2009-07-20 Benjamin Kosnik <bkoz@redhat.com>

[pf3gnuchains/gcc-fork.git] / libstdc++-v3 / doc / xml / manual / containers.xml

<?xml version='1.0'?>
<!DOCTYPE part PUBLIC "-//OASIS//DTD DocBook XML V4.5//EN" 
 "http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd" 
[ ]>

<part id="manual.containers" xreflabel="Containers">
<?dbhtml filename="containers.html"?>
 
<partinfo>
  <keywordset>
    <keyword>
      ISO C++
    </keyword>
    <keyword>
      library
    </keyword>
  </keywordset>
</partinfo>

<title>
  Containers
  <indexterm><primary>Containers</primary></indexterm>
</title>

<!-- Chapter 01 : Sequences -->
<chapter id="manual.containers.sequences" xreflabel="Sequences">
<?dbhtml filename="sequences.html"?>
  <title>Sequences</title>

<sect1 id="containers.sequences.list" xreflabel="list">
<?dbhtml filename="list.html"?>
  <title>list</title>
  <sect2 id="sequences.list.size" xreflabel="list::size() is O(n)">
    <title>list::size() is O(n)</title>
   <para>
     Yes it is, and that's okay.  This is a decision that we preserved
     when we imported SGI's STL implementation.  The following is
     quoted from <ulink
     url="http://www.sgi.com/tech/stl/FAQ.html">their FAQ</ulink>:
   </para>
   <blockquote>
     <para>
       The size() member function, for list and slist, takes time
       proportional to the number of elements in the list.  This was a
       deliberate tradeoff.  The only way to get a constant-time
       size() for linked lists would be to maintain an extra member
       variable containing the list's size.  This would require taking
       extra time to update that variable (it would make splice() a
       linear time operation, for example), and it would also make the
       list larger.  Many list algorithms don't require that extra
       word (algorithms that do require it might do better with
       vectors than with lists), and, when it is necessary to maintain
       an explicit size count, it's something that users can do
       themselves.
     </para>
     <para>
       This choice is permitted by the C++ standard. The standard says
       that size() <quote>should</quote> be constant time, and
       <quote>should</quote> does not mean the same thing as
       <quote>shall</quote>.  This is the officially recommended ISO
       wording for saying that an implementation is supposed to do
       something unless there is a good reason not to.
      </para>
      <para>
        One implication of linear time size(): you should never write
      </para>
         <programlisting>
         if (L.size() == 0)
             ...
         </programlisting>

         <para>
         Instead, you should write
         </para>

         <programlisting>
         if (L.empty())
             ...
         </programlisting>
   </blockquote>
  </sect2>
</sect1>

<sect1 id="containers.sequences.vector" xreflabel="vector">
<?dbhtml filename="vector.html"?>
  <title>vector</title>
  <para>
  </para>
  <sect2 id="sequences.vector.management" xreflabel="Space Overhead Management">
    <title>Space Overhead Management</title>
   <para>
     In <ulink
     url="http://gcc.gnu.org/ml/libstdc++/2002-04/msg00105.html">this
     message to the list</ulink>, Daniel Kostecky announced work on an
     alternate form of <code>std::vector</code> that would support
     hints on the number of elements to be over-allocated.  The design
     was also described, along with possible implementation choices.
   </para>
   <para>
     The first two alpha releases were announced <ulink
     url="http://gcc.gnu.org/ml/libstdc++/2002-07/msg00048.html">here</ulink>
     and <ulink
     url="http://gcc.gnu.org/ml/libstdc++/2002-07/msg00111.html">here</ulink>.
   </para>

  </sect2></sect1>
</chapter>

<!-- Chapter 02 : Associative -->
<chapter id="manual.containers.associative" xreflabel="Associative">
<?dbhtml filename="associative.html"?>
  <title>Associative</title>

