docs/utility/db_codegen.html

<!--$Id: db_codegen.so,v 10.6 2007/05/17 18:29:34 bostic Exp $-->
<!--Copyright (c) 1997,2008 Oracle.  All rights reserved.-->
<!--See the file LICENSE for redistribution information.-->
<html>
<head>
<title>Berkeley DB: db_codegen</title>
<meta name="description" content="Berkeley DB: An embedded database programmatic toolkit.">
<meta name="keywords" content="embedded,database,programmatic,toolkit,btree,hash,hashing,transaction,transactions,locking,logging,access method,access methods,Java,C,C++">
</head>
<body bgcolor=white>
<table width="100%"><tr valign=top>
<td>
<b>db_codegen</b>
</td>
</tr></table>
<hr size=1 noshade>
<tt>
<b><pre>db_codegen [<b>-Vv</b>] [<b>-a c</b>] [<b>-i file</b>] [<b>-o file</b>]</pre></b>
<b>Description</b>
<a name="2"><!--meow--></a>
<p>The db_codegen utility generates application code to create and
configure Berkeley DB database environments and databases based on a simple
description language, and writes it to one or more output files.  The
generated code may need modification, in the case of complicated
applications, but will usually significantly reduce the time required
to create Berkeley DB applications.</p>
<p>The options are as follows:</p>
<br>
<b>-a</b><ul compact><li>Generate code for the specified API (currently, only "c" is accepted).</ul>
<b>-i</b><ul compact><li>Specify an input file; by default, stdin is used.</ul>
<b>-o</b><ul compact><li>Specify an output file prefix; by default, "application" is used.</ul>
<b>-V</b><ul compact><li>Write the library version number to the standard output, and exit.</ul>
<b>-v</b><ul compact><li>Run in verbose mode.</ul>
<br>
<p>The db_codegen utility exits 0 on success, and &gt;0 if an error occurs.</p>
<b>C Language Specific Information</b>
<p>By default, when the db_codegen utility generates C-language
code, the output file is named "application.c".  The output filename
can be specified with <b>-o</b> option.</p>
<p>At the beginning of the output file is a list of public database
environment (<a href="/api_c/env_class.html">DB_ENV</a>) handles and database (<a href="/api_c/db_class.html">DB</a>) handles,
as specified by the description language.  The database environment
handle variables are named "XXX_dbenv", where "XXX" is the name of the
environment in the input specification.  For databases associated with
a database environment, the database handle variables are named
"XXX_YYY", where "XXX" is the name of the environment, and "YYY" is the
name of the database.  For standalone databases, the database handle
variables are named "XXX", where "XXX" is the name of the database.</p>
<p>There are two public functions in the output file: bdb_startup and
bdb_shutdown.  The bdb_startup function should be called to create and
configure the database environments and databases, and the bdb_shutdown
function should be called to gracefully shut down the environments and
databases.</p>
<b>Specification Language</b>
<p>The db_codegen uses a simple description language:</p>
<p><ul type=disc>
<li>Lines in the input consist of white-space separated tokens.
<li>Tokens are case-insensitive.
<li>Empty lines, and lines where the first non-space character
is hash mark ("#"). are ignored.  In addition, hash marks may appear
in lines, in which case the content of the line from the hash mark to
the end of the line is ignored.
</ul>
<p>There are two top-level objects: "environment" and "database", which
correspond to database environments and databases, respectively.  These
top-level objects can be associated with keywords to describe their
configuration and relationships.</p>
<p>For example, the following input would create two standalone databases:</p>
<blockquote><pre>database data_one {
    type btree
}
database data_two {
    type btree
}</pre></blockquote>
<p>In this case, there would be no <a href="/api_c/env_class.html">DB_ENV</a> handle, and the public
<a href="/api_c/db_class.html">DB</a> handles would be:</p>
<blockquote><pre>DB      *data_one;
DB      *data_two;</pre></blockquote>
<p>For example, the following input would create a database environment
which contains three databases:</p>
<blockquote><pre>environment myenv {
    database data_one {
	type btree
    }
    database data_two {
    	type btree
    }
    database data_three {
    	type btree
    }
}</pre></blockquote>
<p>In this case, the public <a href="/api_c/env_class.html">DB_ENV</a> and <a href="/api_c/db_class.html">DB</a> handles would be:</p>
<blockquote><pre>DB_ENV  *myenv_dbenv;
DB      *myenv_data_one;
DB      *myenv_data_two;
DB      *myenv_data_three;</pre></blockquote>
<p>A variety of keywords can be specified for the databases and the
environments.  For example, the cache size can be specified for
the database environment, and the page size can be specified for
the database, as well as secondary relationships:</p>
<blockquote><pre>environment myenv {
    cachesize 2 0 10
    database data_one {
    	type btree
    	pagesize 1024
    }
    database data_two {
	primary data_one
	secondary_offset 10 15
    	type btree
    	pagesize 32768
    }
    database data_three {
    	type btree
    	pagesize 512
    }
}</pre></blockquote>
<b>Environment Keywords</b>
<br>
<b>environment</b><ul compact><li>Start a database environment block.
