format/doc/format.html

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">

<html>
<head>
  <meta http-equiv="Content-Language" content="en-us">
  <meta http-equiv="Content-Type" content="text/html; charset=us-ascii">

  <title>The Boost Format library</title>
</head>

<body bgcolor="white" text="black">
  <h1><img align="middle" alt="boost.png (6897 bytes)" height="86" src=
  "../../../boost.png" width="277">The Boost Format library</h1>

  <p>The <code><a href=
  "../../../boost/format.hpp">&lt;boost/format.hpp&gt;</a></code> format
  class provides printf-like formatting, in a type-safe manner which allows
  output of user-defined types.<br></p>

  <ul>
    <li><a href="#synopsis">Synopsis</a></li>

    <li><a href="#how_it_works">How it works</a></li>

    <li><a href="#examples">Examples</a></li>

    <li>
      <a href="#syntax">Syntax</a>

      <ul>
        <li><a href="#printf_directives">printf format-specification
        syntax</a></li>

        <li><a href="#printf_differences">Incompatibilities with
        printf</a></li>
      </ul>
    </li>

    <li><a href="#manipulators">Manipulators and the internal stream
    state</a></li>

    <li><a href="#user-defined">User-defined types</a></li>

    <li><a href="#alternatives">Alternatives</a></li>

    <li><a href="#exceptions">Exceptions</a></li>

    <li><a href="#performance">Performance</a></li>

    <li><a href="#extract">Class Interface Extract</a></li>

    <li><a href="#rationale">Rationale</a></li>
  </ul><a name="synopsis" id="synopsis"></a>
  <hr>

  <h2>Synopsis</h2>

  <p>A format object is constructed from a format-string, and is then given
  arguments through repeated calls to <i>operator%</i>.<br>
  Each of those arguments are then converted to strings, who are in turn
  combined into one string, according to the format-string.</p>

  <blockquote>
    <pre>
cout &lt;&lt; boost::format("writing %1%,  x=%2% : %3%-th try") % "toto" % 40.23 % 50;
     // prints "writing toto,  x=40.230 : 50-th try"
</pre>
  </blockquote><a name="how_it_works" id="how_it_works"></a>
  <hr>

  <h2>How it works</h2>

  <ol>
    <li>When you call <i>format(s)</i>, where s is the format-string, it
    constructs an object, which parses the format string and look for all
    directives in it and prepares internal structures for the next step.</li>

    <li>Then, either immediately, as in

      <blockquote>
        <pre>
cout &lt;&lt; format("%2% %1%") % 36 % 77;
</pre>
      </blockquote>or later on, as in

      <blockquote>
        <pre>
format fmter("%2% %1%");
fmter % 36; fmter % 77;
</pre>
      </blockquote>you <i>feed</i> variables into the formatter.<br>
      those variables are dumped into an internal stream, which state is set
      according to the given formatting options in the format-string -if
      there are any-, and the format object stores the string results for the
      last step.
    </li>

    <li>Once all arguments have been fed you can dump the format object to a
    stream, or get its string value by using the <i>str()</i> member
    function, or the free function <i>str(const format&amp; )</i> in
    namespace <i>boost</i>. The result string stays accessible in the format
    object until another argument is passed, at which time it is
    reinitialised.

      <blockquote>
        <pre>

// fmter was previously created and fed arguments, it can print the result :
cout &lt;&lt; fmter ;

// You can take the string result :
string s  = fmter.str();

// possibly several times :
s = fmter.str( );

// You can also do all steps at once :
cout &lt;&lt; boost::format("%2% %1%") % 36 % 77;

// using the str free function :
string s2 = str( format("%2% %1%") % 36 % 77 );

</pre>
      </blockquote>
    </li>

    <li>Optionnally, after step 3, you can re-use a format object and restart
    at step2 : <i>fmter % 18 % 39;</i><br>
    to format new variables with the same format-string, saving the expensive
    processing involved at step 1.</li>
  </ol>All in all, the format class translates a format-string (with
  eventually printf-like directives) into operations on an internal stream,
  and finally returns the result of the formatting, as a string, or directly
  into an output stream. <a name="examples" id="examples"></a>
  <hr>

  <h2>Examples</h2>

  <blockquote>
    <pre>
using namespace std;
using boost::format;
using boost::io::group;
</pre>
  </blockquote>

  <ul>
    <li>Simple output, with reordering :

      <blockquote>
        <pre>

cout &lt;&lt; format("%1% %2% %3% %2% %1% \n") % "11" % "22" % "333"; // 'simple' style.

