<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
<head>
<meta name="generator" content=
"HTML Tidy for Linux/x86 (vers 12 April 2005), see www.w3.org" />
<title>Trie-Based Containers</title>
<meta http-equiv="Content-Type" content=
"text/html; charset=us-ascii" />
</head>
<body>
<div id="page">
<h1>Trie Design</h1>
<h2><a name="overview" id="overview">Overview</a></h2>
<p>The trie-based container has the following declaration:</p>
<pre>
<b>template</b><
<b>typename</b> Key,
<b>typename</b> Mapped,
<b>typename</b> Cmp_Fn = std::less<Key>,
<b>typename</b> Tag = <a href="pat_trie_tag.html">pat_trie_tag</a>,
<b>template</b><
<b>typename</b> Const_Node_Iterator,
<b>typename</b> Node_Iterator,
<b>typename</b> E_Access_Traits_,
<b>typename</b> Allocator_>
<b>class</b> Node_Update = <a href=
"null_trie_node_update.html">null_trie_node_update</a>,
<b>typename</b> Allocator = std::allocator<<b>char</b>> >
<b>class</b> <a href=
"trie.html">trie</a>;
</pre>
<p>The parameters have the following meaning:</p>
<ol>
<li><tt>Key</tt> is the key type.</li>
<li><tt>Mapped</tt> is the mapped-policy, and is explained in
<a href="tutorial.html#assoc_ms">Tutorial::Associative
Containers::Associative Containers Others than Maps</a>.</li>
<li><tt>E_Access_Traits</tt> is described in <a href=
"#e_access_traits">Element-Access Traits</a>.</li>
<li><tt>Tag</tt> specifies which underlying data structure
to use, and is described shortly.</li>
<li><tt>Node_Update</tt> is a policy for updating node
invariants. This is described in <a href="#invariants">Node
Invariants</a>.</li>
<li><tt>Allocator</tt> is an allocator
type.</li>
</ol>
<p>The <tt>Tag</tt> parameter specifies which underlying
data structure to use. Instantiating it by <a href=
"pat_trie_tag.html">pat_trie_tag</a>, specifies an
underlying PATRICIA trie (explained shortly); any other tag is
currently illegal.</p>
<hr />
<p>Following is a description of a (PATRICIA) trie
(<tt>pb_ds</tt> follows specifically [<a href=
"references.html#okasaki98mereable">okasaki98mereable</a>] and
[<a href=
"references.html#filliatre2000ptset">filliatre2000ptset</a>]).</p>
<p>A (PATRICIA) trie is similar to a tree, but with the
following differences:</p>
<ol>
<li>It explicitly views keys as a sequence of elements.
<i>E.g.</i>, a trie can view a string as a sequence of
characters; a trie can view a number as a sequence of
bits.</li>
<li>It is not (necessarily) binary. Each node has fan-out <i>n
+ 1</i>, where <i>n</i> is the number of distinct
elements.</li>
<li>It stores values only at leaf nodes.</li>
<li>Internal nodes have the properties that A) each has at
least two children, and B) each shares the same prefix with
any of its descendant.</li>
</ol>
<p><a href="#e_access_traits">Element-Access Traits</a> shows
an example of such a trie.</p>
<p>A (PATRICIA) trie has some useful properties:</p>
<ol>
<li>It can be configured to use large node fan-out, giving it
very efficient find performance (albeit at insertion
complexity and size).</li>
<li>It works well for common-prefix keys.</li>
<li>It can support efficiently queries such as which keys
match a certain prefix. This is sometimes useful in
file systems and routers.</li>
</ol>
<p>(We would like to thank Matt Austern for the suggestion to
include tries.)</p>
<h2><a name="e_access_traits" id=
"e_access_traits">Element-Access Traits</a></h2>
<p>A trie inherently views its keys as sequences of elements.
