Alessandro Bonazzi 5c7aa8c1c0 Patch level : 12.0 no-patch
Files correlati     :
Commento            :

Aggiunta documentazione di sqlite 3
2020-11-29 00:32:36 +01:00

1563 lines
66 KiB
HTML

<!DOCTYPE html>
<html><head>
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<link href="sqlite.css" rel="stylesheet">
<title>The SQLite Query Optimizer Overview</title>
<!-- path= -->
</head>
<body>
<div class=nosearch>
<a href="index.html">
<img class="logo" src="images/sqlite370_banner.gif" alt="SQLite" border="0">
</a>
<div><!-- IE hack to prevent disappearing logo --></div>
<div class="tagline desktoponly">
Small. Fast. Reliable.<br>Choose any three.
</div>
<div class="menu mainmenu">
<ul>
<li><a href="index.html">Home</a>
<li class='mobileonly'><a href="javascript:void(0)" onclick='toggle_div("submenu")'>Menu</a>
<li class='wideonly'><a href='about.html'>About</a>
<li class='desktoponly'><a href="docs.html">Documentation</a>
<li class='desktoponly'><a href="download.html">Download</a>
<li class='wideonly'><a href='copyright.html'>License</a>
<li class='desktoponly'><a href="support.html">Support</a>
<li class='desktoponly'><a href="prosupport.html">Purchase</a>
<li class='search' id='search_menubutton'>
<a href="javascript:void(0)" onclick='toggle_search()'>Search</a>
</ul>
</div>
<div class="menu submenu" id="submenu">
<ul>
<li><a href='about.html'>About</a>
<li><a href='docs.html'>Documentation</a>
<li><a href='download.html'>Download</a>
<li><a href='support.html'>Support</a>
<li><a href='prosupport.html'>Purchase</a>
</ul>
</div>
<div class="searchmenu" id="searchmenu">
<form method="GET" action="search">
<select name="s" id="searchtype">
<option value="d">Search Documentation</option>
<option value="c">Search Changelog</option>
</select>
<input type="text" name="q" id="searchbox" value="">
<input type="submit" value="Go">
</form>
</div>
</div>
<script>
function toggle_div(nm) {
var w = document.getElementById(nm);
if( w.style.display=="block" ){
w.style.display = "none";
}else{
w.style.display = "block";
}
}
function toggle_search() {
var w = document.getElementById("searchmenu");
if( w.style.display=="block" ){
w.style.display = "none";
} else {
w.style.display = "block";
setTimeout(function(){
document.getElementById("searchbox").focus()
}, 30);
}
}
function div_off(nm){document.getElementById(nm).style.display="none";}
window.onbeforeunload = function(e){div_off("submenu");}
/* Disable the Search feature if we are not operating from CGI, since */
/* Search is accomplished using CGI and will not work without it. */
if( !location.origin || !location.origin.match || !location.origin.match(/http/) ){
document.getElementById("search_menubutton").style.display = "none";
}
/* Used by the Hide/Show button beside syntax diagrams, to toggle the */
function hideorshow(btn,obj){
var x = document.getElementById(obj);
var b = document.getElementById(btn);
if( x.style.display!='none' ){
x.style.display = 'none';
b.innerHTML='show';
}else{
x.style.display = '';
b.innerHTML='hide';
}
return false;
}
</script>
</div>
<div class=fancy>
<div class=nosearch>
<div class="fancy_title">
The SQLite Query Optimizer Overview
</div>
<div class="fancy_toc">
<a onclick="toggle_toc()">
<span class="fancy_toc_mark" id="toc_mk">&#x25ba;</span>
Table Of Contents
</a>
<div id="toc_sub"><div class="fancy-toc1"><a href="#introduction">1. Introduction</a></div>
<div class="fancy-toc1"><a href="#where_clause_analysis">2. WHERE clause analysis</a></div>
<div class="fancy-toc2"><a href="#index_term_usage_examples">2.1. Index term usage examples</a></div>
<div class="fancy-toc1"><a href="#the_between_optimization">3. The BETWEEN optimization</a></div>
<div class="fancy-toc1"><a href="#or_optimizations">4. OR optimizations</a></div>
<div class="fancy-toc1"><a href="#the_like_optimization">5. The LIKE optimization</a></div>
<div class="fancy-toc1"><a href="#the_skip_scan_optimization">6. The Skip-Scan Optimization</a></div>
<div class="fancy-toc1"><a href="#joins">7. Joins</a></div>
<div class="fancy-toc2"><a href="#order_of_tables_in_a_join">7.1. Order of tables in a join</a></div>
<div class="fancy-toc2"><a href="#manual_control_of_query_plans_using_sqlite_stat_tables">7.2. Manual Control Of Query Plans Using SQLITE_STAT Tables</a></div>
<div class="fancy-toc2"><a href="#manual_control_of_query_plans_using_cross_join">7.3. Manual Control Of Query Plans Using CROSS JOIN</a></div>
<div class="fancy-toc1"><a href="#choosing_between_multiple_indices">8. Choosing between multiple indices</a></div>
<div class="fancy-toc2"><a href="#disqualifying_where_clause_terms_using_unary_">8.1. Disqualifying WHERE Clause Terms Using Unary-"+"</a></div>
<div class="fancy-toc2"><a href="#range_queries">8.2. Range Queries</a></div>
<div class="fancy-toc1"><a href="#covering_indices">9. Covering Indices</a></div>
<div class="fancy-toc1"><a href="#order_by_optimizations">10. ORDER BY optimizations</a></div>
<div class="fancy-toc2"><a href="#partial_order_by_via_index">10.1. Partial ORDER BY Via Index</a></div>
<div class="fancy-toc1"><a href="#subquery_flattening">11. Subquery flattening</a></div>
<div class="fancy-toc1"><a href="#subquery_co_routines">12. Subquery Co-routines</a></div>
<div class="fancy-toc2"><a href="#using_co_routines_to_defer_work_until_after_the_sorting">12.1. Using Co-routines To Defer Work Until After The Sorting</a></div>
<div class="fancy-toc1"><a href="#the_min_max_optimization">13. The MIN/MAX optimization</a></div>
<div class="fancy-toc1"><a href="#automatic_indexes">14. Automatic Indexes</a></div>
<div class="fancy-toc1"><a href="#the_push_down_optimization">15. The Push-Down Optimization</a></div>
<div class="fancy-toc1"><a href="#the_left_join_strength_reduction_optimization">16. The LEFT JOIN Strength Reduction Optimization</a></div>
<div class="fancy-toc1"><a href="#the_omit_left_join_optimization">17. The Omit LEFT JOIN Optimization</a></div>
</div>
</div>
<script>
function toggle_toc(){
var sub = document.getElementById("toc_sub")
var mk = document.getElementById("toc_mk")
if( sub.style.display!="block" ){
sub.style.display = "block";
mk.innerHTML = "&#x25bc;";
} else {
sub.style.display = "none";
mk.innerHTML = "&#x25ba;";
}
}
</script>
</div>
<h1 id="introduction"><span>1. </span>Introduction</h1>
<p>
This document provides an overview of how the query planner and optimizer
for SQLite works.
</p><p>
Given a single SQL statement, there might be dozens, hundreds, or even
thousands of ways to implement that statement, depending on the complexity
of the statement itself and of the underlying database schema. The
task of the query planner is to select an algorithm from among the many
choices that provides the answer with a minimum of disk I/O and CPU
overhead.
</p><p>
Additional background information is available in the
<a href="queryplanner.html">indexing tutorial</a> document.
</p><p>
With release 3.8.0 (2013-08-26),
the SQLite query planner was reimplemented as the
<a href="queryplanner-ng.html">Next Generation Query Planner</a> or "NGQP". All of the features, techniques,
and algorithms described in this document are applicable to both the
pre-3.8.0 legacy query planner and to the NGQP. For further information on
how the NGQP differs from the legacy query planner, see the
<a href="queryplanner-ng.html">detailed description of the NGQP</a>.