<p>There must be three tokens on the line: the keyword, the name of the
environment and an opening brace ("{").</p></ul>
<b>home</b><ul compact><li>Specify the database environment home directory.
<p>There must be two tokens on the line: the keyword, and the home
directory.</p></ul>
<b>cachesize</b><ul compact><li>Specify the database environment cache size.
<p>There must be two tokens on the line: the keyword, the gigabytes of
cache, the bytes of cache, and the number of caches (the number of
underlying physical areas into which the cache is logically
divided).</p></ul>
<b>private</b><ul compact><li>Specify the database environment is private.
<p>There must be one token on the line: the keyword by itself.</p></ul>
<b>}</b><ul compact><li>End the database environment block.
<p>There must be one token on the line: the keyword by itself.</p></ul>
<br>
<b>Database Keywords</b>
<br>
<b>database</b><ul compact><li>Start a database block.
<p>There must be three tokens on the line: the keyword, the name of the
database and an opening brace ("{").</p></ul>
<b>custom</b><ul compact><li>Specify a custom key-comparison routine.  This is used when the Btree
database requires a specific sort that db_codegen cannot
generate.  A stub key comparison routine will be created and configured
for the database which should be modified as necessary.  See the
"key_type" keyword for more information.
<p>There must be one token on the line: the keyword by itself.</p></ul>
<b>dupsort</b><ul compact><li>Configure the database to support sorted duplicates.
<p>There must be one token on the line: the keyword by itself.</p></ul>
<b>extentsize</b><ul compact><li>Configure the size of the Queue database extent files.
<p>There must be two tokens on the line: the keyword, and the extent file size,
as a number of pages.</p></ul>
<b>key_type</b><ul compact><li>Configure a integral type key-comparison routine.  This is used when the
Btree database Btree database key is an integral type (such as "unsigned
int", or "u_int32_t").  Any C-language integral type may be specified.
See the "custom" keyword for more information.  A Btree comparison routine
based on the type of the key will be created and configured.
<p>There must be two tokens on the line: the keyword, and the type.</p></ul>
<b>pagesize</b><ul compact><li>Configure the database page size.
<p>There must be two tokens on the line: the keyword, and the page size
in bytes.</p></ul>
<b>primary</b><ul compact><li>Configure the database as a secondary index.  A stub secondary callback
routine will be created and configured for the database, which should
be modified as necessary.  See the "secondary_offset" keyword for more
information.
<p>name of the primary database for which this database is a secondary.</p></ul>
<b>recnum</b><ul compact><li>Configure the Btree database to support record number access.
<p>There must be one token on the line: the keyword by itself.</p></ul>
<b>re_len</b><ul compact><li>Configure the record length for a Queue database or a fixed-length Recno
database.
<p>There must be two tokens on the line: the keyword, and the length of
a record, in bytes.</p></ul>
<b>secondary_offset</b><ul compact><li>Configure a secondary callback routine based on a byte string found in
the primary database's data item.
<p>There must be three tokens on the line: the keyword, the byte offset
from the beginning of the primary data item where the secondary key
occurs, and the length of the secondary key in bytes.</p></ul>
<b>transaction</b><ul compact><li>Configure the database (and, by extension, the database environment),
to be transactional.
<p>There must be one token on the line: the keyword by itself.</p></ul>
<b>type</b><ul compact><li>Configure the database type.
<p>There must be two tokens on the line: the keyword, and the type, where
the type is one of "btree", "hash", "queue" or "recno".</p></ul>
<b>}</b><ul compact><li>End the database environment block.
<p>There must be one token on the line: the keyword by itself.</p></ul>
<br>
</tt>
<p><font size=1>Copyright (c) 1996,2008 Oracle.  All rights reserved.</font>
</body>
</html>