Hsqldb User Guide

The HSQLDB Development Group

Edited by

Blaine Simpson

Fred Toussi

Copyright 2002-2007 HSQLDB Development Group. Permission is granted to distribute this document without any alteration under the terms of the HSQLDB license.

$Date: 2007/08/28 12:13:28 $

Table of Contents

Available formats for this document
1. Running and Using Hsqldb
Running Tools
Running Hsqldb
Server Modes
Hsqldb Server
Hsqldb Web Server
Hsqldb Servlet
In-Process (Standalone) Mode
Memory-Only Databases
Closing the Database
Using Multiple Databases in One JVM
Creating a New Database
Using the Database Engine
Different Types of Tables
Constraints and Indexes
SQL Support
JDBC Support
2. SQL Issues
SQL Standard Support
Constraints and Indexes
Primary Key Constraints
Unique Constraints
Unique Indexes
Indexes and Query Speed
Where Condition or Join
Subqueries and Joins
Types and Arithmetic Operations
Integral Types
Other Numeric Types
Bit and Boolean Types
Storage and Handling of Java Objects
Type Size, Precision and Scale
Sequences and Identity
Identity Auto-Increment Columns
Issues with Transactions
New Features and Changes
3. UNIX Quick Start
Setting up Database Instance and Server
Accessing your Database
Create additional Accounts
Running Hsqldb as a System Daemon
Portability of hsqldb init script
Init script Setup Procedure
Troubleshooting the Init Script
4. Advanced Topics
Connection properties
Properties Files
Server and Web Server Properties
Starting a Server from your application
Individual Database Properties
SQL Commands for Database Properties
5. Deployment Issues
Mode of Operation and Tables
Mode of Operation
Large Objects
Deployment context
Memory and Disk Use
Cache Memory Allocation
Managing Database Connections
Upgrading Databases
Upgrading Using the SCRIPT Command
Manual Changes to the .script File
Backing Up Databases
6. Text Tables
The Implementation
Definition of Tables
Scope and Reassignment
Null Values in Columns of Text Tables
Disconnecting Text Tables
Text File Issues
Text File Global Properties
Importing from a Text Table file
7. TLS
Encrypting your JDBC connection
Making a Private-key Keystore
CA-Signed Cert
Non-CA-Signed Cert
Automatic Server or WebServer startup on UNIX
8. SqlTool
Purpose, Coverage, Changes in Behavior
Platforms and SqlTool versions covered
Functional Changes
The Bare Minimum
Non-displayable Types
Desktop shortcuts
Loading sample data
RC File Authentication Setup
Using Inline RC Authentication
Using the current version of SqlTool with an older HSQLDB distribution.
Interactive Usage
Command Types
Command Types
Special Commands
Edit Buffer / History Commands
PL Commands
? Variable
Storing and retrieving binary files
Command History
Shell scripting and command-line piping
Emulating Non-Interactive mode
Giving SQL on the Command Line
SQL Files
Piping and shell scripting
Optimally Compatible SQL Files
Special Commands and Edit Buffer Commands in SQL Files
Getting Interactive Functionality with SQL Files
Character Encoding
Generating Text or HTML Reports
SqlTool Procedural Language
PL Aliases
Logical Expressions
Flow Control
Raw Mode
Using hsqltool.jar and hsqldbutil.jar
Delimiter-Separated-Value Imports and Exports
Simple DSV exports and imports using default settings
Specifying queries and options
Unit Testing SqlTool
9. SQL Syntax
Notational Conventions Used in this Chapter
SQL Commands
Schema object naming
Data Types
SQL Comments
Stored Procedures / Functions
Built-in Functions and Stored Procedures
SQL Expression
A. Building HSQLDB
Building with Ant
Obtaining Ant
Building Hsqldb with Ant
Building with DOS Batch Files
Hsqldb CodeSwitcher
Building documentation
B. First JDBC Client Example
C. Hsqldb Database Files and Recovery
Clean Shutdown
D. Running Hsqldb with OpenOffice.org 1.1.x
Setting up OpenOffice.org
On Windows
On Linux
E. Hsqldb Test Utility
F. Database Manager
Brief Introduction
Auto tree-update
Automatic Connection
RC File
Using the current DatabaseManagers with an older HSQLDB distribution.
DatabaseManagerSwing as an Applet
G. Transfer Tool
Brief Introduction

List of Tables

1. Alternate formats of this document
4.1. Hsqldb URL Components
4.2. Connection Properties
4.3. Hsqldb Server Properties Files
4.4. Property File Properties
4.5. Server Property File Properties
4.6. WebServer Property File Properties
4.7. Database-specific Property File Properties
4.8. SQL command properties
9.1. Data Types

List of Examples

1.1. Java code to connect to the local Server above
2.1. Column values which satisfy a 2-column UNIQUE constraint
2.2. Query comparison
2.3. Numbering returned rows of a SELECT in sequential order
3.1. server.properties fragment
3.2. example sqltool.rc stanza
7.1. Exporting certificate from the server's keystore
7.2. Adding a certificate to the client keystore
7.3. Specifying your own trust store to a JDBC client
7.4. Running an Hsqldb server with TLS encryption
7.5. Getting a pem-style private key into a JKS keystore
8.1. Sample RC File
8.2. Defining and using a PL alias (PL variable)
8.3. Inserting binary data into database from a file
8.4. Downloading binary data from database to a file
8.5. Piping input into SqlTool
8.6. Valid comment example
8.7. Invalid comment example
8.8. Simple SQL file using PL
8.9. SQL File showing use of most PL features
8.10. Single-line chunking example
8.11. Multi-line chunking example
8.12. Interactive Raw Mode example
8.13. PL/SQL Example
8.14. DSV Export Example
8.15. DSV Import Example
8.16. DSV Export of an Arbitrary SELECT Statement
8.17. Sample DSV headerswitch settings
8.18. DSV targettable setting
A.1. Buiding the standard Hsqldb jar file with Ant
A.2. Example source code before CodeSwitcher is run
A.3. CodeSwitcher command line invocation
A.4. Source code after CodeSwitcher processing
A.5. Building HTML User Guides
A.6. Building User Guides in all formats
B.1. JDBC Client source code example


If you notice any mistakes in this document, please email the author listed at the beginning of the chapter. If you have problems with the procedures themselves, please use the HSQLDB support facilities which are listed at http://hsqldb.org/web/hsqlSupport.html.

Available formats for this document

This document is available in several formats.

You may be reading this document right now at http://hsqldb.org/doc/guide, or in a distribution somewhere else. I hereby call the document distribution from which you are reading this, your current distro.

http://hsqldb.org/doc/guide hosts the latest production versions of all available formats. If you want a different format of the same version of the document you are reading now, then you should try your current distro. If you want the latest production version, you should try http://hsqldb.org/doc/guide.

Sometimes, distributions other than http://hsqldb.org/doc/guide do not host all available formats. So, if you can't access the format that you want in your current distro, you have no choice but to use the newest production version at http://hsqldb.org/doc/guide.

Table 1. Alternate formats of this document

formatyour distroat http://hsqldb.org/doc/guide
Chunked HTML index.htmlhttp://hsqldb.org/doc/guide/index.html
All-in-one HTML guide.htmlhttp://hsqldb.org/doc/guide/guide.html
PDF guide.pdfhttp://hsqldb.org/doc/guide/guide.pdf

Chapter 1. Running and Using Hsqldb

Fred Toussi

HSQLDB Development Group

Copyright 2002-2005 Fred Toussi. Permission is granted to distribute this document without any alteration under the terms of the HSQLDB license. Additional permission is granted to the HSQLDB Development Group to distribute this document with or without alterations under the terms of the HSQLDB license.

$Date: 2007/05/30 18:34:46 $


The HSQLDB jar package is located in the /lib directory and contains several components and programs. Different commands are used to run each program.

Components of the Hsqldb jar package


  • HSQLDB JDBC Driver

  • Database Manager (Swing and AWT versions)

  • Query Tool (AWT)

  • Sql Tool (command line)

The HSQLDB RDBMS and JDBC Driver provide the core functionality. The rest are general-purpose database tools that can be used with any database engine that has a JDBC driver.

Running Tools

All tools can be run in the standard way for archived Java classes. In the following example the AWT version of the Database Manager, the hsqldb.jar is located in the directory ../lib relative to the current directory.

    java -cp ../lib/hsqldb.jar org.hsqldb.util.DatabaseManager

If hsqldb.jar is in the current directory, the command would change to:

    java -cp hsqldb.jar org.hsqldb.util.DatabaseManager

Main classes for the Hsqldb tools

  • org.hsqldb.util.DatabaseManager

  • org.hsqldb.util.DatabaseManagerSwing

  • org.hsqldb.util.Transfer

  • org.hsqldb.util.QueryTool

  • org.hsqldb.util.SqlTool

Some tools, such as the Database Manager or SQL Tool, can use command line arguments or entirely rely on them. You can add the command line argument -? to get a list of available arguments for these tools. Database Manager features a graphical user interface and can be explored interactively.

Running Hsqldb

HSQLDB can be run in a number of different ways. In general these are divided into Server Modes and In-Process Mode (also called Standalone Mode). A different sub-program from the jar is used to run HSQLDB in each mode.

Each HSQLDB database consists of between 2 to 5 files, all named the same but with different extensions, located in the same directory. For example, the database named "test" consists of the following files:

  • test.properties

  • test.script

  • test.log

  • test.data

  • test.backup

The properties files contains general settings about the database. The script file contains the definition of tables and other database objects, plus the data for non-cached tables. The log file contains recent changes to the database. The data file contains the data for cached tables and the backup file is a zipped backup of the last known consistent state of the data file. All these files are essential and should never be deleted. If the database has no cached tables, the test.data and test.backup files will not be present. In addition to those files, HSQLDB database may link to any formatted text files, such as CSV lists, anywhere on the disk.

While the "test" database is operational, a test.log file is used to write the changes made to data. This file is removed at a normal SHUTDOWN. Otherwise (with abnormal shutdown) this file is used at the next startup to redo the changes. A test.lck file is also used to record the fact that the database is open. This is deleted at a normal SHUTDOWN. In some circumstances, a test.data.old is created and deleted afterwards.


When the engine closes the database at a shutdown, it creates temporary files with the extension .new which it then renames to those listed above.

Server Modes

Server modes provide the maximum accessibility. The database engine runs in a JVM and listens for connections from programs on the same computer or other computers on the network. Several different programs can connect to the server and retrieve or update information. Applications programs (clients) connect to the server using the HSQLDB JDBC driver. In most server modes, the server can serve up to 10 databases that are specified at the time of running the server.

Server modes can use preset properties or command line arguments as detailed in the Advanced Topics chapter. There are three server modes, based on the protocol used for communications between the client and server.

Hsqldb Server

This is the preferred way of running a database server and the fastest one. A proprietary communications protocol is used for this mode. A command similar to those used for running tools and described above is used for running the server. The following example of the command for starting the server starts the server with one (default) database with files named "mydb.*".

    java -cp ../lib/hsqldb.jar org.hsqldb.Server -database.0 file:mydb -dbname.0 xdb

The command line argument -? can be used to get a list of available arguments.

Hsqldb Web Server

This mode is used when access to the computer hosting the database server is restricted to the HTTP protocol. The only reason for using the Web Server mode is restrictions imposed by firewalls on the client or server machines and it should not be used where there are no such restrictions. The HSQLDB Web Server is a special web server that allows JDBC clients to connect via HTTP. From 1.7.2 this mode also supports transactions.

To run a web server, replace the main class for the server in the example command line above with the following:


The command line argument -? can be used to get a list of available arguments.

Hsqldb Servlet

This uses the same protocol as the Web Server. It is used when a separate servlet engine (or application server) such as Tomcat or Resin provides access to the database. The Servlet Mode cannot be started independently from the servlet engine. The Servlet class, in the HSQLDB jar, should be installed on the application server to provide the connection. The database is specified using an application server property. Refer to the source file org.hsqldb.Servlet.java to see the details.

Both Web Server and Servlet modes can only be accessed using the JDBC driver at the client end. They do not provide a web front end to the database. The Servlet mode can serve only a single database.

Please note that you do not normally use this mode if you are using the database engine in an application server.

Connecting to a Database running as a Server

Once an HSQLDB server is running, client programs can connect to it using the HSQLDB JDBC Driver contained in hsqldb.jar. Full information on how to connect to a server is provided in the Java Documentation for jdbcConnection (located in the /doc/src directory of HSQLDB distribution. A common example is connection to the default port (9001) used for the hsql protocol on the same machine:

Example 1.1. Java code to connect to the local Server above

    try {
        Class.forName("org.hsqldb.jdbcDriver" );
    } catch (Exception e) {
        System.out.println("ERROR: failed to load HSQLDB JDBC driver.");

    Connection c = DriverManager.getConnection("jdbc:hsqldb:hsql://localhost/xdb", "sa", "");

In some circumstances, you may have to use the following line to get the driver.


Note in the above connection URL, there is no mention of the database file, as this was specified when running the server. Instead, the value defined for dbname.0 is used. Also, see the Advanced Topics chapter for the connection URL when there is more than one database per server instance.

Security Considerations

When HSQLDB is run as a server, network access should be adequately protected. Source IP addresses may be restricted by use of TCP filtering or firewall programs, or standalone firewalls. If the traffic will cross an unprotected network (such as the Internet), the stream should be encrypted (for example by VPN, ssh tunneling, or TLS using the SSL enabled HSQLS and HTTPS variants of the server and web server modes). Only secure passwords should be used-- most importantly, the password for the default system user should be changed from the default empty string. If you are purposefully providing data to the public, then the wide-open public network connection should be used exclusively to access the public data via read-only accounts. (I.e., neither secure data nor privileged accounts should use this connection). These considerations also apply to HSQLDB servers run with the HTTP protocol.

In-Process (Standalone) Mode

This mode runs the database engine as part of your application program in the same Java Virtual Machine. For most applications this mode can be faster, as the data is not converted and sent over the network. The main drawback is that it is not possible by default to connect to the database from outside your application. As a result you cannot check the contents of the database with external tools such as Database Manager while your application is running. In 1.8.0, you can run a server instance in a thread from the same virtual machine as your application and provide external access to your in-process database.

The recommended way of using the in-process mode in an application is to use an HSQLDB Server instance for the database while developing the application and then switch to In-Process mode for deployment.

An In-Process Mode database is started from JDBC, with the database file path specified in the connection URL. For example, if the database name is testdb and its files are located in the same directory as where the command to run your application was issued, the following code is used for the connection:

    Connection c = DriverManager.getConnection("jdbc:hsqldb:file:testdb", "sa", "");

The database file path format can be specified using forward slashes in Windows hosts as well as Linux hosts. So relative paths or paths that refer to the same directory on the same drive can be identical. For example if your database path in Linux is /opt/db/testdb and you create an identical directory structure on the C: drive of a Windows host, you can use the same URL in both Windows and Linux:

    Connection c = DriverManager.getConnection("jdbc:hsqldb:file:/opt/db/testdb", "sa", "");

When using relative paths, these paths will be taken relative to the directory in which the shell command to start the Java Virtual Machine was executed. Refer to Javadoc for jdbcConnection for more details.

Memory-Only Databases

It is possible to run HSQLDB in a way that the database is not persistent and exists entirely in random access memory. As no information is written to disk, this mode should be used only for internal processing of application data, in applets or certain special applications. This mode is specified by the mem: protocol.

    Connection c = DriverManager.getConnection("jdbc:hsqldb:mem:aname", "sa", "");

You can also run a memory-only server instance by specifying the same URL in the server.properties. This usage is not common and is limited to special applications where the database server is used only for exchanging information between clients, or for non-persistent data.


Closing the Database

All databases running in different modes can be closed with the SHUTDOWN command, issued as an SQL query. From version 1.7.2, in-process databases are no longer closed when the last connection to the database is explicitly closed via JDBC, a SHUTDOWN is required. In 1.8.0, a connection property, shutdown=true, can be specified on the first connection to the database (the connection that opens the database) to force a shutdown when the last connection closes.

When SHUTDOWN is issued, all active transactions are rolled back. A special form of closing the database is via the SHUTDOWN COMPACT command. This command rewrites the .data file that contains the information stored in CACHED tables and compacts it to size. This command should be issued periodically, especially when lots of inserts, updates or deletes have been performed on the cached tables. Changes to the structure of the database, such as dropping or modifying populated CACHED tables or indexes also create large amounts of unused file space that can be reclaimed using this command.

Using Multiple Databases in One JVM

In the above examples each server serves only one database and only one in-memory database can be created. However, from version 1.7.2, HSQLDB can serve several databases in multiple server modes and allow simultaneous access to multiple in-process and memory-only databases. These capabilities are covered in the Advanced Topics chapter.

Creating a New Database

When a server instance is started, or when a connection is made to an in-process database, a new, empty database is created if no database exists at the given path.

This feature has a side effect that can confuse new users. If a mistake is made in specifying the path for connecting to an existing database, a connection is nevertheless established to a new database. For troubleshooting purposes, you can specify a connection property ifexists=true to allow connection to an existing database only and avoid creating a new database. In this case, if the database does not exist, the getConnection() method will throw an exception.

Using the Database Engine

Once a connection is established to a database in any mode, JDBC methods are used to interact with the database. The Javadoc for jdbcConnection , jdbcDriver , jdbcDatabaseMetadata , jdbcResultSet , jdbcStatement , and jdbcPreparedStatement list all the supported JDBC methods together with information that is specific to HSQLDB. JDBC methods are broadly divided into: connection related methods, metadata methods and database access methods. The database access methods use SQL commands to perform actions on the database and return the results either as a Java primitive type or as an instance of the java.sql.ResultSet class.

You can use Database Manager or other Java database access tools to explore your database and update it with SQL commands. These programs use JDBC internally to submit your commands to the database engine and to display the results in a human readable format.

The SQL dialect used in HSQLDB is as close to the SQL92 and SQL200n standards as it has been possible to achieve so far in a small-footprint database engine. The full list of SQL commands is in the SQL Syntax chapter.

Different Types of Tables

HSQLDB supports TEMP tables and three types of persistent tables.

TEMP tables are not written to disk and last only for the lifetime of the Connection object. The contents of each TEMP table is visible only from the Connection that was used to populate it; other concurrent connections to the database will have access to their own copies of the table. Since 1.8.0 the definition of TEMP tables conforms to the GLOBAL TEMPORARY type in the SQL standard. The definition of the table persists but each new connections sees its own copy of the table, which is empty at the beginning. When the connection commits, the contents of the table are cleared by default. If the table definition statements includes ON COMMIT PRESERVE ROWS, then the contents are kept when a commit takes place.

The three types of persistent tables are MEMORY tables, CACHED tables and TEXT tables.

Memory tables are the default type when the CREATE TABLE command is used. Their data is held entirely in memory but any change to their structure or contents is written to the <dbname>.script file. The script file is read the next time the database is opened, and the MEMORY tables are recreated with all their contents. So unlike TEMP table, the default, MEMORY tables are persistent.

CACHED tables are created with the CREATE CACHED TABLE command. Only part of their data or indexes is held in memory, allowing large tables that would otherwise take up to several hundred megabytes of memory. Another advantage of cached tables is that the database engine takes less time to start up when a cached table is used for large amounts of data. The disadvantage of cached tables is a reduction in speed. Do not use cached tables if your data set is relatively small. In an application with some small tables and some large ones, it is better to use the default, MEMORY mode for the small tables.

TEXT tables are supported since version 1.7.0 and use a CSV (Comma Separated Value) or other delimited text file as the source of their data. You can specify an existing CSV file, such as a dump from another database or program, as the source of a TEXT table. Alternatively, you can specify an empty file to be filled with data by the database engine. TEXT tables are efficient in memory usage as they cache only part of the text data and all of the indexes. The Text table data source can always be reassigned to a different file if necessary. Two commands are needed to set up a TEXT table as detailed in the Text Tables chapter.

With memory-only databases (see above), both MEMORY table and CACHED table declarations are treated as declarations for non-persistent memory tables. TEXT table declarations are not allowed in this mode.

Constraints and Indexes

HSQLDB supports PRIMARY KEY, NOT NULL, UNIQUE, CHECK and FOREIGN KEY constraints. In addition, it supports UNIQUE or ordinary indexes. This support is fairly comprehensive and covers multi-column constraints and indexes, plus cascading updates and deletes for foreign keys.

HSQLDB creates indexes internally to support PRIMARY KEY, UNIQUE and FOREIGN KEY constraints: a unique index is created for each PRIMARY KEY or UNIQUE constraint; an ordinary index is created for each FOREIGN KEY constraint. Because of this, you should not create duplicate user-defined indexes on the same column sets covered by these constraints. This would result in unnecessary memory and speed overheads. See the discussion in the SQL Issues chapter for more information.

Indexes are crucial for adequate query speed. When queries joining multiple tables are used, there must be an index on each joined column of each table. When range or equality conditions are used e.g. SELECT ... WHERE acol >10 AND bcol = 0, an indexe is required on the acol column used in the condition. Indexes have no effect on ORDER BY clauses or some LIKE conditions.

As a rule of thumb, HSQLDB is capable of internal processing of queries at over 100,000 rows per second. Any query that runs into several seconds should be checked and indexes should be added to the relevant columns of the tables if necessary.

SQL Support

The SQL syntax supported by HSQLDB is essentially that specified by the SQL Standard (92 and 200n). Not all the features of the Standard are supported and there are some proprietary extensions. In 1.8.0 the behaviour of the engine is far more compliant with the Standards than with older versions. The main changes are

  • correct treatment of NULL column values in joins, in UNIQUE constraints and in query conditions

  • correct processing of selects with JOIN and LEFT OUTER JOIN

  • correct processing of aggregate functions contained in expressions or containing expression arguments

The supported commands are listed in the SQL Syntax chapter. For a well written basic guide to SQL with examples you can consult PostgreSQL: Introduction and Concepts by Bruce Momjian, which is available on the web. Most of the SQL coverage in the book applies also to HSQLDB. There are some differences in keywords supported by one and not the other engine (OUTER, OID's, etc.) or used differently (IDENTITY/SERIAL, TRIGGER, SEQUENCE, etc.).

JDBC Support

Since 1.7.2, support for JDBC2 has been significantly extended and some features of JDBC3 are also supported. The relevant classes are thoroughly documented. See the JavaDoc for org.hsqldb.jdbcXXXX classes.

Chapter 2. SQL Issues

Fred Toussi

HSQLDB Development Group

Copyright 2002-2005 Fred Toussi. Permission is granted to distribute this document without any alteration under the terms of the HSQLDB license. Additional permission is granted to the HSQLDB Development Group to distribute this document with or without alterations under the terms of the HSQLDB license.

$Date: 2005/07/01 17:06:32 $


Many questions repeatedly asked in Forums and mailing lists are answered in this guide. If you want to use HSQLDB with your application, you should read this guide.

SQL Standard Support

HSQLDB 1.8.0 supports the dialect of SQL defined by SQL standards 92, 99 and 2003. This means where a feature of the standard is supported, e.g. left outer join, the syntax is that specified by the standard text. Many features of SQL92 and 99 up to Advanced Level are supported and there is support for most of SQL 2003 Foundation and several optional features of this standard. However, certain features of the Standards are not supported so no claim is made for full support of any level of the standards.

The SQL Syntax chapter of this guide SQL Syntax lists all the keywords and syntax that is supported. When writing or converting existing SQL DDL (Data Definition Language) and DML (Data Manipulation Language) statements for HSQLDB, you should consult the supported syntax and modify the statements accordingly.

Several words are reserved by the standard and cannot be used as table or column names. For example, the word POSITION is reserved as it is a function defined by the Standards with a similar role as String.indexOf() in Java. HSQLDB does not currently prevent you from using a reserved word if it does not support its use or can distinguish it. For example BEGIN is a reserved words that is not currently supported by HSQLDB and is allowed as a table or column name. You should avoid the use of such words as future versions of HSQLDB are likely to support the words and will reject your table definitions or queries. The full list of SQL reserved words is in the source of the org.hsqldb.Token class.

HSQLDB also supports some keywords and expressions that are not part of the SQL standard as enhancements. Expressions such as SELECT TOP 5 FROM .., SELECT LIMIT 0 10 FROM ... or DROP TABLE mytable IF EXISTS are among such constructs.

All keywords, can be used for database objects if they are double quoted.

Constraints and Indexes

Primary Key Constraints

Before 1.7.0, a CONSTRAINT <name> PRIMARY KEY was translated internally to a unique index and, in addition, a hidden column was added to the table with an extra unique index. From 1.7.0 both single-column and multi-column PRIMARY KEY constraints are supported. They are supported by a unique index on the primary key column(s) specified and no extra hidden column is maintained for these indexes.

Unique Constraints

According to the SQL standards, a unique constraint on a single column means no two values are equal unless one of them is NULL. This means you can have one or more rows where the column value is NULL.

A unique constraint on multiple columns (c1, c2, c3, ..) means that no two sets of values for the columns are equal unless at lease one of them is NULL. Each single column taken by itself can have repeat values. The following example satisfies a UNIQUE constraint on the two columns:

Example 2.1. Column values which satisfy a 2-column UNIQUE constraint


Since version 1.7.2 the behaviour of UNIQUE constraints and indexes with respect to NULL values has changed to conform to SQL standards. A row, in which the value for any of the UNIQUE constraint columns is NULL, can always be added to the table. So multiple rows can contain the same values for the UNIQUE columns if one of the values is NULL.

Unique Indexes

In 1.8.0, user defined UNIQUE indexes can still be declared but they are deprecated. You should use a UNIQUE constraint instead.

CONSTRAINT <name> UNIQUE always creates internally a unique index on the columns, as with previous versions, so it has exactly the same effect as the deprecated UNIQUE index declaration.


From version 1.7.0, HSQLDB features single and multiple column foreign keys. A foreign key can also be specified to reference a target table without naming the target column(s). In this case the primary key column(s) of the target table is used as the referenced column(s). Each pair of referencing and referenced columns in any foreign key should be of identical type. When a foreign key is declared, a unique constraint (or primary key) must exist on the referenced columns in the primary key table. A non-unique index is automatically created on the referencing columns. For example:

    CREATE TABLE child(c1 INTEGER, c2 VARCHAR, FOREIGN KEY (c1, c2) REFERENCES parent(p1, p2));

There must be a UNIQUE constraint on columns (p1,p2) in the table named "parent". A non-unique index is automatically created on columns (c1, c2) in the table named "child". Columns p1 and c1 must be of the same type (INTEGER). Columns p2 and c2 must be of the same type (VARCHAR).

Indexes and Query Speed

HSQLDB does not use indexes to improve sorting of query results. But indexes have a crucial role in improving query speed. If no index is used in a query on a single table, such as a DELETE query, then all the rows of the table must be examined. With an index on one of the columns that is in the WHERE clause, it is often possible to start directly from the first candidate row and reduce the number of rows that are examined.

Indexes are even more important in joins between multiple tables. SELECT ... FROM t1 JOIN t2 ON t1.c1 = t2.c2 is performed by taking rows of t1 one by one and finding a matching row in t2. If there is no index index on t2.c2 then for each row of t1, all the rows of t2 must be checked. Whereas with an index, a matching row can be found in a fraction of the time. If the query also has a condition on t1, e.g., SELECT ... FROM t1 JOIN t2 ON t1.c1 = t2.c2 WHERE t1.c3 = 4 then an index on t1.c3 would eliminate the need for checking all the rows of t1 one by one, and will reduce query time to less than a millisecond per returned row. So if t1 and t2 each contain 10,000 rows, the query without indexes involves checking 100,000,000 row combinations. With an index on t2.c2, this is reduced to 10,000 row checks and index lookups. With the additional index on t2.c2, only about 4 rows are checked to get the first result row.

Indexes are automatically created for primary key and unique columns. Otherwise you should define an index using the CREATE INDEX command.

Note that in HSQLDB a unique index on multiple columns can be used internally as a non-unique index on the first column in the list. For example: CONSTRAINT name1 UNIQUE (c1, c2, c3); means there is the equivalent of CREATE INDEX name2 ON atable(c1);. So you do not need to specify an extra index if you require one on the first column of the list.

In 1.8.0, a multi-column index will speed up queries that contain joins or values on ALL the columns. You need NOT declare additional individual indexes on those columns unless you use queries that search only on a subset of the columns. For example, rows of a table that has a PRIMARY KEY or UNIQUE constraint on three columns or simply an ordinary index on those columns can be found efficiently when values for all three columns are specified in the WHERE clause. For example, SELECT ... FROM t1 WHERE t1.c1 = 4 AND t1.c2 = 6 AND t1.c3 = 8 will use an index on t1(c1,c2,c3) if it exists.

As a result of the improvements to multiple key indexes, the order of declared columns of the index or constraint has less affect on the speed of searches than before. If the column that contains more diverse values appears first, the searches will be slightly faster.

A multi-column index will not speed up queries on the second or third column only. The first column must be specified in the JOIN .. ON or WHERE conditions.

Query speed depends a lot on the order of the tables in the JOIN .. ON or FROM clauses. For example the second query below should be faster with large tables (provided there is an index on TB.COL3). The reason is that TB.COL3 can be evaluated very quickly if it applies to the first table (and there is an index on TB.COL3):

    (TB is a very large table with only a few rows where TB.COL3 = 4)



The general rule is to put first the table that has a narrowing condition on one of its columns.

1.7.3 features automatic, on-the-fly indexes for views and subselects that are used in a query. An index is added to a view when it is joined to a table or another view.

Where Condition or Join

Using WHERE conditions to join tables is likely to reduce execution speed. For example the following query will generally be slow, even with indexes:


The query implies TA.COL1 = TB.COL2 but does not explicitly set this condition. If TA and TB each contain 100 rows, 10000 combinations will be joined with TC to apply the column conditions, even though there may be indexes on the joined columns. With the JOIN keyword, the TA.COL1 = TB.COL2 condition has to be explicit and will narrow down the combination of TA and TB rows before they are joined with TC, resulting in much faster execution with larger tables:


The query can be speeded up a lot more if the order of tables in joins are changed, so that TC.COL1 = 1 is applied first and a smaller set of rows are joined together:


In the above example the engine automatically applies TC.COL4 = 1 to TC and joins only the set of rows that satisfy this condition with other tables. Indexes on TC.COL4, TB.COL2 and TA.COL1 will be used if present and will speed up the query.

Subqueries and Joins

Using joins and setting up the order of tables for maximum performance applies to all areas. For example, the second query below should generally be much faster if there are indexes on TA.COL1 and TB.COL3:

Example 2.2. Query comparison



The second query turns MAX(TB.COL2) into a single row table then joins it with TA. With an index on TA.COL1, this will be very fast. The first query will test each row in TA and evaluate MAX(TB.COL2) again and again.

Types and Arithmetic Operations

Table columns of all types supported by HSQLDB can be indexed and can feature in comparisons. Types can be explicitly converted using the CONVERT() library function, but in most cases they are converted automatically. It is recommended not to use indexes on LONGVARBINARY, LONGVARCHAR and OTHER columns, as these indexes will probably not be allowed in future versions.

Previous versions of HSQLDB featured poor handling of arithmetic operations. For example, it was not possible to insert 10/2.5 into any DOUBLE or DECIMAL column. Since 1.7.0, full operations are possible with the following rules:

TINYINT, SMALLINT, INTEGER, BIGINT, NUMERIC and DECIMAL (without a decimal point) are supported integral types and map to byte, short, int, long and BigDecimal in Java. The SQL type dictates the maximum and minimum values that can be held in a field of each type. For example the value range for TINYINT is -128 to +127, although the actual Java type used for handling TINYINT is java.lang.Integer.

REAL, FLOAT, DOUBLE are all mapped to double in Java.

DECIMAL and NUMERIC are mapped to java.math.BigDecimal and can have very large numbers of digits.

Integral Types

TINYINT, SMALLINT, INTEGER, BIGINT, NUMERIC and DECIMAL (without a decimal point) are fully interchangeable internally, and no data narrowing takes place. Depending on the types of the operands, the result of the operations is returned in a JDBC ResultSet in any of related Java types: Integer, Long or BigDecimal. The ResultSet.getXXXX() methods can be used to retrieve the values so long as the returned value can be represented by the resulting type. This type is deterministically based on the query, not on the actual rows returned. The type does not change when the same query that returned one row, returns many rows as a result of adding more data to the tables.

If the SELECT statement refers to a simple column or function, then the return type is the type corresponding to the column or the return type of the function. For example:


would return a result set where the type of the first column is java.lang.Integer and the second column is java.lang.Long. However,

    SELECT MAX(a) + 1, MAX(b) + 1 FROM t;

would return java.lang.Long and BigDecimal values, generated as a result of uniform type promotion for all the return values.

There is no built-in limit on the size of intermediate integral values in expressions. As a result, you should check for the type of the ResultSet column and choose an appropriate getXXXX() method to retrieve it. Alternatively, you can use the getObject() method, then cast the result to java.lang.Number and use the intValue() or longValue() methods on the result.

When the result of an expression is stored in a column of a database table, it has to fit in the target column, otherwise an error is returned. For example when 1234567890123456789012 / 12345687901234567890 is evaluated, the result can be stored in any integral type column, even a TINYINT column, as it is a small value.

Other Numeric Types

In SQL statements, numbers with a decimal point are treated as DECIMAL unless they are written with an exponent. Thus 0.2 is considered a DECIMAL value but 0.2E0 is considered a DOUBLE value.

When PreparedStatement.setDouble() or setFloat() is used, the value is treated as a DOUBLE automatically.

When a REAL, FLOAT or DOUBLE (all synonymous) is part of an expression, the type of the result is DOUBLE.