<a name="where_clause"></a>
</p><h1 id="where_clause_analysis"><span>2. </span>WHERE clause analysis</h1>
<p>
The WHERE clause on a query is broken up into "terms" where each term
is separated from the others by an AND operator.
If the WHERE clause is composed of constraints separate by the OR
operator then the entire clause is considered to be a single "term"
to which the <a href="#or_opt">OR-clause optimization</a> is applied.
</p><p>
All terms of the WHERE clause are analyzed to see if they can be
satisfied using indices.
To be usable by an index a term must be of one of the following
forms:
<blockquote><pre><b>
</b><i>column</i><b> = </b><i>expression</i><b>
</b><i>column</i><b> IS </b><i>expression</i><b>
</b><i>column</i><b> &gt; </b><i>expression</i><b>
</b><i>column</i><b> &gt;= </b><i>expression</i><b>
</b><i>column</i><b> &lt; </b><i>expression</i><b>
</b><i>column</i><b> &lt;= </b><i>expression</i><b>
</b><i>expression</i><b> = </b><i>column</i><b>
</b><i>expression</i><b> &gt; </b><i>column</i><b>
</b><i>expression</i><b> &gt;= </b><i>column</i><b>
</b><i>expression</i><b> &lt; </b><i>column</i><b>
</b><i>expression</i><b> &lt;= </b><i>column</i><b>
</b><i>column</i><b> IN (</b><i>expression-list</i><b>)
</b><i>column</i><b> IN (</b><i>subquery</i><b>)
</b><i>column</i><b> IS NULL
</b></pre></blockquote>
</p><p>
If an index is created using a statement like this:
</p><div class="codeblock"><pre>CREATE INDEX idx_ex1 ON ex1(a,b,c,d,e,...,y,z);
</pre></div>
<p>
Then the index might be used if the initial columns of the index
(columns a, b, and so forth) appear in WHERE clause terms.
The initial columns of the index must be used with
the *=* or *IN* or *IS* operators.
The right-most column that is used can employ inequalities.
For the right-most
column of an index that is used, there can be up to two inequalities
that must sandwich the allowed values of the column between two extremes.
</p><p>
It is not necessary for every column of an index to appear in a
WHERE clause term in order for that index to be used.
But there cannot be gaps in the columns of the index that are used.
Thus for the example index above, if there is no WHERE clause term
that constraints column c, then terms that constrain columns a and b can
be used with the index but not terms that constraint columns d through z.
Similarly, index columns will not normally be used (for indexing purposes)
if they are to the right of a
column that is constrained only by inequalities.
(See the <a href="optoverview.html#skipscan">skip-scan optimization</a> below for the exception.)
</p><p>
In the case of <a href="expridx.html">indexes on expressions</a>, whenever the word "column" is
used in the foregoing text, one can substitute "indexed expression"
(meaning a copy of the expression that appears in the <a href="lang_createindex.html">CREATE INDEX</a>
statement) and everything will work the same.
<a name="idxexamp"></a>
</p><h2 id="index_term_usage_examples"><span>2.1. </span>Index term usage examples</h2>
<p>
For the index above and WHERE clause like this:
</p><div class="codeblock"><pre>... WHERE a=5 AND b IN (1,2,3) AND c IS NULL AND d='hello'
</pre></div>
<p>
The first four columns a, b, c, and d of the index would be usable since
those four columns form a prefix of the index and are all bound by
equality constraints.
</p><p>
For the index above and WHERE clause like this:
</p><div class="codeblock"><pre>... WHERE a=5 AND b IN (1,2,3) AND c>12 AND d='hello'
</pre></div>
<p>
Only columns a, b, and c of the index would be usable. The d column
would not be usable because it occurs to the right of c and c is
constrained only by inequalities.
</p><p>
For the index above and WHERE clause like this:
</p><div class="codeblock"><pre>... WHERE a=5 AND b IN (1,2,3) AND d='hello'
</pre></div>
<p>
Only columns a and b of the index would be usable. The d column
would not be usable because column c is not constrained and there can
be no gaps in the set of columns that usable by the index.
</p><p>
For the index above and WHERE clause like this:
</p><div class="codeblock"><pre>... WHERE b IN (1,2,3) AND c NOT NULL AND d='hello'
</pre></div>
<p>
The index is not usable at all because the left-most column of the
index (column "a") is not constrained. Assuming there are no other
indices, the query above would result in a full table scan.
</p><p>
For the index above and WHERE clause like this:
</p><div class="codeblock"><pre>... WHERE a=5 OR b IN (1,2,3) OR c NOT NULL OR d='hello'
</pre></div>
<p>
The index is not usable because the WHERE clause terms are connected
by OR instead of AND. This query would result in a full table scan.
However, if three additional indices where added that contained columns
b, c, and d as their left-most columns, then the
<a href="#or_opt">OR-clause optimization</a> might apply.
<a name="between_opt"></a>
</p><h1 id="the_between_optimization"><span>3. </span>The BETWEEN optimization</h1>
<p>
If a term of the WHERE clause is of the following form:
<blockquote><pre><b>
</b><i>expr1</i><b> BETWEEN </b><i>expr2</i><b> AND </b><i>expr3</i><b>
</b></pre></blockquote>
</p><p>
Then two "virtual" terms are added as follows:
<blockquote><pre><b>
</b><i>expr1</i><b> &gt;= </b><i>expr2</i><b> AND </b><i>expr1</i><b> &lt;= </b><i>expr3</i><b>
</b></pre></blockquote>
</p><p>
Virtual terms are used for analysis only and do not cause any byte-code
to be generated.
If both virtual terms end up being used as constraints on an index,
then the original BETWEEN term is omitted and the corresponding test
is not performed on input rows.
Thus if the BETWEEN term ends up being used as an index constraint
no tests are ever performed on that term.
On the other hand, the
virtual terms themselves never causes tests to be performed on
input rows.
Thus if the BETWEEN term is not used as an index constraint and
instead must be used to test input rows, the <i>expr1</i> expression is
only evaluated once.
<a name="or_opt"></a>
</p><h1 id="or_optimizations"><span>4. </span>OR optimizations</h1>
<p>
WHERE clause constraints that are connected by OR instead of AND can
be handled in two different ways.
If a term consists of multiple subterms containing a common column
name and separated by OR, like this:
<blockquote><pre><b>
</b><i>column</i><b> = </b><i>expr1</i><b> OR </b><i>column</i><b> = </b><i>expr2</i><b> OR </b><i>column</i><b> = </b><i>expr3</i><b> OR ...
</b></pre></blockquote>
</p><p>
Then that term is rewritten as follows:
<blockquote><pre><b>
</b><i>column</i><b> IN (</b><i>expr1</i><b>,</b><i>expr2</i><b>,</b><i>expr3</i><b>,...)
</b></pre></blockquote>
</p><p>
The rewritten term then might go on to constrain an index using the
normal rules for *IN* operators. Note that <i>column</i> must be
the same column in every OR-connected subterm,
although the column can occur on either the left or the right side of
the *=* operator.
</p><p>
If and only if the previously described conversion of OR to an IN operator
does not work, the second OR-clause optimization is attempted.
Suppose the OR clause consists of multiple subterms as follows:
<blockquote><pre><b>
</b><i>expr1</i><b> OR </b><i>expr2</i><b> OR </b><i>expr3</i><b>
</b></pre></blockquote>
</p><p>
Individual subterms might be a single comparison expression like
*a=5* or *x>y* or they can be LIKE or BETWEEN expressions, or a subterm
can be a parenthesized list of AND-connected sub-subterms.
Each subterm is analyzed as if it were itself the entire WHERE clause
in order to see if the subterm is indexable by itself.