</pre>
      </blockquote>It prints : "11 22 333 22 11 \n"
    </li>

    <li>More precise formatting, with Posix-printf positional directives :

      <blockquote>
        <pre>

cout &lt;&lt; format("(x,y) = (%1$+5d,%2$+5d) \n") % -23 % 35;     // Posix-Printf style

</pre>
      </blockquote>It prints : "(x,y) = ( -23, +35) \n"
    </li>

    <li>classical printf directive, no reordering :

      <blockquote>
        <pre>

cout &lt;&lt; format("writing %s,  x=%s : %d-th step \n") % "toto" % 40.23 % 50;

</pre>
      </blockquote>It prints : "writing toto, x=40.23 : 50-th step \n"
    </li>

    <li>Several ways to express the same thing :

      <blockquote>
        <pre>

cout &lt;&lt; format("(x,y) = (%+5d,%+5d) \n") % -23 % 35;
cout &lt;&lt; format("(x,y) = (%|+5|,%|+5|) \n") % -23 % 35;

cout &lt;&lt; format("(x,y) = (%1$+5d,%2$+5d) \n") % -23 % 35;
cout &lt;&lt; format("(x,y) = (%|1$+5|,%|2$+5|) \n") % -23 % 35;

</pre>
      </blockquote>all those print : "(x,y) = ( -23, +35) \n"
    </li>

    <li>Using manipulators to modify the format-string :

      <blockquote>
        <pre>

format fmter("_%1$+5d_ %1$d \n");

format fmter2("_%1%_ %1% \n");
fmter2.modify_item(1, group(showpos, setw(5)) );

cout &lt;&lt; fmter % 101 ;
cout &lt;&lt; fmter2 % 101 ;

</pre>
      </blockquote>Both print the same : "_ +101_ 101 \n"
    </li>

    <li>Using manipulators with arguments :

      <blockquote>
        <pre>

cout &lt;&lt; format("_%1%_ %1% \n") % group(showpos, setw(5), 101);

</pre>
      </blockquote>The manipulators are applied at each occurrence of %1%, and
      thus it prints : "_ +101_ +101 \n"
    </li>

    <li>New formatting feature : 'absolute tabulations', useful inside loops,
    to insure a field is printed at the same position from one line to the
    next, even if the widthes of the previous arguments can vary a lot.

      <blockquote>
        <pre>

for(unsigned int i=0; i &lt; names.size(); ++i)
    cout &lt;&lt; format("%1%, %2%, %|40t|%3%\n") % names[i] % surname[i] % tel[i];

</pre>
      </blockquote>For some std::vector <i>names</i>, <i>surnames</i>, and
      <i>tel</i> (see sample_new_features.cpp) it prints :

      <blockquote>
        <pre>
Marc-Fran&ccedil;ois Michel, Durand,           +33 (0) 123 456 789
Jean, de Lattre de Tassigny,            +33 (0) 987 654 321
</pre>
      </blockquote>
    </li>
  </ul>
  <hr>

  <h2>Sample Files</h2>

  <p>The program <a href=
  "../example/sample_formats.cpp">sample_formats.cpp</a> demonstrates simple
  uses of <b>format</b>.<br></p>

  <p><a href="../example/sample_new_features.cpp">sample_new_features.cpp</a>
  illustrates the few formatting features that were added to printf's syntax
  such as simple positional directives, centered alignment, and
  'tabulations'.<br></p>

  <p><a href="../example/sample_advanced.cpp">sample_advanced.cpp</a>
  demonstrates uses of advanced features, like reusing, and modifying, format
  objects, etc..<br></p>

  <p>And <a href="../example/sample_userType.cpp">sample_userType.cpp</a>
  shows the behaviour of the <b>format</b> library on user-defined
  types.</p><a name="syntax" id="syntax"></a>
  <hr>

