Documentation Source Text

Artifact Content
Login

Artifact a8a0df078ef88fcae2c17fc3fd3e07693a07026b:


<title>Architecture of SQLite</title>
<h2>The Architecture Of SQLite</h2>

<h3>Introduction</h3>

<table align="right" border="1" cellpadding="15" cellspacing="1">
<tr><th>Block Diagram Of SQLite</th></tr>
<tr><td><img src="images/arch2.gif"></td></tr>
</table>
<p>This document describes the architecture of the SQLite library.
The information here is useful to those who want to understand or
modify the inner workings of SQLite.
</p>

<p>
A block diagram showing the main components of SQLite
and how they interrelate is shown at the right.  The text that
follows will provide a quick overview of each of these components.
</p>


<p>
This document describes SQLite version 3.0.  Version 2.8 and
earlier are similar but the details differ.
</p>

<h3>Interface</h3>

<p>Much of the public interface to the SQLite library is implemented by
functions found in the <b>main.c</b>, <b>legacy.c</b>, and
<b>vdbeapi.c</b> source files
though some routines are
scattered about in other files where they can have access to data 
structures with file scope.  The
<b>sqlite3_get_table()</b> routine is implemented in <b>table.c</b>.
<b>sqlite3_mprintf()</b> is found in <b>printf.c</b>.
<b>sqlite3_complete()</b> is in <b>tokenize.c</b>.
The Tcl interface is implemented by <b>tclsqlite.c</b>.  More
information on the C interface to SQLite is
<a href="capi3ref.html">available separately</a>.<p>

<p>To avoid name collisions with other software, all external
symbols in the SQLite library begin with the prefix <b>sqlite3</b>.
Those symbols that are intended for external use (in other words,
those symbols which form the API for SQLite) begin
with <b>sqlite3_</b>.</p>

<h3>Tokenizer</h3>

<p>When a string containing SQL statements is to be executed, the
interface passes that string to the tokenizer.  The job of the tokenizer
is to break the original string up into tokens and pass those tokens
one by one to the parser.  The tokenizer is hand-coded in C in 
the file <b>tokenize.c</b>.

<p>Note that in this design, the tokenizer calls the parser.  People
who are familiar with YACC and BISON may be used to doing things the
other way around -- having the parser call the tokenizer.  The author
of SQLite 
has done it both ways and finds things generally work out nicer for
the tokenizer to call the parser.  YACC has it backwards.</p>

<h3>Parser</h3>

<p>The parser is the piece that assigns meaning to tokens based on
their context.  The parser for SQLite is generated using the
<a href="http://www.hwaci.com/sw/lemon/">Lemon</a> LALR(1) parser
generator.  Lemon does the same job as YACC/BISON, but it uses
a different input syntax which is less error-prone.
Lemon also generates a parser which is reentrant and thread-safe.
And lemon defines the concept of a non-terminal destructor so
that it does not leak memory when syntax errors are encountered.
The source file that drives Lemon is found in <b>parse.y</b>.</p>

<p>Because
lemon is a program not normally found on development machines, the
complete source code to lemon (just one C file) is included in the
SQLite distribution in the "tool" subdirectory.  Documentation on
lemon is found in the "doc" subdirectory of the distribution.
</p>

<h3>Code Generator</h3>

<p>After the parser assembles tokens into complete SQL statements,
it calls the code generator to produce virtual machine code that
will do the work that the SQL statements request.  There are many
files in the code generator:
<b>attach.c</b>,
<b>auth.c</b>,
<b>build.c</b>,
<b>delete.c</b>,
<b>expr.c</b>,
<b>insert.c</b>,
<b>pragma.c</b>,
<b>select.c</b>,
<b>trigger.c</b>,
<b>update.c</b>,
<b>vacuum.c</b>
and <b>where.c</b>.
In these files is where most of the serious magic happens.
<b>expr.c</b> handles code generation for expressions.
<b>where.c</b> handles code generation for WHERE clauses on
SELECT, UPDATE and DELETE statements.  The files <b>attach.c</b>,
<b>delete.c</b>, <b>insert.c</b>, <b>select.c</b>, <b>trigger.c</b>
<b>update.c</b>, and <b>vacuum.c</b> handle the code generation
for SQL statements with the same names.  (Each of these files calls routines
in <b>expr.c</b> and <b>where.c</b> as necessary.)  All other
SQL statements are coded out of <b>build.c</b>.
The <b>auth.c</b> file implements the functionality of
<b>sqlite3_set_authorizer()</b>.</p>