If <u>every</u> subterm of an OR clause is separately indexable
then the OR clause might be coded such that a separate index is used
to evaluate each term of the OR clause. One way to think about how
SQLite uses separate indices for each OR clause term is to imagine
that the WHERE clause where rewritten as follows:
<blockquote><pre><b>
rowid IN (SELECT rowid FROM </b><i>table</i><b> WHERE </b><i>expr1</i><b>
UNION SELECT rowid FROM </b><i>table</i><b> WHERE </b><i>expr2</i><b>
UNION SELECT rowid FROM </b><i>table</i><b> WHERE </b><i>expr3</i><b>)
</b></pre></blockquote>
</p><p>
The rewritten expression above is conceptual; WHERE clauses containing
OR are not really rewritten this way.
The actual implementation of the OR clause uses a mechanism that is
more efficient and that works even for <a href="withoutrowid.html">WITHOUT ROWID</a> tables or
tables in which the "rowid" is inaccessible.
But the essence of the implementation is captured by the statement
above: Separate indices are used to find candidate result rows
from each OR clause term and the final result is the union of
those rows.
</p><p>
Note that in most cases, SQLite will only use a single index for each
table in the FROM clause of a query. The second OR-clause optimization
described here is the exception to that rule. With an OR-clause,
a different index might be used for each subterm in the OR-clause.
</p><p>
For any given query, the fact that the OR-clause optimization described
here can be used does not guarantee that it will be used.
SQLite uses a cost-based query planner that estimates the CPU and
disk I/O costs of various competing query plans and chooses the plan
that it thinks will be the fastest. If there are many OR terms in
the WHERE clause or if some of the indices on individual OR-clause
subterms are not very selective, then SQLite might decide that it is
faster to use a different query algorithm, or even a full-table scan.
Application developers can use the
<a href="lang_explain.html">EXPLAIN QUERY PLAN</a> prefix on a statement to get a
high-level overview of the chosen query strategy.
<a name="like_opt"></a>
</p><h1 id="the_like_optimization"><span>5. </span>The LIKE optimization</h1>
<p>
A WHERE-clause term that uses the <a href="lang_expr.html#like">LIKE</a> or <a href="lang_expr.html#glob">GLOB</a> operator
can sometimes be used with an index to do a range search,
almost as if the LIKE or GLOB were an alternative to a <a href="lang_expr.html#between">BETWEEN</a>
operator.
There are many conditions on this optimization:
</p><p>
</p><ol>
<li>The right-hand side of the LIKE or GLOB must be either a string literal
or a <a href="lang_expr.html#varparam">parameter</a> bound to a string literal
that does not begin with a wildcard character.</li>
<li>It must not be possible to make the LIKE or GLOB operator true by
having a numeric value (instead of a string or blob) on the
left-hand side. This means that either:
<ol type="A">
<li> the left-hand side of the LIKE or GLOB operator is the name
of an indexed column with <a href="datatype3.html#affinity">TEXT affinity</a>, or</li>
<li> the right-hand side pattern argument does not begin with a
minus sign ("-") or a digit.</li>
</ol>
This constraint arises from the fact that numbers do not sort in
lexicographical order. For example: 9&lt;10 but '9'&gt;'10'.</li>
<li>The built-in functions used to implement LIKE and GLOB must not
have been overloaded using the <a href="c3ref/create_function.html">sqlite3_create_function()</a> API.</li>
<li>For the GLOB operator, the column must be indexed using the
built-in BINARY collating sequence.</li>
<li>For the LIKE operator, if <a href="pragma.html#pragma_case_sensitive_like">case_sensitive_like</a> mode is enabled then
the column must indexed using BINARY collating sequence, or if
<a href="pragma.html#pragma_case_sensitive_like">case_sensitive_like</a> mode is disabled then the column must indexed
using built-in NOCASE collating sequence.</li>
<li>If the ESCAPE option is used, the ESCAPE character must be ASCII,
or a single-byte character in UTF-8.
</li></ol>
<p>
The LIKE operator has two modes that can be set by a
<a href="pragma.html#pragma_case_sensitive_like">pragma</a>. The
default mode is for LIKE comparisons to be insensitive to differences
of case for latin1 characters. Thus, by default, the following
expression is true:
</p><div class="codeblock"><pre>'a' LIKE 'A'
</pre></div>
<p>
But if the case_sensitive_like pragma is enabled as follows:
</p><div class="codeblock"><pre>PRAGMA case_sensitive_like=ON;
</pre></div>
<p>
Then the LIKE operator pays attention to case and the example above would
evaluate to false. Note that case insensitivity only applies to
latin1 characters - basically the upper and lower case letters of English
in the lower 127 byte codes of ASCII. International character sets
are case sensitive in SQLite unless an application-defined
<a href="datatype3.html#collation">collating sequence</a> and <a href="lang_corefunc.html#like">like() SQL function</a> are provided that
take non-ASCII characters into account.
But if an application-defined collating sequence and/or like() SQL
function are provided, the LIKE optimization described here will never
be taken.
</p><p>
The LIKE operator is case insensitive by default because this is what
the SQL standard requires. You can change the default behavior at
compile time by using the <a href="compile.html#case_sensitive_like">SQLITE_CASE_SENSITIVE_LIKE</a> command-line option
to the compiler.
</p><p>
The LIKE optimization might occur if the column named on the left of the
operator is indexed using the built-in BINARY collating sequence and
case_sensitive_like is turned on. Or the optimization might occur if
the column is indexed using the built-in NOCASE collating sequence and the
case_sensitive_like mode is off. These are the only two combinations
under which LIKE operators will be optimized.
</p><p>
The GLOB operator is always case sensitive. The column on the left side
of the GLOB operator must always use the built-in BINARY collating sequence
or no attempt will be made to optimize that operator with indices.
</p><p>
The LIKE optimization will only be attempted if
the right-hand side of the GLOB or LIKE operator is either
literal string or a <a href="lang_expr.html#varparam">parameter</a> that has been <a href="c3ref/bind_blob.html">bound</a>
to a string literal. The string literal must not
begin with a wildcard; if the right-hand side begins with a wildcard
character then this optimization is not attempted. If the right-hand side
is a <a href="lang_expr.html#varparam">parameter</a> that is bound to a string, then this optimization is
only attempted if the <a href="c3ref/stmt.html">prepared statement</a> containing the expression
was compiled with <a href="c3ref/prepare.html">sqlite3_prepare_v2()</a> or <a href="c3ref/prepare.html">sqlite3_prepare16_v2()</a>.
The LIKE optimization is not attempted if the
right-hand side is a <a href="lang_expr.html#varparam">parameter</a> and the statement was prepared using
<a href="c3ref/prepare.html">sqlite3_prepare()</a> or <a href="c3ref/prepare.html">sqlite3_prepare16()</a>.
</p><p>
Suppose the initial sequence of non-wildcard characters on the right-hand
side of the LIKE or GLOB operator is <i>x</i>. We are using a single
character to denote this non-wildcard prefix but the reader should
understand that the prefix can consist of more than 1 character.
Let <i>y</i> be the smallest string that is the same length as /x/ but which
compares greater than <i>x</i>. For example, if <i>x</i> is *hello* then
<i>y</i> would be *hellp*.
The LIKE and GLOB optimizations consist of adding two virtual terms
like this:
<blockquote><pre><b>
</b><i>column</i><b> &gt;= </b><i>x</i><b> AND </b><i>column</i><b> &lt; </b><i>y</i><b>
</b></pre></blockquote>
</p><p>
Under most circumstances, the original LIKE or GLOB operator is still
tested against each input row even if the virtual terms are used to
constrain an index. This is because we do not know what additional
constraints may be imposed by characters to the right
of the <i>x</i> prefix. However, if there is only a single
global wildcard to the right of <i>x</i>, then the original LIKE or
GLOB test is disabled.
In other words, if the pattern is like this:
<blockquote><pre><b>
</b><i>column</i><b> LIKE </b><i>x</i><b>%
</b><i>column</i><b> GLOB </b><i>x</i><b>*
</b></pre></blockquote>
</p><p>
then the original LIKE or GLOB tests are disabled when the virtual
terms constrain an index because in that case we know that all of the
rows selected by the index will pass the LIKE or GLOB test.