  <h2>Syntax</h2>

  <p><b>boost::format(</b> format-string <b>) %</b> arg1 <b>%</b> arg2
  <b>%</b> ... <b>%</b> argN</p>

  <p>The <i>format-string</i> contains text in which special directives will
  be replaced by strings resulting from the formatting of the given
  arguments.<br>
  The legacy syntax in the C and C++ worlds is the one used by printf, and
  thus format can use directly printf format-strings, and produce the same
  result (in almost all cases. see <a href=
  "#printf_differences">Incompatibilities with printf</a> for details)<br>
  This core syntax was extended, to allow new features, but also to adapt to
  the C++ streams context. Thus, format accepts several forms of directives
  in format-strings :</p>

  <ul>
    <li>Legacy printf format strings : <b>%</b><i>spec</i> where <i>spec</i>
    is a <a href="#printf_directives">printf format specification</a><br>
    <i>spec</i> passes formatting options, like width, alignment, numerical
    base used for formatting numbers, as well as other specific flags. But
    the classical <i>type-specification</i> flag of printf has a weaker
    meaning in format. It merely sets the appropriate flags on the internal
    stream, and/or formatting parameters, but does not require the
    corresponding argument to be of a specific type.<br>
    e.g. : the specification <i>2$x</i>, meaning "print argument number 2,
    which is an integral number, in hexa" for printf, merely means "print
    argument 2 with stream basefield flags set to <i>hex</i>" for
    format.</li>

    <li><b>%|</b><i>spec</i><b>|</b> where <i>spec</i> is a printf format
    specification.<br>
    This pipe-delimited syntax is introduced, to improve the readability of the
    format-string, but primarily, to make the <i>type-conversion
    character</i> optional in <i>spec</i>. This information is not necessary
    with C++ variables, but with direct printf syntax, it is necessary to
    always give a type-conversion character, merely because this character is
    crucial to determine the end of a format-specification.<br>
    e.g. : "%|-5|" will format the next variable with width set to 5, and
    left-alignment just like the following printf directives : "%-5g",
    "%-5f", "%-5s" ..</li>

    <li><b>%</b><i>N</i><b>%</b><br>
    This simple positional notation requests the formatting of the
    <i>N</i>-th argument - wihout any formatting option.<br>
    (It's merely a shortcut to Printf's positional directives (like
    "%<i>N</i>$s"), but a major benefit is that it's much more readable, and
    does not use a "type-conversion" character)</li>
  </ul>On top of the standard printf format specifications, new features were
  implemented, like centered alignment. See <a href="#new_directives">new
  format specification</a> for details. <a name="printf_directives" id=
  "printf_directives"></a>

  <h3>printf format specifications</h3>

  <p>The printf format specifications supported by Boost.format follows the
  Unix98 <a href=
  "http://www.opengroup.org/onlinepubs/7908799/xsh/fprintf.html">Open-group
  printf</a> precise syntax, rather than the standard C printf, which does
  not support positional arguments. (Common flags have the same meaning in
  both, so it should not be a headache for anybody)<br>
  <i>Note that it is an error to use positional format specifications</i>
  (e.g. <i>%3$+d</i>) <i>mixed with non-positional ones</i> (e.g. <i>%+d</i>)
  <i>in the same format string.</i><br>
  In the Open-group specification, referring to the same argument several
  times (e.g. <i>"%1$d %1$d"</i>) has undefined behaviour. Boost.format's
  behaviour in such cases is to allow each argument to be reffered to any
  number of times. The only constraint is that it expects exactly <i>P</i>
  arguments, <i>P</i> being the maximum argument number used in the format
  string. (e.g., for "%1$d %10$d", <i>P</i> == 10 ).<br>
  Supplying more, or less, than <i>P</i> arguments raises an exception.
  (unless it was set otherwise, see <a href="#exceptions">exceptions</a>)</p>

  <p>A specification <i>spec</i> has the form:
      <pre>    [ <i>N</i><b>$</b> ] [ <i>flags</i> ] [ <i>width</i> ] [ <b>.</b> <i>precision</i> ] [ <i>argument-type</i> ] <i>conversion-specifier</i></pre>