Otherwise, when no DOUBLE value exists, if a DECIMAL or NUMERIC value is part an expression, the type of the result is DECIMAL. The result can be retrieved from a ResultSet in the required type so long as it can be represented. This means DECIMAL values can be converted to DOUBLE unless they are beyond the Double.MIN_VALUE - Double.MAX_VALUE range. Similar to integral values, when the result of an expression is stored in a table column, it has to fit in the target column, otherwise an error is returned.

The distinction between DOUBLE and DECIMAL is important when a division takes place. When the terms are DECIMAL, the result is a value with a scale (number of digits to the right of the decimal point) equal to the larger of the scales of the two terms. With a DOUBLE term, the scale will reflect the actual result of the operation. For example, 10.0/8.0 (DECIMAL) equals 1.2 but 10.0E0/8.0E0 (DOUBLE) equals 1.25. Without division operations, DECIMAL values represent exact arithmetic; the resulting scale is the sum of the scales of the two terms when multiplication is performed.

REAL, FLOAT and DOUBLE values are all stored in the database as java.lang.Double objects. Special values such as NaN and +-Infinity are also stored and supported. These values can be submitted to the database via JDBC PreparedStatement methods and are returned in ResultSet objects.

Bit and Boolean Types

Since 1.7.2, BIT is simply an alias for BOOLEAN. The primary representation of BOOLEAN column is 'true' or 'false' either as the boolean type or as strings when used from JDBC. This type of column can also be initialised using values of any numeric type. In this case 0 is translated to false and any other value such as 1 is translated to true.

Since 1.7.3 the BOOLEAN type conforms to the SQL standards and supports the UNDEFINED state in addition to TRUE or FALSE. NULL values are treated as undefined. This improvement affects queries that contain NOT IN. See the test text file, TestSelfNot.txt, for examples of the queries.

Storage and Handling of Java Objects

Since version 1.7.2 this support has improved and any serializable JAVA Object can be inserted directly into a column of type OTHER using any variation of PreparedStatement.setObject() methods.

For comparison purposes and in indexes, any two Java Objects are considered equal unless one of them is NULL. You cannot search for a specific object or perform a join on a column of type OTHER.

Please note that HSQLDB is not an object-relational database. Java Objects can simply be stored internally and no operations should be performed on them other than assignment between columns of type OTHER or tests for NULL. Tests such as WHERE object1 = object2, or WHERE object1 = ? do not mean what you might expect, as any non-null object would satisfy such a tests. But WHERE object1 IS NOT NULL is perfectly acceptable.

The engine does not return errors when normal column values are assigned to Java Object columns (for example assigning an INTEGER or STRING to such a column with an SQL statement such as UPDATE mytable SET objectcol = intcol WHERE ...) but this is highly likely to be disallowed in future. So please use columns of type OTHER only to store your objects and nothing else.

Type Size, Precision and Scale

Prior to 1.7.2, all table column type definitions with a column size, precision or scale qualifier were accepted and ignored.

In 1.8.0, such qualifiers must conform to the SQL standards. For example INTEGER(8) is no longer acceptable. The qualifiers are still ignored unless you set a database property. SET PROPERTY "sql.enforce_strict_size" TRUE will enforce sizes for CHARACTER or VARCHAR columns and pad any strings when inserting or updating a CHARACTER column. The precision and scale qualifiers are also enforced for DECIMAL and NUMERIC types. TIMESTAMP can be used with a precision of 0 or 6 only.

Casting a value to a qualified CHARACTER type will result in truncation or padding as you would expect. So a test such as CAST (mycol AS VARCHAR(2)) = 'xy' will find the values beginning with 'xy'. This is the equivalent of SUBSTRING(mycol FROM 1 FOR 2) = 'xy'.

Sequences and Identity

The SEQUENCE keyword was introduced in 1.7.2 with a subset of the SQL 200n standard syntax. Corresponding SQL 200n syntax for IDENTITY columns has also been introduced.

Identity Auto-Increment Columns

Each table can contain one auto-increment column, known as the IDENTITY column. An IDENTITY column is always treated as the primary key for the table (as a result, multi-column primary keys are not possible with an IDENTITY column present). Support has been added for CREATE TABLE <tablename>(<colname> IDENTITY, ...) as a shortcut.

Since 1.7.2, the SQL standard syntax is used by default, which allows the initial value to be specified. The supported form is(<colname> INTEGER GENERATED BY DEFAULT AS IDENTITY(START WITH n, [INCREMENT BY m])PRIMARY KEY, ...). Support has also been added for BIGINT identity columns. As a result, an IDENTITY column is simply an INTEGER or BIGINT column with its default value generated by a sequence generator.

When you add a new row to such a table using an INSERT INTO <tablename> ...; statement, you can use the NULL value for the IDENTITY column, which results in an auto-generated value for the column. The IDENTITY() function returns the last value inserted into any IDENTITY column by this connection. Use CALL IDENTITY(); as an SQL statement to retrieve this value. If you want to use the value for a field in a child table, you can use INSERT INTO <childtable> VALUES (...,IDENTITY(),...);. Both types of call to IDENTITY() must be made before any additional update or insert statements are issued on the database.

The next IDENTITY value to be used can be set with the

ALTER TABLE ALTER COLUMN <column name> RESTART WITH <new value>;


The SQL 200n syntax and usage is different from what is supported by many existing database engines. Sequences are created with the CREATE SEQUENCE command and their current value can be modified at any time with ALTER SEQUENCE. The next value for a sequence is retrieved with the NEXT VALUE FOR <name> expression. This expression can be used for inserting and updating table rows. You can also use it in select statements. For example, if you want to number the returned rows of a SELECT in sequential order, you can use:

Example 2.3. Numbering returned rows of a SELECT in sequential order

    SELECT NEXT VALUE FOR mysequence, col1, col2 FROM mytable WHERE ...

Please note that the semantics of sequences is not exactly the same as defined by SQL 200n. For example if you use the same sequence twice in the same row insert query, you will get two different values, not the same value as required by the standard.

You can query the SYSTEM_SEQUENCES table for the next value that will be returned from any of the defined sequences. The SEQUENCE_NAME column contains the name and the NEXT_VALUE column contains the next value to be returned.

Issues with Transactions

HSQLDB supports transactions at the READ_UNCOMMITTED level, also known as level 0 transaction isolation. This means that during the lifetime of a transaction, other connections to the database can see the changes made to the data. Transaction support works well in general. Reported bugs concerning transactions being committed if the database is abruptly closed have been fixed. However, the following issues may be encountered only with multiple connections to a database using transactions:

If two transactions modify the same row, no exception is raised when both transactions are committed. This can be avoided by designing your database in such a way that application data consistency does not depend on exclusive modification of data by one transaction. You can set a database property to cause an exception when this happens.

SET PROPERTY "sql.tx_no_multi_rewrite" TRUE
When an ALTER TABLE .. INSERT COLUMN or DROP COLUMN command results in changes to the table structure, the current session is committed. If an uncommitted transaction started by another connections has changed the data in the affected table, it may not be possible to roll it back after the ALTER TABLE command. This may also apply to ADD INDEX or ADD CONSTRAINT commands. It is recommended to use these ALTER commands only when it is known that other connections are not using transactions.

After a CHECKPOINT command is issued, uncommitted transactions can be continued, committed, or rolled back. However, if the database is not subsequently closed properly with the SHUTDOWN command, any such transaction that still remains uncommitted at the time of shutdown, is part committed (to the state at CHECKPOINT) at the next startup. It is recommended to use the CHECKPOINT command either when no uncommitted transactions is in progress, or it is known that any such transaction is not likely to last for such a long time that an abnormal shutdown might affect its data.

New Features and Changes

In recent versions leading to 1.8.0 many enhancements were made for better SQL support. These are listed in the SQL Syntax chapter, in ../changelog_1_8_0.txt and ../changelog_1_7_2.txt. Functions and expressions such as POSITION(), SUBSTRING(), NULLIF(), COALESCE(), CASE ... WHEN .. ELSE, ANY, ALL etc. are among them. Other enhancements may not be very obvious in the documentation but can result in changes of behaviour from previous versions. Most significant among these are handling of NULL values in joins (null columns are no longer joined) and OUTER joins (the results are now correct). You should test your applications with the new version to ensure they do not rely on past incorrect behaviour of the engine. The engine will continue to evolve in future versions towards full SQL standard support, so it is best not to rely on any non-standard feature of the current version.

Chapter 3. UNIX Quick Start

How to quickly get Hsqldb up and running on UNIX, including Mac OS X

Blaine Simpson

HSQLDB Development Group

$Date: 2005/07/25 23:20:53 $


This chapter explains how to quickly install, run, and use HSQLDB on UNIX.

HSQLDB has lots of great optional features. I intend to cover very few of them. I do intend to cover what I think is the most common UNIX setup: To run a multi-user database with permament data persistence. (By the latter I mean that data is stored to disk so that the data will persist across database shutdowns and startups). I also cover how to run HSQLDB as a system daemon.


Go to http://sourceforge.net/projects/hsqldb and click on the "files" link. You want the current version. This will be the highest numbered version under the plain black "hsqldb" heading. See if there's a distribution for the current HSQLDB version in the format that you want.

If you want an rpm, you should still find out the current version of HSQLDB as described in the previous paragraph. Then click "hsqldb" in the "free section" of http://www.jpackage.org/ and see if they have the current HSQLDB version built yet. Hopefully, the JPackage folk will document what JVM versions their rpm will support (currently they document this neither on their site nor within the package itself). (I really can't document how to download from a site that is totally beyond my control).


It could very well happen that some of the file formats which I discuss below are not in fact offered. If so, then we have not gotten around to building them.

Binary installation depends on the package format that you downloaded.

Installing from a .pkg.Z file

This package is only for use by a Solaris super-user. It's a System V package. Download then uncompress the package with uncompress or gunzip

    uncompress filename.pkg.Z
You can read about the package by running
    pkginfo -l -d filename.pkg
Run pkgadd as root to install.

    pkgadd -d filename.pkg
Installing from a .rpm file

This is a Linux rpm package. After you download the rpm, you can read about it by running

    rpm -qip /path/to/file.rpm

Rpms can be installed or upgraded by running

    rpm -Uvh /path/to/file.rpm
as root. Suse users may want to keep Yast aware of installed packages by running rpm through Yast: yast2 -i /path/to/file.rpm.

Installing from a .zip file

Extract the zip file to the parent directory of the new HSQLDB home. You don't need to create the HSQLDB_HOME directory because the extraction will create it for you with the right name)

    cd parent/of/new/hsqldb/home
    unzip /path/to/file.zip

All the files in the zip archive will be extracted to underneath a new hsqldb directory.

Take a look at the files you installed. (Under hsqldb for zip file installations. Otherwise, use the utilities for your packaging system). The most important file of the hsqldb system is hsqldb.jar, which resides in the directory lib.


For the purposes of this chapter, I define HSQLDB_HOME to be the parent directory of the lib directory that contains hsqldb.jar. E.g., if your path to hsqldb.jar is /a/b/hsqldb/lib/hsqldb.jar, then your HSQLDB_HOME is /a/b/hsqldb.

If the description of your distribution says that the hsqldb.jar file will work for your Java version, then you are finished with installation. Otherwise you need to build a new hsqldb.jar file.

If you followed the instructions above and you still don't know what Java version your hsqldb.jar supports, then read HSQLDB_HOME/readme.txt and HSQLDB_HOME/index.html. If that still doesn't help, then you can just try your hsqldb.jar and see if it works, or build your own.

To use the supplied hsqldb.jar, just skip to the next section of this document. Otherwise build a new hsqldb.jar.

Procedure 3.1. Building hsqldb.jar

  1. If you don't already have Ant, download the latest stable binary version from http://ant.apache.org. cd to where you want Ant to live, and extract from the archive with

        unzip /path/to/file.zip
        tar -xzf /path/to/file.tar.gz
        bunzip2 -c /path/to/file.tar.bz2 | tar -xzf -
    Everything will be installed into a new subdirectory named apache-ant- + version. You can rename the directory after the extraction if you wish.

  2. Set the environmental variable JAVA_HOME to the base directory of your Java JRE or SDK, like

        export JAVA_HOME; JAVA_HOME=/usr/java/j2sdk1.4.0
    The location is entirely dependent upon your variety of UNIX. Sun's rpm distributions of Java normally install to /usr/java/something. Sun's System V package distributions of Java (including those that come with Solaris) normally install to /usr/something, with a sym-link from /usr/java to the default version (so for Solaris you will usually set JAVA_HOME to /usr/java).

  3. Remove the existing file HSQLDB_HOME/lib/hsqldb.jar.

  4. cd to HSQLDB_HOME/build. Make sure that the bin directory under your Ant home is in your search path. Run the following command.

        ant hsqldb
    This will build a new HSQLDB_HOME/lib/hsqldb.jar.

See the Building HSQLDB appendix if you want to build anything other than hsqldb.jar with all default settings.

Setting up a Hsqldb Persistent Database Instance and a Hsqldb Server

If you installed from an OS-specific package, you may already have a database instance and server pre-configured. See if your package includes a file named server.properties (make use of your packaging utilities). If you do, then I suggest that you still read this section while you poke around, in order to understand your setup.

  1. Select a UNIX user to run the database as. If this database is for the use of multiple users, or is a production system (or to emulate a production system), you should dedicate a UNIX user for this purpose. In my examples, I use the user name hsqldb. In this chapter, I refer to this user as the HSQLDB_OWNER, since that user will own the database instance files and processes.

    If the account doesn't exist, then create it. On all system-5 UNIXes and most hybrids (including Linux), you can run (as root) something like

        useradd -c 'HSQLDB Database Owner' -s /bin/bash -m hsqldb
    (BSD-variant users can use a similar pw useradd hsqldb... command).

  2. Become the HSQLDB_OWNER. Copy the sample file HSQLDB_HOME/src/org/hsqldb/sample/sample-server.properties to the HSQLDB_OWNER's home directory and rename it to server.properties.

    # Hsqldb Server cfg file.
    # See the Advanced Topics chapter of the Hsqldb User Guide.
    server.database.0   file:db0/db0
    # I suggest that, for every file: database you define, you add the
    # connection property "ifexists=true" after the database instance
    # is created (which happens simply by starting the Server one time).
    # Just append ";ifexists=true" to the file: URL, like so:
    # server.database.0   file:db0/db0;ifexists=true

    Since the value of the first database (server.database.0) begins with file:, the database instance will be persisted to a set of files in the specified directory with names beginning with the specified name. Set the path to whatever you want (relative paths will be relative to the directory containing the properties file). You can read about how to specify other database instances of various types, and how to make settings for the listen port and many other things, in the Advanced Topics chapter.

  3. Set and export the environmental variable CLASSPATH to the value of HSQLDB_HOME (as described above) plus "/lib/hsqldb.jar", like

        export CLASSPATH; CLASSPATH=/path/to/hsqldb/lib/hsqldb.jar
    In HSQLDB_OWNER's home directory, run

        nohup java org.hsqldb.Server &

    This will start the Server process in the background, and will create your new database instance "db0". Continue on when you see the message containing HSQLDB server... is online. nohup just makes sure that the command will not quit when you exit the current shell (omit it if that's what you want to do).

Accessing your Database

Copy the file HSQLDB_HOME/src/org/hsqldb/sample/sqltool.rc to the HSQLDB_OWNER's home directory. Use chmod to make the file readable and writable only to HSQLDB_OWNER.

# $Id: sqltool.rc,v 1.22 2007/08/09 03:22:21 unsaved Exp $

# This is a sample RC configuration file used by SqlTool, DatabaseManager,
# and any other program that uses the org.hsqldb.util.RCData class.

# You can run SqlTool right now by copying this file to your home directory
# and running
#    java -jar /path/to/hsqldb.jar mem
# This will access the first urlid definition below in order to use a 
# personal Memory-Only database.
# "url" values may, of course, contain JDBC connection properties, delimited
# with semicolons.

# If you have the least concerns about security, then secure access to
# your RC file.
# See the documentation for SqlTool for various ways to use this file.

# A personal Memory-Only (non-persistent) database.
urlid mem
url jdbc:hsqldb:mem:memdbid
username sa

# A personal, local, persistent database.
urlid personal
url jdbc:hsqldb:file:${user.home}/db/personal;shutdown=true
username sa
# When connecting directly to a file database like this, you should 
# use the shutdown connection property like this to shut down the DB
# properly when you exit the JVM.

# This is for a hsqldb Server running with default settings on your local
# computer (and for which you have not changed the password for "sa").
urlid localhost-sa
url jdbc:hsqldb:hsql://localhost
username sa

# Template for a urlid for an Oracle database.
# You will need to put the oracle.jdbc.OracleDriver class into your 
# classpath.
# In the great majority of cases, you want to use the file classes12.zip
# (which you can get from the directory $ORACLE_HOME/jdbc/lib of any
# Oracle installation compatible with your server).
# Since you need to add to the classpath, you can't invoke SqlTool with
# the jar switch, like "java -jar .../hsqldb.jar..." or 
# "java -jar .../hsqlsqltool.jar...".
# Put both the HSQLDB jar and classes12.zip in your classpath (and export!)
# and run something like "java org.hsqldb.util.SqlTool...".

#urlid cardiff2
#url jdbc:oracle:thin:@aegir.admc.com:1522:TRAFFIC_SID
#username blaine
#password secretpassword
#driver oracle.jdbc.OracleDriver

# Template for a TLS-encrypted HSQLDB Server.
# Remember that the hostname in hsqls (and https) JDBC URLs must match the
# CN of the server certificate (the port and instance alias that follows 
# are not part of the certificate at all).
# You only need to set "truststore" if the server cert is not approved by
# your system default truststore (which a commercial certificate probably
# would be).

#urlid tls
#url jdbc:hsqldb:hsqls://db.admc.com:9001/lm2
#username blaine
#password asecret
#truststore /home/blaine/ca/db/db-trust.store

# Template for a Postgresql database
#urlid blainedb
#url jdbc:postgresql://idun.africawork.org/blainedb
#username blaine
#password losung1
#driver org.postgresql.Driver

# Template for a MySQL database.  MySQL has poor JDBC support.
#urlid mysql-testdb
#url jdbc:mysql://hostname:3306/dbname
#username root
#username blaine
#password hiddenpwd
#driver com.mysql.jdbc.Driver

# Note that "databases" in SQL Server and Sybase are traditionally used for
# the same purpose as "schemas" with more SQL-compliant databases.

# Template for a Microsoft SQL Server database
#urlid msprojsvr
#url jdbc:microsoft:sqlserver://hostname;DatabaseName=DbName;SelectMethod=Cursor
# The SelectMethod setting is required to do more than one thing on a JDBC
# session (I guess Microsoft thought nobody would really use Java for 
# anything other than a "hello world" program).
# This is for Microsoft's SQL Server 2000 driver (requires mssqlserver.jar
# and msutil.jar).
#driver com.microsoft.jdbc.sqlserver.SQLServerDriver
#username myuser
#password hiddenpwd

# Template for a Sybase database
#urlid sybase
#url jdbc:sybase:Tds:hostname:4100/dbname
#username blaine
#password hiddenpwd
# This is for the jConnect driver (requires jconn3.jar).
#driver com.sybase.jdbc3.jdbc.SybDriver

# Template for Embedded Derby / Java DB.
#urlid derby1
#url jdbc:derby:path/to/derby/directory;create=true
#username ${user.name}
#password any_noauthbydefault
#driver org.apache.derby.jdbc.EmbeddedDriver
# The embedded Derby driver requires derby.jar.
# There'a also the org.apache.derby.jdbc.ClientDriver driver with URL
# like jdbc:derby://<server>[:<port>]/databaseName, which requires
# derbyclient.jar.
# You can use \= to commit, since the Derby team decided (why???)
# not to implement the SQL standard statement "commit"!!
# Note that SqlTool can not shut down an embedded Derby database properly,
# since that requires an additional SQL connection just for that purpose.
# However, I've never lost data by not shutting it down properly.
# Other than not supporting this quirk of Derby, SqlTool is miles ahead of ij.

We will be using the "localhost-sa" sample urlid definition from the config file. The JDBC URL for this urlid is jdbc:hsqldb:hsql://localhost. That is the URL for the default database instance of a HSQLDB Server running on the default port of the local host. You can read about URLs to connect to other instances and other servers in the Advanced Topics chapter.

Run SqlTool.

    java -jar path/to/hsqldb.jar localhost-sa
If you get a prompt, then all is well. If security is of any concern to you at all, then you should change the privileged password in the database. Use the command SET PASSWORD command to change SA's password.
    set password "newpassword";

When you're finished playing, exit with the command \q.

If you changed the SA password, then you need to fix the password in the sqltool.rc file accordingly.

You can, of course, also access the database with any JDBC client program. See the First JDBC Client Example appendix. You will need to modify your classpath to include hsqldb.jar as well as your client class(es). You can also use the other HSQLDB client programs, such as org.hsqldb.util.DatabasManagerSwing, a graphical client with a similar purpose to SqlTool.

You can use any normal UNIX account to run the JDBC clients, including SqlTool, as long as the account has read access to the hsqldb.jar file and to an sqltool.rc file. See the SqlTool chapter about where to put sqltool.rc, how to execute sql files, and other SqlTool features.

Create additional Accounts

Connect to the database as SA (or any other Administrative user) and run CREATE USER to create new accounts for your database instance. HSQLDB accounts are database-instance-specific, not Server-specific.

For the current version of HSQLDB, only users with Role of DBA may create or own database objects. DBA members have privileges to do anything. Non-DBAs may be granted some privileges, but may never create or own database objects. (Before long, non-DBAs will be able to create objects if they have permission to do so in the target schema). When you first create a hsqldb database, it has only one database user-- SA, a DBA account, with an empty string password. You should set a password (as described above). You can create as many additional users as you wish. To make a user a DBA, you can use the "ADMIN" option to the CREATE USER command, or GRANT the DBA Role to the account after creating it.

If you create a user without the ADMIN tag (and without granting the DBA role to them) this user will be able to read the data dictionary tables, but will be able unable to create or own his own objects. He will have only the rights which the pseudo-user PUBLIC has. To give him more permissions, even rights to read objects, you can GRANT permissions for specific objects, grant Roles (which encompass a set of permissions), or grant the DBA Role itself.

Since only people with a database account may do anything at all with the database, it is often useful to permit other database users to view the data in your tables. To optimize performance, reduce contention, and minimize administration, it is often best to grant SELECT to PUBLIC on any object that needs to be accessed by multiple database users (with the significant exception of any data which you want to keep secret).


Do a clean database shutdown when you are finished with the database instance. You need to connect up as SA or some other Admin user, of course. With SqlTool, you can run

    java -jar path/to/hsqldb.jar --sql shutdown localhost-sa
You don't have to worry about stopping the Server because it shuts down automatically when all served database instances are shut down.

Running Hsqldb as a System Daemon

You can, of course, run HSQLDB through inittab on System V UNIXes, but usually an init script is more convenient and manageable. This section explains how to set up and use our UNIX init script. Our init script is only for use by root. (That is not to say that the Server will run as root-- it usually should not).

The main purpose of the init script is to start up a Server with the database instances specified in your server.properties file; and to shut down all of those instances plus additional urlids which you may (optionally) list in your init script config file. These urlids must all have entries in a sqltool.rc file. If, due to firewall issues, you want to run a WebServer instead of a Server, then make sure you have a healthy WebServer with a webserver.properties set up, adjust your URLs in sqltool.rc, and set TARGET_CLASS in the config file. (By following the commented examples in the config file, you can start up any number of Server and/or WebServer listeners with or without TLS ecryption).

After you have the init script set up, root can use it anytime to start or stop HSQLDB. (I.e., not just at system bootup or shutdown).

Portability of hsqldb init script

The primary design criterion of the init script is portability. It does not print pretty color startup/shutdown messages as is common in late-model Linuxes and HPUX; and it does not keep subsystem state files or use the startup/shutdown functions supplied by many UNIXes, because these features are all non-portable.

Offsetting these limitations, this one script does it's intended job great on the UNIX varieties I have tested, and can easily be modified to accommodate other UNIXes. While you don't have tight integration with OS-specific daemon administration guis, etc., you do have a well tested and well behaved script that gives good, utilitarian feedback.

Init script Setup Procedure

The strategy taken here is to get the init script to run your single Server or WebServer first (as specified by TARGET_CLASS). After that's working, you can customize the JVM that is run by running additional Servers in it, running your own application in it (embedding), or even overriding HSQLDB behavior with your own overriding classes.

  1. Copy the init script hsqldb from HSQLDB_HOME/bin into the directory where init scripts live on your variety of UNIX. The most common locations are /etc/init.d or /etc/rc.d/init.d on System V style UNIXes, /usr/local/etc/rc.d on BSD style UNIXes, and /Library/StartupItems/hsqldb on OS X (you'll need to create the directory for the last).

  2. Look at the comment towards the top of the init script which lists recommended locations for the configuration file for various UNIX platforms. Copy the sample config file HSQLDB_HOME/src/org/hsqldb/sample/sample-hsqldb.cfg to one of the listed locations (your choice). Edit the config file according to the instructions in it.

    # $Id: sample-hsqldb.cfg,v 1.16 2005/07/24 18:33:13 unsaved Exp $
    # Sample configuration file for HSQLDB database server.
    # See the "UNIX Quick Start" chapter of the Hsqldb User Guide.
    # N.b.!!!!  You must place this in the right location for your type of UNIX.
    # See the init script "hsqldb" to see where this must be placed and
    # what it should be renamed to.
    # This file is "sourced" by a Bourne shell, so use Bourne shell syntax.
    # This file WILL NOT WORK until you set (at least) the non-commented
    # variables to the appropriate values for your system.
    # Life will be easier if you avoid all filepaths with spaces or any other
    # funny characters.  Don't ask for support if you ignore this advice.
    # Thanks to Meikel Bisping for his contributions.  -- Blaine
    # Unless you copied a hsqldb.jar file from another system, this typically
    # resides at $HSQLDB_HOME/lib/hsqldb.jar, where $HSQLDB_HOME is your HSQLDB
    # software base directory.
    # Where the file "server.properties" resides.
    # What UNIX user the server will run as.
    # (The shutdown client is always run as root or the invoker of the init script).
    # Runs as root by default, but you should take the time to set database file
    # ownerships to another user and set that user name here.
    # The HSQLDB jar file specified in HSQLDB_JAR_PATH above will automatically
    # be in the class path.  This arg specifies additional classpath elements.
    # To embed your own application, add your jar file(s) or class base
    # directories here, and add your main class to the INVOC_ADDL_ARGS setting
    # below.
    # We require all Server/WebServer instances to be accessible within 
    # $MAX_START_SECS from when the Server/WebServer is started.
    # Defaults to 60.
    # Raise this is you are running lots of DB instances or have a slow server.
    # Time to allow for JVM to die after all HSQLDB instances stopped.
    # Defaults to 1.
    # These are "urlid" values from a SqlTool authentication file
    # ** IN ADDITION TO THOSE IN YOUR server.properties OR webserver.properties **
    # file.  All server.urlid.X values from your properties file will automatically
    # be started/stopped/tested.  $SHUTDOWN_URLIDS is for additional urlids which
    # will stopped.  (Therefore, most users will not set this at all).
    # Separate multiple values with white space.  NO OTHER SPECIAL CHARACTERS!
    # Make sure to quote the entire value if it contains white space separator(s).
    # Defaults to none (i.e., only urlids set in properties file will be stopped).
    #SHUTDOWN_URLIDS='sa mygms'
    # SqlTool authentication file used only for shutdown.
    # The default value will be sqltool.rc in root's home directory, since it is 
    # root who runs the init script.
    # (See the SqlTool chapter of the HSQLDB User Guide if you don't understand 
    # this).
    # Set this to either 'WebServer' or 'Server'.  Defaults to Server.
    # The JVM that is started can invoke many classes (see the following item
    # about that), but this is the Server that is used (1) to check status,
    # (2) to shut down the JVM, (3) to get urlids for #1 from the 
    # server's server/webserver.properties file.
    # Note that you don't specify the org.hsqldb package, since you have no 
    # choice in the matter (you can only run org.hsqldb.Server or 
    # org.hsqldb.WebServer).  If you specify additional classes with
    # INVOC_ADDL_ARGS (described next), you do need to specify the
    # full class name with package name.
    # This is where you specify exactly what your HSQLDB JVM will run.
    # The class org.hsqldb.util.MainInvoker will run the TARGET_CLASS
    # specified above with any arguments supplied here + any other classes
    # and arguments.  Every additional class (in addition to the TARGET_CLASS)
    # must be preceded with an empty string, so that MainInvoker will know
    # you are giving a class name.  MainInvoker will invoke the normal 
    # static main(String[]) method of each such class.  
    # By default, MainInvoker will just run TARGET_CLASS with no args.
    # Example that runs just the TARGET_CLASS with the specified arguments:
    #INVOC_ADDL_ARGS='-silent false'
    # Example that runs the TARGET_CLASS plus a WebServer:
    #INVOC_ADDL_ARGS='"" org.hsqldb.WebServer'
    # Note the empty string preceding the class name.
    # Example that starts TARGET_CLASS with an argument + a WebServer +
    # your own application with its args (i.e., the HSQLDB Servers are
    # "embedded" in your application).  (Set SERVER_ADDL_CLASSPATH too).:
    #INVOC_ADDL_ARGS='-silent false "" org.hsqldb.WebServer "" com.acme.Stone --env prod localhost'
    # Example to run a non-TLS server in same JVM with a TLS server.  In this
    # case, TARGET_CLASS is Server which will run in TLS mode by virtue of 
    # setting TLS_KEYSTORE and TLS_PASSWORD above.  The "additional" Server
    # here overrides the 'tls' and 'port' settings:
    #INVOC_ADDL_ARGS="'' org.hsqldb.Server -port 9002 -tls false"
    # Note that you use nested quotes to group arguments and to specify the
    # empty-string delimiter.
    # For TLS encryption for your Server, set these two variables.
    # N.b.:  If you set these, then make this file unreadable to non-root users!!!!
    # See the TLS chapter of the HSQLDB User Guide, paying attention to the 
    # security warning(s).
    # If you are running with a private server cert, then you will also need to 
    # set "truststore" in the your SqlTool config file (location is set by the
    # AUTH_FILE variable in this file, or it must be at the default location for 
    # Any JVM args for the invocation of the JDBC client used to verify DB
    # instances and to shut them down (SqlToolSprayer).
    # This example specifies the location of a private trust store for TLS 
    # encryption.
    # For multiple args, put quotes around entire value.
    # Any JVM args for the server.
    # For multiple args, put quotes around entire value.
  3. Either copy HSQLDB_OWNER's sqltool.rc file into root's home directory, or set the value of AUTH_FILE to the absolute path of HSQLDB_OWNER's sqltool.rc file. This file is read (for stops) directly by root, even if you run hsqldb as non-root (by setting HSQLDB_OWNER in the config file). If you copy the file, make sure to use chmod to restrict permissions on the new copy. (The init script now enforces permissions on this file).

  4. Edit your server.properties file. For every server.database.X that you have defined, set a property of name server.urlid.X to the urlid for an Administrative user for that database instance.

    Example 3.1. server.properties fragment



    Make sure to add a urlid for each and every database instance. If you don't then the init script will never know about databases that become inaccessible and will give false diagnostics.

    For this example, you would need to define the urlid localhostdb1 in your sqltool.rc file.

    Example 3.2. example sqltool.rc stanza

        urlid localhostdb1
        url jdbc:hsqldb:hsql://localhost
        username sa
        password secret
  5. Verify that the init script works.

    Just run

    as root to see the arguments you may use. Notice that you can run

        /path/to/hsqldb status

    at any time to see whether your HSQLDB Server is running.

    Re-run the script with each of the possible arguments to really test it good. If anything doesn't work right, then see the Troubleshooting the Init Script section.

  6. Tell your OS to run the init script upon system startup and shutdown. If you are using a UNIX variant that has /etc/rc.conf or /etc/rc.conf.local (like BSD variants and Gentoo), you must set "hsqldb_enable" to "YES" in either of those files. (Just run cd /etc; ls rc.conf rc.conf.local to see if you have one of these files). For good UNIXes that use System V style init, you must set up hard links or soft links either manually or with management tools (such as chkconfig or insserv) or Gui's (like run level editors).

    This paragraph is for Mac OS X users only. If you followed the instructions above, your init script should reside at /Library/StartupItems/hsqldb/hsqldb. Now copy the file StartupParameters.plist from the directory src/org.hsqldb/sample of your HSQLDB distribution to the same directory as the init script. As long as these two files reside in /Library/StartupItems/hsqldb, your init script is active (for portability reasons, it doesn't check for a setting in /etc/hostconfig). You can run it as a Startup Item by running

        SystemStarter {start|stop|restart} Hsqldb
    Hsqldb is the service name. See the man page for SystemStarter. To disable the init script, wipe out the /Library/StartupItems/hsqldb directory. Hard to believe, but the Mac people tell me that during system shutdown the Startup Items don't run at all. Therefore, if you don't want your data corrupted, make sure to run "SystemStarter stop Hsqldb" before shutting down your Mac.

Follow the examples in the config file to add additional classes to the server JVM's classpath and to execute additional classes in your JVM. (See the SERVER_ADDL_CLASSPATH and INVOC_ADDL_ARGS items).

Troubleshooting the Init Script

Do a ps to look for processes containing the string hsqldb, and try to connect to the database from any client. If the init script starts up your database successfully, but incorrectly reports that it has not, then your problem is with specification of urlid(s) or SqlTool setup. If your database really did not start, then skip to the next paragraph. Verify that the urlid(s) listed in the server.properties or webserver.properties are correct. and verify that you can run SqlTool as root to connect to the instances. (For the latter test, use the --rcfile switch if you are setting AUTH_FILE in the init script config file).