</p><p>
Note that when the right-hand side of a LIKE or GLOB operator is
a <a href="lang_expr.html#varparam">parameter</a> and the statement is prepared using <a href="c3ref/prepare.html">sqlite3_prepare_v2()</a>
or <a href="c3ref/prepare.html">sqlite3_prepare16_v2()</a> then the statement is automatically reparsed
and recompiled on the first <a href="c3ref/step.html">sqlite3_step()</a> call of each run if the binding
to the right-hand side parameter has changed since the previous run.
This reparse and recompile is essentially the same action that occurs
following a schema change. The recompile is necessary so that the query
planner can examine the new value bound to the right-hand side of the
LIKE or GLOB operator and determine whether or not to employ the
optimization described above.
<a name="skipscan"></a>
</p><h1 id="the_skip_scan_optimization"><span>6. </span>The Skip-Scan Optimization</h1>
<p>
The general rule is that indexes are only useful if there are
WHERE-clause constraints on the left-most columns of the index.
However, in some cases,
SQLite is able to use an index even if the first few columns of
the index are omitted from the WHERE clause but later columns
are included.
</p><p>
Consider a table such as the following:
</p><div class="codeblock"><pre>CREATE TABLE people(
name TEXT PRIMARY KEY,
role TEXT NOT NULL,
height INT NOT NULL, -- in cm
CHECK( role IN ('student','teacher') )
);
CREATE INDEX people_idx1 ON people(role, height);
</pre></div>
<p>
The people table has one entry for each person in a large
organization. Each person is either a "student" or a "teacher",
as determined by the "role" field. And we record the height in
centimeters of each person. The role and height are indexed.
Notice that the left-most column of the index is not very
selective - it only contains two possible values.
</p><p>
Now consider a query to find the names of everyone in the
organization that is 180cm tall or taller:
</p><div class="codeblock"><pre>SELECT name FROM people WHERE height>=180;
</pre></div>
<p>
Because the left-most column of the index does not appear in the
WHERE clause of the query, one is tempted to conclude that the
index is not usable here. But SQLite is able to use the index.
Conceptually, SQLite uses the index as if the query were more
like the following:
</p><div class="codeblock"><pre>SELECT name FROM people
WHERE role IN (SELECT DISTINCT role FROM people)
AND height>=180;
</pre></div>
<p>
Or this:
</p><div class="codeblock"><pre>SELECT name FROM people WHERE role='teacher' AND height>=180
UNION ALL
SELECT name FROM people WHERE role='student' AND height>=180;
</pre></div>
<p>
The alternative query formulations shown above are conceptual only.
SQLite does not really transform the query.
The actual query plan is like this:
SQLite locates the first possible value for "role", which it
can do by rewinding the "people_idx1" index to the beginning and reading
the first record. SQLite stores this first "role" value in an
internal variable that we will here call "$role". Then SQLite
runs a query like: "SELECT name FROM people WHERE role=$role AND height>=180".
This query has an equality constraint on the left-most column of the
index and so the index can be used to resolve that query. Once
that query is finished, SQLite then uses the "people_idx1" index to
locate the next value of the "role" column, using code that is logically
similar to "SELECT role FROM people WHERE role>$role LIMIT 1".
This new "role" value overwrites the $role variable, and the process
repeats until all possible values for "role" have been examined.
</p><p>
We call this kind of index usage a "skip-scan" because the database
engine is basically doing a full scan of the index but it optimizes the
scan (making it less than "full") by occasionally skipping ahead to the
next candidate value.
</p><p>
SQLite might use a skip-scan on an index if it knows that the first
one or more columns contain many duplication values.
If there are too few duplicates
in the left-most columns of the index, then it would
be faster to simply step ahead to the next value, and thus do
a full table scan, than to do a binary search on an index to locate
the next left-column value.
</p><p>
The only way that SQLite can know that the left-most columns of an index
have many duplicate is if the <a href="lang_analyze.html">ANALYZE</a> command has been run
on the database.
Without the results of ANALYZE, SQLite has to guess at the "shape" of
the data in the table, and the default guess is that there are an average
of 10 duplicates for every value in the left-most column of the index.
But skip-scan only becomes profitable (it only gets to be faster than
a full table scan) when the number of duplicates is about 18 or more.
Hence, a skip-scan is never used on a database that has not been analyzed.
<a name="joins"></a>
</p><h1 id="joins"><span>7. </span>Joins</h1>
<p>
The ON and USING clauses of an inner join are converted into additional
terms of the WHERE clause prior to WHERE clause analysis described
above in paragraph 1.0. Thus with SQLite, there is no computational
advantage to use the newer SQL92 join syntax
over the older SQL89 comma-join syntax. They both end up accomplishing
exactly the same thing on inner joins.
</p><p>
For a LEFT OUTER JOIN the situation is more complex. The following
two queries are not equivalent:
</p><div class="codeblock"><pre>SELECT * FROM tab1 LEFT JOIN tab2 ON tab1.x=tab2.y;
SELECT * FROM tab1 LEFT JOIN tab2 WHERE tab1.x=tab2.y;
</pre></div>
<p>
For an inner join, the two queries above would be identical. But
special processing applies to the ON and USING clauses of an OUTER join:
specifically, the constraints in an ON or USING clause do not apply if
the right table of the join is on a null row, but the constraints do apply
in the WHERE clause. The net effect is that putting the ON or USING
clause expressions for a LEFT JOIN in the WHERE clause effectively converts
the query to an
ordinary INNER JOIN - albeit an inner join that runs more slowly.
<a name="table_order"></a>
</p><h2 id="order_of_tables_in_a_join"><span>7.1. </span>Order of tables in a join</h2>
<p>
The current implementation of
SQLite uses only loop joins. That is to say, joins are implemented as
nested loops.
</p><p>
The default order of the nested loops in a join is for the left-most
table in the FROM clause to form the outer loop and the right-most
table to form the inner loop.
However, SQLite will nest the loops in a different order if doing so
will help it to select better indices.
</p><p>
Inner joins can be freely reordered. However a left outer join is
neither commutative nor associative and hence will not be reordered.
Inner joins to the left and right of the outer join might be reordered
if the optimizer thinks that is advantageous but the outer joins are
always evaluated in the order in which they occur.
</p><p>
SQLite <a href="lang_select.html#crossjoin">treats the CROSS JOIN operator specially</a>.
The CROSS JOIN operator is commutative in theory. But SQLite chooses to
never reorder tables in a CROSS JOIN. This provides a mechanism
by which the programmer can force SQLite to choose a particular loop nesting
order.
</p><p>
When selecting the order of tables in a join, SQLite uses an efficient
polynomial-time algorithm. Because of this,
SQLite is able to plan queries with 50- or 60-way joins in a matter of
microseconds
</p><p>
Join reordering is automatic and usually works well enough that
programmers do not have to think about it, especially if <a href="lang_analyze.html">ANALYZE</a>
has been used to gather statistics about the available indices.
But occasionally some hints from the programmer are needed.
Consider, for example, the following schema:
</p><div class="codeblock"><pre>CREATE TABLE node(
id INTEGER PRIMARY KEY,
name TEXT
);
CREATE INDEX node_idx ON node(name);
CREATE TABLE edge(
orig INTEGER REFERENCES node,
dest INTEGER REFERENCES node,
PRIMARY KEY(orig, dest)
);
CREATE INDEX edge_idx ON edge(dest,orig);
</pre></div>
<p>
The schema above defines a directed graph with the ability to store a
name at each node. Now consider a query against this schema:
</p><div class="codeblock"><pre>SELECT *
FROM edge AS e,
node AS n1,
node AS n2
WHERE n1.name = 'alice'
AND n2.name = 'bob'
AND e.orig = n1.id
AND e.dest = n2.id;
</pre></div>
<p>
This query asks for is all information about edges that go from
nodes labeled "alice" to nodes labeled "bob".
The query optimizer in SQLite has basically two choices on how to
implement this query. (There are actually six different choices, but
we will only consider two of them here.)
