CWG 616, 1213 - value category of subobject references.

[official-gcc.git] / libstdc++-v3 / doc / xml / manual / strings.xml

<chapter xmlns="http://docbook.org/ns/docbook" version="5.0" 
         xml:id="std.strings" xreflabel="Strings">
<?dbhtml filename="strings.html"?>

<info><title>
  Strings
  <indexterm><primary>Strings</primary></indexterm>
</title>
  <keywordset>
    <keyword>ISO C++</keyword>
    <keyword>library</keyword>
  </keywordset>
</info>

<!-- Sect1 01 : Character Traits -->

<!-- Sect1 02 : String Classes -->
<section xml:id="std.strings.string" xreflabel="string"><info><title>String Classes</title></info>
  

  <section xml:id="strings.string.simple" xreflabel="Simple Transformations"><info><title>Simple Transformations</title></info>
    
    <para>
      Here are Standard, simple, and portable ways to perform common
      transformations on a <code>string</code> instance, such as
      "convert to all upper case." The word transformations
      is especially apt, because the standard template function
      <code>transform&lt;&gt;</code> is used.
   </para>
   <para>
     This code will go through some iterations.  Here's a simple
     version:
   </para>
   <programlisting>
   #include &lt;string&gt;
   #include &lt;algorithm&gt;
   #include &lt;cctype&gt;      // old &lt;ctype.h&gt;

   struct ToLower
   {
     char operator() (char c) const  { return std::tolower(c); }
   };

   struct ToUpper
   {
     char operator() (char c) const  { return std::toupper(c); }
   };

   int main()
   {
     std::string  s ("Some Kind Of Initial Input Goes Here");

     // Change everything into upper case
     std::transform (s.begin(), s.end(), s.begin(), ToUpper());

     // Change everything into lower case
     std::transform (s.begin(), s.end(), s.begin(), ToLower());

     // Change everything back into upper case, but store the
     // result in a different string
     std::string  capital_s;
     capital_s.resize(s.size());
     std::transform (s.begin(), s.end(), capital_s.begin(), ToUpper());
   }
   </programlisting>
   <para>
     <emphasis>Note</emphasis> that these calls all
      involve the global C locale through the use of the C functions
      <code>toupper/tolower</code>.  This is absolutely guaranteed to work --
      but <emphasis>only</emphasis> if the string contains <emphasis>only</emphasis> characters
      from the basic source character set, and there are <emphasis>only</emphasis>
      96 of those.  Which means that not even all English text can be
      represented (certain British spellings, proper names, and so forth).
      So, if all your input forevermore consists of only those 96
      characters (hahahahahaha), then you're done.
   </para>
   <para><emphasis>Note</emphasis> that the
      <code>ToUpper</code> and <code>ToLower</code> function objects
      are needed because <code>toupper</code> and <code>tolower</code>
      are overloaded names (declared in <code>&lt;cctype&gt;</code> and
      <code>&lt;locale&gt;</code>) so the template-arguments for
      <code>transform&lt;&gt;</code> cannot be deduced, as explained in
      <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://gcc.gnu.org/ml/libstdc++/2002-11/msg00180.html">this
      message</link>.
      <!-- section 14.8.2.4 clause 16 in ISO 14882:1998  -->
      At minimum, you can write short wrappers like
   </para>
   <programlisting>
   char toLower (char c)
   {
      // std::tolower(c) is undefined if c &lt; 0 so cast to unsigned char.
      return std::tolower((unsigned char)c);
   } </programlisting>
   <para>(Thanks to James Kanze for assistance and suggestions on all of this.)
   </para>
   <para>Another common operation is trimming off excess whitespace.  Much
      like transformations, this task is trivial with the use of string's
      <code>find</code> family.  These examples are broken into multiple
      statements for readability:
   </para>
   <programlisting>
   std::string  str (" \t blah blah blah    \n ");

   // trim leading whitespace
   string::size_type  notwhite = str.find_first_not_of(" \t\n");
   str.erase(0,notwhite);

   // trim trailing whitespace
   notwhite = str.find_last_not_of(" \t\n");
   str.erase(notwhite+1); </programlisting>
   <para>Obviously, the calls to <code>find</code> could be inserted directly
      into the calls to <code>erase</code>, in case your compiler does not
      optimize named temporaries out of existence.
   </para>

  </section>
  <section xml:id="strings.string.case" xreflabel="Case Sensitivity"><info><title>Case Sensitivity</title></info>
    