<h3>Virtual Machine</h3>

<p>The program generated by the code generator is executed by
the virtual machine.  Additional information about the virtual
machine is <a href="opcode.html">available separately</a>.
To summarize, the virtual machine implements an abstract computing
engine specifically designed to manipulate database files.  The
machine has a stack which is used for intermediate storage.
Each instruction contains an opcode and
up to three additional operands.</p>

<p>The virtual machine itself is entirely contained in a single
source file <b>vdbe.c</b>.  The virtual machine also has
its own header files: <b>vdbe.h</b> that defines an interface
between the virtual machine and the rest of the SQLite library and
<b>vdbeInt.h</b> which defines structure private the virtual machine.
The <b>vdbeaux.c</b> file contains utilities used by the virtual
machine and interface modules used by the rest of the library to
construct VM programs.  The <b>vdbeapi.c</b> file contains external
interfaces to the virtual machine such as the 
<b>sqlite3_bind_...</b> family of functions.  Individual values
(strings, integer, floating point numbers, and BLOBs) are stored
in an internal object named "Mem" which is implemented by
<b>vdbemem.c</b>.</p>

<p>
SQLite implements SQL functions using callbacks to C-language routines.
Even the built-in SQL functions are implemented this way.  Most of
the built-in SQL functions (ex: <b>coalesce()</b>, <b>count()</b>,
<b>substr()</b>, and so forth) can be found in <b>func.c</b>.
Date and time conversion functions are found in <b>date.c</b>.
</p>

<h3>B-Tree</h3>

<p>An SQLite database is maintained on disk using a B-tree implementation
found in the <b>btree.c</b> source file.  A separate B-tree is used for
each table and index in the database.  All B-trees are stored in the
same disk file.  Details of the file format are recorded in a large
comment at the beginning of <b>btree.c</b>.</p>

<p>The interface to the B-tree subsystem is defined by the header file
<b>btree.h</b>.
</p>

<h3>Page Cache</h3>

<p>The B-tree module requests information from the disk in fixed-size
chunks.  The default chunk size is 1024 bytes but can vary between 512
and 65536 bytes.
The page cache is responsible for reading, writing, and
caching these chunks.
The page cache also provides the rollback and atomic commit abstraction
and takes care of locking of the database file.  The
B-tree driver requests particular pages from the page cache and notifies
the page cache when it wants to modify pages or commit or rollback
changes and the page cache handles all the messy details of making sure
the requests are handled quickly, safely, and efficiently.</p>

<p>The code to implement the page cache is contained in the single C
source file <b>pager.c</b>.  The interface to the page cache subsystem
is defined by the header file <b>pager.h</b>.
</p>

<h3>OS Interface</h3>

<p>
In order to provide portability between POSIX and Win32 operating systems,
SQLite uses an abstraction layer to interface with the operating system.
The interface to the OS abstraction layer is defined in
<b>os.h</b>.  Each supported operating system has its own implementation:
<b>os_unix.c</b> for Unix, <b>os_win.c</b> for windows, and so forth.
Each of these operating-specific implements typically has its own
header file: <b>os_unix.h</b>, <b>os_win.h</b>, etc.
</p>

<h3>Utilities</h3>

<p>
Memory allocation and caseless string comparison routines are located
in <b>util.c</b>.
Symbol tables used by the parser are maintained by hash tables found
in <b>hash.c</b>.  The <b>utf.c</b> source file contains Unicode
conversion subroutines.
SQLite has its own private implementation of <b>printf()</b> (with
some extensions) in <b>printf.c</b> and its own random number generator
in <b>random.c</b>.
</p>

<h3>Test Code</h3>

<p>
If you count regression test scripts,
more than half the total code base of SQLite is devoted to testing.
There are many <b>assert()</b> statements in the main code files.
In additional, the source files <b>test1.c</b> through <b>test5.c</b>
together with <b>md5.c</b> implement extensions used for testing
purposes only.  The <b>os_test.c</b> backend interface is used to
simulate power failures to verify the crash-recovery mechanism in
the pager.
</p>