Pseudocode below demonstrating these two choices.
<a name="option1"></a>
</p><p>Option 1:
</p><div class="codeblock"><pre>foreach n1 where n1.name='alice' do:
foreach n2 where n2.name='bob' do:
foreach e where e.orig=n1.id and e.dest=n2.id
return n1.*, n2.*, e.*
end
end
end
</pre></div>
<a name="option2"></a>
<p>Option 2:
</p><div class="codeblock"><pre>foreach n1 where n1.name='alice' do:
foreach e where e.orig=n1.id do:
foreach n2 where n2.id=e.dest and n2.name='bob' do:
return n1.*, n2.*, e.*
end
end
end
</pre></div>
<p>
The same indices are used to speed up every loop in both implementation
options.
The only difference in these two query plans is the order in which
the loops are nested.
</p><p>
So which query plan is better? It turns out that the answer depends on
what kind of data is found in the node and edge tables.
</p><p>
Let the number of alice nodes be M and the number of bob nodes be N.
Consider two scenarios. In the first scenario, M and N are both 2 but
there are thousands of edges on each node. In this case, option 1 is
preferred. With option 1, the inner loop checks for the existence of
an edge between a pair of nodes and outputs the result if found.
But because there are only 2 alice and bob nodes each, the inner loop
only has to run 4 times and the query is very quick. Option 2 would
take much longer here. The outer loop of option 2 only executes twice,
but because there are a large number of edges leaving each alice node,
the middle loop has to iterate many thousands of times. It will be
much slower. So in the first scenario, we prefer to use option 1.
</p><p>
Now consider the case where M and N are both 3500. Alice nodes are
abundant. But suppose each of these nodes is connected by only one
or two edges. In this case, option 2 is preferred. With option 2,
the outer loop still has to run 3500 times, but the middle loop only
runs once or twice for each outer loop and the inner loop will only
run once for each middle loop, if at all. So the total number of
iterations of the inner loop is around 7000. Option 1, on the other
hand, has to run both its outer loop and its middle loop 3500 times
each, resulting in 12 million iterations of the middle loop.
Thus in the second scenario, option 2 is nearly 2000 times faster
than option 1.
</p><p>
So you can see that depending on how the data is structured in the table,
either query plan 1 or query plan 2 might be better. Which plan does
SQLite choose by default? As of version 3.6.18, without running <a href="lang_analyze.html">ANALYZE</a>,
SQLite will choose option 2.
But if the <a href="lang_analyze.html">ANALYZE</a> command is run in order to gather statistics,
a different choice might be made if the statistics indicate that the
alternative is likely to run faster.
<a name="manctrl"></a>
</p><h2 id="manual_control_of_query_plans_using_sqlite_stat_tables"><span>7.2. </span>Manual Control Of Query Plans Using SQLITE_STAT Tables</h2>
<p>
SQLite provides the ability for advanced programmers to exercise control
over the query plan chosen by the optimizer. One method for doing this
is to fudge the <a href="lang_analyze.html">ANALYZE</a> results in the <a href="fileformat2.html#stat1tab">sqlite_stat1</a>,
<a href="fileformat2.html#stat3tab">sqlite_stat3</a>, and/or <a href="fileformat2.html#stat4tab">sqlite_stat4</a> tables. That approach is not
recommended except for the one scenario described in the next paragraph.
</p><p>
For a program that uses an SQLite database as its
<a href="appfileformat.html">application file-format</a>,
when a new database instance is first created the <a href="lang_analyze.html">ANALYZE</a>
command is ineffective because the database contain no data from which
to gather statistics. In that case, one could construct a large prototype
database containing typical data during development and run the
<a href="lang_analyze.html">ANALYZE</a> command on this prototype database to gather statistics,
then save the prototype statistics as part of the application.
After deployment, when the application goes to create a new database file,
it can run the <a href="lang_analyze.html">ANALYZE</a> command in order to create the statistics
tables, then copy the precomputed statistics obtained
from the prototype database into these new statistics tables.
In that way, statistics from large working data sets can be preloaded
into newly created application files.
<a name="crossjoin"></a>
</p><h2 id="manual_control_of_query_plans_using_cross_join"><span>7.3. </span>Manual Control Of Query Plans Using CROSS JOIN</h2>
<p>
Programmers can force SQLite to use a particular loop nesting order
for a join by using the CROSS JOIN operator instead of just JOIN,
INNER JOIN, NATURAL JOIN, or a "," join. Though CROSS JOINs are
commutative in theory, SQLite chooses to never reorder the tables in
a CROSS JOIN. Hence, the left table of a CROSS JOIN will always be
in an outer loop relative to the right table.
</p><p>
In the following query, the optimizer is free to reorder the
tables of FROM clause anyway it sees fit:
</p><div class="codeblock"><pre>SELECT *
FROM node AS n1,
edge AS e,
node AS n2
WHERE n1.name = 'alice'
AND n2.name = 'bob'
AND e.orig = n1.id
AND e.dest = n2.id;
</pre></div>
<p>
But in the following logically equivalent formulation of the same query,
the substitution of "CROSS JOIN" for the "," means that the order
of tables must be N1, E, N2.
</p><div class="codeblock"><pre>SELECT *
FROM node AS n1 CROSS JOIN
edge AS e CROSS JOIN
node AS n2
WHERE n1.name = 'alice'
AND n2.name = 'bob'
AND e.orig = n1.id
AND e.dest = n2.id;
</pre></div>
<p>
In the latter query, the query plan must be
<a href="#option2">option 2</a>. Note that
you must use the keyword "CROSS" in order to disable the table reordering
optimization; INNER JOIN, NATURAL JOIN, JOIN, and other similar
combinations work just like a comma join in that the optimizer is
free to reorder tables as it sees fit. (Table reordering is also
disabled on an outer join, but that is because outer joins are not
associative or commutative. Reordering tables in OUTER JOIN changes
the result.)
</p><p>
See "<a href="queryplanner-ng.html#fossilcasestudy">The Fossil NGQP Upgrade Case Study</a>" for another real-world example
of using CROSS JOIN to manually control the nesting order of a join.
The <a href="queryplanner-ng.html#howtofix">query planner checklist</a> found later in the same document provides
further guidance on manual control of the query planner.
<a name="multi_index"></a>
</p><h1 id="choosing_between_multiple_indices"><span>8. </span>Choosing between multiple indices</h1>
<p>
Each table in the FROM clause of a query can use at most one index
(except when the <a href="#or_opt">OR-clause optimization</a> comes into
play)
and SQLite strives to use at least one index on each table. Sometimes,
two or more indices might be candidates for use on a single table.
For example:
</p><div class="codeblock"><pre>CREATE TABLE ex2(x,y,z);
CREATE INDEX ex2i1 ON ex2(x);
CREATE INDEX ex2i2 ON ex2(y);
SELECT z FROM ex2 WHERE x=5 AND y=6;
</pre></div>
<p>
For the SELECT statement above, the optimizer can use the ex2i1 index
to lookup rows of ex2 that contain x=5 and then test each row against
the y=6 term. Or it can use the ex2i2 index to lookup rows
of ex2 that contain y=6 then test each of those rows against the
x=5 term.
</p><p>
When faced with a choice of two or more indices, SQLite tries to estimate
the total amount of work needed to perform the query using each option.
It then selects the option that gives the least estimated work.
</p><p>
To help the optimizer get a more accurate estimate of the work involved
in using various indices, the user may optionally run the <a href="lang_analyze.html">ANALYZE</a> command.
The <a href="lang_analyze.html">ANALYZE</a> command scans all indices of database where there might
be a choice between two or more indices and gathers statistics on the
selectiveness of those indices. The statistics gathered by
this scan are stored in special database tables names shows names all
begin with "<b>sqlite_stat</b>".
The content of these tables is not updated as the database
changes so after making significant changes it might be prudent to
rerun <a href="lang_analyze.html">ANALYZE</a>.
The results of an ANALYZE command are only available to database connections
that are opened after the ANALYZE command completes.
