepare-agent / report / check; the check rule requires a 90% line coverage ratio (configured with haltOnFailure=false, i.e. reported rather than hard-failing).

10. Versioning & Branches

Branch	JDK baseline	Version pattern	Status
feature/1.0.x	JDK 8	1.0.x.*	Active; current snapshot 1.0.x.20260630-SNAPSHOT
feature/2.0.x	JDK 17	2.0.x.*	Maintained
feature/3.0.x	JDK 21	3.0.x.*	Maintained

Maintenance strategy: the 1.0.x line keeps JDK 8 compatibility for legacy deployments; the 2.0.x and 3.0.x lines are the modern JDK baselines. Release artifacts are published to the project's configured artifact repository (Aliyun Packages) and GitHub Releases; the project has not yet published to Maven Central.

11. Contributing & License

Contributions are welcome — please open an issue or a pull request on the GitHub repository. Code style follows the existing conventions of the repository (4-space indentation, commented Maven plugin/dependency blocks).

This project is licensed under the Apache License 2.0. See LICENSE for details.