  <sect1 id="containers.associative.insert_hints" xreflabel="Insertion Hints">
    <title>Insertion Hints</title>
   <para>
     Section [23.1.2], Table 69, of the C++ standard lists this
     function for all of the associative containers (map, set, etc):
   </para>
   <programlisting>
      a.insert(p,t);
   </programlisting>
   <para>
     where 'p' is an iterator into the container 'a', and 't' is the
     item to insert.  The standard says that <quote><code>t</code> is
     inserted as close as possible to the position just prior to
     <code>p</code>.</quote> (Library DR #233 addresses this topic,
     referring to <ulink
     url="http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2005/n1780.html">N1780</ulink>.
     Since version 4.2 GCC implements the resolution to DR 233, so
     that insertions happen as close as possible to the hint. For
     earlier releases the hint was only used as described below.
   </para>
   <para>
     Here we'll describe how the hinting works in the libstdc++
     implementation, and what you need to do in order to take
     advantage of it.  (Insertions can change from logarithmic
     complexity to amortized constant time, if the hint is properly
     used.)  Also, since the current implementation is based on the
     SGI STL one, these points may hold true for other library
     implementations also, since the HP/SGI code is used in a lot of
     places.
   </para>
   <para>
     In the following text, the phrases <emphasis>greater
     than</emphasis> and <emphasis>less than</emphasis> refer to the
     results of the strict weak ordering imposed on the container by
     its comparison object, which defaults to (basically)
     <quote>&lt;</quote>.  Using those phrases is semantically sloppy,
     but I didn't want to get bogged down in syntax.  I assume that if
     you are intelligent enough to use your own comparison objects,
     you are also intelligent enough to assign <quote>greater</quote>
     and <quote>lesser</quote> their new meanings in the next
     paragraph.  *grin*
   </para>
   <para>
     If the <code>hint</code> parameter ('p' above) is equivalent to:
   </para>
     <itemizedlist>
      <listitem>
        <para>
          <code>begin()</code>, then the item being inserted should
          have a key less than all the other keys in the container.
          The item will be inserted at the beginning of the container,
          becoming the new entry at <code>begin()</code>.
      </para>
      </listitem>
      <listitem>
        <para>
          <code>end()</code>, then the item being inserted should have
          a key greater than all the other keys in the container.  The
          item will be inserted at the end of the container, becoming
          the new entry at <code>end()</code>.  
      </para>
      </listitem>
      <listitem>
        <para>
          neither <code>begin()</code> nor <code>end()</code>, then:
          Let <code>h</code> be the entry in the container pointed to
          by <code>hint</code>, that is, <code>h = *hint</code>.  Then
          the item being inserted should have a key less than that of
          <code>h</code>, and greater than that of the item preceding
          <code>h</code>.  The new item will be inserted between
          <code>h</code> and <code>h</code>'s predecessor.
          </para>
      </listitem>
     </itemizedlist>
   <para>
     For <code>multimap</code> and <code>multiset</code>, the
     restrictions are slightly looser: <quote>greater than</quote>
     should be replaced by <quote>not less than</quote>and <quote>less
     than</quote> should be replaced by <quote>not greater
     than.</quote> (Why not replace greater with
     greater-than-or-equal-to?  You probably could in your head, but
     the mathematicians will tell you that it isn't the same thing.)
   </para>
   <para>
     If the conditions are not met, then the hint is not used, and the
     insertion proceeds as if you had called <code> a.insert(t)
     </code> instead.  (<emphasis>Note </emphasis> that GCC releases
     prior to 3.0.2 had a bug in the case with <code>hint ==
     begin()</code> for the <code>map</code> and <code>set</code>
     classes.  You should not use a hint argument in those releases.)
   </para>
   <para>
     This behavior goes well with other containers'
     <code>insert()</code> functions which take an iterator: if used,
     the new item will be inserted before the iterator passed as an
     argument, same as the other containers.
   </para>
   <para>
     <emphasis>Note </emphasis> also that the hint in this
     implementation is a one-shot.  The older insertion-with-hint
     routines check the immediately surrounding entries to ensure that
     the new item would in fact belong there.  If the hint does not
     point to the correct place, then no further local searching is
     done; the search begins from scratch in logarithmic time.
   </para>
  </sect1>


  <sect1 id="containers.associative.bitset" xreflabel="bitset">
    <?dbhtml filename="bitset.html"?>
    <title>bitset</title>
    <sect2 id="associative.bitset.size_variable" xreflabel="Variable">
      <title>Size Variable</title>
      <para>
        No, you cannot write code of the form
      </para>
      <!-- Careful, the leading spaces in PRE show up directly. -->
   <programlisting>
      #include &lt;bitset&gt;