    <para>
    </para>

   <para>The well-known-and-if-it-isn't-well-known-it-ought-to-be
      <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://www.gotw.ca/gotw/">Guru of the Week</link>
      discussions held on Usenet covered this topic in January of 1998.
      Briefly, the challenge was, <quote>write a 'ci_string' class which
      is identical to the standard 'string' class, but is
      case-insensitive in the same way as the (common but nonstandard)
      C function stricmp()</quote>.
   </para>
   <programlisting>
   ci_string s( "AbCdE" );

   // case insensitive
   assert( s == "abcde" );
   assert( s == "ABCDE" );

   // still case-preserving, of course
   assert( strcmp( s.c_str(), "AbCdE" ) == 0 );
   assert( strcmp( s.c_str(), "abcde" ) != 0 ); </programlisting>

   <para>The solution is surprisingly easy.  The original answer was
   posted on Usenet, and a revised version appears in Herb Sutter's
   book <emphasis>Exceptional C++</emphasis> and on his website as <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://www.gotw.ca/gotw/029.htm">GotW 29</link>.
   </para>
   <para>See?  Told you it was easy!</para>
   <para>
     <emphasis>Added June 2000:</emphasis> The May 2000 issue of C++
     Report contains a fascinating <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://lafstern.org/matt/col2_new.pdf"> article</link> by
     Matt Austern (yes, <emphasis>the</emphasis> Matt Austern) on why
     case-insensitive comparisons are not as easy as they seem, and
     why creating a class is the <emphasis>wrong</emphasis> way to go
     about it in production code.  (The GotW answer mentions one of
     the principle difficulties; his article mentions more.)
   </para>
   <para>Basically, this is "easy" only if you ignore some things,
      things which may be too important to your program to ignore.  (I chose
      to ignore them when originally writing this entry, and am surprised
      that nobody ever called me on it...)  The GotW question and answer
      remain useful instructional tools, however.
   </para>
   <para><emphasis>Added September 2000:</emphasis>  James Kanze provided a link to a
      <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://www.unicode.org/reports/tr21/tr21-5.html">Unicode
      Technical Report discussing case handling</link>, which provides some
      very good information.
   </para>

  </section>
  <section xml:id="strings.string.character_types" xreflabel="Arbitrary Characters"><info><title>Arbitrary Character Types</title></info>
    
    <para>
    </para>

   <para>The <code>std::basic_string</code> is tantalizingly general, in that
      it is parameterized on the type of the characters which it holds.
      In theory, you could whip up a Unicode character class and instantiate
      <code>std::basic_string&lt;my_unicode_char&gt;</code>, or assuming
      that integers are wider than characters on your platform, maybe just
      declare variables of type <code>std::basic_string&lt;int&gt;</code>.
   </para>
   <para>That's the theory.  Remember however that basic_string has additional
      type parameters, which take default arguments based on the character
      type (called <code>CharT</code> here):
   </para>
   <programlisting>
      template &lt;typename CharT,
                typename Traits = char_traits&lt;CharT&gt;,
                typename Alloc = allocator&lt;CharT&gt; &gt;
      class basic_string { .... };</programlisting>
   <para>Now, <code>allocator&lt;CharT&gt;</code> will probably Do The Right
      Thing by default, unless you need to implement your own allocator
      for your characters.
   </para>
   <para>But <code>char_traits</code> takes more work.  The char_traits
      template is <emphasis>declared</emphasis> but not <emphasis>defined</emphasis>.
      That means there is only
   </para>
   <programlisting>
      template &lt;typename CharT&gt;
        struct char_traits
        {
            static void foo (type1 x, type2 y);
            ...
        };</programlisting>
   <para>and functions such as char_traits&lt;CharT&gt;::foo() are not
      actually defined anywhere for the general case.  The C++ standard
      permits this, because writing such a definition to fit all possible
      CharT's cannot be done.
   </para>
   <para>The C++ standard also requires that char_traits be specialized for
      instantiations of <code>char</code> and <code>wchar_t</code>, and it
      is these template specializations that permit entities like
      <code>basic_string&lt;char,char_traits&lt;char&gt;&gt;</code> to work.
   </para>
   <para>If you want to use character types other than char and wchar_t,
      such as <code>unsigned char</code> and <code>int</code>, you will
      need suitable specializations for them.  For a time, in earlier
      versions of GCC, there was a mostly-correct implementation that
      let programmers be lazy but it broke under many situations, so it
      was removed.  GCC 3.4 introduced a new implementation that mostly
      works and can be specialized even for <code>int</code> and other
      built-in types.
   </para>
   <para>If you want to use your own special character class, then you have
      <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://gcc.gnu.org/ml/libstdc++/2002-08/msg00163.html">a lot
      of work to do</link>, especially if you with to use i18n features
      (facets require traits information but don't have a traits argument).
   </para>
   <para>Another example of how to specialize char_traits was given <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://gcc.gnu.org/ml/libstdc++/2002-08/msg00260.html">on the
      mailing list</link> and at a later date was put into the file <code>
      include/ext/pod_char_traits.h</code>.  We agree
      that the way it's used with basic_string (scroll down to main())
      doesn't look nice, but that's because <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://gcc.gnu.org/ml/libstdc++/2002-08/msg00236.html">the
      nice-looking first attempt</link> turned out to <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://gcc.gnu.org/ml/libstdc++/2002-08/msg00242.html">not
      be conforming C++</link>, due to the rule that CharT must be a POD.
      (See how tricky this is?)
   </para>