If your database really is not starting, then verify that you can su to the database owner account and start the database. The command su USERNAME -c ... won't work on most UNIXes unless the target user has a real login shell. Therefore, if you try to tighten up security by disabling this user's login shell, you will break the init script. If these possibilities don't pan out, then debug the init script or seek help, as described below.

To debug the init script, run it in verbose mode to see exactly what is happening (and perhaps manually run the steps that are suspect). To run an init script (in fact, any sh shell script) in verbose mode, use sh with the -x or -v switch, like

    sh -x path/to/hsqldb start
See the man page for sh if you don't know the difference between -v and -x.

If you want troubleshooting help, use the HSQLDB lists/forums or email me at blaine.simpson@admc.com. If you email me, make sure to include the revision number from your hsqldb init script (it's towards the top in the line that starts like "# $Id:"), and the output of a run of

    sh -x path/to/hsqldb start > /tmp/hstart.log 2>&1

Chapter 4. Advanced Topics

Fred Toussi

HSQLDB Development Group

Copyright 2002-2005 Fred Toussi. Permission is granted to distribute this document without any alteration under the terms of the HSQLDB license. Additional permission is granted to the HSQLDB Development Group to distribute this document with or without alterations under the terms of the HSQLDB license.

$Date: 2007/03/24 11:39:08 $


Many questions repeatedly asked in Forums and mailing lists are answered in this guide. If you want to use HSQLDB with your application, you should read this guide. This document covers system related issues. For issues related to SQL see the SQL Issues chapter.


The normal method of accessing an HSQLDB database is via the JDBC Connection interface. An introduction to different methods of providing database services and accessing them can be found in the SQL Issues chapter. Details and examples of how to connect via JDBC are provided in our JavaDoc for jdbcConnection.

Version 1.7.2 introduced a uniform method of distinguishing between different types of connection, alongside new capabilities to provide access to multiple databases. The common driver identifier is jdbc:hsqldb: followed by a protocol identifier (mem: file: res: hsql: http: hsqls: https:) then followed by host and port identifiers in the case of servers, then followed by database identifier.

Table 4.1. Hsqldb URL Components

Driver and ProtocolHost and PortDatabase
not available

Lowercase, single-word identifier creates the in-memory database when the first connection is made. Subsequent use of the same Connection URL connects to the existing DB.

The old form for the URL, jdbc:hsqldb:. creates or connects to the same database as the new form for the URL, jdbc:hsqldb:mem:.

not available

The file path specifies the database file. In the above examples the first one refers to a set of mydb.* files in the directory where the javacommand for running the application was issued. The second and third examples refer to absolute paths on the host machine.

not available
Database files can be loaded from one of the jars specified as part of the Java command the same way as resource files are accessed in Java programs. The /adirectory above stands for a directory in one of the jars.

The host and port specify the IP address or host name of the server and an optional port number. The database to connect to is specified by an alias. This alias is a lowercase string defined in the server.properties file to refer to an actual database on the file system of the server or a transient, in-memory database on the server. The following example lines in server.properties or webserver.properties define the database aliases listed above and accessible to clients to refer to different file and in-memory databases.




The old form for the server URL, e.g., jdbc:hsqldb:hsql//localhost connects to the same database as the new form for the URL, jdbc:hsqldb:hsql//localhost/ where the alias is a zero length string. In the example below, the database files lists.* in the /home/dbmaster/ directory are associated with the empty alias:


Connection properties

Each new JDBC Connection to a database can specify connection properties. The properties user and password are always required. In 1.8.0 the following optional properties can also be used.

Connection properties are specified either by establishing the connection via the:

    DriverManager.getConnection (String url, Properties info);

method call, or the property can be appended to the full Connection URL.

Table 4.2. Connection Properties

get_column_nametruecolumn name in ResultSet

This property is used for compatibility with other JDBC driver implementations. When true (the default), ResultSet.getColumnName(int c) returns the underlying column name

When false, the above method returns the same value as ResultSet.getColumnLabel(int column) Example below:


When a ResultSet is used inside a user-defined stored procedure, the default, true, is always used for this property.

ifexistsfalseconnect only if database already exists

Has an effect only with mem: and file: database. When true, will not create a new database if one does not already exist for the URL.

When false (the default), a new mem: or file: database will be created if it does not exist.

Setting the property to true is useful when troubleshooting as no database is created if the URL is malformed. Example below:

shutdownfalseshut down the database when the last connection is closed

This mimics the behaviour of 1.7.1 and older versions. When the last connection to a database is closed, the database is automatically shut down. The property takes effect only when the first connection is made to the database. This means the connection that opens the database. It has no effect if used with subsequent, simultaneous connections.

This command has two uses. One is for test suites, where connections to the database are made from one JVM context, immediately followed by another context. The other use is for applications where it is not easy to configure the environment to shutdown the database. Examples reported by users include web application servers, where the closing of the last connection conisides with the web app being shut down.

In addition, when a connection to an in-process database creates a new database, or opens an existing database (i.e. it is the first connection made to the database by the application), all the user-defined database properties can be specified as URL properties. This can be used to specify properties to enforce more strict SQL adherence, or to change cache_scale or similar properties before the database files are created. However, for new databases, it is recommended to use the SET PROPERTY command for such settings.

Properties Files

HSQLDB relies on a set of properties files for different settings. Since 1.7.0 property naming has been streamlined and a number of new properties have been introduced.

In all properties files, values are case-sensitive. All values apart from names of files or pages are required in lowercase (e.g. server.silent=FALSE will have no effect, but server.silent=false will work).

The properties files and the settings stored in them are as follows:

Table 4.3. Hsqldb Server Properties Files

File NameLocationFunction
server.propertiesthe directory where the command to run the Server class is issuedsettings for running HSQLDB as a database server communicating with the HSQL protocol
webserver.propertiesthe directory where the command to run the WebServer class is issuedsettings for running HSQLDB as a database server communicating with the HTTP protocol
<dbname>.propertiesthe directory where all the files for a database are locatedsettings for each particular database

Properties files for running the servers are not created automatically. You should create your own files that contain server.property=value pairs for each property.

The properties file for each database is generated by the database engine. This file can be edited after closing the database. In 1.8.0, most of these properties can be changed via SQL commands.

Server and Web Server Properties

In both server.properties and webserver.properties files, supported values and their defaults are as follows:

Table 4.4. Property File Properties

server.database.0testthe path and file name of the first database file to use
server.dbname.0""lowercase server alias for the first database file
server.urlid.0NONESqlTool urlid used by UNIX init script. (This property is not used if your are running Server/Webserver on a platform other than UNIX, or of you are not using our UNIX init script).
server.silenttrueno extensive messages displayed on console
server.tracefalseJDBC trace messages displayed on console

In 1.8.0, each server can serve up to 10 different databases simultaneously. The server.database.0 property defines the filename / path whereas the server.dbname.0 defines the lowercase alias used by clients to connect to that database. The digit 0 is incremented for the second database and so on. Values for the server.database.{0-9} property can use the mem:, file: or res: prefixes and properties as discussed above under CONNECTIONS. For example,


Values specific to server.properties are:

Table 4.5. Server Property File Properties

server.port9001 (normal) or 554 (if TLS encrypted)TCP/IP port used for talking to clients. All databases are served on the same port.
server.no_system_exittrueno System.exit() call when the database is closed

Values specific to webserver.properties are:

Table 4.6. WebServer Property File Properties

server.port80TCP/IP port used for talking to clients
server.default_pageindex.htmlthe default web page for server
server.root./the location of served pages
.<extension>?multiple entries such as .html=text/html define the mime types of the static files served by the web server. See the source for WebServer.java for a list.

All the above values can be specified on the command line to start the server by omitting the server. prefix.

Starting a Server from your application

If you want to start the server from within your application, as opposed to the command line or batch files, you should create an instance of Server or Web Server, then assign the properties in the form of a String and start the Server. An example of this can be found in the org.hsqldb.test.TestBase source.


Upgrading: If you have existing custom properties files, change the values to the new naming convention. Note the use of digits at the end of server.database.n and server.dbname.n properties.

Individual Database Properties

Each database has its own <dbname>.properties file as part of a small group of files which also includes <dbname>.script and <dbname>.data. The properties files contain key/value pairs for some important settings.

In version 1.8.0 a new SQL command allows most database properties to be modified as follows:

    SET PROPERTY "property_name" property_value

Properties that can be modified via SET PROPERTY are indicated in the table below. Other properties are indicated as PROPERTIES FILE ONLY and can be modified only by editing the .properties file after a shutdown and before a restart. Only the user-defined values listed below should ever be modified. Changing any other value could result in unexpected malfunction in database operations. Most of these values have been introduced for the new features since 1.7.0:

Table 4.7. Database-specific Property File Properties

readonlyfalsewhole database is read-only

When true, the database cannot be modified in use. This setting can be changed to true if the database is to be opened from a CD. Prior to changing this setting, the database should be closed with the SHUTDOWN COMPACT command to ensure consistency and compactness of the data. (PROPERTIES FILE ONLY) but can be used as a connection property to open a normal database as readonly.

hsqldb.files_readonlyfalsedatabase files will not be written to

When true, data in MEMORY tables can be modified and new MEMORY tables can be added. However, these changes are not saved when the database is shutdown. CACHED and TEXT tables are always readonly when this setting is true. (PROPERTIES FILE ONLY)

hsqldb.cache_file_scale1Set larger data file limits. Once set, the limit will go up to 8GB.

This property can be set to 8 to increase the size limit of the .data file from 2GB to 8GB. To apply the change to an existing database, SHUTDOWN SCRIPT should be performed first, then the property=value line below should be added to the .properties file before reopening the database.


The property can be set with the SQL command (as opposed to changing the value in the properties file) when the database has no CACHED tables (e.g. a new database). (SET PROPERTY)

sql.enforce_sizefalsetrimming and padding string columns

This property is no longer supported. Use sql.enforce_sctrict_size

sql.enforce_strict_sizefalsesize enforcement and padding string columns

Conforms to SQL standards for size and precision of data types. When true, all CHARACTER, VARCHAR, NUMERIC and DECIMAL values that are in a row affected by an INSERT INTO or UPDATE statement are checked against the size specified in the SQL table definition. An exception is thrown if the value is too long. Also all CHARACTER values that are shorter than the specified size are padded with spaces. TIMESTAMP(0) and TIMESTAMP(6) are also allowed in order to specify the subsecond resolution of the values. When false (default), stores the exact string that is inserted. (SET PROPERTY)

sql.tx_no_multi_rewritefalsetransaction management

In the default READ_UNCOMMITED mode, a transaction can write over rows inserted or updated by another uncommitted transaction. Setting this property to true will raise an exception when such a write is attempted (SET PROPERTY)

hsqldb.cache_scale14memory cache exponent

Indicates the maximum number of rows of cached tables that are held in memory, calculated as 3 *(2**value) (three multiplied by (two to the power value)). The default results in up to 3*16384 rows from all cached tables being held in memory at any time.

The value can range between 8-18. (SET PROPERTY). If the value is set via SET PROPERTY then it becomes effective after the next database SHUTDOWN or CHECKPOINT. (SET PROPERTY)

hsqldb.cache_size_scale10memory cache exponent

Indicates the average size of each row in the memory cache used with cached tables, calculated as 2**value (two to the power value). This result value is multiplied by the maximum number of rows defined by hsqldb.cache_scale to form the maximum number of bytes for all the rows in memory cache. The default results in 1024 bytes per row. This default, combined with the default number of rows, results in approximately 50MB of the .data file to be stored in the memory cache.

The value can range between 6-20. (SET PROPERTY). If the value is set via SET PROPERTY then it becomes effective after the next database SHUTDOWN or CHECKPOINT. (SET PROPERTY)

hsqldb.log_size200size of log when checkpoint is performed

The value is the size in megabytes that the .log file can reach before an automatic checkpoint occurs. A checkpoint and rewrites the .script file and clears the .log file. The value can be changed via the SET LOGSIZE nnn SQL command.

runtime.gc_interval0forced garbage collection

This setting forces garbage collection each time a set number of result set row or cache row objects are created. The default, "0" means no garbage collection is forced by the program.

This should not be set when the database engine is acting as a server inside an exclusive JVM. The setting can be useful when the database is used in-process with the application with some Java Runtime Environments (JRE's). Some JRE's increase the size of the memory heap before doing any automatic garbage collection. This setting would prevent any unnecessary enlargement of the heap. Typical values for this setting would probably be between 10,000 to 100,000. (PROPERTIES FILE ONLY)

hsqldb.nio_data_filetrueuse of nio access methods for the .data file

When HSQLDB is compiled and run in Java 1.4 or higher, setting this property to false will avoid the use of nio access methods, resulting in somewhat reduced speed. If the data file is larger than 256MB when it is first opened, nio access methods are not used. Also, if the file gets larger than the amount of available computer memory that needs to be allocated for nio access, non-nio access methods are used.

(SET PROPERTY). If used before defining any CACHED table, it applies to the current session, otherwise it comes to effect after a SHUTDOWN and restart or CHECKPOINT.

hsqldb.default_table_typememorytype of table created with unqualified CREATE TABLE

The CREATE TABLE command results in a MEMORY table by default. Setting the value "cached" for this property will result in a cached table by default. The qualified forms such as CREATE MEMORY TABLE or CREATE CACHED TABLE are not affected at all by this property. (SET PROPERTY)

hsqldb.applog0application logging level

The default level 0 indicates no logging. Level 1 results in events related to persistence to be logged, including any failures. The events are logged in a file ending with .app.log

textdb.*0default properties for new text tables

Properties that override the database engine defaults for newly created text tables. Settings in the text table SET <tablename> SOURCE <source string> command override both the engine defaults and the database properties defaults. Individual textdb.* properties are listed in the Text Tables chapter. (SET PROPERTY)

When connecting to an in-process database creates a new database, or opens an existing database (i.e. it is the first connection made to the database by the application), all the user-defined database properties listed in this section can be specified as URL properties.


Upgrading: From 1.7.0, the location of the database files can no longer be overridden by paths defined in the properties file. All files belonging to a database should reside in the same directory.

The property sql.compare_in_locale=true is no longer supported. If the line exists in a .properties file, it will switch the database to the collation for the current default. See the SET DATABASE COLLATION[2] command.

When HSQLDB is used in OpenOffice.org, some property values will have a different default. The properties and values are:

hsqldb.default_table_type=cached hsqldb.cache_scale=13 hsqldb.log_size=10; hsqldb.nio_data_file=false sql.enforce_strict_size=true

SQL Commands for Database Properties

There are some database properties that are set with dedicated SQL commands beginning with SET.

Table 4.8. SQL command properties

SET WRITE_DELAY {{TRUE | FALSE} | <seconds> | <milliseconds> MILLIS

The default is TRUE and indicates that the changes to the database that have been logged are synched to the file system once every 20 seconds. FALSE indicates there is no delay and at each commit a file synch operation is performed. Numeric values from 0 can also be specified for the synch delay.

The purpose of this command is to control the amount of data loss in case of a total system crash. A delay of 1 second means at most the data written to disk during the last second before the crash is lost. All data written prior to this has been synced and should be recoverable

This setting should be specified on the basis of the reliability of the hardware used for running the database engine, the type of disk system used, the possibility of power failure etc. Also the nature of the data stored should be considered.

In general, when the system is very reliable, the setting can be left to the default. If it is not very reliable, or the data is critical a setting of 1 or 2 seconds would suffice. Only in the worst case scenario or with the most critical data should a setting of 0 or FALSE be specified as this will slow the engine down to the speed at which the file synch operation can be performed by the disk subsystem.

Values down to 10 millisconds can be specified by adding MILLIS to the command, but in practice a delay of 100 milliseconds provides 99.99999% reliability with an average one system crash per 6 days.

SET LOG_SIZE <numeric value>

The engine writes out a log of all the changes to the database as they occur. This log is synched to the disk based on the WRITE_DELAY property above. The log is never reused unless there is an abnormal termination, i.e. the database process is terminated without SHUTDOWN, or it was terminated using SHUTDOWN IMMEDIATELY.

The default maximum size of the .log file is 200 MB. When the maximum size is reached, a CHECKPOINT operation is performed. This operation will save the other database files in a consistent state and delete the old log. A value of 0 indicates no limit for the .log file.


When rows in CACHED tables are updated or deleted, the spaces are mostly reused. However, in time, some unused spaces are left in the .data file, especially when large tables are dropped or their structure is modified.

A CHECKPOINT operation does not normally reclaim the empty spaces, whereas CHECKPOINT DEFRAG always does.

This property determines when a normal CHECKPOINT, whether initiated by an administrator or when the size of the log exceeds its limit.

The numeric value is the number of megabytes of recorded empty spaces in the .data file that would force a DEFRAG operation. Low values result in more frequent DEFRAG operations. A value of 0 indicates no automatic DEFRAG is performed. The default is 200 megabytes of lost space.


This is TRUE by default. If bulk data needs to be loaded into the database, this property can be set FALSE for the duration of bulk load operation. This allows loading data for related tables in any order. The property should be set TRUE after bulk load. If the loaded data is not guaranteed to conform to the referential integrity constraints, SQL queries should be run after loading to identify and modify any non-conforming rows.

Chapter 5. Deployment Issues

Fred Toussi

HSQLDB Development Group

Copyright 2005 Fred Toussi. Permission is granted to distribute this document without any alteration under the terms of the HSQLDB license. Additional permission is granted to the HSQLDB Development Group to distribute this document with or without alterations under the terms of the HSQLDB license.

$Date: 2005/07/02 09:11:39 $


Many questions repeatedly asked in Forums and mailing lists are answered in this guide. If you want to use HSQLDB with your application, you should read this guide. This document covers system related issues. For issues related to SQL see the SQL Issues chapter.

Mode of Operation and Tables

HSQLDB has many modes of operation and features that allow it to be used in very different scenarios. Levels of memory usage, speed and accessibility by different applications are influenced by how HSQLDB is deployed.

Mode of Operation

The decision to run HSQLDB as a separate server process or as an in-process database should be based on the following:

  • When HSQLDB is run as a server on a separate machine, it is isolated from hardware failures and crashes on the hosts running the application.

  • When HSQLDB is run as a server on the same machine, it is isolated from application crashes and memory leaks.

  • Server connections are slower than in-process connections due to the overhead of streaming the data for each JDBC call.


TEXT tables are designed for special applications where the data has to be in an interchangeable format, such as CSV. TEXT tables should not be used for routine storage of data.

MEMORY tables and CACHED tables are generally used for data storage. The difference between the two is as follows:

  • The data for all MEMORY tables is read from the .script file when the database is started and stored in memory. In contrast the data for cached tables is not read into memory until the table is accessed. Furthermore, only part of the data for each CACHED table is held in memory, allowing tables with more data than can be held in memory.

  • When the database is shutdown in the normal way, all the data for MEMORY tables is written out to the disk. In comparison, the data in CACHED tables that has changed is written out at shutdown, plus a compressed backup of all the data in all cached tables.

  • The size and capacity of the data cache for all the CACHED tables is configurable. This makes it possible to allow all the data in CACHED tables to be cached in memory. In this case, speed of access is good, but slightly slower than MEMORY tables.

  • For normal applications it is recommended that MEMORY tables are used for small amounts of data, leaving CACHED tables for large data sets. For special applications in which speed is paramount and a large amount of free memory is available, MEMORY tables can be used for large tables as well

Large Objects

JDBC Clobs are supported as columns of the type LONGVARCHAR. JDBC Blobs are supported as columns of the type LONGVARBINARY. When large objects (LONGVARCHAR, LONGVARBINARY, OBJECT) are stored with table definitions that contain several normal fields, it is better to use two tables instead. The first table to contain the normal fields and the second table to contain the large object plus an identity field. Using this method has two benefits. (a) The first table can usually be created as a MEMORY table while only the second table is a CACHED table. (b) The large objects can be retrieved individually using their identity, instead of getting loaded into memory for finding the rows during query processing. An example of two tables and a select query that exploits the separation between the two follows:

SELECT * FROM (SELECT * FROM MAINTABLE <join any other table> WHERE <various conditions apply>) JOIN LOBTABLE ON MAINID=LOBID;

The inner SELECT finds the required rows without reference to the LOBTABLE and when it has found all the rows, retrieves the required large objects from the LOBTABLE.

Deployment context

The files used for storing HSQLDB database data are all in the same directory. New files are always created and deleted by the database engine. Two simple principles must be observed:

  • The Java process running HSQLDB must have full privileges on the directory where the files are stored. This include create and delete privileges.

  • The file system must have enough spare room both for the 'permanent' and 'temporary' files. The default maximum size of the .log file is 200MB. The .data file can grow to up to 8GB. The .backup file can be up to 50% of the .data file. The temporary file created at the time of a SHUTDOWN COMPACT can be equal in size to the .data file.

Memory and Disk Use

Memory used by the program can be thought of as two distinct pools: memory used for table data, and memory used for building result sets and other internal operations. In addition, when transactions are used, memory is utilised for storing the information needed for a rollback.

Since version 1.7.1, memory use has been significantly reduced compared to previous versions. The memory used for a MEMORY table is the sum of memory used by each row. Each MEMORY table row is a Java object that has 2 int or reference variables. It contains an array of objects for the fields in the row. Each field is an object such as Integer, Long, String, etc. In addition each index on the table adds a node object to the row. Each node object has 6 int or reference variables. As a result, a table with just one column of type INTEGER will have four objects per row, with a total of 10 variables of 4 bytes each - currently taking up 80 bytes per row. Beyond this, each extra column in the table adds at least a few bytes to the size of each row.

The memory used for a result set row has fewer overheads (fewer variables and no index nodes) but still uses a lot of memory. All the rows in the result set are built in memory, so very large result sets may not be possible. In server mode databases, the result set memory is released from the server once the database server has returned the result set. In-process databases release the memory when the application program releases the java.sql.ResultSet object. Server modes require additional memory for returning result sets, as they convert the full result set into an array of bytes which is then transmitted to the client.

When UPDATE and DELETE queries are performed on CACHED tables, the full set of rows that are affected, including those affected due to ON UPDATE actions, is held in memory for the duration of the operation. This means it may not be possible to perform deletes or updates involving very large numbers of rows of CACHED tables. Such operations should be performed in smaller sets.

When transactions support is enabled with SET AUTOCOMMIT OFF, lists of all insert, delete or update operations are stored in memory so that they can be undone when ROLLBACK is issued. Transactions that span hundreds of modification to data will take up a lot of memory until the next COMMIT or ROLLBACK clears the list.

Most JVM implementations allocate up to a maximum amount of memory (usually 64 MB by default). This amount is generally not adequate when large memory tables are used, or when the average size of rows in cached tables is larger than a few hundred bytes. The maximum amount of allocated memory can be set on the java ... command line that is used for running HSQLDB. For example, with Sun JVM version 1.3.0 the parameter -Xmx256m increases the amount to 256 MB.

1.8.0 uses a fast cache for immutable objects such as Integer or String that are stored in the database. In most circumstances, this reduces the memory footprint still further as fewer copies of the most frequently-used objects are kept in memory.

Cache Memory Allocation

With CACHED tables, the data is stored on disk and only up to a maximum number of rows are held in memory at any time. The default is up to 3*16384 rows. The hsqldb.cache_scale database property can be set to alter this amount. As any random subset of the rows in any of the CACHED tables can be held in the cache, the amount of memory needed by cached rows can reach the sum of the rows containing the largest field data. For example if a table with 100,000 rows contains 40,000 rows with 1,000 bytes of data in each row and 60,000 rows with 100 bytes in each, the cache can grow to contain nearly 50,000 rows, including all the 40,000 larger rows.

An additional property, hsqldb.cache_size_scale can be used in conjunction with the hsqldb.cache_scale property. This puts a limit in bytes on the total size of rows that are cached. When the default values is used for both properties, the limit on the total size of rows is approximately 50MB. (This is the size of binary images of the rows and indexes. It translates to more actual memory, typically 2-4 times, used for the cache because the data is represented by Java objects.)

If memory is limited, the hsqldb.cache_scale or hsqldb.cache_size_scale database properties can be reduced. In the example above, if the hsqldb.cache_size_scale is reduced from 10 to 8, then the total binary size limit is reduced from 50MB to 12.5 MB. This will allow the number of cached rows to reach 50,000 small rows, but only 12,500 of the larger rows.

Managing Database Connections

In all running modes (server or in-process) multiple connections to the database engine are supported. In-process (standalone) mode supports connections from the client in the same Java Virtual Machine, while server modes support connections over the network from several different clients.

Connection pooling software can be used to connect to the database but it is not generally necessary. With other database engines, connection pools are used for reasons that may not apply to HSQLDB.

  • To allow new queries to be performed while a time-consuming query is being performed in the background. This is not possible with HSQLDB 1.8.0 as it blocks while performing the first query and deals with the next query once it has finished it. This capability is under development and will be introduced in a future version.

  • To limit the maximum number of simultaneous connections to the database for performance reasons. With HSQLDB this can be useful only if your application is designed in a way that opens and closes connections for each small task.

  • To control transactions in a multi-threaded application. This can be useful with HSQLDB as well. For example, in a web application, a transaction may involve some processing between the queries or user action across web pages. A separate connection should be used for each HTTP session so that the work can be committed when completed or rolled back otherwise. Although this usage cannot be applied to most other database engines, HSQLDB is perfectly capable of handling over 100 simultaneous HTTP sessions as individual JDBC connections.

An application that is not both multi-threaded and transactional, such as an application for recording user login and logout actions, does not need more than one connection. The connection can stay open indefinitely and reopened only when it is dropped due to network problems.

When using an in-process database with versions prior to 1.7.2 the application program had to keep at least one connection to the database open, otherwise the database would have been closed and further attempts to create connections could fail. This is not necessary since 1.7.2, which does not automatically close an in-process database that is opened by establishing a connection. An explicit SHUTDOWN command, with or without an argument, is required to close the database. In version 1.8.0 a connection property can be used to revert to the old behaviour.

When using a server database (and to some extent, an in-process database), care must be taken to avoid creating and dropping JDBC Connections too frequently. Failure to observe this will result in unsuccessful connection attempts when the application is under heavy load.

Upgrading Databases

Any database not produced with the release version of HSQLDB 1.8.0 must be upgraded to this version. This includes databases created with the RC versions of 1.8.0. The instructions under the Upgrading Using the SCRIPT Command section should be followed in all cases.

Once a database is upgraded to 1.8.0, it can no longer be used with Hypersonic or previous versions of HSQLDB.

There may be some potential legacy issues in the upgrade which should be resolved by editing the .script file:

  • Version 1.8.0 does not accept duplicate names for indexes that were allowed before 1.7.2.

  • Version 1.8.0 does not accept duplicate names for table columns that were allowed before 1.7.0.

  • Version 1.8.0 does not create the same type of index for foreign keys as versions before 1.7.2.

  • Version 1.8.0 does not accept table or column names that are SQL identifiers without double quoting.

Upgrading Using the SCRIPT Command

To upgrade from 1.7.2 or 1.7.3 to 1.8.0, simply issue the SET SCRIPTFORMAT TEXT and SHUTDOWN SCRIPT commands with the old version, then open with the new version of the engine. The upgrade is then complete.

To upgrade from older version database files (1.7.1 and older) that do not contain CACHED tables, simple SHUTDOWN with the older version and open with the new version. If there is any error in the .script file, try again after editing the .script file.

To upgrade from older version database files (1.7.1 and older) that contain CACHED tables, use the SCRIPT procedure below. In all versions of HSQLDB and Hypersonic 1.43, the SCRIPT 'filename' command (used as an SQL query) allows you to save a full record of your database, including database object definitions and data, to a file of your choice. You can export a script file using the old version of the database engine and open the script as a database with 1.8.0.

Procedure 5.1. Upgrade Using SCRIPT procedure

  1. Open the original database in the old version of DatabaseManager

  2. Issue the SCRIPT command, for example SCRIPT 'newversion.script' to create a script file containing a copy of the database.

  3. Use the 1.8.0 version of DatabaseManager to create a new database, in this example 'newversion' in a different directory.

  4. SHUTDOWN this database.

  5. Copy the newversion.script file from step 2 over the file of the same name for the new database created in 4.

  6. Try to open the new database using DatabaseManager.

  7. If there is any inconsistency in the data, the script line number is reported on the console and the opening process is aborted. Edit and correct any problems in the newversion.script before attempting to open again. Use the guidelines in the next section (Manual Changes to the .script File). Use a programming editor that is capable of handling very large files and does not wrap long lines of text.

Manual Changes to the .script File

In 1.8.0 the full range of ALTER TABLE commands is available to change the data structures and their names. However, if an old database cannot be opened due to data inconsistencies, or the use of index or column names that are not compatible with 1.8.0, manual editing of the SCRIPT file can be performed.

The following changes can be applied so long as they do not affect the integrity of existing data.

  • Names of tables, columns and indexes can be changed.

  • CREATE UNIQUE INDEX ... to CREATE INDEX ... and vice versa

    A unique index can always be converted into a normal index. A non-unique index can only be converted into a unique index if the table data for the column(s) is unique in each row.


    A not-null constraint can always be removed. It can only be added if the table data for the column has no null values.


    A primary key constraint can be removed or added. It cannot be removed if there is a foreign key referencing the column(s).


    Some changes to column types are possible. For example an INTEGER column can be changed to BIGINT, or DATE, TIME and TIMESTAMP columns can be changed to VARCHAR.

After completing the changes and saving the modified *.script file, you can open the database as normal.

Backing Up Databases

The data for each database consists of up to 5 files in the same directory. The endings are *.properties, *.script, *.data, *.backup and *.log (a file with the *.lck ending is used for controlling access to the database and should not be backed up). These should be backed up together. The files can be backed up while the engine is running but care should be taken that a CHECKPOINT or SHUTDOWN operation does not take place during the backup. It is more efficient to perform the backup immediately after a CHECKPOINT. The *.data file can be excluded from the backup. In this case, when restoring, a dummy *.data file is needed which can be an empty, 0 length file. The engine will expand the *.backup file to replace this dummy file if the backup is restored. If the *.data file is not backed up, the *.properties file may have to be modified to ensure it contain modified=yes instead of modified=no prior to restoration. If a backup immediately follows a checkpoint, then the *.log file can also be excluded, reducing the significant files to *.properties, *.script and *.backup. Normal backup methods, such as archiving the files in a compressed bundle can be used.

Chapter 6. Text Tables

Text Tables as a Standard Feature of Hsqldb

Bob Preston

HSQLDB Development Group

Fred Toussi

HSQLDB Development Group

Copyright 2002-2005 Bob Preston and Fred Toussi. Permission is granted to distribute this document without any alteration under the terms of the HSQLDB license. Additional permission is granted to the HSQLDB Development Group to distribute this document with or without alterations under the terms of the HSQLDB license.

$Date: 2007/08/28 13:19:09 $

Text Table support for HSQLDB was originally developed by Bob Preston independently from the Project. Subsequently Bob joined the Project and incorporated this feature into version 1.7.0, with a number of enhancements, especially the use of conventional SQL commands for specifying the files used for Text Tables.

In a nutshell, Text Tables are CSV or other delimited files treated as SQL tables. Any ordinary CSV or other delimited file can be used. The full range of SQL queries can be performed on these files, including SELECT, INSERT, UPDATE and DELETE. Indexes and unique constraints can be set up, and foreign key constraints can be used to enforce referential integrity between Text Tables themselves or with conventional tables.

HSQLDB with Text Table support is the only comprehensive solution that employs the power of SQL and the universal reach of JDBC to handle data stored in text files and will have wide-ranging use way beyond the currently established Java realm of HSQLDB.

Goals of the Implementation

  1. We aimed to finalise the DDL for Text Tables so that future releases of HSQLDB use the same DDL scripts.

  2. We aimed to support Text Tables as GLOBAL TEMPORARY or GLOBAL BASE tables in the SQL domain.

The Implementation

Definition of Tables

Text Tables are defined similarly to conventional tables with the added TEXT keyword:

    CREATE TEXT TABLE <tablename> (<column definition> [<constraint definition>])

In addition, a SET command specifies the file and the separator character that the Text table uses:

    SET TABLE <tablename> SOURCE <quoted_filename_and_options> [DESC]

Text Tables cannot be created in memory-only databases (databases that have no script file).

Scope and Reassignment

  • A Text table without a file assigned to it is READ ONLY and EMPTY.

  • A Temporary Text table has the scope and the lifetime of the SQL session (a JDBC Connection).

  • Reassigning a Text Table definition to a new file has implications in the following areas:

    1. The user is required to be an administrator.

    2. Existing transactions are committed at this point.

    3. Constraints, including foreign keys referencing this table, are kept intact. It is the responsibility of the administrator to ensure their integrity.

    From version 1.7.2 the new source file is scanned and indexes are built when it is assigned to the table. At this point any violation of NOT NULL, UNIQUE or PRIMARY KEY constrainst are caught and the assignment is aborted. However, foreign key constraints are not checked at the time of assignment or reassignment of the source file.

Null Values in Columns of Text Tables

This has changed since 1.7.2 to support both null values and empty strings.

  • Empty fields are treated as NULL. These are fields where there is nothing or just spaces between the separators.

  • Quoted empty strings are treated as empty strings.