      void foo (size_t n)
      {
          std::bitset&lt;n&gt;   bits;
          ....
      } 
   </programlisting>
   <para>
     because <code>n</code> must be known at compile time.  Your
     compiler is correct; it is not a bug.  That's the way templates
     work.  (Yes, it <emphasis>is</emphasis> a feature.)
   </para>
   <para>
     There are a couple of ways to handle this kind of thing.  Please
     consider all of them before passing judgement.  They include, in
     no particular order:
   </para>
      <itemizedlist>
        <listitem><para>A very large N in <code>bitset&lt;N&gt;</code>.</para></listitem>
        <listitem><para>A container&lt;bool&gt;.</para></listitem>
        <listitem><para>Extremely weird solutions.</para></listitem>
      </itemizedlist>
   <para>
     <emphasis>A very large N in
     <code>bitset&lt;N&gt;</code>.&nbsp;&nbsp;</emphasis> It has been
     pointed out a few times in newsgroups that N bits only takes up
     (N/8) bytes on most systems, and division by a factor of eight is
     pretty impressive when speaking of memory.  Half a megabyte given
     over to a bitset (recall that there is zero space overhead for
     housekeeping info; it is known at compile time exactly how large
     the set is) will hold over four million bits.  If you're using
     those bits as status flags (e.g.,
     <quote>changed</quote>/<quote>unchanged</quote> flags), that's a
     <emphasis>lot</emphasis> of state.
   </para>
   <para>
     You can then keep track of the <quote>maximum bit used</quote>
     during some testing runs on representative data, make note of how
     many of those bits really need to be there, and then reduce N to
     a smaller number.  Leave some extra space, of course.  (If you
     plan to write code like the incorrect example above, where the
     bitset is a local variable, then you may have to talk your
     compiler into allowing that much stack space; there may be zero
     space overhead, but it's all allocated inside the object.)
   </para>
   <para>
     <emphasis>A container&lt;bool&gt;.&nbsp;&nbsp;</emphasis> The
     Committee made provision for the space savings possible with that
     (N/8) usage previously mentioned, so that you don't have to do
     wasteful things like <code>Container&lt;char&gt;</code> or
     <code>Container&lt;short int&gt;</code>.  Specifically,
     <code>vector&lt;bool&gt;</code> is required to be specialized for
     that space savings.
   </para>
   <para>
     The problem is that <code>vector&lt;bool&gt;</code> doesn't
     behave like a normal vector anymore.  There have been recent
     journal articles which discuss the problems (the ones by Herb
     Sutter in the May and July/August 1999 issues of C++ Report cover
     it well).  Future revisions of the ISO C++ Standard will change
     the requirement for <code>vector&lt;bool&gt;</code>
     specialization.  In the meantime, <code>deque&lt;bool&gt;</code>
     is recommended (although its behavior is sane, you probably will
     not get the space savings, but the allocation scheme is different
     than that of vector).
   </para>
   <para>
     <emphasis>Extremely weird solutions.&nbsp;&nbsp;</emphasis> If
     you have access to the compiler and linker at runtime, you can do
     something insane, like figuring out just how many bits you need,
     then writing a temporary source code file.  That file contains an
     instantiation of <code>bitset</code> for the required number of
     bits, inside some wrapper functions with unchanging signatures.
     Have your program then call the compiler on that file using
     Position Independent Code, then open the newly-created object
     file and load those wrapper functions.  You'll have an
     instantiation of <code>bitset&lt;N&gt;</code> for the exact
     <code>N</code> that you need at the time.  Don't forget to delete
     the temporary files.  (Yes, this <emphasis>can</emphasis> be, and
     <emphasis>has been</emphasis>, done.)
   </para>
   <!-- I wonder if this next paragraph will get me in trouble... -->
   <para>
     This would be the approach of either a visionary genius or a
     raving lunatic, depending on your programming and management
     style.  Probably the latter.
   </para>
   <para>
     Which of the above techniques you use, if any, are up to you and
     your intended application.  Some time/space profiling is
     indicated if it really matters (don't just guess).  And, if you
     manage to do anything along the lines of the third category, the
     author would love to hear from you...
   </para>
   <para>
     Also note that the implementation of bitset used in libstdc++ has
     <link linkend="manual.ext.containers.sgi">some extensions</link>.
   </para>