</p><p>
The various <b>sqlite_stat</b><i>N</i> tables contain information on how
selective the various indices are. For example, the <a href="fileformat2.html#stat1tab">sqlite_stat1</a>
table might indicate that an equality constraint on column x reduces the
search space to 10 rows on average, whereas an equality constraint on
column y reduces the search space to 3 rows on average. In that case,
SQLite would prefer to use index ex2i2 since that index is more selective.
<a name="uplus"></a>
</p><h2 id="disqualifying_where_clause_terms_using_unary_"><span>8.1. </span>Disqualifying WHERE Clause Terms Using Unary-"+"</h2>
<p>
Terms of the WHERE clause can be manually disqualified for use with
indices by prepending a unary *+* operator to the column name. The
unary *+* is a no-op and will not generate any byte code in the prepared
statement.
But the unary *+* operator will prevent the term from constraining an index.
So, in the example above, if the query were rewritten as:
</p><div class="codeblock"><pre>SELECT z FROM ex2 WHERE +x=5 AND y=6;
</pre></div>
<p>
The *+* operator on the *x* column will prevent that term from
constraining an index. This would force the use of the ex2i2 index.
</p><p>
Note that the unary *+* operator also removes
<a href="datatype3.html#affinity">type affinity</a> from
an expression, and in some cases this can cause subtle changes in
the meaning of an expression.
In the example above,
if column *x* has <a href="datatype3.html#affinity">TEXT affinity</a>
then the comparison "x=5" will be done as text. But the *+* operator
removes the affinity. So the comparison "+x=5" will compare the text
in column *x* with the numeric value 5 and will always be false.
<a name="rangequery"></a>
</p><h2 id="range_queries"><span>8.2. </span>Range Queries</h2>
<p>
Consider a slightly different scenario:
</p><div class="codeblock"><pre>CREATE TABLE ex2(x,y,z);
CREATE INDEX ex2i1 ON ex2(x);
CREATE INDEX ex2i2 ON ex2(y);
SELECT z FROM ex2 WHERE x BETWEEN 1 AND 100 AND y BETWEEN 1 AND 100;
</pre></div>
<p>
Further suppose that column x contains values spread out
between 0 and 1,000,000 and column y contains values
that span between 0 and 1,000. In that scenario,
the range constraint on column x should reduce the search space by
a factor of 10,000 whereas the range constraint on column y should
reduce the search space by a factor of only 10. So the ex2i1 index
should be preferred.
</p><p>
SQLite will make this determination, but only if it has been compiled
with <a href="compile.html#enable_stat3">SQLITE_ENABLE_STAT3</a> or <a href="compile.html#enable_stat4">SQLITE_ENABLE_STAT4</a>.
The <a href="compile.html#enable_stat3">SQLITE_ENABLE_STAT3</a> and <a href="compile.html#enable_stat4">SQLITE_ENABLE_STAT4</a> options causes
the <a href="lang_analyze.html">ANALYZE</a> command to collect a histogram of column content in the
<a href="fileformat2.html#stat3tab">sqlite_stat3</a> or <a href="fileformat2.html#stat4tab">sqlite_stat4</a> tables and to use this histogram to
make a better guess at the best query to use for range constraints
such as the above. The main difference between STAT3 and STAT4 is
that STAT3 records histogram data for only the left-most column of
an index whereas STAT4 records histogram data for all columns of an
index. For single-column indexes, STAT3 and STAT4 work the same.
</p><p>
The histogram data is only useful if the right-hand side of the constraint
is a simple compile-time constant or <a href="lang_expr.html#varparam">parameter</a> and not an expression.
</p><p>
Another limitation of the histogram data is that it only applies to the
left-most column on an index. Consider this scenario:
</p><div class="codeblock"><pre>CREATE TABLE ex3(w,x,y,z);
CREATE INDEX ex3i1 ON ex2(w, x);
CREATE INDEX ex3i2 ON ex2(w, y);
SELECT z FROM ex3 WHERE w=5 AND x BETWEEN 1 AND 100 AND y BETWEEN 1 AND 100;
</pre></div>
<p>
Here the inequalities are on columns x and y which are not the
left-most index columns. Hence, the histogram data which is collected no
left-most column of indices is useless in helping to choose between the
range constraints on columns x and y.
<a name="covidx"></a>
</p><h1 id="covering_indices"><span>9. </span>Covering Indices</h1>
<p>
When doing an indexed lookup of a row, the usual procedure is to
do a binary search on the index to find the index entry, then extract
the <a href="lang_createtable.html#rowid">rowid</a> from the index and use that <a href="lang_createtable.html#rowid">rowid</a> to do a binary search on
the original table. Thus a typical indexed lookup involves two
binary searches.
If, however, all columns that were to be fetched from the table are
already available in the index itself, SQLite will use the values
contained in the index and will never look up the original table
row. This saves one binary search for each row and can make many
queries run twice as fast.
</p><p>
When an index contains all of the data needed for a query and when the
original table never needs to be consulted, we call that index a
"covering index".
<a name="order_by"></a>
</p><h1 id="order_by_optimizations"><span>10. </span>ORDER BY optimizations</h1>
<p>
SQLite attempts to use an index to satisfy the ORDER BY clause of a
query when possible.
When faced with the choice of using an index to satisfy WHERE clause
constraints or satisfying an ORDER BY clause, SQLite does the same
cost analysis described above
and chooses the index that it believes will result in the fastest answer.
</p><p>
SQLite will also attempt to use indices to help satisfy GROUP BY clauses
and the DISTINCT keyword. If the nested loops of the join can be arranged
such that rows that are equivalent for the GROUP BY or for the DISTINCT are
consecutive, then the GROUP BY or DISTINCT logic can determine if the
current row is part of the same group or if the current row is distinct
simply by comparing the current row to the previous row.
This can be much faster than the alternative of comparing each row to
all prior rows.
<a name="partsort"></a>
</p><h2 id="partial_order_by_via_index"><span>10.1. </span>Partial ORDER BY Via Index</h2>
<p>
If a query contains an ORDER BY clause with multiple terms, it might
be that SQLite can use indices to cause rows to come out in the order
of some prefix of the terms in the ORDER BY but that later terms in
the ORDER BY are not satisfied. In that case, SQLite does block sorting.
Suppose the ORDER BY clause has four terms and the natural order of the
query results in rows appearing in order of the first two terms. As
each row is output by the query engine and enters the sorter, the
outputs in the current row corresponding to the first two terms of
the ORDER BY are compared against the previous row. If they have
changed, the current sort is finished and output and a new sort is
started. This results in a slightly faster sort. But the bigger
advantages are that many fewer rows need to be held in memory,
reducing memory requirements, and outputs can begin to appear before
the core query has run to completion.
<a name="flattening"></a>
</p><h1 id="subquery_flattening"><span>11. </span>Subquery flattening</h1>
<p>
When a subquery occurs in the FROM clause of a SELECT, the simplest
behavior is to evaluate the subquery into a transient table, then run
the outer SELECT against the transient table. But such a plan
can be suboptimal since the transient table will not have any indices
and the outer query (which is likely a join) will be forced to do a
full table scan on the transient table.
</p><p>
To overcome this problem, SQLite attempts to flatten subqueries in
the FROM clause of a SELECT.
This involves inserting the FROM clause of the subquery into the
FROM clause of the outer query and rewriting expressions in
the outer query that refer to the result set of the subquery.
For example:
</p><div class="codeblock"><pre> SELECT t1.a, t2.b FROM t2, (SELECT x+y AS a FROM t1 WHERE z<100) WHERE="1" a="1">5
</100)>
</pre></div>
<p>
Would be rewritten using query flattening as:
</p><div class="codeblock"><pre> SELECT t1.x+t1.y AS a, t2.b FROM t2, t1 WHERE z<100 AND="1" a="1">5
</100>
</pre></div>
<p>
There is a long list of conditions that must all be met in order for
query flattening to occur. Some of the constraints are marked as
obsolete by italic text. These extra constraints are retained in the
documentation to preserve the numbering of the other constraints.
