Account for prologue spills in reg_pressure scheduling
[official-gcc.git] / libstdc++-v3 / doc / html / manual / iterators.html
blob70dbe491be5ab3efc80a6311dd6857616a1ca87d
1 <?xml version="1.0" encoding="UTF-8" standalone="no"?>
2 <!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"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>Chapter 10.  Iterators</title><meta name="generator" content="DocBook XSL-NS Stylesheets V1.78.1" /><meta name="keywords" content="ISO C++, library" /><meta name="keywords" content="ISO C++, runtime, library" /><link rel="home" href="../index.html" title="The GNU C++ Library" /><link rel="up" href="std_contents.html" title="Part II.  Standard Contents" /><link rel="prev" href="containers_and_c.html" title="Interacting with C" /><link rel="next" href="algorithms.html" title="Chapter 11.  Algorithms" /></head><body><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="3" align="center">Chapter 10
3 Iterators
5 </th></tr><tr><td width="20%" align="left"><a accesskey="p" href="containers_and_c.html">Prev</a> </td><th width="60%" align="center">Part II. 
6 Standard Contents
7 </th><td width="20%" align="right"> <a accesskey="n" href="algorithms.html">Next</a></td></tr></table><hr /></div><div class="chapter"><div class="titlepage"><div><div><h2 class="title"><a id="std.iterators"></a>Chapter 10
8 Iterators
9 <a id="id-1.3.4.8.1.1.1" class="indexterm"></a>
10 </h2></div></div></div><div class="toc"><p><strong>Table of Contents</strong></p><dl class="toc"><dt><span class="section"><a href="iterators.html#std.iterators.predefined">Predefined</a></span></dt><dd><dl><dt><span class="section"><a href="iterators.html#iterators.predefined.vs_pointers">Iterators vs. Pointers</a></span></dt><dt><span class="section"><a href="iterators.html#iterators.predefined.end">One Past the End</a></span></dt></dl></dd></dl></div><div class="section"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="std.iterators.predefined"></a>Predefined</h2></div></div></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="iterators.predefined.vs_pointers"></a>Iterators vs. Pointers</h3></div></div></div><p>
11 The following
12 FAQ <a class="link" href="../faq.html#faq.iterator_as_pod" title="7.1.">entry</a> points out that
13 iterators are not implemented as pointers. They are a generalization
14 of pointers, but they are implemented in libstdc++ as separate
15 classes.
16 </p><p>
17 Keeping that simple fact in mind as you design your code will
18 prevent a whole lot of difficult-to-understand bugs.
19 </p><p>
20 You can think of it the other way 'round, even. Since iterators
21 are a generalization, that means
22 that <span class="emphasis"><em>pointers</em></span> are
23 <span class="emphasis"><em>iterators</em></span>, and that pointers can be used
24 whenever an iterator would be. All those functions in the
25 Algorithms section of the Standard will work just as well on plain
26 arrays and their pointers.
27 </p><p>
28 That doesn't mean that when you pass in a pointer, it gets
29 wrapped into some special delegating iterator-to-pointer class
30 with a layer of overhead. (If you think that's the case
31 anywhere, you don't understand templates to begin with...) Oh,
32 no; if you pass in a pointer, then the compiler will instantiate
33 that template using T* as a type, and good old high-speed
34 pointer arithmetic as its operations, so the resulting code will
35 be doing exactly the same things as it would be doing if you had
36 hand-coded it yourself (for the 273rd time).
37 </p><p>
38 How much overhead <span class="emphasis"><em>is</em></span> there when using an
39 iterator class? Very little. Most of the layering classes
40 contain nothing but typedefs, and typedefs are
41 "meta-information" that simply tell the compiler some
42 nicknames; they don't create code. That information gets passed
43 down through inheritance, so while the compiler has to do work
44 looking up all the names, your runtime code does not. (This has
45 been a prime concern from the beginning.)
46 </p></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="iterators.predefined.end"></a>One Past the End</h3></div></div></div><p>This starts off sounding complicated, but is actually very easy,
47 especially towards the end. Trust me.
48 </p><p>Beginners usually have a little trouble understand the whole
49 'past-the-end' thing, until they remember their early algebra classes
50 (see, they <span class="emphasis"><em>told</em></span> you that stuff would come in handy!) and
51 the concept of half-open ranges.