For example, a trie can view a string as a sequence of
characters. A trie needs to map each of <i>n</i> elements to a
number in <i>{0, n - 1}</i>. For example, a trie can map a
character <tt>c</tt> to
<tt>static_cast<size_t>(c)</tt>.</p>
<p>Seemingly, then, a trie can assume that its keys support
(const) iterators, and that the <tt>value_type</tt> of this
iterator can be cast to a <tt>size_t</tt>. There are several
reasons, though, to decouple the mechanism by which the trie
accesses its keys' elements from the trie:</p>
<ol>
<li>In some cases, the numerical value of an element is
inappropriate. Consider a trie storing DNA strings. It is
logical to use a trie with a fan-out of <i>5 = 1 + |{'A', 'C',
'G', 'T'}|</i>. This requires mapping 'T' to 3, though.</li>
<li>In some cases the keys' iterators are different than what
is needed. For example, a trie can be used to search for
common <u>suffixes</u>, by using strings'
<tt>reverse_iterator</tt>. As another example, a trie mapping
UNICODE strings would have a huge fan-out if each node would
branch on a UNICODE character; instead, one can define an
iterator iterating over 8-bit (or less) groups.</li>
</ol>
<p><a href=
"trie.html">trie</a> is,
consequently, parametrized by <tt>E_Access_Traits</tt> -
traits which instruct how to access sequences' elements.
<a href=
"string_trie_e_access_traits.html"><tt>string_trie_e_access_traits</tt></a>
is a traits class for strings. Each such traits define some
types, <i>e.g.</i>,</p>
<pre>
<b>typename</b> E_Access_Traits::const_iterator
</pre>
<p>is a const iterator iterating over a key's elements. The
traits class must also define methods for obtaining an iterator
to the first and last element of a key.</p>
<p>Figure <a href="#pat_trie">A PATRICIA trie</a> shows a
(PATRICIA) trie resulting from inserting the words: "I wish
that I could ever see a poem lovely as a trie" (which,
unfortunately, does not rhyme).</p>
<p>The leaf nodes contain values; each internal node contains
two <tt><b>typename</b> E_Access_Traits::const_iterator</tt>
objects, indicating the maximal common prefix of all keys in
the sub-tree. For example, the shaded internal node roots a
sub-tree with leafs "a" and "as". The maximal common prefix is
"a". The internal node contains, consequently, to const
iterators, one pointing to <tt>'a'</tt>, and the other to
<tt>'s'</tt>.</p>
<h6 class="c1"><a name="pat_trie" id="pat_trie"><img src=
"pat_trie.png" alt="no image" /></a></h6>
<h6 class="c1">A PATRICIA trie.</h6>
<h2><a name="invariants" id="invariants">Node
Invariants</a></h2>
<p>Trie-based containers support node invariants, as do
tree-based containers (see <a href=
"tree_based_containers.html#invariants">Tree-Based
Containers::Node Invariants</a>). There are two minor
differences, though, which, unfortunately, thwart sharing them
sharing the same node-updating policies:</p>
<ol>
<li>A trie's <tt>Node_Update</tt> template-template
parameter is parametrized by <tt>E_Access_Traits</tt>, while
a tree's <tt>Node_Update</tt> template-template parameter is
parametrized by <tt>Cmp_Fn</tt>.</li>
<li>Tree-based containers store values in all nodes, while
trie-based containers (at least in this implementation) store
values in leafs.</li>
</ol>
<p>Figure <a href="#trie_node_update_cd">A trie and its update
policy</a> shows the scheme, as well as some predefined
policies (which are explained below).</p>
<h6 class="c1"><a name="trie_node_update_cd" id=
"trie_node_update_cd"><img src=
"trie_node_update_policy_cd.png" alt="no image" /></a></h6>
<h6 class="c1">A trie and its update policy.</h6>
<p><tt>pb_ds</tt> offers the following pre-defined trie node
updating policies:</p>
<ol>
<li><a href=
"trie_order_statistics_node_update.html"><tt>trie_order_statistics_node_update</tt></a>
supports order statistics.</li>
<li><a href=
"trie_prefix_search_node_update.html"><tt>trie_prefix_search_node_update</tt></a>
supports searching for ranges that match a given prefix. See
<a href=
"http://gcc.gnu.org/viewcvs/*checkout*/trunk/libstdc%2B%2B-v3/testsuite/ext/pb_ds/example/trie_prefix_search.cc"><tt>trie_prefix_search.cc</tt></a>.</li>
<li><a href=
"null_trie_node_update.html"><tt>null_trie_node_update</tt></a>
is the null node updater.</li>
</ol>
<h2><a name="add_methods" id="add_methods">Additional
Methods</a></h2>
<p>Trie-based containers support split and join methods; the
rationale is equal to that of tree-based containers supporting
these methods (see <a href=
"tree_based_containers.html#add_methods">Tree-Based
Containers::Additional Methods</a>).</p>
</div>
</body>
</html>