  </section>

  <section xml:id="strings.string.token" xreflabel="Tokenizing"><info><title>Tokenizing</title></info>
    
    <para>
    </para>
   <para>The Standard C (and C++) function <code>strtok()</code> leaves a lot to
      be desired in terms of user-friendliness.  It's unintuitive, it
      destroys the character string on which it operates, and it requires
      you to handle all the memory problems.  But it does let the client
      code decide what to use to break the string into pieces; it allows
      you to choose the "whitespace," so to speak.
   </para>
   <para>A C++ implementation lets us keep the good things and fix those
      annoyances.  The implementation here is more intuitive (you only
      call it once, not in a loop with varying argument), it does not
      affect the original string at all, and all the memory allocation
      is handled for you.
   </para>
   <para>It's called stringtok, and it's a template function. Sources are
   as below, in a less-portable form than it could be, to keep this
   example simple (for example, see the comments on what kind of
   string it will accept).
   </para>

<programlisting>
#include &lt;string&gt;
template &lt;typename Container&gt;
void
stringtok(Container &amp;container, string const &amp;in,
          const char * const delimiters = " \t\n")
{
    const string::size_type len = in.length();
          string::size_type i = 0;

    while (i &lt; len)
    {
        // Eat leading whitespace
        i = in.find_first_not_of(delimiters, i);
        if (i == string::npos)
          return;   // Nothing left but white space

        // Find the end of the token
        string::size_type j = in.find_first_of(delimiters, i);

        // Push token
        if (j == string::npos)
        {
          container.push_back(in.substr(i));
          return;
        }
        else
          container.push_back(in.substr(i, j-i));

        // Set up for next loop
        i = j + 1;
    }
}
</programlisting>


   <para>
     The author uses a more general (but less readable) form of it for
     parsing command strings and the like.  If you compiled and ran this
     code using it:
   </para>


   <programlisting>
   std::list&lt;string&gt;  ls;
   stringtok (ls, " this  \t is\t\n  a test  ");
   for (std::list&lt;string&gt;const_iterator i = ls.begin();
        i != ls.end(); ++i)
   {
       std::cerr &lt;&lt; ':' &lt;&lt; (*i) &lt;&lt; ":\n";
   } </programlisting>
   <para>You would see this as output:
   </para>
   <programlisting>
   :this:
   :is:
   :a:
   :test: </programlisting>
   <para>with all the whitespace removed.  The original <code>s</code> is still
      available for use, <code>ls</code> will clean up after itself, and
      <code>ls.size()</code> will return how many tokens there were.
   </para>
   <para>As always, there is a price paid here, in that stringtok is not
      as fast as strtok.  The other benefits usually outweigh that, however.
   </para>

   <para><emphasis>Added February 2001:</emphasis>  Mark Wilden pointed out that the
      standard <code>std::getline()</code> function can be used with standard
      <code>istringstreams</code> to perform
      tokenizing as well.  Build an istringstream from the input text,
      and then use std::getline with varying delimiters (the three-argument
      signature) to extract tokens into a string.
   </para>


  </section>
  <section xml:id="strings.string.shrink" xreflabel="Shrink to Fit"><info><title>Shrink to Fit</title></info>
    
    <para>
    </para>
   <para>From GCC 3.4 calling <code>s.reserve(res)</code> on a
      <code>string s</code> with <code>res &lt; s.capacity()</code> will
      reduce the string's capacity to <code>std::max(s.size(), res)</code>.
   </para>
   <para>This behaviour is suggested, but not required by the standard. Prior
      to GCC 3.4 the following alternative can be used instead
   </para>
   <programlisting>
      std::string(str.data(), str.size()).swap(str);
   </programlisting>
   <para>This is similar to the idiom for reducing
      a <code>vector</code>'s memory usage
      (see <link linkend="faq.size_equals_capacity">this FAQ
      entry</link>) but the regular copy constructor cannot be used
      because libstdc++'s <code>string</code> is Copy-On-Write in GCC 3.
   </para>
   <para>In <link linkend="status.iso.2011">C++11</link> mode you can call
      <code>s.shrink_to_fit()</code> to achieve the same effect as
      <code>s.reserve(s.size())</code>.
   </para>