  Fields inside square brackets are optional.  Each of those fields are
  explained one by one in the following list:</p>

  <ul>
    <li><i>N</i> <b>$</b> (optional field) specifies that the format
    specification applies to the <i>N</i>-th argument (it is called a
    <i>positional format specification</i>).<br>
    If this is not present, arguments are taken one by one. (and it is then
    an error to later supply an argument number)</li><br />

    <li>
      <i>flags</i> is a sequence of any of these:

      <blockquote>
        <table border="1" cellpadding="5" summary="">
          <tr>
            <td><b>Flag</b></td>

            <td><b>Meaning</b></td>

            <td><b>effect on internal stream</b></td>
          </tr>

          <tr>
            <td><b>'-'</b></td>

            <td>left alignment</td>

            <td>N/A (applied later on the string)</td>
          </tr>

          <tr>
            <td><b>'='</b></td>

            <td>centered alignment</td>

            <td>N/A (applied later on the string)<br>
            <i>- note : added feature, not in printf -</i></td>
          </tr>

          <tr>
            <td><b>'_'</b></td>

            <td>internal alignment</td>

            <td>sets internal alignment<br>
            <i>- note : added feature, not in printf -</i></td>
          </tr>

          <tr>
            <td><b>'+'</b></td>

            <td>show sign even for positive numbers</td>

            <td>sets <i>showpos</i></td>
          </tr>

          <tr>
            <td><b>'#'</b></td>

            <td>show numerical base, and decimal point</td>

            <td>sets <i>showbase</i> and <i>showpoint</i></td>
          </tr>

          <tr>
            <td><b>'0'</b></td>

            <td>pad with 0's (inserted after sign or base indicator)</td>

            <td>if not left-aligned, calls <i>setfill('0')</i> and sets
            <i>internal</i><br>
            Extra actions are taken after stream conversion to handle
            <a href="#user-defined">user-defined output</a>.</td>
          </tr>

          <tr>
            <td><b>' '</b></td>

            <td>if the string does not begin with <i>+</i> or <i>-</i>,
            insert a <i>space</i> before the converted string</td>

            <td>N/A (applied later on the string)<br>
            Different to printf's behaviour : it is not affected by internal
            alignment</td>
          </tr>
        </table>
      </blockquote><br />
    </li>

    <li><i>width</i> specifies a minimal width for the string resulting form
    the conversion. If necessary, the string will be padded with alignment
    and fill characters either set on the stream via manipulators, or
    specified by the format-string (e.g. flags '0', '-', ..)<br>
    Note that width is not just set on the conversion stream. To support
    output of <a href="#user-defined">user-defined types</a> (that might call
    <i>operator&lt;&lt;</i> many times on several members), the width is
    handled after stream conversion of the whole argument object, in the
    format class code.</li><br />

    <li>
      <i>precision</i> (preceded by a point), sets the stream's
      <i>precision</i>

      <ul>
        <li>When outputting a floatting type number, it sets the maximum
        number of digits

          <ul>
            <li>after decimal point when in fixed or scientific mode</li>

            <li>in total when in default mode ('<i>general mode</i>', like
            <i>%g</i>)</li>
          </ul>
        </li>

        <li>When used with type-char <b>s</b> or <b>S</b> it takes another
        meaning : the conversion string is truncated to the <i>precision</i>
        first chars. (Note that the eventual padding to <i>width</i> is done
        after truncation.)</li>
      </ul><br />
    </li>

    <li>
        <i>argument-type</i> is used by the printf family to properly process
        the arguments being passed in through varargs.  With <code>boost::format</code>
        the arguments are fed into format through <code>operator %</code> which
        allows the template to carry the argument type.  Therefore the classical
        printf style argument type is consumed and ignored.
        Argument-types <code>hh</code>, <code>h</code>, <code>l</code>, <code>ll</code>, <code>j</code>,
        <code>z</code>, and <code>L</code> are recognized, as are the Microsoft extensions <code>w</code>,
        <code>I</code> (capital i), <code>I32</code>, and <code>I64</code>.  Argument-type <code>t</code>
        from the ISO C99 standard is not recognized as an argument type, as it has been in use as a
        conversion specifier for tabular output for many years in <code>boost::format</code>.
    </li><br />

