SchemaCrawler is a free database schema discovery and comprehension tool. SchemaCrawler has a good mix of useful features for data governance. You can search for database schema objects using regular expressions, and output the schema and data in a readable text format. The output serves for database documentation, and is designed to be diff-ed against other database schemas. SchemaCrawler also generates schema diagrams. You can execute scripts in any standard scripting language against your database. You can find potential schema design issues with lint.
SchemaCrawler supports almost any database that has a JDBC driver, but for convenience is bundled with drivers for some commonly used RDBMS systems. SchemaCrawler works with any operating system that supports Java SE 8, Compact Profile 2 or better.
SchemaCrawler is integrated with templating engines, such as Apache Velocity, Thymeleaf, <FreeMarker> and Mustache. You can write templates to generate SQL scripts or any other text output. However, you will need to download the templating engine separately, since these are not part of the SchemaCrawler download.
Complete SchemaCrawler command-line help is available with the -h or -help command-line options.
SchemaCrawler is also a Java API that makes working with database metadata as easy as working with plain old Java objects. Java programmers need to access database metadata
- in order to dynamically generate SQL statements
- when programmatically determining the capabilities of a given RDBMS
- when finding the names and types of tables and columns in the database
Programmers can obtain database metadata using JDBC, but with the raw JDBC API database metadata is returned as result sets, not Java objects. Also, programmers are still responsible for managing resources, mapping into object structures, and handling exceptions. This makes using the JDBC API very cumbersome when it comes to metadata. Furthermore, the JDBC API is not very consistent. For example, to find the type of a table, you would look at the
TABLE_TYPE , which has a string value, but for procedures,
PROCEDURE_TYPE is an integer. An another example, is the
getCatalogs() call, which returns a result set with exactly one column, in contrast to
getStringFunctions() which returns a string containing the list of function names, separated by commas.
SchemaCrawler attempts to solve some of these problems by providing an API that is consistent and usable. Database metadata is provided in the form of plain old Java objects (POJOs). Some examples of the consistency and usability of the SchemaCrawler API are that:
Tableis an object that has a collection of
Columnobjects, without requiring you to make additional calls
- Booleans are Java booleans, not an integer, a string, or null versus not-null, and enumerated values are Java enums, not integers
- Lists are always returned as
- You don’t worry about database resources or exception handling
- You can use standard Java programming idioms - for example, you can access the
Tableobject from a
SchemaCrawler goes beyond what is available using JDBC, and can provide information on database triggers, sequences and synonyms as well.
SchemaCrawler is free and open-source API, available under a number of licenses. SchemaCrawler is written in Java, making it operating system agnostic. Since it leverages JDBC, it is also database independent. It deliberately doesn’t have any RDBMS-specific code. SchemaCrawler allows you to compare structures between two different database servers, or even two different database systems, from different vendors.
SchemaCrawler provides metadata for the following database objects:
- Column data types
- Tables and views
- Primary keys
- Table constraints
- Foreign keys
- Routines, including functions and stored procedures
- Privileges and grants
The sample code below demonstrates just how easy it is to use SchemaCrawler. For more details, please refer to the javadocs.
SchemaCrawler is grateful to