Release 10.1A: OpenEdge Data Management:
SQL Development

Using the SQLDUMP utility

The SQLDUMP utility is a command-line utility that dumps application data from SQL tables into one or more files. You can load the data from the files into another database with the SQLLOAD utility. The SQLDUMP utility does not dump data from Progress 4GL tables. The utility is available for only those tables that were created with SQL.

The SQLDUMP utility has the following syntax:

Syntax

Operating system	Syntax
UNIX Windows	sqldump -u user_name [ -a password ] [ -C code-page-name ] -t [ owner_name.]table_name1        [ [,owner_name.]table_name2, ... ] database_name

The SQLDUMP utility writes user data in row order into ASCII records with variable-length format. The column order in the files is identical to the column order in the tables. The utility writes both format and content header records to the dump file. You can dump multiple tables in a single execution by specifying multiple table names, separated by commas. Make sure there are no spaces before or after commas in the table list.

Data for one table always goes to a single dump file. Each dump file corresponds to one database table. For example, if you specify 200 tables in the SQLDUMP command, you will create 200 dump files. The SQLDUMP utility assigns the filenames that correspond to the owner_name and table_name in the database, with the file extension .dsql. If a dump file for a specified table already exists, it will be overwritten and replaced. Dump files are created in the current working directory.

The format of each record in a dump file is similar to the Progress 4GL .d file format:

Converts all values to character representation.
Delimits CHARACTER values with double quotes.
Can contain any embedded characters except NULL values, allowing commas, new lines, and other control characters.
Uses two sets of double quotes to escape embedded double quotes.
Delimits NUMERIC and other noncharacter data types using a space.
Processes TIMESTAMP data as if it were CHARACTER data.
Has a size limit of 2K for a single column value.
Has a maximum record length of 32K for dump file records.

Any error is a fatal error, and SQLDUMP halts the dumping process so that data integrity will not be compromised. SQLDUMP reports errors to standard output.

After successful processing, SQLDUMP writes a summary report to standard output. For each table SQLDUMP processes, the report shows:

Table name.
Dump filename.
Number of records dumped.
Number of bytes dumped.
Number of seconds required for processing.

Example 5–26 directs the SQLDUMP utility to write the data from two tables to two dump files. The user_name and password for connecting to the database are tucker and sulky. The tucker account must have the authority to access the customers and products tables in database salesdb with owner_name martin.

Example 5–26: SQLDUMP from selected tables

sqldump -u tucker -a sulky -t martin.customers,martin.products progress:T:thunder:4077:salesdb

Example 5–27 directs the SQLDUMP utility to write the data from all tables in the salesdb database that begin with any of these strings: cust, invent, and sales, and having any owner name that the user tucker has authority to access. The user_name and password for connecting to the database are tucker and sulky.

Example 5–27: SQLDUMP based on table names

sqldump -u tucker -a sulky -t%.cust%,%.invent%,%.sales% progress:T:thunder:4077:salesdb

Example 5–28 directs the SQLDUMP utility to write the data from all tables for all owner names in the salesdb database.

Example 5–28: SQLDUMP of entire database

sqldump -u tucker -a sulky s-t %.% progress:T:thunder:4077:salesdb

Notes: The database_name must be the last parameter given. Each dump file records character set information in the identifier section of each file.

Example 5–29 depicts a dump file.

Example 5–29: Dump file

A^B^CProgress     sqlschema       v1.0       Quote fmt A^B^CTimestamp    1999-10-19   19:06:49:0000 A^B^CDatabase     dumpdb.db A^B^CProgress Character Set: iso8859-1 A^B^CJava Charcter Set: Unicode UTF-8 A^B^CDate Format: MM/DD/YYYY

The character set recorded in the dump file is the client character set. The default character set for all non-JDBC clients is taken from the local operating system through the operating system APIs. JDBC clients use the Unicode UTF-8 character set.

To use a character set different than that used by the operating system, set the SQL_CLIENT_CHARSET environment variable to the name of the preferred character set. You can define any Progress 4GL-supported character set name. The name is not case-sensitive.

SQLDUMP does not support the following characters in schema names:

Double quote (")
Forward slash (/)
Backslash (\)

SQLDUMP, however, does support schema names that contain special characters such as a blank space, a hyphen (-), or a pound sign (#). These names must be used as delimited identifiers. Therefore, when specifying names with special characters on a UNIX command line, follow these rules:

Use double quotes to delimit identifiers.
So that the command line does not strip the quotes, use a backslash (\) to escape the double quotes used for delimited identifiers.
Use double quotes to enclose any names with embedded spaces, commas, or characters special to a command shell (such as the Bourne shell). This use of quotes is in addition to quoting delimited identifiers.

For example, to dump the table Yearly Profits, use the following UNIX command-line syntax:

Syntax

sqldump -t "\"Yearly Profits\"" -u xxx -a yyy database_name

In Windows, the command interpreter rules for the use of double quotation marks varies from UNIX.

By default, SQLDUMP displays promsgs messages using the code page corresponding to code-page-name. That is, if you are dumping a Russian database, and code-page-name specifies the name of a Russian code page, the client displays promsgs messages using the Russian code page (unless you specify a different code page by setting the client’s SQL_CLIENT_CHARSET_PROMSGS environment variable).