    <li>
      <i>conversion-specifier</i> does <b>not</b> impose the concerned argument
      to be of a restricted set of types, but merely sets the ios flags that are
      associated with a type specification:

      <blockquote>
        <table border="1" cellpadding="5" summary="">
          <tr>
            <td><b>conversion-specifier</b></td>

            <td><b>Meaning</b></td>

            <td><b>effect on stream</b></td>
          </tr>

          <tr>
            <td><b>b</b></td>

            <td>boolean string output</td>

            <td>sets <i>boolalpha</i>; only works for type <code>bool</code><br />
            To customize the resulting string, see
            <a href="http://en.cppreference.com/w/cpp/locale/numpunct/truefalsename">std::numpunct</a>.
          </tr>

          <tr>
            <td><b>p or x</b></td>

            <td>hexadecimal output</td>

            <td>sets <i>hex</i></td>
          </tr>

          <tr>
            <td><b>o</b></td>

            <td>octal output</td>

            <td>sets <i>oct</i></td>
          </tr>

          <tr>
            <td><b>a</b></td>

            <td>hexadecimal exponent notation</td>

            <td>sets floatfield bits to <i>scientific</i> | <i>fixed</i> (which is equivalent to <i>hexfloat</i>)</td>
          </tr>

          <tr>
            <td><b>e</b></td>

            <td>scientific float format</td>

            <td>sets floatfield bits to <i>scientific</i></td>
          </tr>

          <tr>
            <td><b>f</b></td>

            <td>fixed float format</td>

            <td>sets floatfield bits to <i>fixed</i></td>
          </tr>

          <tr>
            <td><b>g</b></td>

            <td>general -default- float format</td>

            <td><b>unset</b> all floatfield bits</td>
          </tr>

          <tr>
            <td><b>X, A, E, F</b> or <b>G</b></td>

            <td>same effect as their lowercase counterparts, but using
            uppercase letters for number outputs. (exponents, hex digits,
            ..)</td>

            <td>same effects as <i>'x'</i>, <i>'a'</i>, <i>'e'</i>, <i>'f'</i>, or <i>'g'</i> respectively,
            <b>plus</b> <i>uppercase</i></td>
          </tr>

          <tr>
            <td><b>d, i</b> or <b>u</b></td>

            <td><b>decimal</b> type output</td>

            <td>sets basefield bits to <i>dec</i></td>
          </tr>

          <tr>
            <td><b>s</b> or <b>S</b></td>

            <td>string output</td>

            <td><i>precision</i> specification is unset, and its value goes
            to an internal field for later 'truncation'. (see
            <i>precision</i> explanation above)</td>
          </tr>

          <tr>
            <td><b>c</b> or <b>C</b></td>

            <td>1-character output</td>

            <td>only the first character of the conversion string is
            used.</td>
          </tr>

          <tr>
            <td><b>%</b></td>

            <td>print the character <i>%</i></td>

            <td>N/A</td>
          </tr>
        </table>
      </blockquote>

      <p>Note that the 'n' conversion-specifier is ignored (and so is the
      corresponding argument), because it does not fit in this context.</p>
    </li>
  </ul><a name="new_directives" id="new_directives"></a>

  <h3>new format-specifications</h3>

  <ul>
    <li>as stated in the flags table, centered and internal alignment flags
    (' <i>=</i> ', and ' <i>_</i> ') were added.</li>

    <li><i><b>%{</b>n</i><b>t}</b> , where <i>n</i> is a positive number,
    inserts an <i>absolute tabulation</i>. It means that format will, if
    needed, fill the string with characters, until the length of the string
    created so far reaches <i>n</i> characters. (see <a href=
    "#examples">examples</a> )</li>

    <li><b>%|</b><i>n</i><b>T</b><i>X</i><b>|</b> inserts a tabulation in the
    same way, but using <i>X</i> as fill character instead of the current
    'fill' char of the stream (which is <i>space</i> for a stream in default
    state)</li>
  </ul><a name="printf_differences" id="printf_differences"></a>