</p><p>
Casual readers are not expected to understand all of these rules.
A key take-away from this section is that the rules for determining
when query flatting is safe and when it is unsafe are subtle and
complex. There have been multiple bugs over the years caused by
over-aggressive query flattening. On the other hand, performance
of complex queries and/or queries involving views tends to suffer
if query flattening is more conservative.
</p><p>
</p><ol>
<li value="1"> <i>(Obsolete. Query flattening is no longer
attempted for aggregate subqueries.)</i>
</li><li value="2"> <i>(Obsolete. Query flattening is no longer
attempted for aggregate subqueries.)</i>
</li><li value="3">
If the subquery is the right operand of a LEFT JOIN then
<ol type="a"><li> the subquery may not be a join, and
</li><li> the FROM clause of the subquery may
not contain a virtual table, and
</li><li> the outer query may not be an aggregate.</li></ol></li>
<li value="4"> The subquery is not DISTINCT.
</li><li value="5"> <i>(Subsumed into constraint 4)</i>
</li><li value="6"> <i>(Obsolete. Query flattening is no longer
attempted for aggregate subqueries.)</i>
</li><li value="7">
The subquery has a FROM clause.
</li><li value="8">
The subquery does not use LIMIT or the outer query is not a join.
</li><li value="9">
The subquery does not use LIMIT or the outer query does not use
aggregates.
</li><li value="10"> <i>(Restriction relaxed in 2005)</i>
</li><li value="11">
The subquery and the outer query do not both have ORDER BY clauses.
</li><li value="12"> <i>(Subsumed into constraint 3)</i>
</li><li value="13"> The subquery and outer query do not both use LIMIT.
</li><li value="14"> The subquery does not use OFFSET.
</li><li value="15">
If the outer query is part of a compound select, then the
subquery may not have a LIMIT clause.
</li><li value="16">
If the outer query is an aggregate, then the subquery may
not contain ORDER BY.
</li><li value="17">
If the sub-query is a compound SELECT, then
<ol type="'a'">
<li> all compound operators must be UNION ALL, and
</li><li> no terms with the subquery compound may be aggregate
or DISTINCT, and
</li><li> every term within the subquery must have a FROM clause, and
</li><li> the outer query may not be an aggregate, DISTINCT query, or join.
</li></ol>
The parent and sub-query may contain WHERE clauses. Subject to
rules (11), (12) and (13), they may also contain ORDER BY,
LIMIT and OFFSET clauses.
</li><li value="18">
If the sub-query is a compound select, then all terms of the
ORDER by clause of the parent must be simple references to
columns of the sub-query.
</li><li value="19">
If the subquery uses LIMIT then the outer query may not
have a WHERE clause.
</li><li value="20">
If the sub-query is a compound select, then it must not use
an ORDER BY clause.
</li><li value="21">
If the subquery uses LIMIT, then the outer query may not be
DISTINCT.
</li><li value="22"> The subquery may not be a recursive CTE.
</li><li value="23"> <i>(Subsumed into constraint 17d.)</i>
</li><li value="24"> <i>(Obsolete. Query flattening is no longer
attempted for aggregate subqueries.)</i>
</li></ol>
<p>
The casual reader is not expected to understand or remember any part of
the list above. The point of this list is to demonstrate
that the decision of whether or not to flatten a query is complex.
</p><p>
Query flattening is an important optimization when views are used as
each use of a view is translated into a subquery.
<a name="coroutines"></a>
</p><h1 id="subquery_co_routines"><span>12. </span>Subquery Co-routines</h1>
<p>
Prior to SQLite 3.7.15 (2012-12-12),
a subquery in the FROM clause would be
either flattened into the outer query, or else the subquery would be run
to completion
before the outer query started, the result set from the subquery
would be stored in a transient table,
and then the transient table would be used in the outer query. Newer
versions of SQLite have a third option, which is to implement the subquery
using a co-routine.
</p><p>
A co-routine is like a subroutine in that it runs in the same thread
as the caller and eventually returns control back to the caller. The
difference is that a co-routine also has the ability to return
before it has finished, and then resume where it left off the next
time it is called.
</p><p>
When a subquery is implemented as a co-routine, byte-code is generated
to implement the subquery as if it were a standalone query, except
instead of returning rows of results back to the application, the
co-routine yields control back to the caller after each row is computed.
The caller can then use that one computed row as part of its computation,
then invoke the co-routine again when it is ready for the next row.
</p><p>
Co-routines are better than storing the complete result set of the subquery
in a transient table because co-routines use less memory. With a co-routine,
only a single row of the result needs to be remembered, whereas all rows of
the result must be stored for a transient table. Also, because the
co-routine does not need to run to completion before the outer query
begins its work, the first rows of output can appear much sooner, and if
the overall query is aborted, less work is done overall.
</p><p>
On the other hand, if the result of the subquery must be scanned multiple
times (because, for example, it is just one table in a join) then it
is better to use a transient table to remember the entire result of the
subquery, in order to avoid computing the subquery more than once.
<a name="deferred_work"></a>
</p><h2 id="using_co_routines_to_defer_work_until_after_the_sorting"><span>12.1. </span>Using Co-routines To Defer Work Until After The Sorting</h2>
<p>
As of SQLite version 3.21.0 (2017-10-24), the query planner will
always prefer to use a co-routine to implement FROM-clause subqueries
that contains an ORDER BY clause and that are not part of a join when
the result set of the outer query is "complex". This feature allows
applications to shift expensive computations from before the
sorter until after the sorter, which can result in faster operation.
For example, consider this query:
</p><div class="codeblock"><pre>SELECT expensive_function(a) FROM tab ORDER BY date DESC LIMIT 5;
</pre></div>
<p>
The goal of this query is to compute some value for the five most
recent entries in the table. But in the query above, the
"expensive_function()" is invoked prior to the sort and thus is
invoked on every row of the table, even
rows that are ultimately omitted due to the LIMIT clause.
A co-routine can be used to work around this:
</p><div class="codeblock"><pre>SELECT expensive_function(a) FROM (
SELECT a FROM tab ORDER BY date DESC LIMIT 5
);
</pre></div>
<p>
In the revised query, the subquery implemented by a co-routine computes
the five most recent values for "a". Those five values are passed from the
co-routine up into the outer query where the "expensive_function()" is
invoked on only the specific rows that the application cares about.
</p><p>
The query planner in future versions of SQLite might grow smart enough
to make transformations such as the above automatically, in both directions.
That is to say, future versions of SQLite might transform queries of the
first form into the second, or queries written the second way into the
first. As of SQLite version 3.22.0 (2018-01-22), the query planner
will flatten the subquery if the outer query does not make use of any
user-defined functions or subqueries in its result set. For the examples
shown above, however, SQLite implements each of the queries as
written.
<a name="minmax"></a>
</p><h1 id="the_min_max_optimization"><span>13. </span>The MIN/MAX optimization</h1>
<p>
Queries that contain a single MIN() or MAX() aggregate function whose
argument is the left-most column of an index might be satisfied
by doing a single index lookup rather than by scanning the entire table.
Examples:
</p><div class="codeblock"><pre>SELECT MIN(x) FROM table;
SELECT MAX(x)+1 FROM table;
</pre></div>
<a name="autoindex"></a>
<h1 id="automatic_indexes"><span>14. </span>Automatic Indexes</h1>
<p>
When no indices are available to aid the evaluation of a query, SQLite
might create an automatic index that lasts only for the duration
of a single SQL statement.
Since the cost of constructing the automatic index is
O(NlogN) (where N is the number of entries in the table) and the cost of
doing a full table scan is only O(N), an automatic index will
only be created if SQLite expects that the lookup will be run more than
logN times during the course of the SQL statement. Consider an example:
</p><div class="codeblock"><pre>CREATE TABLE t1(a,b);
CREATE TABLE t2(c,d);
-- Insert many rows into both t1 and t2
SELECT * FROM t1, t2 WHERE a=c;
</pre></div>
<p>
In the query above, if both t1 and t2 have approximately N rows, then
without any indices the query will require O(N*N) time. On the other
hand, creating an index on table t2 requires O(NlogN) time and then use
that index to evaluate the query requires an additional O(NlogN) time.