  </section>

  <section xml:id="strings.string.Cstring" xreflabel="CString (MFC)"><info><title>CString (MFC)</title></info>
    
    <para>
    </para>

   <para>A common lament seen in various newsgroups deals with the Standard
      string class as opposed to the Microsoft Foundation Class called
      CString.  Often programmers realize that a standard portable
      answer is better than a proprietary nonportable one, but in porting
      their application from a Win32 platform, they discover that they
      are relying on special functions offered by the CString class.
   </para>
   <para>Things are not as bad as they seem.  In
      <link xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="http://gcc.gnu.org/ml/gcc/1999-04n/msg00236.html">this
      message</link>, Joe Buck points out a few very important things:
   </para>
      <itemizedlist>
         <listitem><para>The Standard <code>string</code> supports all the operations
             that CString does, with three exceptions.
         </para></listitem>
         <listitem><para>Two of those exceptions (whitespace trimming and case
             conversion) are trivial to implement.  In fact, we do so
             on this page.
         </para></listitem>
         <listitem><para>The third is <code>CString::Format</code>, which allows formatting
             in the style of <code>sprintf</code>.  This deserves some mention:
         </para></listitem>
      </itemizedlist>
   <para>
      The old libg++ library had a function called form(), which did much
      the same thing.  But for a Standard solution, you should use the
      stringstream classes.  These are the bridge between the iostream
      hierarchy and the string class, and they operate with regular
      streams seamlessly because they inherit from the iostream
      hierarchy.  An quick example:
   </para>
   <programlisting>
   #include &lt;iostream&gt;
   #include &lt;string&gt;
   #include &lt;sstream&gt;

   string f (string&amp; incoming)     // incoming is "foo  N"
   {
       istringstream   incoming_stream(incoming);
       string          the_word;
       int             the_number;

       incoming_stream &gt;&gt; the_word        // extract "foo"
                       &gt;&gt; the_number;     // extract N

       ostringstream   output_stream;
       output_stream &lt;&lt; "The word was " &lt;&lt; the_word
                     &lt;&lt; " and 3*N was " &lt;&lt; (3*the_number);

       return output_stream.str();
   } </programlisting>
   <para>A serious problem with CString is a design bug in its memory
      allocation.  Specifically, quoting from that same message:
   </para>
   <programlisting>
   CString suffers from a common programming error that results in
   poor performance.  Consider the following code:

   CString n_copies_of (const CString&amp; foo, unsigned n)
   {
           CString tmp;
           for (unsigned i = 0; i &lt; n; i++)
                   tmp += foo;
           return tmp;
   }

   This function is O(n^2), not O(n).  The reason is that each +=
   causes a reallocation and copy of the existing string.  Microsoft
   applications are full of this kind of thing (quadratic performance
   on tasks that can be done in linear time) -- on the other hand,
   we should be thankful, as it's created such a big market for high-end
   ix86 hardware. :-)

   If you replace CString with string in the above function, the
   performance is O(n).
   </programlisting>
   <para>Joe Buck also pointed out some other things to keep in mind when
      comparing CString and the Standard string class:
   </para>
      <itemizedlist>
         <listitem><para>CString permits access to its internal representation; coders
             who exploited that may have problems moving to <code>string</code>.
         </para></listitem>
         <listitem><para>Microsoft ships the source to CString (in the files
             MFC\SRC\Str{core,ex}.cpp), so you could fix the allocation
             bug and rebuild your MFC libraries.
             <emphasis><emphasis>Note:</emphasis> It looks like the CString shipped
             with VC++6.0 has fixed this, although it may in fact have been
             one of the VC++ SPs that did it.</emphasis>
         </para></listitem>
         <listitem><para><code>string</code> operations like this have O(n) complexity
             <emphasis>if the implementors do it correctly</emphasis>.  The libstdc++
             implementors did it correctly.  Other vendors might not.
         </para></listitem>
         <listitem><para>While parts of the SGI STL are used in libstdc++, their
             string class is not.  The SGI <code>string</code> is essentially
             <code>vector&lt;char&gt;</code> and does not do any reference
             counting like libstdc++'s does.  (It is O(n), though.)
             So if you're thinking about SGI's string or rope classes,
             you're now looking at four possibilities:  CString, the
             libstdc++ string, the SGI string, and the SGI rope, and this
             is all before any allocator or traits customizations!  (More
             choices than you can shake a stick at -- want fries with that?)
         </para></listitem>
      </itemizedlist>

  </section>
</section>

<!-- Sect1 03 : Interacting with C -->

</chapter>