The default field separator is a comma (,). A different field separator can be specified within the SET TABLE SOURCE statement. For example, to change the field separator for the table mytable to a vertical bar, place the following in the SET TABLE SOURCE statement, for example:

    SET TABLE mytable SOURCE "myfile;fs=|"

Since HSQLDB treats CHAR's, VARCHARs, and LONGVARCHARs the same, the ability to assign different separators to the latter two is provided. When a different separator is assigned to a VARCHAR or LONGVARCHAR field, it will terminate any CSV field of that type. For example, if the first field is CHAR, and the second field LONGVARCHAR, and the separator fs has been defined as the pipe (|) and vs as the period (.) then the data in the CSV file for a row will look like:

    First field data|Second field data.Third field data

The following example shows how to change the default separator to the pipe (|), VARCHAR separator to the period (.) and the LONGVARCHAR separator to the tilde (~). Place the following within the SET TABLE SOURCE statement, for example:

    SET TABLE mytable SOURCE "myfile;fs=|;vs=.;lvs=~"

HSQLDB also recognises the following special indicators for separators:

special indicators for separators






space character




newline - Used as an end anchor (like $ in regular expressions)


carriage return






a Unicode character specified in hexadecimal

Furthermore, HSQLDB provides csv file support with three additional boolean options: ignore_first, quoted and all_quoted. The ignore_first option (default false) tells HSQLDB to ignore the first line in a file. This option is used when the first line of the file contains column headings. The all_quoted option (default false) tells the program that it should use quotes around all character fields when writing to the source file. The quoted option (default true) uses quotes only when necessary to distinguish a field that contains the separator character. It can be set to false to prevent the use of quoting altogether and treat quote characters as normal characters. These options may be specified within the SET TABLE SOURCE statement:

    SET TABLE mytable SOURCE "myfile;ignore_first=true;all_quoted=true"

When the default options all_quoted= false and quoted=true are in force, fields that are written to a line of the csv file will be quoted only if they contain the separator or the quote character. The quote character is doubled when used inside a string. When all_quoted=false and quoted=false the quote character is not doubled. With this option, it is not possible to insert any string containing the separator into the table, as it would become impossible to distinguish from a separator. While reading an existing data source file, the program treats each individual field separately. It determines that a field is quoted only if the first character is the quote character. It interprets the rest of the field on this basis.

The character encoding for the source file is ASCII by default. To support UNICODE or source files preprared with different encodings this can be changed to UTF-8 or any other encoding. The default is encoding=ASCII and the option encoding=UTF-8 or other supported encodings can be used.

Finally, HSQLDB provides the ability to read a text file from the bottom up and making them READ ONLY, by placing the keyword "DESC" at the end of the SET TABLE SOURCE statement:

    SET TABLE mytable SOURCE "myfile" DESC

This feature provides functionality similar to the Unix tail command, by re-reading the file each time a select is executed. Using this feature sets the table to read-only mode. Afterwards, it will no longer be possible to change the read-only status with SET TABLE <tablename> READONLY TRUE.

Text table source files are cached in memory. The maximum number of rows of data that are in memory at any time is controlled by the textdb.cache_scale property. The default value for textdb.cache_scale is 10 and can be changed by setting the property in the .properties file for the database. The number of rows in memory is calculated as 3*(2**scale), which translates to 3072 rows for the default textdb.cache_scale setting (10). The property can also be set for individual text tables:

    SET TABLE mytable SOURCE "myfile;ignore_first=true;all_quoted=true;cache_scale=12"

Disconnecting Text Tables

The following describes behaviour present in and later.

Text tables may be disconnected from their underlying data source, i.e. the text file.

You can explicitly disconnect a text table from its file by issuing the following statement:


Subsequently, mytable will be empty and read-only. However, the data source description will be preserved, and the table can be re-connected to it with


When a database is opened, if the source file for an existing text table is missing the table remains disconnected from its data source, but the source sescription is preserved. This allows the missing source file to be added to the directory and the table re-connected to it with the above command.

Text File Issues

Text File Issues

  • File locations are restricted to below the directory that contains the database, unless the textdb.allow_full_path property is set true in the database properties file.

  • Blank lines are allowed anywhere in the text file, and are ignored.

  • The file location for a text table created with

        SELECT <select list> INTO TEXT <tablename> FROM

    is the directory that contains the database and the file name is based on the table name. The table name is converted into the file name by replacing all the non-alphanumeric characters with the underscore character, conversion into lowercase, and adding the ".csv" suffix.

  • It is possible to define a primay key or identity column for text tables.

  • When a table source file is used with the ignore_first=true option, the first, ignored line is replaced with a blank line after a SHUTDOWN COMPACT.

  • An existing table source file may include CHARACTER fields that do not begin with the quote character but contain instances of the quote character. These fields are read as literal strings. Alternatively, if any field begins with the quote character, then it is interpreted as a quoted string that should end with the quote character and any instances of the quote character within the string is doubled. When any field containing the quote character or the separator is written out to the source file by the program, the field is enclosed in quote character and any instance of the quote character inside the field is doubled.

  • Inserts or updates of CHARACTER type field values are allowed with strings that contains the linefeed or the carriage return character. This feature is disabled when both quoted and all_quoted properties are false.

  • ALTER TABLE commands that add or drop columns are not supported with non-empty text tables.

Text File Global Properties

Complete list of supported global properties in *.properties files

  • textdb.fs

  • textdb.lvs

  • textdb.quoted

  • textdb.all_quoted

  • textdb.ignore_first

  • textdb.encoding

  • textdb.cache_scale

  • textdb.allow_full_path

Importing a Text Table file in to a Traditional (non-Text Table) Table

The directory src/org/hsqldb/sample in your HSQLDB distibution contains a file named load_binding_lu.sql. This is a working SQL file which imports a pipe-delimited text file from the database's file directory into an existing normal table. You can edit a copy of this file and use it directly with SqlTool, or you can use the SQL therein as a model (using any SQL client at all).

Chapter 7. TLS

TLS Support (a.k.a. SSL)

Blaine Simpson

HSQLDB Development Group

$Date: 2006/07/27 21:08:21 $

The instructions in this document are liable to change at any time. In particular, we will be changing the method to supply the server-side certificate password.


Hsqldb TLS Support Requirements

  • Sun Java 2.x and up. (This is probably possible with IBM's Java, but I don't think anybody has attempted to run HSQLDB with TLS under IBM's Java, and I'm sure that nobody in the HSQLDB Development Group has documented how to set up the environment).

  • If Java 2.x or 3.x, then you will need need to install JSSE. Your server and/or client will start up much slower than that of Java 4.x users. Client-side users will not be able to use the https: JDBC protocol (because the https protocol handler is not implemented in 2.x/3.x Java JSSE; if there is demand, we could work around this).

  • A JKS keystore containing a private key, in order to run a server.

  • If you are running the server side, then you'll need to run a HSQLDB Server or WebServer. It doesn't matter if the underlying database instances are new, and it doesn't matter if you are making a new Server configuration or encrypting an existing Server configuration. (You can turn encryption on and off at will).

  • You need a HSQLDB jar file that was built with JSSE present. If you got your HSQLDB 1.7.2 distribution from us, you are all set, because we build with Java 1.4 (which contains JSSE). If you build your own jar file with Java 1.3, make sure to install JSSE first.

Encrypting your JDBC connection

At this time, only 1-way, server-cert encryption is tested.


Just use one of the following protocol prefixes.

Hsqldb TLS URL Prefixes

  • jdbc:hsqldb:hsqls://

  • jdbc:hsqldb:https://

At this time, the latter will only work for clients running with Java 1.4.

If the server you wish to connect to is using a certificate approved by your default trust keystores, then there is nothing else to do. If not, then you need to tell Java to "trust" the server cert. (It's a slight over-simplification to say that if the server certificate was purchased, then you are all set; if somebody "signed their own" certificate by self-signing or using a private ca certificate, then you need to set up trust).

First, you need to obtain the cert (only the "public" part of it). Since this cert is passed to all clients, you could obtain it by writing a java client that dumps it to file, or perhaps by using openssl s_client. Since in most cases, if you want to trust a non-commercial cert, you probably have access to the server keystore, I'll show an example of how to get what you need from the server-side JKS keystore.

You may already have an X509 cert for your server. If you have a server keystore, then you can generate a X509 cert like this.

Example 7.1. Exporting certificate from the server's keystore

    keytool -export -keystore server.store -alias existing_alias -file server.cer
In this example, server.cer is the X509 certificate that you need for the next step.

Now, you need to add this cert to one of the system trust keystores or to a keystore of your own. See the Customizing Stores section in JSSERefGuide.html to see where your system trust keystores are. You can put private keystores anywhere you want to. The following command will add the cert to an existing keystore, or create a new keystore if client.store doesn't exist.

Example 7.2. Adding a certificate to the client keystore

keytool -import -trustcacerts -keystore trust.store -alias new_alias -file server.cer

If you are making a new keystore, you probably want to start with a copy of your system default keystore which you can find somewhere under your JAVA_HOME directory (typically jre/lib/security/cacerts for a JDK, but I forget exactly where it is for a JRE).

Unless your OS can't stop other people from writing to your files, you probably do not want to set a password on the trust keystore.

If you added the cert to a system trust store, then you are finished. Otherwise you will need to specify your custom trust keystore to your client program. The generic way to set the trust keystore is to set the sytem property javax.net.ssl.trustStore every time that you run your client program. For example

Example 7.3. Specifying your own trust store to a JDBC client

    java -Djavax.net.ssl.trustStore=/home/blaine/trust.store -jar /path/to/hsqldb.jar dest-urlid
This example runs the program SqlTool. SqlTool has built-in TLS support, however, so, for SqlTool you can set truststore on a per-urlid basis in the SqlTool configuration file.

N.b. The hostname in your database URL must match the Common Name of the server's certificate exactly. That means that if a site certificate is admc.com, you can not use jdbc:hsqldb:hsqls://localhost or jdbc:hsqldb:hsqls://www.admc.com:1100 to connect to it.

If you want more details on anything, see JSSERefGuide.html on Sun's site, or in the subdirectory docs/guide/security/jsse of your Java SE docs.


Get yourself a JKS keystore containing a private key. Then set the system property javax.net.ssl.keyStore to the path to that file, and javax.net.ssl.keyStorePassword to the password of the keystore (and to the private key-- they have to be the same).

Example 7.4. Running an Hsqldb server with TLS encryption

    java -Djavax.net.ssl.keyStorePassword=secret  \
        -Djavax.net.ssl.keyStore=/usr/hsqldb/db/db3/server.store  \
        -cp /path/to/hsqldb.jar org.hsqldb.Server

(This is a single command that I have broken into 2 lines using my shell's \ line-continuation feature. In this example, I'm using a server.properties file so that I don't need to give arguments to specify database instances or the server endpoint).


Specifying a password on the command-line is definitely not secure. It's really only appropriate when untrusted users do not have any access to your computer.

If there is any user demand, we will have a more secure way to supply the password before long.


If you are running Java 4.x, then you are all set. Java 1.x users, you are on your own (Sun does not provide a JSSE that will work with 1.x). Java 2.x and 3.x users continue...

Go to http://java.sun.com/products/jsse/index-103.html. If you agree to the terms and meet the requirements, download the domestic or global JSSE software. All you need from the software distro is the three jar files. If you have a JDK installation, then move the 3 jar files into the directory $JAVA_HOME/jre/lib/ext. If you have a JRE installation, then move the 3 jar files into the directory $JAVA_HOME/lib/ext.

Pretty painless.

Making a Private-key Keystore

There are two main ways to do this. Either you can use a certificate signed by a certificate authority, or you can make your own. One thing that you need to know in both cases is, the Common Name of the cert has to be the exact hostname that JDBC clients will use in their database URL.

CA-Signed Cert

I'm not going to tell you how to get a CA-signed SSL certificate. That is well documented at many other places.

Assuming that you have a standard pem-style private key certificate, here's how you can use openssl and the program DERImport to get it into a JKS keystore.

Because I have spent a lot of time on this document already, I am just giving you an example.

Example 7.5. Getting a pem-style private key into a JKS keystore

openssl pkcs8 -topk8 -outform DER -in Xpvk.pem -inform PEM -out Xpvk.pk8 -nocrypt

openssl x509 -in Xcert.pem -out Xcert.der -outform DER

java DERImport new.keystore NEWALIAS Xpvk.pk8 Xcert.der


Make sure to set the password of the key exactly the same as the password for the keystore!

You need the program DERImport.class of course. Do some internet searches to find DERImport.java or DERImport.class and download it.

If DERImport has become difficult to obtain, I can write a program to do the same thing-- just let me know.

Non-CA-Signed Cert

Run man keytool or see the Creating a Keystore section of JSSERefGuide.html.

Automatic Server or WebServer startup on UNIX

If you are on UNIX and want to automatically start and stop a Server or WebServer running with encryption, follow the instructions in the UNIX Quick Start chapter, and remember to make the init script configuration file readable only to root and to set the variables TLS_PASSWORD and TLS_KEYSTORE.

If you are using a private server certificate, make sure to also set the trust store filepath as shown in the sample init script configuration file.

The cautionary warning above still applies. The password will be visible to any minimally competent local UNIX user who wants to see it.

Chapter 8. SqlTool

SqlTool Manual

Blaine Simpson

HSQLDB Development Group

$Date: 2007/08/09 23:53:01 $

Purpose, Coverage, Changes in Behavior

Purpose, Coverage, Changes in Behavior

This document explains how to use SqlTool, the main purpose of which is to read your SQL text file or stdin, and execute the SQL commands therein against a JDBC database. There are also a great number of features to facilitate both interactive use and automation. The following paragraphs explain in a general way why SqlTool is better than any existing tool for text-mode interactive SQL work, and for automated SQL tasks. Two important benefits which SqlTool shares with other pure Java JDBC tools is that users can use a consistent interface and syntax to interact with a huge variety of databases-- any database which supports JDBC; plus the tool itself runs on any Java platform. Instead of using isql for Sybase, psql for Postgresql, Sql*plus for Oracle, etc., you can use SqlTool for all of them. As far as I know, SqlTool is the only production-ready, pure Java, command-line, generic JDBC client. Several databases come with a command-line client with limited JDBC abilities (usually designed for use with their specific database).

SqlTool is purposefully not a Gui tool like Toad or DatabaseManager. There are many use cases where a Gui SQL tool would be better. Where automation is involved in any way, you really need a text client to at least test things properly and usually to prototype and try things out. A command-line tool is really better for executing SQL scripts, any form of automation, direct-to-file fetching, and remote client usage. To clarify this last, if you have to do your SQL client work on a work server on the other side of a VPN connection, you will quickly appreciate the speed difference between text data transmission and graphical data transmission, even if using VNC or Remote Console. Another case would be where you are doing some repetitive or very structured work where variables or language features would be useful. Gui proponents may disagree with me, but scripting (of any sort) is more efficient than repetitive copy & pasting with a Gui editor. SqlTool starts up very quickly, and it takes up a tiny fraction of the RAM required to run a comparably complex Gui like Toad.

SqlTool is superior for interactive use because over many years it has evolved lots of features proven to be efficient for day-to-day use. Three concise help commands (\?, :?, and *?) list all available commands of the corresponding type. SqlTool doesn't support up-arrow or other OOB escapes (due to basic Java I/O limitations), but it more than makes up for this limitation with aliases, user variables, command-line history and recall, and command-line editing with extended Perl/Java regular expressions. The \d commands deliver JDBC metadata information as consistently as possible (in several cases, database-specific work-arounds are used to obtain the underlying data even though the database doesn't provide metadata according to the JDBC specs). Unlike server-side language features, the same feature set works for any database server. Database access details may be supplied on the command line, but day-to-day users will want to centralize JDBC connection details into a single, protected RC file. You can put connection details (username, password, URL, and other optional settings) for scores of target databases into your RC file, then connect to any of them whenever you want by just giving SqlTool the ID ("urlid") for that database. When you Execute SqlTool interactively, it behaves by default exactly as you would want it to. If errors occur, you are given specific error messages and you can decide whether to roll back your session. You can easily change this behavior to auto-commit, exit-upon-error, etc., for the current session or for all interactive invocations. You can import or export delimiter-separated-value files.

When you Execute SqlTool with a SQL script, it behaves by default exactly as you would want it to. If any error is encountered, the connection will be rolled back, then SqlTool will exit with an error exit value. If you wish, you can detect and handle error (or other) conditions yourself. For scripts expected to produce errors (like many scripts provided by database vendors), you can have SqlTool continue-upon-error. For SQL script-writers, you will have access to portable scripting features which you've had to live without until now. You can use variables set on the command line or in your script. You can handle specific errors based on the output of SQL commands or of your variables. You can chain SQL scripts, invoke external programs, dump data to files, use prepared statements, Finally, you have a procedural language with if, foreach, while, continue, and break statements.

Platforms and SqlTool versions covered

SqlTool runs on any Java 1.4 or later platform. I haven't run it with a non-Sun JVM in years (like Blackdown, IBM, JRockit, etc.), but I've had no reports of problems with them, and SqlTool uses none of the Sun-proprietary classes directly. Some of the examples below use quoting which works exactly as-is for any Bourne-compatible UNIX shell. (Only line-continuation would need to be changed for C-compatible UNIX shells). I have not yet tested these commands on Windows, and I doubt whether the quoting will work just like this (though it is possible). SqlTool is still a very useful tool even if you have no quoting capability at all.

If you are using SqlTool from a HSQLDB distribution before version final, you should use the documentation with that distribution, because this manual documents many new features, several significant changes to interactive-only commands, and a few changes effecting backwards-compatibility (see next section about that). This document is now updated for the current versions of SqlTool and SqlFile at the time I am writing this (versions 333 and 354 correspondingly, SqlFile is the class which does most of the work for SqlTool). Therefore, if you are using a version of SqlTool or SqlFile that is more than a couple revisions greater, you should find a newer version of this document. (The imprecision is due to content-independent revision increments at build time, and the likelihood of one or two behavior-independent bug fixes after public releases). The startup banner will report both versions when you run SqlTool interactively. (Dotted version numbers of SqlTool and SqlFile are older than 333 and 354).

This guide covers SqlTool bundled with series 1.8 and 1.9 of HSQLDB. [1]

Functional Changes

This section lists changes to SqlTool since the last major release of HSQLDB which may effect the portability of SQL scripts. For this revision of this document, this list consists of script-impacting changes made to SqlTool after the final HSQLDB release. I'm specifically not listing changes to interactive-only commands (":" commands, with one legacy exception which is listed below), since these commands can't be used in SQL scripts; and I'm specifically not listing backwards-compatible feature additions and enhancements. The reason for limiting the change list to only portability- impacting changes is that a list of all enhancements since just would be pages long.

  • SqlTool now consistently outputs \r\n line breaks when on \r\n-linebreak platforms, like Windows. This includes output written to stdout, \w files, and \o files.
  • Time type values are always output with the date as well as the time. This was required in order to produce consistent output for the wildly varying formats provided by different database vendors.
  • DSV input now takes JDBC Timestamp format with date and optionally time of day.
  • The command ":;" is now strictly an interactive command. If you want to repeat a command in an SQL scripts, just repeat the exact text of the command. Non-interactive use now has no dependency on command history.
  • The command ":w" has replace the command \w. Unlike writing "output" to a file with \w, :w is used to write SQL "commands", and this is an interactive feature.
  • Shell scripts using raw mode (e.g. PL/SQL scripts) must terminate the raw code with a line containing ".;", which will also send the code to the database for execution. (The old "." command has been changed to ":." to make it very clear that the command is now an interactive command).
  • The --sql argument will never automatically append a semicolon to the text you provide. If you want to execute a command ending with a semi-- then type a semi.

Although it doesn't effect scripts, I will mention a significant recent change to interactive commands. Special and PL commands are not stored to the edit buffer and to command history, so they can be recalled and edited just like SQL commands. Now, only edit/history : commands are not stored to the buffer and history.

The Bare Minimum You Need to Know to Run SqlTool


If you are using an Oracle database server, it will commit your current transaction if you cleanly disconnect, regardless of whether you have set auto-commit or not. This will occur if you exit SqlTool (or any other client) in the normal way (as opposed to killing the process or using Ctrl-C, etc.). This is mentioned in this section only for brevity, so I don't need to mention it in the main text in the many places where auto-commit is discussed. This behavior has nothing to do with SqlTool. It is a quirk of Oracle.

If you want to use SqlTool, then you either have an SQL text file, or you want to interactively type in SQL commands. If neither case applies to you, then you are looking at the wrong program.

Procedure 8.1. To run SqlTool...

  1. Copy the file sqltool.rc from the directory sample [1] of your HSQLDB distribution to your home directory and secure access to it if your computer is accessible to anybody else (most likely from the network). This file will work as-is for a Memory Only database instance; or if your target is a HSQLDB Server running on your local computer with default settings and the password for the "sa" account is blank (the sa password is blank when new HSQLDB database instances are created). Edit the file if you need to change the target Server URL, username, password, character set, JDBC driver, or TLS trust store as documented in the RC File Authentication Setup section. (You could, alternatively, use the --inlineRc command-line switch to specify your connection parameters as documented in the Using Inline RC Authentication section).

  2. Find out where your hsqldb.jar file resides. It typically resides at HSQLDB_HOME/lib/hsqldb.jar where HSQLDB_HOME is the base directory of your HSQLDB software installation. For this reason, I'm going to use "$HSQLDB_HOME/lib/hsqldb.jar" as the path to hsqldb.jar for my examples, but understand that you need to use the actual path to your own hsqldb.jar file.

  3. Run

        java -jar $HSQLDB_HOME/lib/hsqldb.jar --help
    to see what command-line arguments are available. Note that you don't need to worry about setting the CLASSPATH when you use the -jar switch to java. Assuming that you set up your SqlTool RC file at the default location and you want to use the HSQLDB JDBC driver, you will want to run something like
        java -jar $HSQLDB_HOME/lib/hsqldb.jar mem
    for interactive use, or
        java -jar $HSQLDB_HOME/lib/hsqldb.jar --sql 'SQL statement;' mem
        java -jar $HSQLDB_HOME/lib/hsqldb.jar mem filepath1.sql...
    where mem is an urlid, and the following arguments are paths to text SQL files. For the filepaths, you can use whatever wildcards your operating system shell supports.

    The urlid mem in these commands is a key into your RC file, as explained in the RC File Authentication Setup section. Since this is a Memory Only database, you can use SqlTool with this urlid immediately with no database setup whatsoever (however, you can't persist any changes that you make to this database). The sample sqltool.rc file also defines the urlid "localhost-sa" for a local HSQLDB Server. At the end of this section, I explain how you can load some sample data to play with, if you want to.


SqlTool does not commit SQL changes by default. This leaves it to the user's disgression whether to commit or rollback their modifications. Remember to either run \= to commit before quitting SqlTool (most databases also support the SQL command commit;, or use the --autoCommit command-line switch.

If you put a file named auto.sql into your home directory, this file will be executed automatically every time that you run SqlTool interactively and without the --noAutoFile switch.

To use a JDBC Driver other than the HSQLDB driver, you can't use the -jar switch because you need to modify the classpath. You must add the hsqldb.jar file and your JDBC driver classes to your classpath, and you must tell SqlTool what the JDBC driver class name is. The latter can be accomplished by either using the "--driver" switch, or setting "driver" in your config file. The RC File Authentication Setup section. explains the second method. Here's an example of the first method (after you have set the classpath appropriately).

java org.hsqldb.util.SqlTool --driver oracle.jdbc.OracleDriver urlid


If the tables of query output on your screen are all messy because of lines wrapping, the best and easiest solution is usually to resize your terminal emulator window to make it wider. (With some terms you click & drag the frame edges to resize, with others you use a menu system where you can enter the number of columns).

If you are using SqlTool to connect to a HSQLDB network server or any non-HSQLDB database, you may prefer to use the jar file hsqltool.jar or hsqldbutil.jar instead of hsqldb.jar. These alternative jar files contain all of SqlTool without stuff you don't need, but you will have to follow a simple procedure to generate these jars. See the Using hsqltool.jar and hsqldbutil.jar section.

Non-displayable Types

There are many SQL types which SqlTool (being a text-based program) can't display properly. This includes the SQL types BLOB, JAVA_OBJECT, STRUCT, and OTHER. When you run a query that returns any of these, SqlTool will save the very first such value obtained to the binary buffer and will not display any output from this query. You can then save the binary value to a file, as explained in the Storing and retrieving binary files section.

There are other types, such as BINARY, which JDBC can make displayable (by using ResultSet.getString()), but which you may very well want to retrieve in raw binary format. You can use the \b command to retrieve any-column-type-at-all in raw binary format (so you can later store the value to a binary file).

Another restriction which all text-based database clients have is the practical inability for the user to type in binary data such as photos, audio streams, and serialized Java objects. You can use SqlTool to load any binary object into a database by telling SqlTool to get the insert/update datum from a file. This is also explained in the Storing and retrieving binary files section.

Desktop shortcuts

Desktop shortcuts and quick launch icons are useful, especially if you often run SqlTool with the same set of arguments. It's really easy to set up several of them-- one for each way that you invoke SqlTool (i.e., each one would start SqlTool with all the arguments for one of your typical startup needs). One typical setup is to have one shortcut for each database account which you normally use (use a different urlid argument in each shortcut's Target specification.

Desktop icon setup varies depending on your Desktop manager, of course. I'll explain how to set up a SqlTool startup icon in Windows XP. Linux and Mac users should be able to take it from there, since it's easier with the common Linux and Mac desktops.

Procedure 8.2. Creating a Desktop Shortcut for SqlTool

  1. Right click in the main Windows background.

  2. New

  3. Shortcut

  4. Browse

  5. Navigate to where your good JRE lives. For recent Sun JRE's, it installs to C:\Program Files\Java\*\bin by default (the * will be a JDK or JRE name and version number).

  6. Select java.exe.

  7. OK

  8. Next

  9. Enter any name

  10. Finish

  11. Right click the new icon.

  12. Properties

  13. Edit the Target field.

  14. Leave the path to java.exe exactly as it is, including the quotes, but append to what is there. Beginning with a space, enter the command-line that you want run.

  15. Change Icon... to a pretty icon.

  16. If you want a quick-launch icon instead of (or in addition to) a desktop shortcut icon, click and drag it to your quick launch bar. (You may or may not need to edit the Windows Toolbar properties to let you add new items).

Loading sample data

If you want some sample database objects and data to play with, execute the sampledata.sql SQL file. sampledata.sql resides in the sample directory of your HSQLDB distribution [1]. To separate the sample data from your regular data, you can put it into its own schema by running this before you import:

    SET SCHEMA sampledata;
Run it like this from an SqlTool session
\i HSQLDB_HOME/sample/sampledata.sql
where HSQLDB_HOME is the base directory of your HSQLDB software installation [1].

For memory-only databases, you'll need to run this every time that you run SqlTool. For other (persistent) databases, the data will reside in your database until you drop the tables.

RC File Authentication Setup

RC file authentication setup is accomplished by creating a text RC configuration file. In this section, when I say configuration or config file, I mean an RC configuration file. RC files can be used by any JDBC client program that uses the org.hsqldb.util.RCData class-- this includes SqlTool, DatabaseManager, DatabaseManagerSwing. You can use it for your own JDBC client programs too.

The following sample RC file resides at sample/sqltool.rc in your HSQLDB distribution [1].

Example 8.1. Sample RC File

# $Id: sqltool.rc,v 1.22 2007/08/09 03:22:21 unsaved Exp $

# This is a sample RC configuration file used by SqlTool, DatabaseManager,
# and any other program that uses the org.hsqldb.util.RCData class.

# You can run SqlTool right now by copying this file to your home directory
# and running
#    java -jar /path/to/hsqldb.jar mem
# This will access the first urlid definition below in order to use a 
# personal Memory-Only database.
# "url" values may, of course, contain JDBC connection properties, delimited
# with semicolons.

# If you have the least concerns about security, then secure access to
# your RC file.
# See the documentation for SqlTool for various ways to use this file.

# A personal Memory-Only (non-persistent) database.
urlid mem
url jdbc:hsqldb:mem:memdbid
username sa

# A personal, local, persistent database.
urlid personal
url jdbc:hsqldb:file:${user.home}/db/personal;shutdown=true
username sa
# When connecting directly to a file database like this, you should 
# use the shutdown connection property like this to shut down the DB
# properly when you exit the JVM.

# This is for a hsqldb Server running with default settings on your local
# computer (and for which you have not changed the password for "sa").
urlid localhost-sa
url jdbc:hsqldb:hsql://localhost
username sa

# Template for a urlid for an Oracle database.
# You will need to put the oracle.jdbc.OracleDriver class into your 
# classpath.
# In the great majority of cases, you want to use the file classes12.zip
# (which you can get from the directory $ORACLE_HOME/jdbc/lib of any
# Oracle installation compatible with your server).
# Since you need to add to the classpath, you can't invoke SqlTool with
# the jar switch, like "java -jar .../hsqldb.jar..." or 
# "java -jar .../hsqlsqltool.jar...".
# Put both the HSQLDB jar and classes12.zip in your classpath (and export!)
# and run something like "java org.hsqldb.util.SqlTool...".

#urlid cardiff2
#url jdbc:oracle:thin:@aegir.admc.com:1522:TRAFFIC_SID
#username blaine
#password secretpassword
#driver oracle.jdbc.OracleDriver

# Template for a TLS-encrypted HSQLDB Server.
# Remember that the hostname in hsqls (and https) JDBC URLs must match the
# CN of the server certificate (the port and instance alias that follows 
# are not part of the certificate at all).
# You only need to set "truststore" if the server cert is not approved by
# your system default truststore (which a commercial certificate probably
# would be).

#urlid tls
#url jdbc:hsqldb:hsqls://db.admc.com:9001/lm2
#username blaine
#password asecret
#truststore /home/blaine/ca/db/db-trust.store

# Template for a Postgresql database
#urlid blainedb
#url jdbc:postgresql://idun.africawork.org/blainedb
#username blaine
#password losung1
#driver org.postgresql.Driver

# Template for a MySQL database.  MySQL has poor JDBC support.
#urlid mysql-testdb
#url jdbc:mysql://hostname:3306/dbname
#username root
#username blaine
#password hiddenpwd
#driver com.mysql.jdbc.Driver

# Note that "databases" in SQL Server and Sybase are traditionally used for
# the same purpose as "schemas" with more SQL-compliant databases.

# Template for a Microsoft SQL Server database
#urlid msprojsvr
#url jdbc:microsoft:sqlserver://hostname;DatabaseName=DbName;SelectMethod=Cursor
# The SelectMethod setting is required to do more than one thing on a JDBC
# session (I guess Microsoft thought nobody would really use Java for 
# anything other than a "hello world" program).
# This is for Microsoft's SQL Server 2000 driver (requires mssqlserver.jar
# and msutil.jar).
#driver com.microsoft.jdbc.sqlserver.SQLServerDriver
#username myuser
#password hiddenpwd

# Template for a Sybase database
#urlid sybase
#url jdbc:sybase:Tds:hostname:4100/dbname
#username blaine
#password hiddenpwd
# This is for the jConnect driver (requires jconn3.jar).
#driver com.sybase.jdbc3.jdbc.SybDriver

# Template for Embedded Derby / Java DB.
#urlid derby1
#url jdbc:derby:path/to/derby/directory;create=true
#username ${user.name}
#password any_noauthbydefault
#driver org.apache.derby.jdbc.EmbeddedDriver
# The embedded Derby driver requires derby.jar.
# There'a also the org.apache.derby.jdbc.ClientDriver driver with URL
# like jdbc:derby://<server>[:<port>]/databaseName, which requires
# derbyclient.jar.
# You can use \= to commit, since the Derby team decided (why???)
# not to implement the SQL standard statement "commit"!!
# Note that SqlTool can not shut down an embedded Derby database properly,
# since that requires an additional SQL connection just for that purpose.
# However, I've never lost data by not shutting it down properly.
# Other than not supporting this quirk of Derby, SqlTool is miles ahead of ij.

You can put this file anywhere you want to, and specify the location to SqlTool/DatabaseManager/DatabaseManagerSwing by using the --rcfile argument. If there is no reason to not use the default location (and there are situations where you would not want to), then use the default location and you won't have to give --rcfile arguments to SqlTool/DatabaseManager/DatabaseManagerSwing. The default location is sqltool.rc or dbmanager.rc in your home directory (corresponding to the program using it). If you have any doubt about where your home directory is, just run SqlTool with a phony urlid and it will tell you where it expects the configuration file to be.

    java -jar $HSQLDB_HOME/lib/hsqldb.jar x

The config file consists of stanza(s) like this:

    urlid web
    url jdbc:hsqldb:hsql://localhost
    username web
    password webspassword

These four settings are required for every urlid. (There are optional settings also, which are described a couple paragraphs down). The URL may contain JDBC connection properties. You can have as many blank lines and comments like

    # This comment

in the file as you like. The whole point is that the urlid that you give in your SqlTool/DatabaseManager command must match a urlid in your configuration file.


Use whatever facilities are at your disposal to protect your configuration file.

It should be readable, both locally and remotely, only to users who run programs that need it. On UNIX, this is easily accomplished by using chmod/chown commands and making sure that it is protected from anonymous remote access (like via NFS, FTP or Samba).

You can also put the following optional settings into a urlid stanza. The setting will, of course, only apply to that urlid.

This is used by the SqlTool program, but not by the DatabaseManager programs. See the Character Encoding section of the Non-Interactive section. You can, alternatively, set this for one SqlTool invocation by setting the system property sqlfile.charset . Defaults to US-ASCII.
Sets the JDBC driver class name. You can, alternatively, set this for one SqlTool/DatabaseManager invocation by using the command line switch --driver. Defaults to org.hsqldb.jdbcDriver.
TLS trust keystore store file path as documented in the TLS chapter. You usually only need to set this if the server is using a non-publicly-certified certificate (like a self-signed self-ca'd cert).

Property and SqlTool command-line switches override settings made in the configuration file.

Using Inline RC Authentication

Inline RC authentication setup is accomplished by using the --inlineRc command-line switch on SqlTool. The --inlineRc command-line switch takes a comma-separated list of key/value elements. The url and user elements are required. The rest are optional.