52 </p><p>First, some history, and a reminder of some of the funkier rules in
53 C and C++ for builtin arrays. The following rules have always been
54 true for both languages:
55 </p><div class="orderedlist"><ol class="orderedlist" type="1"><li class="listitem"><p>You can point anywhere in the array, <span class="emphasis"><em>or to the first element
56 past the end of the array</em></span>. A pointer that points to one
57 past the end of the array is guaranteed to be as unique as a
58 pointer to somewhere inside the array, so that you can compare
59 such pointers safely.
60 </p></li><li class="listitem"><p>You can only dereference a pointer that points into an array.
61 If your array pointer points outside the array -- even to just
62 one past the end -- and you dereference it, Bad Things happen.
63 </p></li><li class="listitem"><p>Strictly speaking, simply pointing anywhere else invokes
64 undefined behavior. Most programs won't puke until such a
65 pointer is actually dereferenced, but the standards leave that
66 up to the platform.
67 </p></li></ol></div><p>The reason this past-the-end addressing was allowed is to make it
68 easy to write a loop to go over an entire array, e.g.,
69 while (*d++ = *s++);.
70 </p><p>So, when you think of two pointers delimiting an array, don't think
71 of them as indexing 0 through n-1. Think of them as <span class="emphasis"><em>boundary
72 markers</em></span>:
73 </p><pre class="programlisting">
75 beginning end
76 | |
77 | | This is bad. Always having to
78 | | remember to add or subtract one.
79 | | Off-by-one bugs very common here.
80 V V
81 array of N elements
82 |---|---|--...--|---|---|
83 | 0 | 1 | ... |N-2|N-1|
84 |---|---|--...--|---|---|
86 ^ ^
87 | |
88 | | This is good. This is safe. This
89 | | is guaranteed to work. Just don't
90 | | dereference 'end'.
91 beginning end
93 </pre><p>See? Everything between the boundary markers is chapter of the array.
94 Simple.
95 </p><p>Now think back to your junior-high school algebra course, when you
96 were learning how to draw graphs. Remember that a graph terminating
97 with a solid dot meant, "Everything up through this point,"
98 and a graph terminating with an open dot meant, "Everything up
99 to, but not including, this point," respectively called closed
100 and open ranges? Remember how closed ranges were written with
101 brackets, <span class="emphasis"><em>[a,b]</em></span>, and open ranges were written with parentheses,
102 <span class="emphasis"><em>(a,b)</em></span>?
103 </p><p>The boundary markers for arrays describe a <span class="emphasis"><em>half-open range</em></span>,
104 starting with (and including) the first element, and ending with (but
105 not including) the last element: <span class="emphasis"><em>[beginning,end)</em></span>. See, I
106 told you it would be simple in the end.
107 </p><p>Iterators, and everything working with iterators, follows this same
108 time-honored tradition. A container's <code class="code">begin()</code> method returns
109 an iterator referring to the first element, and its <code class="code">end()</code>
110 method returns a past-the-end iterator, which is guaranteed to be
111 unique and comparable against any other iterator pointing into the
112 middle of the container.
113 </p><p>Container constructors, container methods, and algorithms, all take
114 pairs of iterators describing a range of values on which to operate.
115 All of these ranges are half-open ranges, so you pass the beginning
116 iterator as the starting parameter, and the one-past-the-end iterator
117 as the finishing parameter.
118 </p><p>This generalizes very well. You can operate on sub-ranges quite
119 easily this way; functions accepting a <span class="emphasis"><em>[first,last)</em></span> range
120 don't know or care whether they are the boundaries of an entire {array,
121 sequence, container, whatever}, or whether they only enclose a few
122 elements from the center. This approach also makes zero-length
123 sequences very simple to recognize: if the two endpoints compare
124 equal, then the {array, sequence, container, whatever} is empty.
125 </p><p>Just don't dereference <code class="code">end()</code>.
126 </p></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="containers_and_c.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="std_contents.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="algorithms.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">Interacting with C </td><td width="20%" align="center"><a accesskey="h" href="../index.html">Home</a></td><td width="40%" align="right" valign="top"> Chapter 11
127 Algorithms
129 </td></tr></table></div></body></html>