schemacrawler-hive

English | 简体中文

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

schemacrawler-hive is a SchemaCrawler database plug-in that registers Apache Hive (HiveServer2) as a supported database type. It is discovered through the META-INF/services/schemacrawler.tools.databaseconnector.DatabaseConnector service loader entry and makes SchemaCrawler recognize jdbc:hive2:* URLs.

Is	Is not
A SchemaCrawler DatabaseConnector plug-in for Hive	A Hive client SDK
Auto-loads org.apache.hive.jdbc.HiveDriver	A replacement for Beeline / HiveServer2
Ships information-schema SQL views for Hive	A full Hive metadata repository

Typical scenarios:

Scenario	Description
Schema introspection	Use SchemaCrawler commands (-command=list, -command=schema, ...) against Hive
Metadata export	Generate schema diagrams/text reports from Hive databases
CI documentation	Keep Hive schema docs in sync with the warehouse

2. Features & Status

Capability	Status	Notes
Database server type registration (hive)	Implemented	HiveDatabaseConnector; SchemaCrawler reports it as "Apache Hive DataBase"
JDBC URL recognition	Implemented	supportsUrlPredicate() matches jdbc:hive2:.*
Driver loading	Implemented	Loads org.apache.hive.jdbc.HiveDriver at connector construction
Information-schema SQL views	Implemented	hive.information_schema/ resource folder (tables, views, routines, triggers, constraints, ...)
Connection defaults	Implemented	schemacrawler-hive.config.properties: host/port/database/url
Help command	Implemented	--server=hive help text with host/port/database options
Unit tests	Partial	TestBundledDistributions asserts the registry knows hive; HiveJdbcClient is a runnable JDBC example

Known limitation (carried over from the original README — Assumption: not re-verified against this branch's driver):

Hive's JDBC driver historically implements only part of java.sql.DatabaseMetaData. Standard SchemaCrawler commands that rely on DatabaseMetaData may fail with Database access exception: Method not supported. Verify against your Hive version before relying on full schema introspection.

3. Requirements & Compatibility

Requirement	Version
JDK	8+
Maven	3.0+ (wrapper included)
SchemaCrawler	16.7.2
Hive JDBC	3.1.2 (hive-jdbc)
Hadoop	3.2.1 (hadoop-common, provided)

Version lines of the easy4j project:

Branch	JDK	Version pattern	Notes
feature/1.0.x	8	1.0.x.*	This README, current branch
feature/2.0.x	17	2.0.x.*	JDK 17 line
feature/3.0.x	21	3.0.x.*	JDK 21 line

4. Architecture & Modules

  SchemaCrawler (16.7.2)
        |
        v
  ServiceLoader -> HiveDatabaseConnector
        |
        v
  supportsUrlPredicate: jdbc:hive2:*
        |
        v
  HiveDriver (org.apache.hive.jdbc) -> HiveServer2
        |
        v
  /hive.information_schema SQL views + config properties

Single-module Maven project (jar packaging):

Package	Responsibility
schemacrawler.server.hive	HiveDatabaseConnector (the plug-in entry)
META-INF/services	Service registration for SchemaCrawler
resources/hive.information_schema	Metadata SQL views for Hive
resources/schemacrawler-hive.config.properties	Connection defaults

5. Installation

Artifacts are published to the aliyun repository and GitHub Releases; they are not on Maven Central yet.

<dependency>
    <groupId>io.github.easy4j</groupId>
    <artifactId>schemacrawler-hive</artifactId>
    <version>2.0.x.x.20260630-SNAPSHOT</version>
</dependency>

implementation 'io.github.easy4j:schemacrawler-hive:2.0.x.x.20260630-SNAPSHOT'

The plug-in must be on the same classpath as SchemaCrawler (SchemaCrawler discovers DatabaseConnector implementations via ServiceLoader).

6. Quick Start

# List tables in the "default" database
./sc.sh -server=hive -database=default -host localhost -port 10000 \
        -infolevel=standard -command=list

or programmatically:

DatabaseConnectorRegistry registry = DatabaseConnectorRegistry.getDatabaseConnectorRegistry();
System.out.println(registry.hasDatabaseSystemIdentifier("hive")); // true

// Connect via the connector's configuration (host/port/database -> url)
Connection connection = DriverManager.getConnection(
        "jdbc:hive2://localhost:10000/default", "user", "password");

Expected result: hasDatabaseSystemIdentifier("hive") returns true; standard introspection commands work to the extent the Hive JDBC driver supports DatabaseMetaData (see the limitation note in Section 2).

7. Configuration

Defaults live in schemacrawler-hive.config.properties:

Property	Default	Meaning
host	localhost	HiveServer2 host
port	10000	HiveServer2 port
database	default	Database name
url	jdbc:hive2://${host}:${port}/${database}	JDBC URL template

Command-line options (HiveDatabaseConnector.getHelpCommand()): -host (default localhost), -port (default 3306 in the help text), -database, -user, -password.

8. Core Usage / API

Type	Role
schemacrawler.server.hive.HiveDatabaseConnector	Registers server type hive, loads HiveDriver, provides help and URL matching

The information-schema folder (hive.information_schema) supplies the SQL views SchemaCrawler uses to read tables, columns, constraints, routines, triggers and sequences from the Hive catalog.

9. Testing & Build

./mvnw clean verify

• TestBundledDistributions verifies the plug-in is registered (registry.hasDatabaseSystemIdentifier("hive")).
• HiveJdbcClient (in src/test) is a runnable JDBC example against a live HiveServer2.
• JaCoCo is configured with a line-coverage rule of 90% (haltOnFailure=false).
• Note: hive-jdbc pulls a large dependency tree; hadoop-common is declared provided and must come from your Hadoop environment.

10. Versioning & Branches

Branch	JDK	Version pattern	Maintenance
feature/1.0.x	8	1.0.x.*	Current branch (Hive JDBC 3.1.2)
feature/2.0.x	17	2.0.x.*	JDK 17 line
feature/3.0.x	21	3.0.x.*	JDK 21 line

Artifacts are distributed via the aliyun Maven repository and GitHub Releases. Use the branch matching your JDK baseline.

11. Contributing & License

Contributions are welcome — especially verification of DatabaseMetaData behavior against newer Hive versions. Please open an issue before larger changes.

This project is licensed under the Apache License, Version 2.0.