The JDBC URL of the database you wish to connect to.
The username to connect to the database as.
Sets the character encoding. Defaults to US-ASCII.
The TLS trust keystore file path as documented in the TLS chapter.
You may only use this element to set empty password, like
. For any other password value, omit the password element and you will be prompted for the value.

(Use the --driver switch instead of --inlineRc to specify a JDBC driver class). Here is an example of invoking SqlTool to connect to a standalone database.

    java -jar $HSQLDB_HOME/lib/hsqldb.jar
         --inlineRc URL=jdbc:hsqldb:file:/home/dan/dandb,USER=dan

For security reasons, you cannot specify a non-empty password as an argument. You will be prompted for a password as part of the login process.

Using the current version of SqlTool with an older HSQLDB distribution.

This procedure will allow users of a legacy version of HSQLDB to use all of the new features of SqlTool. You will also get the new versions of the DatabaseManagers! This procedure works for distros going back to at least, probably much farther.

Follow the instructions in the See the Using hsqltool.jar and hsqldbutil.jar section to build the jar file hsqldbutil.jar.

For now on, whenever you are going to run SqlTool, make sure that you have this hsqldbutil.jar as the first item in your CLASSPATH. You can't run SqlTool with the "-jar" switch (because the -jar switch doesn't permit setting your own class path).

Here's a UNIX example where somebody wants to use the new SqlTool with their older HSQLDB database, as well as with Postgresql and a local application.

java org.hsqldb.util.SqlTool urlid

Interactive Usage

Do read the The Bare Minimum section before you read this section.

You run SqlTool interactively by specifying no SQL filepaths on the SqlTool command line. Like this.

    java -jar $HSQLDB_HOME/lib/hsqldb.jar urlid

Procedure 8.3. What happens when SqlTool is run interactively (using all default settings)

  1. SqlTool starts up and connects to the specified database, using your SqlTool configuration file (as explained in the RC File Authentication Setup section).

  2. SQL file auto.sql in your home directory is executed (if there is one),

  3. SqlTool displays a banner showing the SqlTool and SqlFile version numbers and describes the different command types that you can give, as well as commands to list all of the specific commands available to you.

You exit your session by using the "\q" special command or ending input (like with Ctrl-D or Ctrl-Z).


Every command (regardless of type) and comment must begin at the beginning of a line (or immediately after a comment ends with "*/").

You can't nest commands or comments. You can only start new commands (and comments) after the preceding statement has been terminated. (Remember that if you're running SqlTool interactively, you can terminate an SQL statement without executing it by entering a blank line).

(Special Commands, Edit Buffer Commands and PL Commands always consist of just one line. Any of these commands or comments may be preceded by space characters.)

These rules do not apply at all to Raw Mode. Raw mode is for use by advanced users when they want to completely bypass SqlTool processing in order to enter a chunk of text for direct transmission to the database engine.

Command Types

When you are typing into SqlTool, you are always typing part of the immediate command. You execute the immediate command by hitting ENTER after a semi-colon (for SQL commands) or by just hitting ENTER (after any other non-empty command-- see next section about this distinction). The interactive : commands can perform actions with or on the edit buffer. The edit buffer usually contains a copy of the last command executed, and you can always view it with the :b command. If you never use any : commands, you can entirely ignore the edit buffer. If you want to repeat commands or edit previous commands, you will need to work with the edit buffer. The immediate command contains whatever (and exactly what) you type. The command history and edit buffer may contain any type of command other than comments and : commands (i.e., : commands and comments are just not copied to the history or to the edit buffer).

Hopefully an example will clarify the difference between the immediate command and the edit buffer. If you type in the edit buffer Substitution command ":s/tbl/table/", the :s command that you typed is the immediate command (and it will never be stored to the edit buffer or history, since it is a : command), but the purpose of the substitution command is to modify the contents of the edit buffer (perform a substitution on it)-- the goal being that after your substitutions you would execute the buffer with the ":;" command. The ":a" command is special in that when you hit ENTER to execute it, it copies the contents of the edit buffer to a new immediate command and leaves you in a state where you are appending to that immediate command (nearly) exactly as if you had just typed it in.

Command Types

Command types


Above, we said that if you enter an SQL command, one SQL command corresponds to one SqlTool command. This is the most typical usage, however, you can actually put multiple SQL statements into one SQL command. One example would be

This is one SqlTool command containing two SQL statements. See the Chunking section to see why you may want to chunk SQL commands, how, and the implications.

SQL Statement

Any command that you enter which does not begin with "\", ":", or "* " is an SQL Statement. The command is not terminated when you hit ENTER, like most OS shells. You terminate SQL Statements with either ";" at the end of a line, or with a blank line. In the former case, the SQL Statement will be executed against the SQL database and the command will go into the edit buffer and SQL command history for editing or viewing later on. In the former case, execute against the SQL database means to transmit the SQL text to the database engine for execution. In the latter case (you end an SQL Statement with a blank line), the command will go to the edit buffer and SQL history, but will not be executed (but you can execute it later from the edit buffer). (See the note immediately above about multiple SQL statements in one SqlTool command).

(Blank lines are only interpreted this way when SqlTool is run interactively. In SQL files, blank lines inside of SQL statements remain part of the SQL statement).

As a result of these termination rules, whenever you are entering text that is not a Special Command, Edit Buffer / History Command, or PL Command, you are always appending lines to an SQL Statement or comment. (In the case of the first line, you will be appending to an empty SQL statement. I.e. you will be starting a new SQL Statement or comment).

Special Command
Run the command "\?" to list the Special Commands. All of the Special Commands begin with "\". I'll describe some of the most useful Special Commands below.
Edit Buffer / History Command
Run the command ":?" to list the Edit-Buffer/History Commands. All of these commands begin with ":". These commands use commands from the command history, or operate upon the edit "buffer", so that you can edit and/or (re-)execute previously entered commands.
PL Command

Procedural Langage commands. Run the command "*?" to list the PL Commands. All of the PL Commands begin with "*". PL commands are for setting and using scripting variables and conditional and flow control statements like * if and * while. A few PL features (such as PL aliases and updating and selecing data directly from/to files) can be a real convenience for nearly all users, so these features will be discussed briefly in this section. More detailed explanation of PL variables and the other PL features, with examples, are covered in the SqlTool Procedural Language section.

Raw Mode
The descriptions of command-types above do not apply to Raw Mode. In raw mode, SqlTool doesn't interpret what you type at all. It all just goes into the edit buffer which you can send to the database engine. Beginners can safely ignore raw mode. You will never encounter it unless you run the "\." special command, or enter a PL/SQL command. See the Raw Mode section for the details.

Special Commands

Essential Special Commands

\i path/to/script.sql
execute the specified SQL script, then continue again interactively.
commit the current SQL transaction. Most users are used to typing the SQL statement commit;, but this command is crucial for those databases which don't support the statement. It's obviously unnecessary if you have auto-commit mode on.
List a summary of DSV eXporting, and all available DSV options.
List a summary of DSV iMporting, and all available DSV options.
List a summary of the \d commands below.
\dt [filter_substring]
\dv [filter_substring]
\ds [filter_substring]
\di [table_name]
\dS [filter_substring]
\da [filter_substring]
\dn [filter_substring]
\du [filter_substring]
\dr [filter_substring]
\d* [filter_substring]

Lists available objects of the given type.

  • t: non-system Tables
  • v: Views
  • s: Sequences
  • i: Indexes
  • S: System tables
  • a: Aliases
  • n: schema Names
  • u: database Users
  • r: Roles
  • *: all table-like objects
If your database supports schemas, then the schema name will also be listed.

If you supply an optional filter substring, then only items which contain the given substring (in the object name or schema name) will be listed.


The substring test is case-sensitive! Even though in SQL queries and for the "\d objectname" command object names are usually case-insensitive, for the \dX commands, you must capitalize the filter substring exactly as it will appear in the special command output. This is an inconvenience, since the database engine will change names in SQL to default case unless you double-quote the name, but that is server-side functionality which cannot (portably) be reproduced by SqlTool. You can use spaces and other special characters in the string.


Filter substrings ending with "." are special. If a substring ends with ".", then this means to narrow the search by the exact, case-sensitive schema name given. For example, if I run "\d* BLAINE.", this will list all table-like database objects in the "BLAINE" schema. The capitalization of the schema must be exactly the same as how the schema name is listed by the "\dn" command. You can use spaces and other special characters in the string. (I.e., enter the name exactly how you would enter it inside of double-quotes in an SQL command). This is an inconvenience, since the database engine will change names in SQL to default case unless you double-quote the name, but that is server-side functionality which cannot (portably) be reproduced by SqlTool.


Indexes may not be searched for by substring, only by exact target table name. So if I1 is an index on table T1, then you list this index by running "\di T1". In addition, many database vendors will report on indexes only if a target table is identified. Therefore, "\di" with no argument will fail if your database vendor does not support it.

\d objectname [filter]

Lists names of columns in the specified table or view. objectname may be a base table name or a schema.object name.

If you supply a filter string, then only columns with a name containing the given filter will be listed. The objectname is nearly always case-insensitive (depends on your database), but the filter is always case-sensitive. You'll find this filter is a great convenience compared to other database utilities, where you have to list all columns of large tables when you are only interested in one of them.


When working with real data (as opposed to learning or playing), I often find it useful to run two SqlTool sessions in two side-by-side terminal emulator windows. I do all of my real work in one window, and use the other mostly for \d commands. This way I can refer to the data dictionary while writing SQL commands, without having to scroll.

This list here includes only the essential Special Commands, but n.b. that there are other useful Special Commands which you can list by running \?. (You can, for example, execute SQL from external SQL files, and save your interactive SQL commands to files). Some specifics of these other commands are specified immediately below, and the Generating Text or HTML Reports section explains how to use the "\o" and "\H" special commands to generate reports.

Be aware that the \! Special Command does not work for external programs that read from standard input. You can invoke non-interactive and graphical interactive programs, but not command-line interactive programs.

SqlTool executes \! programs directly, it does not run an operating system shell (this is to avoid OS-specific code in SqlTool). Because of this, you can give as many command-line arguments as you wish, but you can't use shell wildcards or redirection.

The \w command can be used to store any command in your SQL history to a file. Just restore the command to the edit buffer with a command like "\-4" before you give the \w command.

Edit Buffer / History Commands

Edit Buffer / History Commands

List the current contents of the edit buffer.
Shows the Command History. For each command which has been executed (up to the max history length), the SQL command history will show the command; its command number (#); and also how many commands back it is (as a negative number). : commands are never added to the history list. You can then use either form of the command identifier to recall a command to the edit buffer (the command described next) or as the target of any of the following : commands. This last is accomplished in a manner very similar to the vi editor. You specify the target command number between the colon and the command. As an example, if you gave the command :s/X/Y/, that would perform the substitution on the contents of the edit buffer; but if you gave the command :-3 s/X/Y/, that would perform the substitution on the command 3 back in the command history (and copy the output to the edit buffer). Also, just like vi, you can identify the command to recall by using a regular expression inside of slashes, like :/blue/ s/X/Y/ to operate on the last command you ran which contains "blue".
:13 OR :-2 OR :/blue/

Recalls a command from Command history to the edit buffer. Enter ":" followed by the positive command number from Command history, like ":13"... or ":" followed by a negative number like ":-2" for two commands back in the Command history... or ":" followed by a regular expression inside slashes, like ":/blue/" to recall the last command which contains "blue". The specified command will be written to the edit buffer so that you can execute it or edit it using the commands below.

As described under the :h command immediately above, you can follow the command number here with any of the commands below to perform the given operation on the specified command from history instead of on the edit buffer contents. So, for example, ":4;" would load command 4 from history then execute it (see the ":;" command below).

Executes the SQL, Special or PL statement in the edit buffer (by default). This is an extremely useful command. It's easy to remember because it consists of ":", meaning Edit Buffer Command, plus a line-terminating ";", (which generally means to execute an SQL statement, though in this case it will also execute a special or PL command).

Enter append mode with the contents of the edit buffer (by default) as the current command. When you hit ENTER, things will be exactly as if you physically re-typed the command that is in the edit buffer. Whatever lines you type next will be appended to the immediate command. As always, you then have the choice of hitting ENTER to execute a Special or PL command, entering a blank line to store back to the edit buffer, or end a SQL statement with semi-colon and ENTER to execute it.

You can, optionally, put a string after the :a, in which case things will be exactly as just described except the additional text will also be appended to the new immediate command. If you put a string after the :a which ends with ;, then the resultant new immediate command will just be executed right away, as if you typed in and entered the entire thing.

If your edit buffer contains SELECT x FROM mytab and you run a:le, the resultant command will be SELECT x FROM mytable. If your edit buffer contains SELECT x FROM mytab and you run a: ORDER BY y, the resultant command will be SELECT x FROM mytab ORDER BY y. Notice that in the latter case the append text begins with a space character.

:s/from regex/to string/switches

The Substitution Command is the primary method for SqlTool command editing-- it operates upon the current edit buffer by default. The "to string" and the "switches" are both optional (though the final "/" is not). To start with, I'll discuss the use and behavior if you don't supply any substitution mode switches.

Don't use "/" if it occurs in either "from string" or "to string". You can use any character that you want in place of "/", but it must not occur in the from or to strings. Example

    :s@from string@to string@

The to string is substituted for the first occurrence of the (case-specific) from string. The replacement will consider the entire SQL statement, even if it is a multi-line statement.

In the example above, the from regex was a plain string, but it is interpreted as a regular expression so you can do all kinds of powerful substitutions. See the perlre man page, or the java.util.regex.Pattern API Spec for everything you need to know about extended regular expressions.

Don't end a to string with ";" in attempt to make a command execute. There is a substitution mode switch to use for that purpose.

You can use any combination of the substitution mode switches.

  • Use "i" to make the searches for from regex case insensitive.

  • Use "g" to substitute Globally, i.e., to subsitute all occurrences of the from regex instead of only the first occurrence found.

  • Use ";" to execute the command immediately after the substitution is performed.

  • Use "m" for ^ and $ to match each line-break in a multi-line edit buffer, instead of just at the very beginning and every end of the entire buffer.

If you specify a command number (from the command history), you end up with a feature very reminiscent of vi, but even more powerful, since the Perl/Java regular expression are a superset of the vi regular expressions. As an example,

                    :24 s/pin/needle/g;
would start with command number 24 from command history, substitute "needle" for all occurrences of "pin", then execute the result of that substitution (and this final statement will of course be copied to the edit buffer and to command history).

:w /path/to/file.sql
This appends the contents of the current buffer (by default) to the specified file. Since what is being written are Special, PL, or SQL commands, you are effectively creating an SQL script.

I find the ":/regex/" and ":/regex/;" constructs particularly handy for every-day usage.

re-executes the last \d command that you gave (The extra "\" is needed to escape the special meaning of "\" in regular expressions). It's great to be able to recall and execute the last "insert" command, for example, without needing to check the history or keep track of how many commands back it was. To re-execute the last insert command, just run ":/insert/;". If you want to be safe about it, do it in two steps to verify that you didn't accidentally recall some other command which happened to contain the string "insert", like
(Executing the last only if you are satisfied when SqlTool reports what command it restored). Often, of course, you will want to change the command before re-executing, and that's when you combine the :s and :a commands.

We'll finish up with a couple fine points about Edit/Buffer commands. You generally can't use PL variables in Edit/Buffer commands, to eliminate possible ambiguities and complexities when modifying commands. The :w command is an exception to this rule, since it can be useful to use variables to determine the output file, and this command does not do any "editing".

The :? help explains how you can change the default regular expression matching behavior (case sensitivity, etc.), but you can always use syntax like "(?i)" inside of your regular expression, as described in the Java API spec for class java.util.regex.Pattern, found here. History-command-matching with the /regex/ construct is purposefully liberal, matching any portion of the command, case sensitive, etc., but you can still use the method just described to modify this behavior. In this case, you could use "(?-i)" at the beginning of your regular expression to be case-sensitive.

PL Commands

Essential PL Command

* VARNAME = value

Set the value of a variable. If the variable doesn't exist yet, it will be created. The most common use for this is so that you can later use it in SQL statements, print statements, and PL conditionals, by using the *{VARNAME} or *{:VARNAME} construct. The only difference between *{VARNAME} and *{:VARNAME} is that the former produces an error if VARNAME is not set, whereas the latter will expand to a zero-length string if VARNAME is not set.

If you set a variable to an SQL statement (without the terminating ";") you can then use it as a PL alias like /VARNAME, as shown in this example.

Example 8.2. Defining and using a PL alias (PL variable)

    * qry = SELECT COUNT(*) FROM mytable
    \p The stored query is '*{qry}'
    /qry WHERE mass > 200;

If you put variable definitions into the SQL file auto.sql in your home directory, those aliases/variables will always be available for interactive use.

PL variables can be expanded within all commands other than : edit/history commands.

* load VARNAME /file/path.txt
Sets VARNAME to the content of the specified ASCII file.
* prepare VARNAME
Indicate that next command should be a SQL INSERT or UPDATE command containing one question mark. The value of VARNAME will be substuted for the ? variable. This does work for CLOB columns.
When next SQL command is run, instead of displaying the rows, just store the very first column value to variable VARNAME. This works for CLOB columns too. It also works with Oracle XML type columns if you use column labels and the getclobval function.
Exactly the same as
except that the fetched results will be displayed in addition to setting the variable.
* dump VARNAME /file/path.txt
Store the value of VARNAME to the specified ASCII file.

? Variable

You don't set the ? variable. It is just like the Bourne shell variable ? in that it is always automatically set to the first value of a result set (or the return value of other SQL commands). It works just like the

command described above, but it all happens automatically. You can, of course, dereference ? like any PL variable, but it does not list with the
commands. You can see the value whenever you want by running
    \p  *{?}

Note that PL commands are used to upload and download column values to/from local ASCII files, but the corresponding actions for binary files use the special \b commands. This is because PL variables are used for ASCII values and you can store any number of column values in PL variables. This is not true for binary column values. The \b commands work with a single binary byte buffer.

See the SqlTool Procedural Language section below for information on using variables in other ways, and information on the other PL commands and features.

Storing and retrieving binary files

You can upload binary files such as photographs, audio files, or serialized Java objects into database columns. SqlTool keeps one binary buffer which you can load from files with the \bl command, or from a database query by doing a one-row query for any non-displayable type (including BLOB, OBJECT, and OTHER). In the latter case, the data returned for the first non-displayable column of the first result row will be stored into the binary buffer.

Once you have data in the binary buffer, you can upload it to a database column (including BLOB, OBJECT, and OTHER type columns), or save it to a file. The former is accomplished by the special command \bp followed by a prepared SQL query containing one question mark place-holder to indicate where the data gets inserted. The latter is accomplished with the \bd command.

You can also store the output from normal, displayable column into the binary buffer by using the special command \b. The very first column value from the first result row of the next SQL command will be stored to the binary byte buffer.

Example 8.3. Inserting binary data into database from a file

    \bl /tmp/favoritesong.mp3
    INSERT INTO musictbl (id, stream) VALUES(3112, ?);

Example 8.4. Downloading binary data from database to a file

    SELECT stream FROM musictbl WHERE id = 3112;
    \bd /tmp/favoritesong.mp3

You can also store and retrieve text column values to/from ASCII files, as documented in the Essential PL Command section.

Command History

The SQL history shown by the \h command, and used by other commands, is truncated to 100 entries, since its utility comes from being able to quickly view the history list. You can change the history length by setting the system property sqltool.historyLength to an integer like

java -Dsqltool.historyLength=100 -jar $HSQLDB_HOME/lib/hsqldb.jar urlid
If there is any demand, I'll make the setting of this value more convenient.

The SQL history list contains all executed commands other than Edit Buffer commands and comments, even if the command has a syntax error or fails upon execution. The reason for including bad commands is so that you can recall and fix them if you wish to. The same applies to the edit buffer. If you copy a command to the edit buffer by entering blank line, or if you edit the edit buffer, that edit buffer value will never make it into the command history until and if you execute it.

Shell scripting and command-line piping

You normally use non-interactive mode for input piping. You specify "-" as the SQL file name. See the Piping and shell scripting subsection of the Non-Interactive chapter.

Emulating Non-Interactive mode

You can run SqlTool interactively, but have SqlTool behave exactly as if it were processing an SQL file (i.e., no command-line prompts, error-handling that defaults to fail-upon-error, etc.). Just specify "-" as the SQL file name in the command line. This is a good way to test what SqlTool will do when it encounters any specific command in an SQL file. See the Piping and shell scripting subsection of the Non-Interactive chapter for an example.


Read the Interactive Usage section if you have not already, because much of what is in this section builds upon that. You can skip all discussion about Command History and the edit buffer if you will not use those interactive features.


If you're doing data updates, remember to issue a commit command or use the --autoCommit switch.

As you'll see, SqlTool has many features that are very convenient for scripting. But what really makes it superior for automation tasks (as compared to SQL tools from other vendors) is the ability to reliably detect errors and to control JDBC transactions. SqlTool is designed so that you can reliably determine if errors occurred within SQL scripts themselves, and from the invoking environment (for example, from a perl, Bash, or Python script, or a simple cron tab invocation).

Giving SQL on the Command Line

If you just have a couple Commands to run, you can run them directly from the comand-line or from a shell script without an SQL file, like this.

    java -jar $HSQLDB_HOME/lib/hsqldb.jar --sql 'SQL statement;' urlid


The --sql automatically implies --noinput, so if you want to execute the specified SQL before and in addition to an interactive session (or stdin piping), then you must also give the --stdinput switch.

Since SqlTool transmits SQL statements to the database engine only when a line is terminated with ";", if you want feedback from multiple SQL statements in an --sql expression, you will need to use functionality of your OS shell to include linebreaks after the semicolons in the expression. With any Bourne-compatible shell, you can include linebreaks in the SQL statements like this.

    java -jar $HSQLDB_HOME/lib/hsqldb.jar --sql '
        SQL statement number one;
        SQL statement
            number two;
        SQL statement three;
    ' urlid
If you don't need feedback, just separate the SQL commands with semicolons and the entire expression will be chunked.

The --sql switch is very useful for setting shell variables to the output of SQL Statements, like this.

    # A shell script
    USERCOUNT=`java -jar $HSQLDB_HOME/lib/hsqldb.jar --sql '
        select count(*) from usertbl;
    ' urlid` || {
        # Handle the SqlTool error
    echo "There are $USERCOUNT users registered in the database."
    [ "$USECOUNT" -gt 3 ] && {   # If there are more than 3 users registered
        # Some conditional shell scripting

SQL Files

Just give paths to sql text file(s) on the command line after the urlid.

Often, you will want to redirect output to a file, like

java -jar $HSQLDB_HOME/lib/hsqldb.jar sql... > /tmp/log.sql 2>&1

(Skip the "2>&1" if you're on Windows).

You can also execute SQL files from an interactive session with the "\i"' Special Command, but be aware that the default behavior in an interactive session is to continue upon errors. If the SQL file was written without any concern for error handling, then the file will continue to execute after errors occur. You could run \c false before \i filename, but then your SqlTool session will exit if an error is encountered in the SQL file. If you have an SQL file without error handling, and you want to abort that file when an error occurs, but not exit SqlTool, the easiest way to accomplish this is usually to add \c false to the top of the script.

If you specify multiple SQL files on the command-line, the default behavior is to exit SqlTool immediately if any of the SQL files encounters an error.

SQL files themselves have ultimate control over error handling. Regardless of what command-line options are set, or what commands you give interactively, if a SQL file gives error handling statements, they will take precedence.

You can also use \i in SQL files. This results in nested SQL files.

You can use the following SQL file, sample.sql, which resides in the sample directory of your HSQLDB distribution [1]. It contains SQL as well as Special Commands making good use of most of the Special Commands documented below.

    $Id: sample.sql,v 1.5 2005/05/02 15:07:27 unsaved Exp $
    Examplifies use of SqlTool.
    PCTASK Table creation

/* Ignore error for these two statements */
\c true
DROP TABLE pctasklist;
DROP TABLE pctask;
\c false

\p Creating table pctask
    id integer identity,
    name varchar(40),
    description varchar,
    url varchar,
    UNIQUE (name)

\p Creating table pctasklist
CREATE TABLE pctasklist (
    id integer identity,
    host varchar(20) not null,
    tasksequence int not null,
    pctask integer,
    assigndate timestamp default current_timestamp,
    completedate timestamp,
    show bit default true,
    FOREIGN KEY (pctask) REFERENCES pctask,
    UNIQUE (host, tasksequence)

\p Granting privileges
GRANT select ON pctask TO public;
GRANT all ON pctask TO tomcat;
GRANT select ON pctasklist TO public;
GRANT all ON pctasklist TO tomcat;

\p Inserting test records
INSERT INTO pctask (name, description, url) VALUES (
    'task one', 'Description for task 1', 'http://cnn.com');
INSERT INTO pctasklist (host, tasksequence, pctask) VALUES (
    'admc-masq', 101, SELECT id FROM pctask WHERE name = 'task one');


You can execute this SQL file with a Memory Only database with a command like

    java -jar $HSQLDB_HOME/lib/hsqldb.jar  --sql '
        create user tomcat password "x";
    ' mem path/to/hsqldb/sample/sample.sql

(The --sql "create...;" arguments create an account which the script uses). You should see error messages betwen the Continue-on-error...true and Continue-on-error...false. The script purposefully runs commands that might fail there. The reason the script does this is to perform database-independent conditional table removals. (The SQL clause IF EXISTS is more graceful and succinct, and should be used if you don't need to support databases which don't support IF EXISTS). If an error occurs when continue-on-error is false, the script would abort immedately.

Piping and shell scripting

You can of course, redirect output from SqlTool to a file or another program.

    java -jar $HSQLDB_HOME/lib/hsqldb.jar urlid file.sql > file.txt 2>&1

    java -jar $HSQLDB_HOME/lib/hsqldb.jar urlid file.sql 2>&1 | someprogram...

You can type commands in to SqlTool while being in non-interactive mode by supplying "-" as the file name. This is a good way to test how SqlTool will behave when processing your SQL files.

        java -jar $HSQLDB_HOME/lib/hsqldb.jar urlid -

This is how you have SqlTool read its input from another program:

Example 8.5. Piping input into SqlTool

        echo "Some SQL commands with '$VARIABLES';" |
        java -jar $HSQLDB_HOME/lib/hsqldb.jar urlid -

Make sure that you also read the Giving SQL on the Command Line section. The --sql switch is a great facility to use with shell scripts.

Optimally Compatible SQL Files

If you want your SQL scripts optimally compatible among other SQL tools, then don't use any Special or PL Commands. SqlTool has default behavior which I think is far superior to the other SQL tools, but you will have to disable these defaults in order to have optimally compatible behavior.

These switches provide compatibilty at the cost of poor control and error detection.

  • --continueOnErr

    The output will still contain error messages about everything that SqlTool doesn't like (malformatted commands, SQL command failures, empty SQL commands), but SqlTool will continue to run. Errors will not cause rollbacks (but that won't matter because of the following setting).

  • --autoCommit

You don't have to worry about accidental expansion of PL variables, since SqlTool will never expand PL variables if you don't set any variables on the command line, or give any "* " PL commands. (And you could not have "* " commands in a compatible SQL file).


SQL comments of the form /*...*/ must begin where a (SQL/Special/Edit-Buffer/PL) Command could begin, and they end with the very first "*/" (regardless of quotes, nesting, etc. You may have as many blank lines as you want inside of a comment.

Example 8.6. Valid comment example

    SELECT count(*) FROM atable;
    /* Lots of
     comments interspersed among
     several lines */   SELECT count(*)
    FROM btable;

Notice that a command can start immediate after the comment ends.

Example 8.7. Invalid comment example

    SELECT count(*) FROM
    /* atable */

This comment is invalid because you could not start another command at the comment location (because it is within an SQL Statement).

You can try using /*...*/ in other locations, and -- style SQL comments, but SqlTool will not treat them as comments. If they occur within an SQL Statment, SqlTool will pass them to the database engine, and the DB engine will determine whether to parse them as comments.

Special Commands and Edit Buffer Commands in SQL Files

Don't use Edit Buffer / History Commands in your sql files, because they won't work. Edit Buffer / History Commands are for interactive use only. (But, see the Raw Mode section for an exception). You can, of course, use any SqlTool command at all interactively. I just wanted to group together the commands most useful to script-writers.

\q [abort message]

Be aware that the \q command will cause SqlTool to completely exit. If a script x.sql has a \q command in it, then it doesn't matter if the script is executed like

    java -jar .../hsqldb.jar urlid a.sql x.sql z.sql
or if you use \i to read it in interactively, or if another SQL file uses \i to nest it. If \q is encountered, SqlTool will quit. See the SqlTool Procedural Language section for commands to abort an SQL file (or even parts of an SQL file) without causing SqlTool to exit.

\q takes an optional argument, which is an abort message. If you give an abort message, the message is displayed to the user and SqlTool will exit with a failure status. If you give no abort message, then SqlTool will exit quietly with successful status. As a result,

means to make an immediate but graceful exit, whereas
\q Message
means to abort immediately.

\p [text to print]
Print the given string to stdout. Just give "\p" alone to print a blank line.
\i /path/to/file.sql
Include another SQL file at this location. You can use this to nest SQL files. For database installation scripts I often have a master SQL file which includes all of the other SQL files in the correct sequence. Be aware that the current continue-upon-error behavior will apply to included files until such point as the SQL file runs its own error handling commands.
\o [file/path.txt]

Tee output to the specified file (or stop doing so). See the Generating Text or HTML Reports section.

A database-independent way to commit your SQL session.
\a [true|false]
This turns on and off SQL transaction autocommits. Auto-commit defaults to false, but you can change that behavior by using the --autoCommit command-line switch.
\c [true|false]

A "true" setting tells SqlTool to Continue when errors are encountered. The current transaction will not be rolled back upon SQL errors, so if \c is true, then run the ROLLCACK; command yourself if that's what you want to happen. The default for interactive use is to continue upon error, but the default for non-interactive use is to abort upon error. You can override this behavior by using the --continueOnErr or the --abortOnErr command-line switch.

With database setup scripts, I usually find it convenient to set "true" before dropping tables (so that things will continue if the tables aren't there), then set it back to false so that real errors are caught. DROP TABLE tablename IF EXISTS; is a more elegant, but less portable, way to accomplish the same thing.


It depends on what you want your SQL files to do, of course, but I usually want my SQL files to abort when an error is encountered, without necessarily killing the SqlTool session. If this is the behavior that you want, then put an explicit \c false at the top of your SQL file and turn on continue-upon-error only for sections where you really want to permit errors, or where you are using PL commands to handle errors manually. This will give the desired behavior whether your script is called by somebody interactively, from the SqlTool command-line, or included in another SQL file (i.e. nested).


The default settings are usually best for people who don't want to put in any explicit \c or error handling code at all. If you run SQL files from the SqlTool command line, then any errors will cause SqlTool to roll back and abort immediately. If you run SqlTool interactively and invoke SQL files with \i commands, the scripts will continue to run upon errors (and will not roll back). This behavior was chosen because there are lots of SQL files out there that produce errors which can be ignored; but we don't want to ignore errors that a user won't see. I reiterate that any and all of this behavior can (and often should) be changed by Special Commands run in your interactive shell or in the SQL files. Only you know whether errors in your SQL files can safely be ignored.


SqlTool is ideal for mission-critical automation because, unlike other SQL tools, SqlTool returns a dependable exit status and gives you control over error handling and SQL transactions. Autocommit is off by default, so you can build a completely dependable solution by intelligently using \c commands (Continue upon Errors) and commit statements, and by verifying exit statuses.

Using the SqlTool Procedural Language, you have ultimate control over program flow, and you can use variables for database input and output as well as for many other purposes. See the SqlTool Procedural Language section.

Getting Interactive Functionality with SQL Files

Some script developers may run into cases where they want to run with sql files but they alwo want SqlTool's interactive behavior. For example, they may want to do command recall in the sql file, or they may want to log SqlTool's command-line prompts (which are not printed in non-interactive mode). In this case, do not give the sql file(s) as an argument to SqlTool, but pipe them in instead, like

java -jar $HSQLDB_HOME/lib/hsqldb.jar urlid < filepath1.sql > /tmp/log.html 2>&1
cat filepath1.sql... |
java -jar $HSQLDB_HOME/lib/hsqldb.jar urlid > /tmp/log.html 2>&1

Character Encoding

SqlTool defaults to the US-ASCII character set (for reading). You can use another character set by setting the system property sqlfile.charset, like

java -Dsqlfile.charset=UTF-8 -jar $HSQLDB_HOME/lib/hsqldb.jar urlid file.sql...

You can also set this per urlid in the SqlTool configuration file. See the RC File Authentication Setup section about that.

Generating Text or HTML Reports

This section is about making a file containing the output of database queries. You can generate reports by using operating system facilities such as redirection, tee, and cutting and pasting. But it is much easier to use the "\o" and "\H" special commands.

Procedure 8.4. Writing query output to an external file

  1. By default, everthing will be done in plain text. If you want your report to be in HTML format, then give the special command \H. If you do so, you will probably want to use filenames with an suffix of ".html" or ".htm" instead of ".txt" in the next step.

  2. Run the command \o path/to/reportfile.txt. From this point on, output from your queries will be appended to the specified file. (I.e. another copy of the output is generated.) This way you can continue to monitor or use output as usual as the report is generated.