In the absence of <a href="lang_analyze.html">ANALYZE</a> information, SQLite guesses that N is one
million and hence it believes that constructing the automatic index will
be the cheaper approach.
</p><p>
An automatic index might also be used for a subquery:
</p><div class="codeblock"><pre>CREATE TABLE t1(a,b);
CREATE TABLE t2(c,d);
-- Insert many rows into both t1 and t2
SELECT a, (SELECT d FROM t2 WHERE c=b) FROM t1;
</pre></div>
<p>
In this example, the t2 table is used in a subquery to translate values
of the t1.b column. If each table contains N rows, SQLite expects that
the subquery will run N times, and hence it will believe it is faster
to construct an automatic, transient index on t2 first and then use
that index to satisfy the N instances of the subquery.
</p><p>
The automatic indexing capability can be disabled at run-time using
the <a href="pragma.html#pragma_automatic_index">automatic_index pragma</a>. Automatic indexing is turned on by
default, but this can be changed so that automatic indexing is off
by default using the <a href="compile.html#default_automatic_index">SQLITE_DEFAULT_AUTOMATIC_INDEX</a> compile-time option.
The ability to create automatic indices can be completely disabled by
compiling with the <a href="compile.html#omit_automatic_index">SQLITE_OMIT_AUTOMATIC_INDEX</a> compile-time option.
</p><p>
In SQLite <a href="releaselog/3_8_0.html">version 3.8.0</a> (2013-08-26) and later,
an <a href="rescode.html#warning_autoindex">SQLITE_WARNING_AUTOINDEX</a> message is sent
to the <a href="errlog.html">error log</a> every time a statement is prepared that uses an
automatic index. Application developers can and should use these warnings
to identify the need for new persistent indices in the schema.
</p><p>
Do not confuse automatic indexes with the <a href="fileformat2.html#intschema">internal indexes</a> (having names
like "sqlite_autoindex_<i>table</i>_<i>N</i>") that are sometimes
created to implement a <a href="lang_createtable.html#primkeyconst">PRIMARY KEY constraint</a> or <a href="lang_createtable.html#uniqueconst">UNIQUE constraint</a>.
The automatic indexes described here exist only for the duration of a
single query, are never persisted to disk, and are only visible to a
single database connection. Internal indexes are part of the implementation
of PRIMARY KEY and UNIQUE constraints, are long-lasting and persisted
to disk, and are visible to all database connections. The term "autoindex"
appears in the names of <a href="fileformat2.html#intschema">internal indexes</a> for legacy reasons and does
not indicate that internal indexes and automatic indexes are related.
<a name="pushdown"></a>
</p><h1 id="the_push_down_optimization"><span>15. </span>The Push-Down Optimization</h1>
<p>
If a subquery cannot be <a href="optoverview.html#flattening">flattened</a> into the outer query, it might
still be possible to enhance performance by "pushing down" WHERE clause
terms from the outer query into the subquery. Consider an example:
</p><div class="codeblock"><pre>CREATE TABLE t1(a INT, b INT);
CREATE TABLE t2(x INT, y INT);
CREATE VIEW v1(a,b) AS SELECT DISTINCT a, b FROM t1;
SELECT x, y, b
FROM t2 JOIN v1 ON (x=a)
WHERE b BETWEEN 10 AND 20;
</pre></div>
<p>
The view v1 cannot be <a href="optoverview.html#flattening">flattened</a> because it is DISTINCT. It must
instead be run as a subquery with the results being stored in a
transient table, then the join is performed between t2 and the
transient table. The push-down optimization pushes down the
"b BETWEEN 10 AND 20" term into the view. This makes the transient
table smaller, and helps the subquery to run faster if there
is an index on t1.b. The resulting evaluation is like this:
</p><div class="codeblock"><pre>SELECT x, y, b
FROM t2
JOIN (SELECT DISTINCT a, b FROM t1 WHERE b BETWEEN 10 AND 20)
WHERE b BETWEEN 10 AND 20;
</pre></div>
<p>
The push-down optimization cannot always be used. For example,
if the subquery contains a LIMIT, then pushing down any part of
the WHERE clause from the outer query could change the result of
the inner query. There are other restrictions, explained in a
comment in the source code on the pushDownWhereTerms() routine
that implements this optimization.
<a name="leftjoinreduction"></a>
</p><h1 id="the_left_join_strength_reduction_optimization"><span>16. </span>The LEFT JOIN Strength Reduction Optimization</h1>
<p>
Sometimes a LEFT JOIN can be converted into an ordinary JOIN,
if there are terms in the WHERE clause that guarantee that the
two joins will give identical results. In particular, if any
column in the right-hand table of the LEFT JOIN must be non-NULL
in order for the WHERE clause to be true, then the LEFT JOIN is
demoted to an ordinary JOIN.
</p><p>
The prover that determines whether any column of the right-hand
table of a LEFT JOIN must be non-NULL in the WHERE clause is
imperfect. It sometimes returns a false negative. In other words,
it sometimes fails to reduce the strength of a LEFT JOIN when doing
so was in fact possible. For example, the prover does not know
the <a href="lang_datefunc.html">datetime() SQL function</a> will always return NULL if its first
argument is NULL, and so it will not recognize that the LEFT JOIN
in the following query could be strength-reduced:
</p><div class="codeblock"><pre>SELECT urls.url
FROM urls
LEFT JOIN
(SELECT *
FROM (SELECT url_id AS uid, max(retrieval_time) AS rtime
FROM lookups GROUP BY 1 ORDER BY 1)
WHERE uid IN (358341,358341,358341)
) recent
ON u.source_seed_id = recent.xyz OR u.url_id = recent.xyz
WHERE
DATETIME(recent.rtime) > DATETIME('now', '-5 days');
</pre></div>
<p>
It is possible that future enhancements to the prover might enable it
to recognize that NULL inputs to certain built-in functions
always result in a NULL answer. But not all built-in
functions have that property (for example <a href="lang_corefunc.html#coalesce">coalesce()</a>) and, of
course, the prover will never be able to reason about
<a href="appfunc.html">application-defined SQL functions</a>.
<a name="omitnoopjoin"></a>
</p><h1 id="the_omit_left_join_optimization"><span>17. </span>The Omit LEFT JOIN Optimization</h1>
<p>
Sometimes a LEFT JOIN can be completely omitted from a query without
changing the result. This can happen if all of the following are
true:
</p><p>
</p><ol>
<li> The query is not an aggregate
</li><li> Either the query is DISTINCT or else the ON or USING clause
on the LEFT JOIN constrains the join such that it matches
only a single row
</li><li> The right-hand table of the LEFT JOIN is not be used anywhere
in the query outside of its own USING or ON clause.
</li></ol>
<p>
LEFT JOIN elimination often comes up when LEFT JOINs are used
inside of views, and then the view is used in such as way that
none of the columns of the right-hand table of the LEFT JOIN are
referenced.
</p><p>
Here is a simple example of omitting a LEFT JOIN:
</p><div class="codeblock"><pre>CREATE TABLE t1(ipk INTEGER PRIMARY KEY, v1);
CREATE TABLE t2(ipk INTEGER PRIMARY KEY, v2);
CREATE TABLE t3(ipk INTEGER PRIMARY KEY, v3);
SELECT v1, v3 FROM t1
LEFT JOIN t2 USING (t1.ipk=t2.ipk)
LEFT JOIN t3 USING (t1.ipk=t3.ipk)
</pre></div>
<p>
The t2 table is completely unused in the query above, and so the
query planner is able to implement the query as if it were written:
</p><div class="codeblock"><pre>SELECT v1, v3 FROM t1
LEFT JOIN t3 USING (t1.ipk=t3.ipk)
</pre></div>