    </sect2>
    <sect2 id="associative.bitset.type_string" xreflabel="Type String">
      <title>Type String</title>
      <para>
      </para>
   <para>
     Bitmasks do not take char* nor const char* arguments in their
     constructors.  This is something of an accident, but you can read
     about the problem: follow the library's <quote>Links</quote> from
     the homepage, and from the C++ information <quote>defect
     reflector</quote> link, select the library issues list.  Issue
     number 116 describes the problem.
   </para>
   <para>
     For now you can simply make a temporary string object using the
     constructor expression:
   </para>
   <programlisting>
      std::bitset&lt;5&gt; b ( std::string(<quote>10110</quote>) );
   </programlisting>
   
   <para>
     instead of
   </para>

    <programlisting>
      std::bitset&lt;5&gt; b ( <quote>10110</quote> );    // invalid
    </programlisting>
    </sect2>  
  </sect1>

</chapter>

<!-- Chapter 03 : Interacting with C -->
<chapter id="manual.containers.c" xreflabel="Interacting with C">
<?dbhtml filename="containers_and_c.html"?>
  <title>Interacting with C</title>

  <sect1 id="containers.c.vs_array" xreflabel="Containers vs. Arrays">
    <title>Containers vs. Arrays</title>
   <para>
     You're writing some code and can't decide whether to use builtin
     arrays or some kind of container.  There are compelling reasons
     to use one of the container classes, but you're afraid that
     you'll eventually run into difficulties, change everything back
     to arrays, and then have to change all the code that uses those
     data types to keep up with the change.
   </para>
   <para>
     If your code makes use of the standard algorithms, this isn't as
     scary as it sounds.  The algorithms don't know, nor care, about
     the kind of <quote>container</quote> on which they work, since
     the algorithms are only given endpoints to work with.  For the
     container classes, these are iterators (usually
     <code>begin()</code> and <code>end()</code>, but not always).
     For builtin arrays, these are the address of the first element
     and the <link linkend="iterators.predefined.end">past-the-end</link> element.
   </para>
   <para>
     Some very simple wrapper functions can hide all of that from the
     rest of the code.  For example, a pair of functions called
     <code>beginof</code> can be written, one that takes an array,
     another that takes a vector.  The first returns a pointer to the
     first element, and the second returns the vector's
     <code>begin()</code> iterator.
   </para>
   <para>
     The functions should be made template functions, and should also
     be declared inline.  As pointed out in the comments in the code
     below, this can lead to <code>beginof</code> being optimized out
     of existence, so you pay absolutely nothing in terms of increased
     code size or execution time.
   </para>
   <para>
     The result is that if all your algorithm calls look like
   </para>
   <programlisting>
   std::transform(beginof(foo), endof(foo), beginof(foo), SomeFunction);
   </programlisting>
   <para>
     then the type of foo can change from an array of ints to a vector
     of ints to a deque of ints and back again, without ever changing
     any client code.
   </para>

<programlisting>
// beginof
template&lt;typename T&gt;
  inline typename vector&lt;T&gt;::iterator
  beginof(vector&lt;T&gt; &amp;v)
  { return v.begin(); }

template&lt;typename T, unsigned int sz&gt;
  inline T* 
  beginof(T (&amp;array)[sz]) { return array; }

// endof
template&lt;typename T&gt;
  inline typename vector&lt;T&gt;::iterator 
  endof(vector&lt;T&gt; &amp;v)
  { return v.end(); }

template&lt;typename T, unsigned int sz&gt;
  inline T* 
  endof(T (&amp;array)[sz]) { return array + sz; }

// lengthof
template&lt;typename T&gt;
  inline typename vector&lt;T&gt;::size_type 
  lengthof(vector&lt;T&gt; &amp;v)
  { return v.size(); }

template&lt;typename T, unsigned int sz&gt;
  inline unsigned int 
  lengthof(T (&amp;)[sz]) { return sz; }
</programlisting>

   <para>
     Astute readers will notice two things at once: first, that the
     container class is still a <code>vector&lt;T&gt;</code> instead
     of a more general <code>Container&lt;T&gt;</code>.  This would
     mean that three functions for <code>deque</code> would have to be
     added, another three for <code>list</code>, and so on.  This is
     due to problems with getting template resolution correct; I find
     it easier just to give the extra three lines and avoid confusion.
   </para>
   <para>
     Second, the line
   </para>
   <programlisting>
    inline unsigned int lengthof (T (&amp;)[sz]) { return sz; } 
   </programlisting>
   <para>
     looks just weird!  Hint:  unused parameters can be left nameless.
   </para>
  </sect1>
  
</chapter>

</part>