  3. When you want SqlTool to stop writing to the file, run \o (or just quit SqlTool if you have no other work to do).

  4. If you turned on HTML mode with \H before, you can run \H again to turn it back off, if you wish.

It is not just the output of "SELECT" statements that will make it into the report file, but

Kinds of output that get teed to \o files

  • Output of SELECT statements.
  • Output of all "\d" Special Commands. (I.e., "\dt", "\dv", etc., and "\d OBJECTNAME").
  • Output of "\p" Special Commands. You will want to use this to add titles, and perhaps spacing, for the output of individual queries.
Other output will go to your screen or stdout, but will not make it into the report file. Be aware that no error messages will go into the report file. If SqlTool is run non-interactively (including if you give any SQL file(s) on the command line), SqlTool will abort with an error status if errors are encountered. The right way to handle errors is to check the SqlTool exit status. (The described error-handling behavior can be modified with SqlTool command-line switches and Special Commands).


Remember that \o appends to the named file. If you want a new file, then use a new file name or remove the pre-existing target file ahead of time.


So that I don't end up with a bunch of junk in my report file, I usually leave \o off while I perfect my SQL. With \o off, I perfect the SQL query until it produces on my screen exactly what I want saved to file. At this point I turn on \o and run ":;" to repeat the last SQL command. If I have several complex queries to run, I turn \o off and repeat until I'm finished. (Every time you turn \o on, it will append to the file, just like we need).

Usually it doesn't come to mind that I need a wider screen until a query produces lines that are too long. In this case, stretch your window and repeat the last command with the ":;" Edit Buffer Command.

SqlTool Procedural Language

Aka PL

Most importantly, run SqlTool interactively and give the "*?" command to see what PL commands are available to you. I've tried to design the language features to be intuitive. Readers experience with significant shell scripting in any language can probably learn everything they need to know by looking at (and running!) the sample script sample/pl.sql in your HSQLDB distribution [1] and using the *? command from within an interactive SqlTool session as a reference. (By significant shell scripting, I mean to the extent of using variables, for loops, etc.).

PL variables will only be expanded after you run a PL command (or set variable(s) from the command-line). We only want to turn on variable expansion if the user wants variable expansion. People who don't use PL don't have to worry about strings getting accidentally expanded.

All other PL commands imply the "*" command, so you only need to use the "*" statement if your script uses PL variables and it is possible that no variables may be set before-hand (and no PL commands have been run previously). In this case, without "*", your script would silently use a literal value like "*{x}" instead of trying to expand it. With a preceding "*" command, PL will notice that the variable x has not been set and will generate an error. (If x had been set here will be no issue because setting a variable automatically turns on PL variable expansion).

PL is also used to upload and download column values to/from local ASCII files, analogously to the special \b commands for binary files. This is explained above in the Interactive Essential PL Command section above.


  • Use the * list command to list some or all variables; or * listvalues to also see the values.
  • You can set variables using the * VARNAME = value command. This document explains elsewhere how you can set a values to the contents of files, and to the return value of SQL statements and fetches.
  • You can also set variables using the --setvar command-line switch. I give a very brief but useful example of this below.
  • Variables are always expanded in SQL, Special, and PL commands if they are written like *{VARNAME} (assuming that a PL command has been run previously). Your SQL scripts can give good feedback by echoing the value of variables with the "\p" special command. Use the construct *{:VARNAME} to expand the variable, but to expand to a zero-length string instead of fail if VARNAME is not set.
  • A variable written like /VARNAME is expanded if it begins an SQL Statement. This usage is called PL Aliasing. See the PL Aliases section below.

  • Variables are normally written like *VARNAME in logical expressions to prevent them from being evaluated too early. See below about logical expressions.
  • You can't do math with expression variables, but you can get functionality like the traditional for (i = 0; i < x; i++) by appending to a variable and testing the string length, like

        * while (*i < ${x})
            * i = *{i}.
    i will be a growing line of dots.

  • Variable names must not contain white space, or the characters "}" or "=".

PL Aliases

PL Aliasing just means the use of a PL variable as the first thing in an SQL statement, with the shortcut notation /VARNAME.

/VARNAME must be followed by whitespace or terminate the Statement, in order for SqlFile to tell where the variable name ends.


Note that PL aliases are a very different thing from SQL aliases or HSQLDB aliases, which are features of databases, not SqlFile.

If the value of a variable is an entire SQL command, you generally do not want to include the terminating ";" in the value. There is an example of this above.

PL aliasing may only be used for SQL statements. You can define variables for everything in a Special or PL Command, except for the very first character ("\" or "*"). Therefore, you can use variables other than alias variables in Special and PL Commands. Here is a hyperbolically impractical example to show the extent to which PL variables can be used in Special commands even though you can not use them as PL aliases.

        sql> * qq = p Hello Butch
        sql> \*{qq} done now
        Hello Butch done now
(Note that the \* here is not the special command "\*", but is the special command "\p" because "*{qq}" resolves to "p").

Here is a short SQL file that gives the specified user write permissions on some application tables.

Example 8.8. Simple SQL file using PL


       Run SqlTool like this:
           java -jar path/to/hsqldb.jar -setvar USER=debbie grantwrite.sql

    /* Explicitly turn on PL variable expansion, in case no variables have
       been set yet.  (Only the case if user did not set USER).

    GRANT all ON book TO *{USER};
    GRANT all ON category TO *{USER};

Note that this script will work for any (existing) user just by supplying a different user name on the command-line. I.e., no need to modify the tested and proven script. There is no need for a commit statement in this SQL file since no DML is done. If the script is accidentally run without setting the USER variable, SqlTool will give a very clear notificaton of that.

The purpose of the plain "*" command is just so that the *{USER} variables will be expanded. (This would not be necessary if the USER variable, or any other variable, were set, but we don't want to depend upon that).

Logical Expressions

Logical expressions occur only inside of logical expression parentheses in PL statements. For example, if (*var1 > astring) and while (*checkvar). (The parentheses after "foreach" do not enclose a logical expression, they just enclose a list).

There is a critical difference between *{VARNAME} and *VARNAME inside logical expressions. *{VARNAME} is expanded one time when the parser first encounters the logical expression. *VARNAME is re-expanded every time that the expression is evaluated. So, you would never want to code * while (*{X} < 5) because the statement will always be true or always be false. (I.e. the following block will loop infinitely or will never run).

Don't use quotes or whitespace of any kind in *{VARNAME} variables in expressions. (They would expand and then the expression would most likely no longer be a valid expression as listed in the table below). Quotes and whitespace are fine in *VARNAME variables, but it is the entire value that will be used in evaluations, regardless of whether quotes match up, etc. I.e. quotes and whitespace are not special to the token evaluator.

Logical Operators

The token may be a literal, a *{VARNAME} which is expanded early, or a *VARNAME which is expanded late. (You usually do not want to use *{VARNAME} in logical expressions). False if the token is not set, empty, or "0". True otherwise.
True if the two tokens are equivalent "strings".
True if the TOKEN1 string is longer than TOKEN2 or is the same length but is greater according to a string sort.
Similarly to TOKEN1 > TOKEN2.
Logical negation of any of the expressions listed above.

*VARNAMEs in logical expressions, where the VARNAME variable is not set, evaluate to an empty string. Therefore (*UNSETVAR = 0) would be false, even though (*UNSETVAR) by itself is false and (0) by itself is false. Another way of saying this is that *VARNAME in a logical expression is equivalent to *{:VARNAME} out of a logical expression.

When developing scripts, you definitely use SqlTool interactively to verify that SqlTool evaluates logical expressions as you expect. Just run * if commands that print something (i.e. \p) if the test expression is true.

Flow Control

Flow control works by conditionally executing blocks of Commands according to conditions specified by logical expressions.

The conditionally executed blocks are called PL Blocks. These PL Blocks always occur between a PL flow control statement (like * foreach, *while, * if) and a corresponding * end PL Command (like * end foreach).


Be aware that the PL block reader is ignorant about SQL statements and comments when looking for the end of the block. It just looks for lines beginning with some specific PL commands. Therefore, if you put a comment line before a PL statement, or if a line of a multi-line SQL statement has a line beginning with a PL command, things may break.

I am not saying that you shouldn't use PL commands or SQL commands inside of PL blocks-- you definitely should! I'm saying that in PL blocks you should not have lines inside of SQL statments or comments which could be mistaken for PL commands. (Especially, "commenting out" PL end statements will not work if you leave * end at the beginning of the line).

(This limitation will very likely be removed in a future version of SqlTool).

The values of control variables for foreach and while PL blocks will change as expected.

There are * break and * continue, which work as any shell scripter would expect them to. The * break command can also be used to quit the current SQL file without triggering any error processing. (I.e. processing will continue with the next line in the including SQL file or interactive session, or with the next SQL file if you supplied multiple on the command-line).

Below is an example SQL File that shows how to use most PL features. If you have a question about how to use a particular PL feature, check this example before asking for help. This file resides in the sample directory with the name pl.sql [1]. Definitely give it a run, like

java -jar $HSQLDB_HOME/lib/hsqldb.jar mem $HSQLDB_HOME/pl.jar

Example 8.9. SQL File showing use of most PL features

    $Id: pl.sql,v 1.4 2005/05/02 15:07:26 unsaved Exp $
    SQL File to illustrate the use of SqlTool PL features.
    Invoke like
        java -jar .../hsqldb.jar .../pl.sql mem
                                                         -- blaine

* if (! *MYTABLE)
    \p MYTABLE variable not set!
    /* You could use \q to Quit SqlTool, but it's often better to just
       break out of the current SQL file.
       If people invoke your script from SqlTool interactively (with
       \i yourscriptname.sql) any \q will kill their SqlTool session. */
    \p Use arguments "--setvar MYTABLE=mytablename" for SqlTool
    * break
* end if

/* Turning on Continue-upon-errors so that we can check for errors ourselves.*/
\c true

\p Loading up a table named '*{MYTABLE}'...

/* This sets the PL variable 'retval' to the return status of the following
   SQL command */
* retval ~
    i int,
    s varchar
\p CREATE status is *{retval}

/* Validate our return status.  In logical expressions, unset variables like
   *unsetvar are equivalent to empty string, which is not equal to 0
   (though both do evaluate to false on their own, i.e. (*retval) is false
   and (0) is false */
* if (*retval != 0)
    \p Our CREATE TABLE command failed.
    * break
* end if

/* Default Continue-on-error behavior is what you usually want */
\c false

/* Insert data with a foreach loop.
   These values could be from a read of another table or from variables
   set on the command line like
\p Inserting some data int our new table (you should see 3 row update messages)
* foreach VALUE (12 22 24 15)
    * if (*VALUE > 23)
        \p Skipping *{VALUE} because it is greater than 23
        * continue
        \p YOU WILL NEVER SEE THIS LINE, because we just 'continued'.
    * end if
* end foreach

* themax ~
/* Can put Special Commands and comments between "* VARNAME ~" and the target 
   SQL statement. */
\p We're saving the max value for later.  You'll still see query output here:

/* This is usually unnecessary because if the SELECT failed, retval would
   be undefined and the following print statement would make SqlTool exit with
   a failure status */
* if (! *themax)
    \p Failed to get the max value.
    /* It's possible that the query succeeded but themax is "0".
       You can check for that if you need to. */
    * break
    \p YOU WILL NEVER SEE THIS LINE, because we just 'broke'.
* end if

\p ##############################################################
\p The results of our work:
\p MAX value is *{themax}

\p Everything worked.


We hereby call the ability to transmit multiple SQL commands to the database in one transmission chunking. Unless you are in Raw mode, SqlTool only transmits commands to the database engine when it reads in a ";" at the end of a line of an SQL command. Therefore, you normally want to end each and every SQL command with ";" at the end of a line. This is because the database can only send one status reply to each JDBC transmission. So, while you could run

SqlTool can only display the results from the last query. This is a limitation of the client/server nature of JDBC, and applies to any JDBC client. There are, however, situations where you don't need immediate feedback from every SQL command. For example,

Example 8.10. Single-line chunking example

It's useful because the output of the second SQL command will tell you whether the first SQL command succeeded. So, you won't miss the status output from the first command.


The first general reason to chunk SQL commands is performance. For standalone databases, the most common performance bottleneck is network latency. Chunking SQL commands can dramatically reduce network traffic.

The second general reason to chunk SQL commands is if your database requires you to send multiple commands in one transmission. This is often the case when you need to tell the database the SQL or PL/SQL commands that comprise a stored procedure, function, trigger, etc.


The most simple way is enter as many SQL commands as you want, but just do not end a line with ";" until you want the chunk to transmit.

Example 8.11. Multi-line chunking example

    ; SELECT * FROM t1;
If you list your command history with \s, you will see that all 3 SQL commands in 3 lines are in one SqlTool command. You can recall this SqlTool command from history to re-execute all three SQL commands.

The other method is by using Raw Mode. Go to the Raw Mode section to see how. You can enter any text at all, exactly how you want it to be sent to the database engine. Therefore, in addition to chunking SQL commands, you can give commands for non-SQL extensions to the database. For example, you could enter JavaScript code to be used in a stored procedure.

Raw Mode

You begin raw mode by issuing the Special Command "\.". You can then enter as much text in any format you want. When you are finished, enter a line consisting of only ".;" to store the input to the edit buffer and send it to the database server for execution.

This paragraph applies only to interactive usage. Interactive users may may end the raw input with ":." instead of ".;". This will just save the input to the edit buffer so that you can edit it and send it to the database manually. You can look at the edit buffer with the ":b" Buffer Command. You would normally use the command ":;" to send the buffer to the database after you are satisfied with it. You'll notice that your prompt will be the continuation prompt between entering "\." and terminating the raw input with ".;" or ":.".

Example 8.12. Interactive Raw Mode example

    sql> \.
    Enter RAW SQL.  No \, :, * commands.
    End with a line containing only ".;" to send to database,
    or ":." to store to edit buffer for editing or saving.
    raw> line one;
      +> line two;
      +> line three;
      +> :.
    Raw SQL chunk moved into buffer.  Run ":;" to execute the chunk.
    sql> :;
    Executing command from buffer:
    line one;
    line two;
    line three;

    SQL Error at 'stdin' line 13:
    "line one;
    line two;
    line three;"
    Unexpected token: LINE in statement [line]
The error message "Unexpected token: LINE in statement [line]" comes from the database engine, not SqlTool. All three lines were transmitted to the database engine.

Edit Buffer Commands are not available when running SqlTool non-interactively.



PL/SQL is not the same as PL. PL is the procedural language of SqlFile and is independent of your back-end database. PL commands always begin with *. PL/SQL is processed on the server side and you can only use it of your database supports it. You can not intermix PL and PL/SQL (except for setting a PL variable to the output of PL/SQL execution), because when you enter PL/SQL to SqlTool that input is not processed by SqlFile.

Use Raw Mode to send PL/SQL code blocks to the database engine. You do not need to enter the "\." command to enter raw mode. Just begin a new SqlTool command line with "DECLARE" or "BEGIN", and SqlTool will automatically put you into raw mode. See the Raw Mode section for details.

The following sample SQL file resides at sample/plsql.sql in your HSQLDB distribution [1]. This script will only work if your database engine supports standard PL/SQL, if you have permission to create the table "T1" in the default schema, and if that object does not already exist.

Example 8.13. PL/SQL Example

 * $Id: plsql.sql,v 1.4 2007/08/09 03:22:21 unsaved Exp $
 * This example is copied from the "Simple Programs in PL/SQL"
 * example by Yu-May Chang, Jeff Ullman, Prof. Jennifer Widom at
 * the Standord University Database Group's page
 * http://www-db.stanford.edu/~ullman/fcdb/oracle/or-plsql.html .
 * I have only removed some blank lines (in case somebody wants to
 * copy this code interactively-- because you can't use blank
 * lines inside of SQL commands in non-raw mode SqlTool when running
 * it interactively); and, at the bottom I have  replaced the
 * client-specific, non-standard command "run;" with SqlTool's
 * corresponding command ".;" and added a plain SQL SELECT command
 * to show whether the PL/SQL code worked.  - Blaine

    e INTEGER,




/* Above is plain SQL; below is the PL/SQL program. */

    a NUMBER;

    b NUMBER;


    SELECT e,f INTO a,b FROM T1 WHERE e>1;



/** The statement on the previous line, ".;" is SqlTool specific.
 *  This command says to save the input up to this point to the
 *  edit buffer and send it to the database server for execution.
 *  I added the SELECT statement below to give imm

/* This should show 3 rows, one containing values 4 and 2 (in this order)...*/
Note that, inside of raw mode, you can use any kind of formatting you want: Whatever you enter-- blank lines, comments, everything-- will be transmitted to the database engine.

Using hsqltool.jar and hsqldbutil.jar

This section is for those users who want to use SqlTool but without the overhead of hsqldb.jar (or who want to use a new SqlTool build with an older HSQLDB distribution).

If you do not need to directly use JDBC URLs like jdbc:hsqldb:mem: + something, jdbc:hsqldb:file: + something, or jdbc:hsqldb:res: + something, then you can use hsqltool.jar in place of the much larger hsqldb.jar file. hsqltool.jar will work for all JDBC databases other than HSQLDB Memory-only and In-process databases (the latter are fine if you access them via a HSQLB Server or WebServer). You will have to supply the JDBC driver for non-HSQLDB URLs, of course.

hsqltool.jar includes the HSQLDB JDBC driver. If you do not need to connect to HSQLDB databases at all, then hsqldbutil.jar is what you want. hsqldbutil.jar contains everything you need to run SqlTool and DatabaseManagerSwing against non-HSQLDB databases... well, besides the JDBC drivers for the target databases.

The HSQLDB distribution doesn't "come with" pre-built hsqltool.jar and hsqldbutil.jar files. You need to "build" them, but that is very easy to do.

These instructions assume that you are capable of running an Ant build. See the Building HSQLDB chapter if you need more details than what you see here.

  1. Download and extract a current HSQLDB distribution. If you don't want to use the source code, documentation, etc., you can use a temporary directory and remove it afterwards.

  2. Cd to the build directory under the root directory where you extracted the distribution to.

  3. Run ant hsqldbutil or ant hsqltool according to the criteria above. (If your goal is to use this jar with an older HSQLDB distribution, then you definitely need to build hsqlbutil.jar).

  4. If you're going to clean up afterwards, copy the jar that you built out of lib to a safe location first.

If you are using the HSQLDB JDBC driver (i.e., you're connecting up to a URL like jdbc:hsqldb:hsql + something or jdbc:hsqldb:http + something), you invoke SqlTool exactly as with hsqldb.jar except you use the file path to your new jar file instead of the path to hsqldb.jar.

If you are using a non-HSQLDB JDBC driver, you must set your CLASSPATH to include this new jar file and your JDBC driver, then run SqlTool like

        java org.hsqldb.util.SqlTool ...
You can specify your JDBC driver class either with the --driver switch to SqlTool, or in your RC file stanza (the last method is usually more convenient).

Delimiter-Separated-Value Imports and Exports


This feature is independent of HSQLDB Text Tables, a server-side feature of HSQLDB. It makes no difference to SqlTool whether the source or target table of your export/import is a memory, cache, or text table. Indeed, like all features of SqlTool, it works fine with other JDBC databases. It works great, for example to migrate data from a table of one type to a table of another type, or to another schema, or to another database instance, or to another database system.

This feature is what most business people call "CSV", but these files are more accurately called Delimiter Separated Value files because the delimiter is usually not a comma, and, more importantly, we purposefully choose an effective delimiter instead of the CSV method of using a delimiter which works in some cases and then use quotes and back-slashes to escape occurrence of the delimiter in the actual data. Just by choosing a delimiter which never needs escaping, we eliminate the whole mess, and the data in our files always looks just like the corresponding data in the database. To make this CSV / Delimiter-separated-value dintinction clear, I use the suffix ".dsv" for my data files. This leads me to stipulate the abbreviation DSV for the Delimiter Separated Value feature of HSQLDB.

Use the \x command to eXport a table to a DSV file, and the \m command to iMport a DSV file into a pre-existing table.

The row and column delimiters may be any String, not just a single character. And just as the delimiter capability is more general than traditional CSV delimiters, the export function is also more general than just a table data exporter. Besides the trivial generalization that you may specify a view or other virtual table name in place of a table name, you can alternatively export the output of any query which produces normal text output. A benefit to this approach is that it allows you to export only some columns of a table, and to specify a WHERE clause to narrow down the rows to be exported (or perform any other SQL transformation, mapping, join, etc.). One specific use for this would be to exclude columns of binary data (which can be exported by other means, such as a PL loop to store binary values to files with the \bd command).

Note that the import command will not create a new table. This is because of the impossibility of guessing appropriate types and constraints based only on column names and a data sampling (which is all that a DSV-importer has access to). Therefore, if you wish to populate a new table, create the table before running the import. The import file does not need to have data for all columns of a table. The only required columns are those required by database constraints (non-null, indexes, keys, etc.) One specific reason to omit columns is if you want values of some columns to be created automatically by column DEFAULT settings, triggers, HSQLDB identity sequences, etc. Another reason would be to skip binary columns.

Simple DSV exports and imports using default settings

Even if you need to change delimiters, table names, or file names from the defaults, I suggest that you run one export and import with default settings as a practice run. A memory-only HSQLDB instance is ideal for test runs like this.

This command exports the table icf.projects to the file projects.dsv in the current directory (where you invoked SqlTool from). By default, the output file name will be the specified source table name plus the extension .dsv.

Example 8.14. DSV Export Example

    SET SCHEMA icf;
    \x projects
We could also have run \x icf.projects (which would have created a file named icf.projects.dsv) instead of changing the session schema. In this example we have chosen to make the export file name independent of the schema to facilitate importing it into a different schema.

Take a look at the output file. Notice that the first line consists of column names, not data. This line is present because it will be needed if the file is to used for a DSV import. Notice the following characterstics about the export data. The column delimiter is the pipe character "|". The record delimiter is the default line delimiter character(s) for your operating system. The string used to represent database NULLs is [null]. See the next section for how to change these from their default values.

This command imports the data from the file projects.dsv in the current directory (where you invoked SqlTool from) into the table newschema.projects. By default, the output table name will be the input filename after removing optional leading directory and trailing final extension.

Example 8.15. DSV Import Example

    SET SCHEMA newschema;
    \m projects.dsv
If the DSV file was named with the target schema, you would have skipped the SET SCHEMA command, like \m newschema.projects.dsv.

Specifying queries and options

For a hands on example of a DSM import which generates an import report and uses some other options, change to directory HSQLDB/sample and play with the working script dsv-sample.sql [1]. You can execute it like

    java -jar ../lib/hsqldb.jar mem dsv-sample.sql
(assuming that you are using the supplied sqltool.rc file or have have urlid mem set up.

The header line in the DSV file is required at this time. (If there is user demand, it can be made optional for exporting, but it will remain required for importing).

Your export will fail if the column or record delimiter, or the null representation value occurs in the data being exported. You change these values by setting the PL variables *DSV_COL_DELIM, *DSV_ROW_DELIM, *DSV_NULL_REP. Notice that the asterisk is part of the variable names, to indicate that these variables are used by SqlTool internally. When specifying delimiters, you can use the escape seqpences \n, \r, \f, \t, \\, and decimal, octal or hex specifications like \20, \020, \0x20. For example, to change the column delimiter to the tab character, you would give the command

    * *DSV_COL_DELIM = \t

For imports, you must always specify the source DSV file path. If you want to export to a different file than one in the current directory named according to the source table, set the PL variable *DSV_TARGET_FILE, like

    * *DSV_TARGET_FILE = /tmp/dtbl.dsv

For exports, you must always specify the source table name or query. If you want to import to a table other than that derived from the input DSV file name, set the PL variable *DSV_TARGET_TABLE. The table name may contain a schema name prefix.

You don't need to import all of the columns in a data file. To designate the fields to be skipped, iether set the PL PL variable *DSV_SKIP_COLUMNS, or replace the column names in the header line to "-" (hyphen). The value of *DSV_SKIP_COLUMNS is case-insensitive, and multiple column names are separated with white space and/or commas.

You can specify a query instead of a tablename with the \x command in order to filter or transform data from a table or view, or to export the output of a join, etc. You must set the PL variable *DSV_TARGET_FILE, as explained above (since there is no table name from which to automatically map a file name).

Example 8.16. DSV Export of an Arbitrary SELECT Statement

    * *DSV_TARGET_FILE = outfile.txt
    \x SELECT entrydate, 2 * aval "Double aval", modtime FROM bs.dtbl
Note that I specified the column label alias "Double aval" so that the label for that column in the DSV file header will not be blank.

By default, imports will abort as soon as a error is encountered during parsing the file or inserting data. If you invoke SqlTool with a SQL script on the command line, the failure will cause SqlTool to roll back and exit. If run interactively, you can decide whether to commit or roll back the rows that inserted before the failure. You can modify this behavior with the \a and \c settings.

If you set either a reject dsv file or a reject report file, then failures during imports will be reported but will not cause the import to abort. When run in this way, SqlTool will give you a report at the end about how many records were skipped, rejected, and successfully inserted. The reject dsv file is just a dsv file with exact copies of the dsv records that failed to insert. The reject report file is a HTML report which lists, for every rejected record, why that record was rejected.

To allow for user-friendly entry of headers, we require that tables for DSV import/exports use standard column names. I.e., no column names that require quoting. The DSV import and export parsers are very smart and user-friendly. The data types of columns are checked so that the parser can make safe assumptions about white space and blank entries in the data. If a column is a JDBC Boolean type, for example, then we know that a field value of " True " obviously means "True", and that a field value of "" obviously means null. Since we require vanilla style column names, we allow white space anywhere in the header column. We allow blank lines anywhere (where "lines" are delimited by *DSV_ROW_DELIM). By default, commented lines are ignored, and the comment character can be changed from its default value.

Run the command "\x?" or "\m?" to see the several system PL variables which you can set to adjust reject file behavior, commenting behavior, and other DSV features.

You can also define some settings right in the DSV file, and you can even specify multiple header lines in a single DSV file. I use this last feature to import data from one data set into multple tables that are joined. Since I don't have any more time to dedicate to explaining all of these features, I'll give you some examples from working DSV files and let you take it from there.

Example 8.17. Sample DSV headerswitch settings

    # RCS keyword was here.

I'll just note that the prefixes for the header rows must be of format target-table-name + :. You can use * for target-table-name here, for the obvious purpose.

Example 8.18. DSV targettable setting

This last example is from the SqlTool unit test file dsv-trimming.dsv. These special commands must be at the top of the file (before any normal data or header lines).

There is also the *DSV_CONST_COLS setting, which you can use to automatically write static, constant values to the specified columns of all inserted rows.

Unit Testing SqlTool

The SqlTool unit tests reside at testrun/sqltool in the HSQLDB source code repository. Just run the runtests.bash script from that directory to execute all of the tests. Read the file README.txt to find out all about it, including everything you'd need to know to test your own scripts or to add more unit test scripts for SqlTool.

[1] To reduce the time I will need to spend maintaining this document, in this chapter I am giving the path to the sample directory as it is in HSQLDB 1.9.x distributions, namely, HSQLDB_HOME/sample. HSQLDB 1.8.x users should translate these sample directory paths to use HSQLDB_HOME/src/org/hsqldb/sample/....

Chapter 9. SQL Syntax

The Hypersonic SQL Group

Fred Toussi

HSQLDB Development Group

Peter Hudson

HSQLDB Development Group

Joe Maher

HSQLDB Development Group

Edited by

Blaine Simpson

$Date: 2007/02/19 21:15:47 $

HSQLDB version 1.8.0 supports the SQL statements and syntax described in this chapter.

Notational Conventions Used in this Chapter

[A] means A is optional.

{ B | C } means either B or C must be used.

[{ B | C }] means either B or C may optionally be used, or nothing at all.

( and ) are the actual characters '(' and ')' used in statements.

UPPERCASE words are keywords

SQL Commands


ALTER INDEX <indexname> RENAME TO <newname>;

Index names can be changed so long as they do not conflict with other user-defined or sytem-defined names.


ALTER SEQUENCE <sequencename> RESTART WITH <value>;

Resets the next value to be returned from the sequence.


ALTER SCHEMA <schemaname> RENAME TO <newname>;

Renames the schema as specified. All objects of the schema will hereafter be accessible only with the new schema name.

Requires Administrative privileges.


ALTER TABLE <tablename> ADD [COLUMN] <columnname> Datatype
    [(columnSize[,precision])] [{DEFAULT <defaultValue> |
    [BEFORE <existingcolumn>];

Adds the column to the end of the column list. The optional BEFORE <existingcolumn> can be used to specify the name of an existing column so that the new column is inserted in a position just before the <existingcolumn>.

It accepts a columnDefinition as in a CREATE TABLE command. If NOT NULL is specified and the table is not empty, then a default value must be specified. In all other respects, this command is the equivalent of a column definition statement in a CREATE TABLE statement.

If an SQL view includes a SELECT * FROM <tablename> in its select statement, the new column is added to the view. This is a non-standard feature which is likely to change in the future.

ALTER TABLE <tablename> DROP [COLUMN] <columnname>;

Drops the column from the table. Will drop any single-column primary key or unique constraint on the column as well. The command will not work if there is any multiple key constraint on the column or the column is referenced in a check constraint or a foreign key.

It will also fail if an SQL view includes the column.

ALTER TABLE <tablename> ALTER COLUMN <columnname> RENAME TO <newname> 

Changes a column name.

ALTER TABLE <tablename> ALTER COLUMN <columnname> SET DEFAULT <defaultvalue>};

Adds the specified default value to the column. Use NULL to remove a default.

ALTER TABLE <tablename> ALTER COLUMN <columnname> SET [NOT] NULL

Sets or removes a NOT NULL constraint for the column.

ALTER TABLE <tablename> ALTER COLUMN <columnDefinition>;

This form of ALTER TABLE ALTER COLUMN accepts a columnDefinition as in a CREATE TABLE command, with the following restrictions.


  • The column must be already be a PK column to accept an IDENTITY definition.
  • If the column is already an IDENTITY column and there is no IDENTITY definition, the existing IDENTITY attribute is removed.
  • The default expression will be that of the new definition, meaning an existing default can be dropped by ommission, or a new default added.
  • The NOT NULL attribute will be that of the new definition (similar to previous item).
  • Depending on the type of change, the table may have to be empty for the command to work. It always works when the type of change is possible in general and the individual existing values can all be converted.
ALTER TABLE <tablename> ALTER COLUMN <columnname>
    RESTART WITH <new sequence value>

This form is used exclusively for IDENTITY columns and changes the next automatic value for the identity sequence.

ALTER TABLE <tablename> ADD [CONSTRAINT <constraintname>]
    CHECK (<search condition>);

Adds a check constraint to the table. In the current version, a check constraint can reference only the row being inserted or updated.

ALTER TABLE <tablename> ADD [CONSTRAINT <constraintname>] UNIQUE (<column list>);

Adds a unique constraint to the table. This will not work if there is already a unique constraint covering exactly the same <column list>.

This will work only if the values of the column list for the existing rows are unique or include a null value.

ALTER TABLE <tablename> ADD [CONSTRAINT <constraintname>]
    PRIMARY KEY (<column list>);

Adds a primary key constraint to the table, using the same constraint syntax as when the primary key is specified in a table definition.

ALTER TABLE <tablename>
    ADD [CONSTRAINT <constraintname>] FOREIGN KEY (<column list>)
    REFERENCES <exptablename> (<column list>)

Adds a foreign key constraint to the table, using the same constraint syntax as when the foreign key is specified in a table definition.

This will fail if for each existing row in the referring table, a matching row (with equal values for the column list) is not found in the referenced tables.

ALTER TABLE <tablename> DROP CONSTRAINT <constraintname>;

Drop a named unique, check or foreign key constraint from the table.

ALTER TABLE <tablename> RENAME TO <newname>;


ALTER USER <username> SET PASSWORD <password>;

Changes the password for an existing user. Password must be double quoted. Use "" for an empty password.

DBA's may change users' base default schema name with the comand

ALTER USER <username> SET INITIAL SCHEMA <schemaname>;
This is the schema which database object names will resolve to for this user, unless overridden as explained in Schema object naming. For reasons of backwards compatibility, the initial schema value will not be persisted across database shutdowns until HSQLDB version 1.8.1. (I.e., INITIAL SCHEMA settings will be lost upon database shutdown with HSQLDB versions lower than version 1.8.1).

Only an administrator may use these commands.


CALL Expression;

Any expression can be called like a stored procedure, including, but not only Java stored procedures or functions. This command returns a ResultSet with one column and one row (the result) just like a SELECT statement with one row and one column.

See also: Stored Procedures / Functions, SQL Expression.



Closes the database files, rewrites the script file, deletes the log file and opens the database.

If DEFRAG is specified, this command also shrinks the .data file to its minimal size.




Ends a transaction and makes the changes permanent.



CONNECT USER <username> PASSWORD <password>;

Connects to the database as a different user. Password should be double quoted. Use "" for an empty password.

See also: GRANT, REVOKE.


CREATE ALIAS <function> FOR <javaFunction>;

Creates an alias for a static Java function to be used as a Stored Procedure. The function must be accessible from the JVM in which the database runs. Example:

    CREATE ALIAS ABS FOR "java.lang.Math.abs";


The CREATE ALIAS command just defines the alias. It does not validate existence of the target method or its containing class. To validate the alias, use it.

See also: CALL, Stored Procedures / Functions.


CREATE [UNIQUE] INDEX <index> ON <table> (<column> [DESC] [, ...]) [DESC];

Creates an index on one or more columns in a table.

Creating an index on searched columns may improve performance. The qualifier DESC can be present for command compatibility with other databases but it has no effect. Unique indexes can be defined but this is deprecated. Use UNIQUE constraints instead. The name of an index must be unique within the whole database.



CREATE ROLE <rolename>;

Creates the named role with no members. Requires Administrative privileges.


CREATE SCHEMA <schemaname> AUTHORIZATION <grantee>
    [<createStatement> [<grantStatement>] [...];

Creates the named schema, with ownership of the specified authorization. The authorization grantee may be a database user or a role.

Optional (nested) CREATE and GRANT statements can be given only for new objects in this new schema. Only the last nested statement should be terminated with a semicolon, because the first semicolon encountered after "CREATE SCHEMA" will end the CREATE SCHEMA command. In the example below, a new schema, ACCOUNTS, is created, then two tables and a view are added to this schma and some rights on these objects are granted.

Note that this example consists of one CREATE SCHEMA statement which is terminated by a semicolon.

Requires Administrative privileges.


    [START WITH <startvalue>] [INCREMENT BY <incrementvalue>];

Creates a sequence. The default type is INTEGER. The default start value is 0 and the increment 1. Negative values are not allowed. If a sequence goes beyond Integer.MAXVALUE or Long.MAXVALUE, the next result is determined by 2's complement arithmetic.

The next value for a sequence can be included in SELECT, INSERT and UPDATE statements as in the following example:

SELECT [...,] NEXT VALUE FOR <sequencename> [, ...] FROM <tablename>;

In the proposed SQL 200n and in the current version, there is no way of retreiving the last returned value of a sequence.


    ( <columnDefinition> [, ...] [, <constraintDefinition>...] )

Creates a tables in memory (default) or on disk and only cached in memory. If the database is all-in-memory, both MEMORY and CACHED forms of CREATE TABLE return a MEMORY table while the TEXT form is not allowed.

Components of a CREATE TABLE command

columnname Datatype [(columnSize[,precision])]
    [{DEFAULT <defaultValue> |
    (START WITH <n>[, INCREMENT BY <m>])}] |

Default values that are allowed are constant values or certain SQL datetime functions.

Allowed Default Values in Column Definitions

  • For character column, a single-quoted string or NULL. The only SQL function that can be used is CURRENT_USER.

  • For datetime columns, a single-quoted DATE, TIME or TIMESTAMP value or NULL. Or a datetime SQL function such as CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP, TODAY, NOW. Each function is allowed for a certain datetime type.

  • For BOOLEAN columns, the literals FALSE, TRUE, NULL.

  • For numeric columns, any valid number or NULL.

  • For binary columns, any valid hex string or NULL.

Only one identity column is allowed in each table. Identity columns are autoincrement columns. They must be of INTEGER or BIGINT type and are automatically primary key columns (as a result, multi-column primary keys are not possible with an IDENTITY column present). Using the long SQL syntax the (START WITH <n>) clause specifies the first value that will be used. The last inserted value into an identity column for a connection is available using the function IDENTITY(), for example (where Id is the identity column):

INSERT INTO Test (Id, Name) VALUES (NULL,'Test');
    UNIQUE ( <column> [,<column>...] ) |
    PRIMARY KEY ( <column> [,<column>...] ) |
    FOREIGN KEY ( <column> [,<column>...] )
    REFERENCES <refTable> ( <column> [,<column>...]) 
    CHECK(<search condition>)[2]

Both ON DELETE and ON UPDATE clauses can be used in a single foreign key definition.

search condition

A search condition is similar to the set of conditions in a WHERE clause. In the current version of HSQLDB, the conditions for a CHECK constraint can only reference the current row, meaning there should be no SELECT statement. Sample table definitions with CHECK constraints are in TestSelfCheckConstraints.txt. This file is in the /hsqldb/testrun/hsqldb/ directory of the zip.

General syntax limitations

HSQLDB databases are initially created in a legacy mode that does not enforce column size and precision. You can set the property: sql.enforce_strict_size=true to enable this feature. When this property has been set, Any supplied column size and precision for numeric and character types (CHARACTER and VARCHAR) are enforced. Use the command, SET PROPERTY "sql.enforce_strict_size" TRUE once before defining the tables.

NOT NULL constraints can be part of the column definition only. Other constraints cannot be part of the column definition and must appear at the end of the column definition list.

TEMPORARY TABLE contents for each session (connection) are emptied by default at each commit or rollback. The optional qualifier ON COMMIT PRESERVE ROWS can be used to keep the rows while the session is open. The default is ON COMMIT DELETE ROWS.

See also: DROP TABLE.


    [FOR EACH ROW] [QUEUE n] [NOWAIT] CALL <TriggerClass>;

TriggerClass is an application-supplied class that implements the org.hsqldb.Trigger interface e.g. "mypackage.TrigClass". It is the fire method of this class that is invoked when the trigger event occurs. You should provide this class, which can have any name, and ensure that this TriggerClass is present in the classpath which you use to start hsqldb.

Since 1.7.2 the implementation has been changed and enhanced. When the 'fire' method is called, it is passed the following arguments:

    fire (String name, String table, Object row1[], Object row2[])

where 'row1' and 'row2' represent the 'before' and 'after' states of the row acted on, with each column being a member of the array. The mapping of members of the row arrays to database types is specified in Data Types. For example, BIGINT is represented by a java.lang.Long Object. Note that the number of elements in the row arrays could be larger than the number of columns by one or two elements. Never modify the last elements of the array, which are not part of the actual row.

If the trigger method wants to access the database, it must establish its own JDBC connection. This can cause data inconsistency and other problems so it is not recommended. The jdbc:default:connection: URL is not currently supported.

Implementation note:

If QUEUE 0 is specified, the fire method is execued in the same thread as the database engine. This allows trigger action to alter the data that is about to be stored in the database. Data can be checked or modified in BEFORE INSERT / UPDATE + FOR EACH ROW triggers. All table constraints are then enforced by the database engine and if there is a violation, the action is rejected for the SQL command that initiated the INSERT or UPDATE. There is an exception to this rule, that is with UPDATE queries, referential integrity and cascading actions resulting from ON UPDATE CASCASE / SET NULL / SET DEFAULT are all performed prior to the invocation of the trigger method. If an invalid value that breaks referential integrity is inserted in the row by the trigger method, this action is not checked and results in inconsistent data in the table.

Alternatively, if the trigger is used for external communications and not for checking or altering the data, a queue size larger than zero can be specified. This is in the interests of not blocking the database's main thread as each trigger will run in a thread that will wait for its firing event to occur. When this happens, the trigger's thread calls TriggerClass.fire. There is a queue of events waiting to be run by each trigger thread. This is particularly useful for 'FOR EACH ROW' triggers, when a large number of trigger events occur in rapid succession, without the trigger thread getting a chance to run. If the queue becomes full, subsequent additions to it cause the database engine to suspend awaiting space in the queue. Take great care to avoid this situation if the trigger action involves accessing the database, as deadlock will occur. This can be avoided either by ensuring the QUEUE parameter makes a large enough queue, or by using the NOWAIT parameter, which causes a new trigger event to overwrite the most recent event in the queue. The default queue size is 1024. Note also that the timing of trigger method calls is not guaranteed, so applications should implement their own synchronization measures if necessary.

With a non-zero QUEUE parameter, if the trigger methods modifies the 'row2' values, these changes may or may not affect the database and will almost certainly result in data inconsistency.

Please refer to the code for org.hsqldb.sample.Trigger and org.hsqldb.sample.TriggerSample for more information on how to write a trigger class.



CREATE USER <username> PASSWORD <password> [ADMIN];

Creates a new user or new administrator in this database. Password must be double quoted. Empty password can be made using "". You can change a password afterwards using a ALTER USER[2] command.

Only an administrator can do this.



CREATE VIEW <viewname>[(<viewcolumn>,..) AS SELECT ... FROM ... [WHERE Expression]
[ORDER BY orderExpression [, ...]]
[LIMIT <limit> [OFFSET <offset>]];

A view can be thought of as either a virtual table or a stored query. The data accessible through a view is not stored in the database as a distinct object. What is stored in the database is a SELECT statement. The result set of the SELECT statement forms the virtual table returned by the view. A user can use this virtual table by referencing the view name in SQL statements the same way a table is referenced. A view is used to do any or all of these functions:

  • Restrict a user to specific rows in a table. For example, allow an employee to see only the rows recording his or her work in a labor-tracking table.

  • Restrict a user to specific columns. For example, allow employees who do not work in payroll to see the name, office, work phone, and department columns in an employee table, but do not allow them to see any columns with salary information or personal information.

  • Join columns from multiple tables so that they look like a single table.

  • Aggregate information instead of supplying details. For example, present the sum of a column, or the maximum or minimum value from a column.

Views are created by defining the SELECT statement that retrieves the data to be presented by the view. The data tables referenced by the SELECT statement are known as the base tables for the view. In this example, is a view that selects data from three base tables to present a virtual table of commonly needed data:

    CREATE VIEW mealsjv AS
      SELECT m.mid mid, m.name name, t.mealtype mt, a.aid aid,
             a.gname + ' ' + a.sname author, m.description description,
             m.asof asof
        FROM meals m, mealtypes t, authors a
       WHERE m.mealtype = t.mealtype
        AND m.aid = a.aid;

You can then reference mealsjv in statements in the same way you would reference a table:

    SELECT * FROM mealsjv;

A view can reference another view. For example, mealsjv presents information that is useful for long descriptions that contain identifiers, but a short list might be all a web page display needs. A view can be built that selects only specific mealsjv columns:

    CREATE VIEW mealswebv AS SELECT name, author FROM mealsjv;

The SELECT statement in a VIEW definition should return columns with distinct names. If the names of two columns in the SELECT statement are the same, use a column alias to distinguish between them. A list of new column names can always be defined for a view.

    CREATE VIEW aview (new_name, new_author) AS
      SELECT name, author
      FROM mealsjv

See also: SQL Expression, SELECT[2], DROP VIEW[2].


DELETE FROM table [WHERE Expression];

Removes rows in a table.

See also: SQL Expression, INSERT, SELECT[2].



Closes this connection. It is not required to call this command when using the JDBC interface: it is called automatically when the connection is closed. After disconnecting, it is not possible to execute other queries (including CONNECT) with this connection.

See also: CONNECT.



Removes the specified index from the database. Will not work if the index backs a UNIQUE of FOREIGN KEY constraint.



DROP ROLE <rolename>;

Removes all members from specified role, then removes the role itself.



Removes the specified sequence from the database. When IF EXIST is used, the statement returns without an error if the sequence does not exist. The RESTRICT option is in effect by default, meaning that DROP will fail if any view reference the sequence. Specify the CASCADE option to silently drop all dependent database objects.



Removes the specified schema from the database. The RESTRICT option is in effect by default, meaning that DROP will fail if any objects such as tables or sequences have been defined in the schema. Specify the CASCADE option to silently drop all database objects in the schema.

Requires Administrative privileges.



Removes a table, the data and indexes from the database. When IF EXIST is used, the statement returns without an error even if the table does not exist.

The RESTRICT option is in effect by default, meaning that DROP will fail if any tables or views refer to this table. Specify the CASCADE option to silently drop all dependent views, and to drop any foreign key constraint that links this table with other tables.

See also:



DROP TRIGGER <trigger>;

Removes a trigger from the database.

See also: CREATE TRIGGER[2].


DROP USER <username>;

Removes a user from the database.

Only an administrator do this.

See also: CREATE USER.



Removes a view from the database. When IF EXIST is used, the statement returns without an error if the view does not exist. The RESTRICT option is in effect by default, meaning that DROP will fail if any other view refers to this view. Specify the CASCADE option to silently drop all dependent views.

See also: CREATE VIEW[2].



EXPLAIN PLAN FOR can be used with any query to get a detailed list of the elements in the execution plan.

This list includes the indexes used for performing the query and can be used to optimise the query or to add indexes to tables.


ON { table | CLASS "package.class" } TO <grantee>;
GRANT <rolename> [,...] TO <grantee>[2];

<grantee> is either a user name, a role name, or PUBLIC. PUBLIC means all users.

The first form of the GRANT command assigns privileges to a grantee for a table or for a class. To allow a user to call a Store Procedure static function, the right ALL must be used. Examples:

    GRANT ALL ON CLASS "java.lang.Math.abs" TO PUBLIC;


Even though the command is GRANT ALL ON CLASS, you must specify a static method name. You are actually granting access to a static method, not to a class.

The second form of the GRANT command gives the specified <grantee> membership in the specified role.

Requires Administrative privileges.



INSERT INTO table [( column [,...] )]
{ VALUES(Expression [,...]) | SelectStatement};

Adds one or more new rows of data into a table.


ON { table | CLASS "package.class" } FROM <grantee>;
REVOKE <rolename> [,...] FROM <grantee>[2];

<grantee> is either a user name, a role name, or PUBLIC. PUBLIC means all users.

The first form of the REVOKE command withdraws privileges from a grantee for a table or for a class.

The second form of the REVOKE command withdraws membership of the specified <grantee> from the specified role.

Both forms require Administrative privileges.

See also: GRANT.


ROLLBACK [TO SAVEPOINT <savepoint name>[2] |  WORK}];

ROLLBACK used on its own, or with WORK, undoes changes made since the last COMMIT or ROLLBACK.

ROLLBACK TO SAVEPOINT <savepoint name> undoes the change since the named savepoint. It has no effect if the savepoint is not found.

See also: COMMIT.


SAVEPOINT <savepoint name>;


See also: COMMIT.


SCRIPT ['file'];

Creates an SQL script describing the database. If the file is not specified, a result set containing only the DDL script is returned. If the file is specified then this file is saved with the path relative to the machine where the database engine is located.

Only an administrator may do this.


SELECT [{LIMIT <offset> <limit> | TOP <limit>}[2]][ALL | DISTINCT]
{ selectExpression | table.* | * } [, ...]
[INTO [CACHED | TEMP  | TEXT][2] newTable]
FROM tableList
[WHERE Expression]
[GROUP BY Expression [, ...]]
[HAVING Expression]
INTERSECT [DISTINCT] } selectStatement]
[ORDER BY orderExpression [, ...]]
[LIMIT <limit> [OFFSET <offset>]];

Retrieves information from one or more tables in the database.

Components of a SELECT command

    JOIN table ON Expression] [, ...]
{ (selectStatement) [AS] label | tableName}
{ Expression | COUNT(*) | {
} ([ALL | DISTINCT][2]] Expression) } [[AS] label]

If DISTINCT is specified, only one instance of several equivalent values is used in the aggregate function. Except COUNT(*), all aggregate functions exclude NULL values. The type of the returned value for SUM is subject to deterministic widenning to ensure lossless results. The returned value type for COUNT is INTEGER, for MIN, MAX and AVG it is the same type as the column, for SOME and EVERY it is BOOLEAN. For VAR_POP, VAR_SAMP, STDDEV_POP and STDDEV_SAMP statistical functions, the type is always DOUBLE. These statistical functions do not allow ALL or DISTINCT qualifiers.

If CROSS JOIN is specified no ON expression is allowed for the join.

{ columnNr | columnAlias | selectExpression }
    [ASC | DESC]

Creates the result set for the SELECT statement first and then discards the first n rows (OFFSET) and returns the first m rows of the remaining result set (LIMIT). Special cases: LIMIT 0 m is equivalent to TOP m or FIRST m in other RDBMS's; LIMIT n 0 discards the first n rows and returns the rest of the result set.


This form is used at the end of the SELECT statement. The OFFSET term is optional.


Equivalent to LIMIT 0 m.

UNION and other set operations

Multiple SELECT statements joined with UNION, EXCEPT and INTERSECT are possible. Each SELECT is then treated as a term, and the set operation as an operator in an expression. The expression is evaluated from left to right but INTERSECT takes precedence over the rest of the operators and is applied first. You can use parentheses around any number of SELECT statements to change the evaluation order.




Switches on or off the connection's auto-commit mode. If switched on, then all statements will be committed as individual transactions. Otherwise, the statements are grouped into transactions that are terminated by either COMMIT or ROLLBACK. By default, new connections are in auto-commit mode. This command should not be used directly. Use the JDBC equivalent method, Connection.setAutoCommit(boolean autocommit).


SET DATABASE COLLATION <double quoted collation name>;

Each database can have its own collation. Sets the collation from the set of collations in the source for org.hsqldb.Collation.

Once this command has been issued, the database can be opened in any JVM and will retain its collation.



The parameter size is the megabytes of abandoned space in the .data file. When a CHECKPOINT is performed either as a result of the .log file reaching the limit set by "SET LOGSIZE size", or by the user issuing a CHECKPOINT command, the amount of space abandoned during the session is checked and if it is larger than size, a CHECKPOINT DEFRAG is performed instead of a checkpoint.



Disables (ignorecase = true) or enables (ignorecase = false) the case sensitivity of text comparison and indexing for new tables. By default, character columns in new databases are case sensitive. The sensitivity must be switched before creating tables. Existing tables and their data are not affected. When switched on, the data type VARCHAR is set to VARCHAR_IGNORECASE in new tables. Alternatively, you can specify the VARCHAR_IGNORECASE type for the definition of individual columns. So it is possible to have some columns case sensitive and some not, even in the same table.

Only an administrator may do this.


Users may change their base default schema name with the comand

SET INITIAL SCHEMA <schemaname>;
This is the schema which database object names will resolve to for the current user, unless overridden as explained in Schema object naming. For reasons of backwards compatibility, the initial schema value will not be persisted across database shutdowns until HSQLDB version 1.8.1. (I.e., INITIAL SCHEMA settings will be lost upon database shutdown with HSQLDB versions lower than version 1.8.1).



Sets the maximum size in MB of the .log file. Default is 200 MB. The database will be closed and opened (just like using CHECKPOINT) if the .log file gets over this limit, and so the .log file will shrink. 0 means no limit.



SET MAXROWS <maxwors>;

Describe me!


SET PASSWORD <password>;

Changes the password of the currently connected user. Password must be double quotedEmpty password can be set using "".


SET PROPERTY <double quoted name> <value>;

Sets a database property. Properties that can be set using this command are either boolean or integral and are listed in the Advanced Topics chapter.



Describe me!



This commands enables / disables the referential integrity checking (foreign keys). Normally it should be switched on (this is the default) but when importing data (and the data is imported in the 'wrong' order) the checking can be switched off.


Note that when referential integrity is switched back on, no check is made that the changes to the data are consistent with the existing referential integrity constraints. You can verify consistency using SQL queries and take appropriate actions.

Only an administrator may do this.



SET SCHEMA <schemaname>;

Sets the current JDBC session's schema. The sole purpose for the session schema is to provide a default schema name for schema objects that do not have the schema name specified explicitly in the SQL command, or by association with another object of known schema. For example, if you run SELECT * FROM atbl;, HSQLDB will look for the table or view named atbl in the session's current schema.

Session schemas last only for the duration of the current session. When a new JDBC session is obtained, the new session will have the default schema.



Changes the format of the script file. BINARY and COMPRESSED formats are slightly faster and more compact than the default TEXT. Recommended only for very large script files.


SET TABLE tableName INDEX 'index1rootPos index2rootPos ... ';

This command is only used internally to store the position of index roots in the .data file. It appears only in database script files; it should not be used directly.



Sets the table as read only.


SET TABLE <tablename> SOURCE <file and options> [DESC];

For details see the Text Tables chapter.

This command is used exclusively with TEXT tables to specify which file is used for storage of the data. The optional DESC qualifier results in the text file indexed from the end and opened as readonly. The <file and options> argument is a double quoted string that consists of:

    <file and options>::= <doublequote> <filepath>
        [<semicolon> <option>...] <doublequote>


    SET TABLE mytable SOURCE "myfile;fs=|;vs=.;lvs=~"

Supported Properties

quoted = { true | false }

default is true. If false, treats double quotes as normal characters

all_quoted = { true | false }

default is false. If true, adds double quotes around all fields.

encoding = <encoding name>

character encoding for text and character fields, for example, encoding=UTF-8

ignore_first = { true | false }

default is false. If true ignores the first line of the file

cache_scale= <numeric value>

exponent to calculate rows of the text file in cache. Default is 8, equivalent to nearly 800 rows

cache_size_scale = <numeric value>r

exponent to calculate average size of each row in cache. Default is 8, equivalent to 256 bytes per row.

fs = <unquoted character>

field separator

vs = <unquoted character>

varchar separator

lvs = <unquoted character>

long varchar separator

Special indicators for Hsqldb Text Table separators






space character




newline - Used as an end anchor (like $ in regular expressions)


carriage return






a Unicode character specified in hexadecimal

Only an administrator may do this.


SET WRITE_DELAY {{ TRUE | FALSE } | <seconds> | <milliseconds> MILLIS};

This controls the frequency of file sync for the log file. When WRITE_DELAY is set to FALSE or 0, the sync takes place immediately at each COMMIT. WRITE_DELAY TRUE performs the sync once every 20 seconds (which is the default). A numeric value can be specified instead.

The purpose of this command is to control the amount of data loss in case of a total system crash. A delay of 1 second means at most the data written to disk during the last second before the crash is lost. All data written prior to this has been synced and should be recoverable.

A write delay of 0 impacts performance in high load situations, as the engine has to wait for the file system to catch up.

To avoid this, you can set write delay down to 10 milliseconds. In practice, a write delay of 100 milliseconds provides better than 99.9999% reliability with an average one system crash per day, or 99.99999% with an average one system crash per 6 days.

Each time a SET WRITE_DELAY is issued with any value, a sync is immediately performed.

Only an administrator may do this.



Closes the current database.

Varieties of the SHUTDOWN command


Performs a checkpoint to creates a new .script file that has the minimum size and contains the data for memory tables only. It then backs up the .data file containing the CACHED TABLE data in zipped format to the .backup file and closes the database.


Just closes the database files (like when the Java process for the database is terminated); this command is used in tests of the recovery mechanism. This command should not be used as the routine method of closing the database.


Writes out a new .script file which contains the data for all the tables, including CACHED and TEXT tables. It then deletes the existing text table files and the .data file before rewriting them. After this, it backs up the .data file in the same way as normal SHUTDOWN. This operation shrinks all files to the minimum size.


Similar to SHUTDOWN COMPACT but after writing the script and deleting the existing files, it does not rewrite the .data and text table files. After SHUTDOWN SCRIPT, only the .script and .properties file remain. At the next startup, these files are processed and the .data and .backup files are created. This command in effect performs part of the job of SHUTDOWN COMPACT, leaving the other part to be performed automatically at the next startup.

This command produces a full script of the database which can be edited for special purposes prior to the next startup.

Only an administrator may use the SHUTDOWN command.


UPDATE table SET column = Expression [, ...] [WHERE Expression];

Modifies data of a table in the database.


Schema object naming

Schema objects are database objects that are always scoped to a specific schema. Each schema has a namespace. There may be multiple schema objects of the same name, each in the namespace of a different schema. A particular schema object may nearly always be uniquely identified using the notation schemaname.objectname. All HSQLDB database objects are schema objects, other than the following.



Store Procedure Java Classes

HSQL Aliases

Our current Java-class-based Triggers are not fully schema objects. However, we are in the process of implementing SQL-conformant triggers which will encompass our Java-class-based triggers. When this work is completed, HSQLDB triggers will be schema objects.

Sequences are schema objects with creation and removal permission governed by schema authorization (as described hereafter), but GRANT and REVOKE command do not work yet for sequences. In a future version of HSQLDB, sequence GRANTs and REVOKEs will work similarly to the current GRANT and REVOKE commands for table access.

Most of the time, you do not need to specify the schema for the desired schema object, because the implicit schema is usually the only one that can be used. For example, when creating an index, the target schema will default to that of the table which is the target of the index. Named constraints are an extreme example of this. There is never a need to specify a schema name for a constraint, since constraint names are only specified in a CREATE or ALTER TABLE command, and the schema must be that of the target table. If the implicit schema is not determined by a related object, then the default comes from your JDBC session's current schema setting. The session schema value will be your login user's initial schema, or whatever you last set it to with SET SCHEMA[2] in your current JDBC session with the SET SCHEMA command. (Your initial schema is "PUBLIC" unless changed with the ALTER USER SET INITIAL SCHEMA or the SET INITIAL SCHEMA [2] command).

In addition to namespace scoping, there are permission aspects to the schema of a database object. The authorization of a schema is a role or user that is basically the owner of the schema. Only a user with the DBA role (an admin user) or the owner of a schema may create objects, or modify the DDL of objects, in the namespace of that schema. In this way, a schema authorization is said to "own" the objects of that schema. A schema authorization/owner can be a role or a user (even a role with no members). The two schemas automatically created when a database is initialized are both owned by the role DBA.

An important implication to database objects being owned by the schema owner is, if a non-DBA database user is to have permission to create any database object, they must have ownership of a schema. To allow a user to create (or modify DDL of) objects in their own personal schema, you would create a new schema with that user as the authorization. To allow a non-DBA user to share create and DDL privileges in some schema, you would create this schema with a role as the authorization, then GRANT this role to all of the desired users.

The INFORMATION_SCHEMA is a system defined schema that contains the system tables for the database. This schema is read-only. When a database is created, a shema named PUBLIC is automatically created as the default schma. This schema has the authorization DBA. You can change the name of this schema. If all non-system schemas are dropped from a database, an empty PUBLIC schema is created again. So each database always has at least one non-system schema.

Data Types

Table 9.1. Data Types. The types on the same line are equivalent.

NameRangeJava Type
INTEGER | INTas Java typeint | java.lang.Integer
DOUBLE [PRECISION] | FLOATas Java typedouble | java.lang.Double
VARCHARas Integer.MAXVALUEjava.lang.String
CHAR | CHARACTERas Integer.MAXVALUEjava.lang.String
LONGVARCHARas Integer.MAXVALUEjava.lang.String
DATEas Java typejava.sql.Date
TIMEas Java typejava.sql.Time
TIMESTAMP | DATETIMEas Java typejava.sql.Timestamp
DECIMALNo limitjava.math.BigDecimal
NUMERICNo limitjava.math.BigDecimal
BOOLEAN | BITas Java typeboolean | java.lang.Boolean
TINYINTas Java typebyte | java.lang.Byte
SMALLINTas Java typeshort | java.lang.Short
BIGINTas Java typelong | java.lang.Long
REALas Java typedouble | java.lang.Double[2]
BINARYas Integer.MAXVALUEbyte[]
OTHER | OBJECTas Integer.MAXVALUEjava.lang.Object

The uppercase names are the data types names defined by the SQL standard or commonly used by RDMS's. The data types in quotes are the Java class names - if these type names are used then they must be enclosed in quotes because in Java names are case-sensitive. Range indicates the maximum size of the object that can be stored. Where Integer.MAXVALUE is stated, this is a theoretical limit and in practice the maximum size of a VARCHAR or BINARY object that can be stored is dictated by the amount of memory available. In practice, objects of up to a megabyte in size have been successfully used in production databases.

The recommended Java mapping for the JDBC datatype FLOAT is as a Java type "double". Because of the potential confusion it is recommended that DOUBLE is used instead of FLOAT.

VARCHAR_IGNORECASE is a special case-insensitive type of VARCHAR. This type is not portable.

In table definition statements, HSQLDB accepts size, precision and scale qualifiers only for certain types: CHAR(s), VARCHAR(s), DOUBLE(p), NUMERIC(p), DECIMAL(p,s) and TIMESTAMP(p).

TIMESTAMP(p) can take only 0 or 6 as precision. Zero indicates no subsecond part. Without the precision, the default is 6.

By default specified precision and scale for the column is simply ignored by the engine. Instead, the values for the corresponding Java types are always used, which in the case of DECIMAL is an unlimited precision and scale. If a size is specified, it is stored in the database definition but is not enforeced by default. Once you have created the database (before adding data), you can add a database property value to enforce the sizes:

    SET PROPERTY "sql.enforce_strict_size" true

This will enforce the specified size and pad CHAR fields with spaces to fill the size. This complies with SQL standards by throwing an exception if an attempt is made to insert a string longer than the maximum size. It also results in all DECIMAL values conforming to the specified precision and scale.

CHAR and VARCHAR and LONGVARCHAR columns are by default compared and sorted according to POSIX standards. See the SET DATABASE COLLATION[2] section above to modify this behavior. The property sql.compare_in_locale is no longer supported. Instead, you can define a collation to be used for all character comparisons.

Columns of the type OTHER or OBJECT contain the serialized form of a Java Object in binary format. To insert or update such columns, a binary format string (see below under Expression) should be used. Using PreparedStatements with JDBC automates this transformation.

SQL Comments

-- SQL style line comment
// Java style line comment
/* C style line comment */

All these types of comments are ignored by the database.

Stored Procedures / Functions

Stored procedures are static Java functions that are called directly from the SQL language or using an alias. Calling Java functions (directly or using the alias) requires that the Java class can be reached by the database (server). The syntax is:


This means the package must be provided, and the name must be written as one word, and inside " because otherwise it is converted to uppercase (and not found).

An alias can be created using the command CREATE ALIAS:

    CREATE ALIAS SQRT FOR "java.lang.Math.sqrt";

When an alias is defined, then the function can be called additionally using this alias:


Only static java methods can be used as stored procedures. If, within the same class, there are overloaded methods with the same number of arguments, then the first one encountered by the program will be used. If you want to use Java library methods, it is recommended that you create your own class with static methods that act as wrappers around the Java library methods. This will allow you to control which method signature is used to call each Java library method.

Built-in Functions and Stored Procedures

Numerical built-in Functions / Stored Procedures


returns the absolute value of a double value


returns the arc cosine of an angle


returns the arc sine of an angle


returns the arc tangent of an angle


returns the tangent of a/b


return a & b


returns a | b


returns the smallest integer that is not less than d


returns the cosine of an angle


returns the cotangent of an angle


converts radians to degrees


returns e (2.718...) raised to the power of d


returns the largest integer that is not greater than d


returns the natural logarithm (base e)


returns the logarithm (base 10)


returns a modulo b


returns pi (3.1415...)


returns a raised to the power of b


converts degrees to radians


returns a random number x bigger or equal to 0.0 and smaller than 1.0


rounds a to b digits after the decimal point


solves rounding problems such as 3.11-3.1-0.01


returns -1 if d is smaller than 0, 0 if d==0 and 1 if d is bigger than 0


returns the sine of an angle


returns the square root


returns the trigonometric tangent of an angle


truncates a to b digits after the decimal point

String built-in Functions / Stored Procedures


returns the ASCII code of the leftmost character of s


returns the length of the string in bits


returns a character that has the ASCII code c


returns the length of the string in characters


returns str1 + str2


returns the difference between the sound of s1 and s2


returns translated string


returns a string where len number of characters beginning at start has been replaced by s2


converts s to lower case


returns the leftmost count of characters of s) - requires double quoting - use SUBSTRING() instead


returns the number of characters in s


returns the first index (1=left, 0=not found) where search is found in s, starting at start


removes all leading blanks in s


returns the length of the string in bytes (twice the number of characters)


returns translated string


returns s repeated count times


replaces all occurrences of replace in s with s2


returns the rightmost count of characters of s


removes all trailing spaces


returns a four character code representing the sound of s


returns a string consisting of count spaces


alias for substring


returns the substring starting at start (1=left) with length len


converts s to upper case


converts s to lower case


converts s to upper case

Date/Time built-in Functions / Stored Procedures


returns the current date


returns the current time

DATEDIFF(string, datetime1, datetime2)[2]

returns the count of units of time elapsed from datetime1 to datetime2. The string indicates the unit of time and can have the following values 'ms'='millisecond', 'ss'='second','mi'='minute','hh'='hour', 'dd'='day', 'mm'='month', 'yy' = 'year'. Both the long and short form of the strings can be used.


returns the name of the day


returns the day of the month (1-31)


returns the day of the week (1 means Sunday)


returns the day of the year (1-366)


return the hour (0-23)


returns the minute (0-59)


returns the month (1-12)


returns the name of the month


returns the current date and time as a timestamp) - use CURRENT_TIMESTAMP instead


returns the quarter (1-4)


returns the second (0-59)


returns the week of this year (1-53)


returns the year


returns the current date


returns the current time


returns the current timestamp

System/Connection built-in Functions / Stored Procedures


returns the name of the database of this connection


returns the user name of this connection


SQL standard function, returns the user name of this connection


returns the last identity values that was inserted by this connection

System built-in Functions / Stored Procedures


if exp is null, value is returned else exp) - use COALESCE() instead


if exp is true, v1 is returned, else v2) - use CASE WHEN instead


converts exp to another data type

CAST(term AS type)[2]

converts exp to another data type


if expr1 is not null then it is returned else, expr2 is evaluated and if not null it is returned and so on


if v1 equals v2 return null, otherwise v1

CASE v1 WHEN...[2]


when v1 equals v2 return v3 [otherwise v4 or null if there is no ELSE]


CASE WHEN expr1 THEN v1[WHEN expr2 THEN v2] [ELSE v4] END

when expr1 is true return v1 [optionally repeated for more cases] [otherwise v4 or null if there is no ELSE]



POSITION (... IN ..)[2]

POSITION(<string expression> IN <string expression>)

if the first string is a sub-string of the second one, returns the position of the sub-string, counting from one; otherwise 0

SUBSTRING(... FROM ... FOR ...)[2]

SUBSTRING(<string expression> FROM <numeric expression> [FOR <numeric expression>])

TRIM( LEDING ... FROM ...)[2]

TRIM([{LEADING | TRAILING | BOTH}] FROM <string expression>)


SQL Expression

[NOT] condition [{ OR | AND } condition]

Components of SQL Expressions

   { value [|| value]
  | value { = | < | <= | > | >= | <> | != } value
  | value IS [NOT] NULL
  | EXISTS(selectStatement)
  | value BETWEEN value AND value
  | value [NOT] IN ( {value [, ...] | selectStatement } )
  | value [NOT] LIKE value [ESCAPE] value }
   [+ | -] { term [{ + | - | * | / | || } term]
  | ( condition )
  | function ( [parameter] [,...] )
  | selectStatement giving one value
  | {ANY|ALL} (selectStatement giving single column)
   { 'string' | number | floatingpoint
  | [table.]column | TRUE | FALSE | NULL }
   NEXT VALUE FOR <sequence>

HSQLDB does not currently enforce the SQL 200n proposed rules on where sequence generated values are allowed to be used. In general, these values can be used in insert and update statements but not in CASE statements, order by clauses, search conditions, aggregate functions, or grouped queries.


Strings in HSQLDB are Unicode strings. A string starts and ends with a single ' (singlequote). In a string started with ' (singlequote) use '' (two singlequotes) to create a ' (singlequote).

String contatenation should be performed with the standard SQL operator || rather than the non-standard + operator.

The LIKE keyword uses '%' to match any (including 0) number of characters, and '_' to match exactly one character. To search for '%' or '_' itself an escape character must also be specified using the ESCAPE clause. For example, if the backslash is the escaping character, '\%' and '\_' can be used to find the '%' and '_' characters themselves. For example, SELECT .... LIKE '\_%' ESCAPE '\' will find the strings beginning with an underscore.


The character set for quoted identifiers (names) in HSQLDB is Unicode.

A unquoted identifier (name) starts with a letter and is followed by any number of ASCII letters or digits. When an SQL statement is issued, any lowercase characters in unquoted identifiers are converted to uppercase. Because of this, unquoted names are in fact ALL UPPERCASE when used in SQL statements. An important implication of this is the for accessing columns names via JDBC DatabaseMetaData: the internal form, which is the ALL UPPERCASE must be used if the column name was not quoted in the CREATE TABLE statement.

Quoted identifiers can be used as names (for tables, columns, constraints or indexes). Quoted identifiers start and end with " (one doublequote). A quoted identifier can contain any Unicode character, including space. In a quoted identifier use "" (two doublequotes) to create a " (one doublequote). With quoted identifiers it is possible to create mixed-case table and column names. Example:

    CREATE TABLE "Address" ("Nr" INTEGER,"Name" VARCHAR);
    SELECT "Nr", "Name" FROM "Address";

The equivalent quoted identifier can be used for an unquoted identifer by converting the identifier to all uppercase and quoting it. For example, if a table name is defined as Address2 (unquoted), it can be referred to by its quoted form, "ADDRESS2", as well as address2, aDDress2 and ADDRESS2. Quoted identifiers should not be confused with SQL strings.

Quoting can sometimes be used for identifiers, aliases or functions when there is an ambiguity. For example:


Although HSQLDB 1.8.0 does not force unquoted identifiers to contain only ASCII characters, the use of non-ASCII characters in these identifiers does not comply with SQL standards. Portability between different JRE locales could be an issue when accented characters (or extended unicode characters) are used in unquoted identifiers. Because native Java methods are used to convert the identifier to uppercase, the result may vary not be expected in different locales. It is recommended that accented characters are used only in quoted identifiers.

When using JDBC DatabaseMetaData methods that take table, column, or index identifiers as arguments, treat the names as they are registered in the database. With these methods, unquoted identifiers should be used in all-uppercase to get the correct result. Quoted identifiers should be used in the exact case combination as they were defined - no quote character should be included around the name. JDBC methods that return a result set containing such identifiers return unquoted identifiers as all-uppercase and quoted identifiers in the exact case they are registered in the database (a change from 1.6.1 and previous versions).

Please also note that the JDBC getXXX(String columnName) methods interpret the columnName as case-independent. This is a general feature of JDBC and not specific to HSQLDB.


Passwords must be double quoted and used consistently. Passwords are case insensitive only for backward compatibility. This may change in future versions.

  • A DATE literal starts and ends with ' (singlequote), the format is yyyy-mm-dd (see java.sql.Date.

  • A TIME liteal starts and ends with ' (singlequote), the format is hh:mm:ss (see java.sql.Time).

  • A TIMESTAMP or DATETIME literal starts and ends with ' (singlequote), the format is yyyy-mm-dd hh:mm:ss.SSSSSSSSS (see java.sql.Timestamp).

When specifying default values for date / time columns in CREATE TABLE statements, or in SELECT,INSERT, and UPDATE statements, special SQL functions: NOW, SYSDATE, TODAY, CURRENT_TIMESTAMP, CURRENT_TIME and CURRENT_DATE (case independent) can be used. NOW is used for TIME and TIMESTAMP columns, TODAY is used for DATE columns. The data and time variants CURRENT_* are SQL standard versions and should be used in preference to others. Example:


Binary data starts and ends with ' (singlequote), the format is hexadecimal. '0004ff' for example is 3 bytes, first 0, second 4 and last 255 (0xff).

Any number of commands may be combined. With combined commands, ';' (semicolon) must be used at the end of each command to ensure data integrity, despite the fact that the engine may understand the end of commands and not return an error when a semicolon is not used.

[2] These features were added by HSQL Development Group since April 2001

Appendix A. Building HSQLDB

Fred Toussi

HSQLDB Development Group

$Date: 2005/05/26 23:22:06 $


From 1.8.0, the supplied hsqldb.jar file is built with Java 1.5. If you want to run the engine under JDK1.3 or earlier, you should rebuild the jar with Ant.

Building with Ant, from the Apache Jakarta Project

Ant (Another Neat Tool) is used for building hsqldb. The version currently used to test the build script is 1.6.1 but versions since 1.5.1 should also be compatible.

Obtaining Ant

Ant is a part of the Jakarta/Apache Project.

Building Hsqldb with Ant

Once you have unpacked the zip package for hsqldb, under the /hsqldb folder, in /build there is a build.xml file that builds the hsqldb.jar with Ant (Ant must be already installed). To use it, change to /build then type:

 ant -projecthelp

This displays the available ant targets, which you can supply as command line arguments to ant. These include


to make the hsqldb.jar


Lists all targets which build jar files, with an explanation of the purposes of the different jars.


to clean up the /classes directory that is created


to remove the old jar as well


to build javadoc


to build a smaller jar for HSQLDB that does not contain utilities


to build an extremely small jar containing only the client-side JDBC driver (does not support direct connection to HSQLDB URLs of the form jdbc:hsldb:mem:*, jdbc:hsqldb:file:*, nor jdbc:hsqldb:res:*).


to build a small jar that supports HSQLDB URLs of the form jdbc:hsqldb:mem:*, jdbc:hsqld:file*, jdbc:hsqldb:res:*; but not network URLs like jdbc:hsql* or jdbc:http*.


to build a larger jar for hsqldb that contains tests


Many more targets are available. Run ant -projecthelp and ant explainjars.

HSQLDB can be built in any combination of three JRE (Java Runtime Environment) versions and many jar file sizes. The smallest jar size(hsqljdbc.jar) contains only the HSQLDB JDBC Driver client. The default size (hsqldb.jar) also contains server mode support and the utilities. The largest size (hsqldbtest.jar)includes some test classes as well. Before building the hsqldbtest.jar package, you should download the junit jar from http://www.junit.org and put it in the /lib directory, alongside servlet.jar, which is included in the .zip package.

Just run ant explainjars for a concise list of all available jar files.

If you want your code built for debugging, as opposed to high performance, make a file named build.properties in your build directory with the contents

build.debug: true
The resulting Java binaries will be larger and slower, but exception stack traces will contain source code line numbers, which can be extremely useful for debugging.

The preferred method of rebuilding the jar is with Ant. After installing Ant on your system use the following command from the /build directory:

ant explainjars

The command displays a list of different options for building different sizes of the HSQLDB Jar. The default is built using:

Example A.1. Buiding the standard Hsqldb jar file with Ant

ant hsqldb

The Ant method always builds a jar with the JDK that is used by Ant and specified in its JAVA_HOME environment variable. Building with JDK 1.4.x or 1.5.x will result in a jar that is not backward compatible.

From version 1.7.2, use of JDK 1.1.x is not recommended for building the JAR, even for running under JDK 1.1.x -- use JDK 1.3.1 for compatibility with 1.1.x. This is done in the following way. JDK 1.3.1 should be used as the JAVA_HOME for ant. You then issue the following commands. The first command will make the sources compatible with JDK 1.3, the second command modifies the sources further so that the compiled result can run under jdk 1.1 as well. The third command builds the jar.

ant switchtojdk12
ant switchtojava1target
ant hsqldb

Building with DOS Batch Files

UNIX users must use Ant to build hsqldb.

For DOS/Windows users, a set of MSDOS batch files is provided as an example. These files produce only the default jar size. The path and classpath variables for the JDK should of course be set before running any of the batch files. These files are not currently maintained and will probably need some additions and changes to work correctly. Please see the build.xml file for up-to-date file

If you are compiling for JDK's other than 1.4.x, you should use the appropriate switchToJDK11.bat or switchToJDK12.bat to adapt the source files to the target JDK before running the appropriate buildJDK11.bat or buildJDK12.bat JDK and JRE versions.

Hsqldb CodeSwitcher

CodeSwitcher is a tool to manage different version of Java source code. It allows to compile HSQLDB for different JDKs. It is something like a precompiler in C but it works directly on the source code and does not create intermediate output or extra files.

CodeSwitcher is used internally in HSQLDB build scripts. You do not have to use it separately to compile HSQLDB.

CodeSwitcher reads the source code of a file, removes comments where appropriate and comments out the blocks that are not used for a particular version of the file. This operation is done for all files of a defined directory, and all subdirectories.

Example A.2. Example source code before CodeSwitcher is run


    //#ifdef JAVA2

        properties.store(out,"hsqldb database");



        properties.save(out,"hsqldb database");




The next step is to run CodeSwitcher.

Example A.3. CodeSwitcher command line invocation

    java org.hsqldb.util.CodeSwitcher . -JAVA2

The '.' means the program works on the current directory (all subdirectories are processed recursively). -JAVA2 means the code labelled with JAVA2 must be switched off.

Example A.4. Source code after CodeSwitcher processing


    //#ifdef JAVA2


        pProperties.store(out,"hsqldb database");



        pProperties.save(out,"hsqldb database");



For detailed information on the command line options run java org.hsqldb.util.CodeSwitcher. Usage examples can be found in the switchtojdk1*.bat files in the /build directory.

Building documentation

To build the User Guide in HTML format, you must have the Docbook stylesheets installed locally. The Docbook stylesheets are available on the Internet. On Linux, just install the docbook-xsl-stylesheets rpm. Then add an entry to build.properties in your build directory with contents like

docbook.xsl.home: /usr/share/sgml/docbook/docbook-xsl-stylesheets
Where you specify your local path to the base directory of your Docbook stylesheet installation. Build like

Example A.5. Building HTML User Guides

ant docbooks-html
ant docbooks-chunk

To build the User Guide in PDF format, you must also have the Java FOP system installed locally. FOP is available for free download on the Internet. Add an entry to build.properties in your build directory with contents like

fop.home /usr/local/fop-0.20.5
Where you specify your local path to the base directory of your FOP installation.

Example A.6. Building User Guides in all formats

ant docbook
Don't pay too much attention to error messages by FOP, because they are really warnings, but do check the output. If there are problems with the PDF output, try using a newer version of FOP.


By default, your docs will fail to build if you do not have Internet connectivity. This is because our primary Docbook source file references the Docbook DTDs via Internet URL. You can build without Internet connectivity by installing the Docbook DTDs and editing our primary Docbook source file. Docbook is available on the Internet. On Linux, just install the docbook-dtds or docbook rpm. Then make one edit to the file docsrc/guide/guide.xml in your HSQLDB distribution. Change the line containing

"http://www.oasis-open.org/docbook/xml/4.2CR1/docbookx.dtd" [
"file:///usr/share/xml/docbook/schema/dtd/4.2/docbookx.dtd" [
where the second filepath is the path to the docbookx.dtd file within your Docbook installation.

Appendix B. First JDBC Client Example

There is a copy of Testdb.java in the directory src/org/hsqldb/sample of your HSQLDB distribution.

Example B.1. JDBC Client source code example

/* Copyright (c) 2001-2005, The HSQL Development Group
 * All rights reserved.
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 * Redistributions of source code must retain the above copyright notice, this
 * list of conditions and the following disclaimer.
 * Redistributions in binary form must reproduce the above copyright notice,
 * this list of conditions and the following disclaimer in the documentation
 * and/or other materials provided with the distribution.
 * Neither the name of the HSQL Development Group nor the names of its
 * contributors may be used to endorse or promote products derived from this
 * software without specific prior written permission.

package org.hsqldb.sample;

import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.ResultSet;
import java.sql.ResultSetMetaData;
import java.sql.SQLException;
import java.sql.Statement;

 * Title:        Testdb
 * Description:  simple hello world db example of a
 *               standalone persistent db application
 *               every time it runs it adds four more rows to sample_table
 *               it does a query and prints the results to standard out
 * Author: Karl Meissner karl@meissnersd.com
public class Testdb {

    Connection conn;                                                //our connnection to the db - presist for life of program

    // we dont want this garbage collected until we are done
    public Testdb(String db_file_name_prefix) throws Exception {    // note more general exception

        // Load the HSQL Database Engine JDBC driver
        // hsqldb.jar should be in the class path or made part of the current jar

        // connect to the database.   This will load the db files and start the
        // database if it is not alread running.
        // db_file_name_prefix is used to open or create files that hold the state
        // of the db.
        // It can contain directory names relative to the
        // current working directory
        conn = DriverManager.getConnection("jdbc:hsqldb:"
                                           + db_file_name_prefix,    // filenames
                                           "sa",                     // username
                                           "");                      // password

    public void shutdown() throws SQLException {

        Statement st = conn.createStatement();

        // db writes out to files and performs clean shuts down
        // otherwise there will be an unclean shutdown
        // when program ends
        conn.close();    // if there are no other open connection

//use for SQL command SELECT
    public synchronized void query(String expression) throws SQLException {

        Statement st = null;
        ResultSet rs = null;

        st = conn.createStatement();         // statement objects can be reused with

        // repeated calls to execute but we
        // choose to make a new one each time
        rs = st.executeQuery(expression);    // run the query

        // do something with the result set.
        st.close();    // NOTE!! if you close a statement the associated ResultSet is

        // closed too
        // so you should copy the contents to some other object.
        // the result set is invalidated also  if you recycle an Statement
        // and try to execute some other query before the result set has been
        // completely examined.

//use for SQL commands CREATE, DROP, INSERT and UPDATE
    public synchronized void update(String expression) throws SQLException {

        Statement st = null;

        st = conn.createStatement();    // statements

        int i = st.executeUpdate(expression);    // run the query

        if (i == -1) {
            System.out.println("db error : " + expression);

    }    // void update()

    public static void dump(ResultSet rs) throws SQLException {

        // the order of the rows in a cursor
        // are implementation dependent unless you use the SQL ORDER statement
        ResultSetMetaData meta   = rs.getMetaData();
        int               colmax = meta.getColumnCount();
        int               i;
        Object            o = null;

        // the result set is a cursor into the data.  You can only
        // point to one row at a time
        // assume we are pointing to BEFORE the first row
        // rs.next() points to next row and returns true
        // or false if there is no next row, which breaks the loop
        for (; rs.next(); ) {
            for (i = 0; i < colmax; ++i) {
                o = rs.getObject(i + 1);    // Is SQL the first column is indexed

                // with 1 not 0
                System.out.print(o.toString() + " ");

            System.out.println(" ");
    }                                       //void dump( ResultSet rs )

    public static void main(String[] args) {

        Testdb db = null;

        try {
            db = new Testdb("db_file");
        } catch (Exception ex1) {
            ex1.printStackTrace();    // could not start db

            return;                   // bye bye

        try {

            //make an empty table
            // by declaring the id column IDENTITY, the db will automatically
            // generate unique values for new rows- useful for row keys
                "CREATE TABLE sample_table ( id INTEGER IDENTITY, str_col VARCHAR(256), num_col INTEGER)");
        } catch (SQLException ex2) {

            //ex2.printStackTrace();  // second time we run program
            //  should throw execption since table
            // already there
            // this will have no effect on the db

        try {

            // add some rows - will create duplicates if run more then once
            // the id column is automatically generated
                "INSERT INTO sample_table(str_col,num_col) VALUES('Ford', 100)");
                "INSERT INTO sample_table(str_col,num_col) VALUES('Toyota', 200)");
                "INSERT INTO sample_table(str_col,num_col) VALUES('Honda', 300)");
                "INSERT INTO sample_table(str_col,num_col) VALUES('GM', 400)");

            // do a query
            db.query("SELECT * FROM sample_table WHERE num_col < 250");

            // at end of program
        } catch (SQLException ex3) {
    }    // main()
}    // class Testdb

Appendix C. Hsqldb Database Files and Recovery

This text is based on HypersonicSQL documentation, updated to reflect the latest version 1.8.0 of HSQLDB.

$Date: 2005/07/01 17:06:32 $

The Standalone and Client/Server modes will in most cases use files to store all data to disk in a persistent and safe way. This document describes the meaning of the files, the states and the procedures followed by the engine to recover the data.

A database named 'test' is used in this description. The database files will be as follows.

Database Files


Contains the entry 'modified'. If the entry 'modified' is set to 'yes' then the database is either running or was not closed correctly (because the close algorithm sets 'modified' to 'no' at the end).


This file contains the SQL statements that makes up the database up to the last checkpoint - it is in synch with test.backup.


This file contains the (binary) data records for CACHED tables only.


This is compressed file that contains the complete backup of the old test.data file at the time of last checkpoint.


This file contains the extra SQL statements that have modified the database since the last checkpoint (something like the 'Redo-log' or 'Transaction-log', but just text).

In the above list, a checkpoint results from both a CHECKPOINT command and a SHUTDOWN command.


Database is closed correctly

State after using the command SHUTDOWN

  • The test.data file is fully updated.

  • The test.backup contains the compressed test.data file.

  • The test.script contains the information in the database, excluding data for CACHED and TEXT tables.

  • The test.properties contains the entry 'modified' set to 'no'.

  • There is no test.log file.

Database is closed correctly with SHUTDOWN SCRIPT

State after using the command SHUTDOWN SCRIPT

  • The test.data file does not exist; all CACHED table data is in the test.script file

  • The test.backup does not exist.

  • The test.script contains the information in the database, including data for CACHED and TEXT tables.

  • The test.properties contains the entry 'modified' set to 'no'.

  • There is no test.log file.

Database is aborted

This may happen by sudden power off, Ctrl+C in Windows, but may be simulated using the command SHUTDOWN IMMEDIATELY.

Aborted Database state

  • The test.properties still containes 'modified=yes'.

  • The test.script contains a snapshot of the database at the last checkpoint and is OK.

  • The test.data file may be corrupt because the cache in memory was not written out completely.

  • The test.backup file contains a snapshot of test.data that corresponds to test.script.

  • The test.log file contain all information to re-do all changes since the snanapshot. As a result of abnormal termination, this file may be partially corrupt.


The database engine performs the following procedures internally in different circumstances.

Clean Shutdown

Procedure C.1. Clean Hsqldb database shutdown

  1. The test.data file is written completely (all the modified cached table rows are witten out) and closed.

  2. The test.backup.new is created (containing the compressed test.data file)

  3. The file test.script.new is created using the information in the database (and thus shrinks because no UPDATE and DELETE statements; only INSERT).

  4. The entry 'modified' in the properties file is set to 'yes-new-files'

  5. The file test.script is deleted

  6. The file test.script.new is renamed to test.script

  7. The file test.backup is deleted

  8. The file test.backup.new is renamed to test.backup

  9. The entry 'modified' in the properties file is set to 'no'

  10. The file test.log is deleted


Procedure C.2. Database is opened

  1. Check if the database files are in use (by checking a special test.lck file).

  2. See if the test.properties file exists, otherwise create it.

  3. If the test.properties did not exist, then this is a new database. Create the empty test.log to append new commands.

  4. If it is an existing database, check in the test.properties file if 'modified=yes'. This would mean last time it was not closed correctly, and thus the test.data file may be corrupted or incomplete. In this case the 'REPAIR' algorithm is executed (see below), before the database is opened normally.

  5. Otherwise, if in the test.properties file 'modified=yes-new-files', then the (old) test.backup and test.script files are deleted and the new test.script.new file is renamed to test.script.

  6. Open the test.script file and execute the commands.

  7. Create the empty test.log to append new commands.


The current test.data file is corrupt, but with the old test.data (from the test.backup file and test.script) and the current test.log, the database is made up-to-date. The database engine takes these steps:

Procedure C.3. Database Repair

  1. Restore the old test.data file from the backup (uncompress the test.backup and overwrite test.data).

  2. Execute all commands in the test.script file.

  3. Execute all commands in the test.log file. If due to corruption, an exception is thrown, the rest of the lines of command in the test.log file are ignored.

  4. Close the database correctly (including a backup).

Appendix D. Running Hsqldb with OpenOffice.org 1.1.x

Hermann Kienlein

EDV - Systeme Kienlein

Copyright 2003-2004 Hermann Kienlein. Permission is granted to distribute this document without any alteration under the terms of the HSQLDB license. Additional permission is granted to the HSQLDB Development Group to distribute this document with or without alterations under the terms of the HSQLDB license.

$Date: 2005/06/08 16:02:34 $


HSQLDB can now act as a Database with OpenOffice.org. This document is written to help you connecting and running HSQLDB out of OpenOffice.org in a simple way. Without user-managment and only for your single-system.

If you have problems read the other available documents, because I will not write them here again. If you need a real DB-System with user-management and different rights for different users, read the other documents.

HSQLDB is included with OpenOffice.org 2.0 and is used by default. Please refer to standard OpenOffice.org 2.0 documentation on how to use HSQLDB with this version.


I assume you have a running OpenOffice.org (OOo) and a JavaRuntimeEnvironment. So place the hsqldb_*.zip file where you want on your disk and unpack it (I assume you have done this already).

Setting up OpenOffice.org

Start OOo with a text document and go to the Database-Explorer (simply by pressing F4). In the left frame you see a tree-view with all known databases in OOo.

A right mouse-click opens a menu where you can manage your databases. So click on New Database and choose a name that you want to have inside OOo. I chose HSQLDB as name.

As connection-type choose JDBC and then switch to the JDBC-tab.

As Driver-Class insert org.hsqldb.jdbcDriver and as URL choose the following:

On Windows

You can specify a directory where HSQLDB should store the info and data. Something like jdbc:hsqldb:file:c:\javasrc\hsqldb-dev\databasename (where jdbc: is written by OOo). The string c:\javasrc\hsqldb-dev\databasename works only on windows, but you can write this down as linux-path like /javasrc/hsqldb-dev/databasename too. Then HSQLDB takes the c:\ drive as root. This means this works only on c:\ for you.

The first is the directory-path and the databasename is the identifier for the database.

On Linux

Choose a path as said for windows like /opt/db/data

As username take sa, this is the standard-administrator for HSQLDB.

Now click the OK-Button

Now OOo has to find your hsqldb.jar file. So go to options => security and insert the path to the .jar file. If you have problems, search the Online-help for JDBC. You then get help in your own language (this is generally quite better than my English, I think ;-)

If you cannot write to your Tables, OOo thinks that you don't have permission to write to HSQLDB. Then we tell OOo to ignore the DriverPrivileges because on our single-user-system we do not need them.

Because OOo is working on this, the next Step is only needed for systems without write - permission.

So we go to http://dba.openoffice.org and look at the IgnoreDriverPrivileges.html file in the HowTo-section. You find here a macro-code.

Open tools => macro in OOo to get the Basic-IDE. Here simple copy and paste the code and run the macro. You see a input-box where you only have to insert the name of your DB, in my example I have to insert HSQLDB, because I took this as name in OOo.

Note that if you change your OOo-DB name, you have to run this macro again!

Now we only have to stop and restart OOo. Be sure that you exit Quickstarter and all running processes too. On next OOo-Start you should have a running Database in OpenOffice.org.

Appendix E. Hsqldb Test Utility

$Date: 2005/05/27 12:41:50 $

The org.hsqldb.test package contains a number of tests for various functions of the database engine. Among these, the TestSelf class performs the tests that are based on scripts. To run the tests, you should compile the hsqldbtest.jar target with Ant.

For TestSelf, a batch file is provided in the testrun/hsqldb directory, together with a set of TestSelf*.txt files. To start the application in Windows, change to the directory and type:

    runtest TestSelf

In Unix / Linux, type:

    ./runTest.sh TestSelf

The new version of TestSelf runs multiple SQL test files to test different SQL operations of the database. All files in the working directory with names matching TestSelf*.txt are processed in alphabetical order.

You can add your own scripts to test different series of SQL queries. The format of the TestSelf*.txt file is simple text, with some indentation and prefixes in the form of Java-style comments. The prefixes indicate what the expected result should be.

  • Comment lines must start with -- and are ignored

  • Lines starting with spaces are the continuation of the previous line

  • SQL statements with no prefix are simply executed.

  • The remaining items in this list exemplify use of the available command line-prefixes.

  • The /*s*/ option stands for silent. It is used for executing quries regardless of results. Used for preparation of tests, not for actual tests.

    /*s*/ Any SQL statement - errors are ignored
  • The /*c<rows>*/ option is for SELECT queries and asserts the number of rows in the result matches the given count.

    /*c<rows>*/ SQL statement returning count of <rows>
  • The /*u*/ option is for queries that return an update count, such as DELETE and UPDATE. It asserts the update count matches.

    /*u<count>*/ SQL statement returning an update count equal to <count>
  • The /*e*/ option asserts that the given query results in an erros. It is mainly used for testing the error detection capabilities of the engine. It can also be used with syntactically valid queries to assert a certain state in the database. For example a CREATE TABLE can be used to assert the table of the same name already exists.

    /*e*/ SQL statement that should produce an error when executing
  • The /*r....*/ option asserts the SELECT query returns a single row containing the given set of field values.

    /*r<string1>,<string2>*/ SQL statement returning a single row ResultSet equal to the specified value
  • The extended /*r...*/ option asserts the SELECT query returns the given rows containing the given set of field values.

    */ SQL statement returning a multiple row ResultSet equal to the specified values

    (note that the result set lines are indented).

  • All the options are lowercase letters. During development, an uppercase can be used for a given test to exclude a test from the test run. The utility will just report the test blocks that have been excluded without running them. Once the code has been developed, the option can be turned into lowercase to perform the actual test.

See the TestSelf*.txt files in the /testrun/hsqldb/ directory for actual examples.

Appendix F. Database Manager

Fred Toussi

HSQLDB Development Group

Blaine Simpson

HSQLDB Development Group

$Date: 2006/07/27 21:08:21 $

Brief Introduction

The Database Manager tool is a simple GUI database query tool with a tree display of the tables. Both AWT and SWING versions of the tool are available and work almost identically. The AWT version class name is org.hsqldb.util.DatabaseManager; the SWING version, org.hsqldb.util.DatabaseManagerSwing.

The AWT version of the database manager can be deployed as an applet in a browser. A demo HTML file with an embedded Database Manager is included in the /demo directory.

When the Database Manager is started, a dialogue allows you to enter the JDBC driver, URL, user and password for the new connection. A drop-down box, Type, offers preset values for JDBC driver and URL for most popular database engines, including HSQLDB. Once you have selected an item from this drop-down box, you should edit the URL to specify the details of the database or any additional properties to pass. You should also enter the username and password before clicking on the OK button.

The connection dialogue allows you to save the settings for the connection you are about to make. You can then access the connection in future sessions. To save a connection setting, enter a name in the Setting Name box before clicking on the OK button. Next time the connection dialogue is displayed, the drop-down box labeled Recent will include the name for all the saved connection settings. When you select a name, the individual settings are displayed in the appropriate boxes.

The small Clr button next to the drop-down box allows you to clear all the saved settings. If you want to modify an existing setting, first select it from the drop-down box then modify any of the text boxes before making the connection. The modified values will be saved.

Most menu items have context-sensitive tool tip help text which will appear if you hold the mouse cursor still over the desired menu item. (Assuming that you don't turn Tooltips off under the Help menu.

The DatabaseManagers do work with HSQLDB servers serving TLS-encrypted JDBC data. See the TLS chapter and the RC File Authentication Setup section of this Guide.


If you are using DatabaseManagerSwing with Oracle, you will want to make sure that Show row counts and Show row counts are both off before connecting to the database. You may also want to turn off Auto tree-update, as described in the next section.

Auto tree-update

By default, the object tree in the left panel is refreshed when you execute DDL which may update those objects. If you are on a slow network or performance-challenged PC, use the view / Auto-refresh tree menu item to turn it off. You will then need to use the viewRefresh tree menu item every time that you want to refresh the tree.


Auto-refresh tree does not automatically show all updates to database objects, it only refreshes when you submit DDL which may update database objects. (This behavior is a compromise between utility and performance).

Automatic Connection

You can use command-line switches to supply connection information. If you use these switch(es), then the connection dialog window will be skipped and a JDBC connection will be established immediately. Assuming that the hsqldb.jar (or an alternative jar) are in your CLASSPATH, this command will list the available command-line options.

    java org.hsqldb.util.DatabaseManagerSwing --help

It's convenient to skip the connection dialog window if you always work with the same database account.


Use of the --password switch is not secure. Everything typed on command-lines is generally available to other users on the computer. The problem is compounded if you use a network connection to obtain your command line. The RC File section explains how you can set up automatic connections without supplying a password on the command line.

RC File

You can skip the connection dialog window securely by putting the connection information into an RC file and then using the --urlid switch to DatabaseManager or DatabaseManagerSwing. This strategy is great for adding launch menu items and/or launch icons to your desktop. You can set up one icon for each of the database accounts which you regularly use.

The default location for the RC file is dbmanager.rc in your home directory. The RC File Authentication Setup section explains how to put the connection information into this text file. If you also run SqlTool, then you can share the RC file with SqlTool by using a sym-link (if your operating system supports sym links), or by using the --rcfile switch for either SqlTool or DatabaseManagerSwing.


Use your operating system facilities to prevent others from reading your RC file, since it contains passwords.

To set up launch items/icons, first experiment on your command line to find exactly what command works. For example,

    java -cp /path/to/hsqldb.jar org.hsqldb.util.DatabaseManagerSwing --urlid mem
Then, use your window manager to add an item that runs this command.

Using the current DatabaseManagers with an older HSQLDB distribution.

This procedure will allow users of a legacy version of HSQLDB to use all of the new features of the DatabaseManagers. You will also get the new version of the SqlTool! This procedure works for distros going back to at least, probably much farther.

These instructions assume that you are capable of running an Ant build. See the Building HSQLDB chapter.

  1. Download and extract a current HSQLDB distribution. If you don't want to use the source code, documentation, etc., you can use a temporary directory and remove it afterwards.

  2. Cd to the build directory under the root directory where you extracted the distribution to.

  3. Run ant hsqldbutil.

  4. If you're going to wipe out the build directory, copy hsqldbutil.jar to a safe location first.

  5. For now on, whenver you are going to run DatabaseManager*, make sure that you have this hsqldbutil.jar as the first item in your CLASSPATH.

Here's a UNIX example where somebody wants to use the new DatabaseManagerSwing with their older HSQLDB database, as well as with Postgresql and a local application.

java org.hsqldb.util.DatabaseManagerSwing --urlid urlid

DatabaseManagerSwing as an Applet

DatabaseManagerSwing is also an applet. You can use it in HTML, JSPs, etc. Be aware that in Applet mode, actions to load or save local files will be disabled, and attempts to access any server other than the HTML-serving-host will be fail.

Since the Applet can not store or load locally saved preferences, the only way to have persistent preference settings is by using Applet parameters.

DatabaseManagerSwing Applet Parameters

URL of a data source to auto-connect to. String value.
URL of a data source to auto-connect to. String value. Defaults to org.hsqldb.jdbcDriver.
User name for data source to auto-connect to. String value.
Password for data source to auto-connect to. String value. Defaults to zero-length string.
Display only object from this schema in the object navigator. String value.
Look-and-feel. String value.
Auto-load sample data. Boolean value. Defaults to false.
Auto-refresh the object navigator when DDL modifications detected in user SQL commands. Boolean value. Defaults to true.
Show number of rows in each table in the object navigator. Boolean value. Defaults to false.
Show system tables in the object navigator. Boolean value. Defaults to false.
Show object names like schema.name in object navigator. Boolean value. Defaults to true.
Show query results in Gui grid (as opposed to in plain text). Boolean value. Defaults to true.
Show help hover-text. Boolean value. Defaults to true.

Appendix G. Transfer Tool

Fred Toussi

HSQLDB Development Group

$Date: 2005/06/29 23:15:13 $

Table of Contents

Brief Introduction

Brief Introduction

Transfer Tool is a GUI program for transferring SQL schema and data from one JDBC source to another. Source and destination can be different database engines or different databases on the same server.

Transfer Tool works in two different modes. Direct transfer maintains a connection to both source and destination and performs the transfer. Dump and Restore mode is invoked once to transfer the data from the source to a text file (Dump), then again to transfer the data from the text file to the destination (Restore). With Dump and Restore, it is possible to make any changes to database object definitions and data prior to restoring it to the target.

Dump and Restore modes can be set via the command line with -d (--dump) or -r (--restore) options. Alternatively the Transfer Tool can be started with any of the three modes from the Database Manager's Tools menu.

The connection dialogue allows you to save the settings for the connection you are about to make. You can then access the connection in future sessions. These settings are shared with those from the Database Manager tool. See the appendix on Database Manager for details of the connection dialogue box.

In version 1.8.0 Transfer Tool is no longer part of the hsqldb.jar. You can build the hsqldbutil.jar using the Ant command of the same name, to build a jar that includes Transfer Tool and the Database Manager.

When collecting meta-data, Transfer Tool performs SELECT * FROM <table> queries on all the tables in the source database. This may take a long time with some database engines. When the source database is HSQLDB, this means memory should be available for the result sets returned from the queries. Therefore, the memory allocation of the java process in which Transfer Tool is executed may have to be high.