  <h2>Differences of behaviour vs printf</h2>Suppose you have variables
  <i>x1, x2</i> (built_in types, supported by C's printf),<br>
  and a format string <i>s</i> intended for use with a printf function this
  way :

  <blockquote>
    <pre>
printf(s, x1, x2);
</pre>
  </blockquote><br>
  In almost all cases, the result will be the same as with this command :

  <blockquote>
    <pre>
cout &lt;&lt; format(s) % x1 % x2;
</pre>
  </blockquote>

  <p>But because some printf format specifications don't translate well into
  stream formatting options, there are a few notable imperfections in the way
  Boost.format emulates printf.<br>
  In any case, the <i>format</i> class should quietly ignore the unsupported
  options, so that printf format-strings are always accepted by format and
  produce almost the same output as printf.</p><br>
  Here is the full list of such differences :

  <ul>
    <li><b>'0'</b> and <b>' '</b> options : printf ignores these options for
    non numeric conversions, but format applies them to all types of
    variables. (so it is possible to use those options on user-defined types,
    e.g. a Rational class, etc..)</li>

    <li><b>precision</b> for integral types arguments has a special meaning
    for printf :<br>
    <i>printf( "(%5.3d)" , 7 ) ;</i> prints &laquo; ( 007) &raquo;<br>
    While format, like streams, ignores the precision parameter for integral
    types conversions.</li>

    <li>the <b>'</b> printf option (<i>format with thousands grouping
    characters)</i>) has no effect in format.</li>

    <li>Width or precision set to asterisk (<i>*</i>) are used by printf to
    read this field from an argument. e.g.
    <i>printf("%1$d:%2$.*3$d:%4$.*3$d\n", hour, min, precision, sec);</i><br>
    This class does not support this mechanism for now. so such precision or
    width fields are quietly ignored by the parsing.</li>

    <li>argument-type is ignored</li>
  </ul>Also, note that the special <b>'n'</b> type-specification (used to
  tell printf to save in a variable the number of characters output by the
  formatting) has no effect in format.<br>
  Thus format strings containing this type-specification should produce the
  same converted string by printf or format. It will not cause differences in
  the formatted strings between printf and format.<br>
  To get the number of characters in the formatted string using Boost.Format,
  you can use the <i>size()</i> member function :

  <blockquote>
    <pre>
format formatter("%+5d");
cout &lt;&lt; formatter % x;
unsigned int n = formatter.size();
</pre>
  </blockquote><a name="user-defined" id="user-defined"></a>
  <hr>

  <h2>User-defined types output</h2>

  <p>All flags which are translated into modification to the stream state act
  recursively within user-defined types. ( the flags remain active, and so
  does the desired format option, for each of the '&lt;&lt;' operations that
  might be called by the user-defined class)</p>e.g., with a Rational class,
  we would have something like :

  <blockquote>
    <pre>
Rational ratio(16,9);
cerr &lt;&lt; format("%#x \n")  % ratio;  // -&gt; "0x10/0x9 \n"
</pre>
  </blockquote>

  <p>It's a different story for other formatting options. For example,
  setting width applies to the final output produced by the object, not to
  each of its internal outputs, and that's fortunate :</p>

  <blockquote>
    <pre>
cerr &lt;&lt; format("%-8d")  % ratio;  // -&gt; "16/9    "      and not    "16      /9       "
cerr &lt;&lt; format("%=8d")  % ratio;  // -&gt; "  16/9  "      and not    "   16   /    9   "
</pre>
  </blockquote>

  <p><br>
  But so does the 0 and ' ' options (contrarily to '+' which is directly
  translated to the stream state by <i>showpos</i>. But no such flags exist
  for the zero and space printf options)<br>
  and that is less natural :</p>

  <blockquote>
    <pre>
cerr &lt;&lt; format("%+08d \n")  % ratio;  // -&gt; "+00016/9"
cerr &lt;&lt; format("% 08d \n")  % ratio;  // -&gt; "000 16/9"
</pre>
  </blockquote>It is possible to obtain a better behaviour by carefully
  designing the Rational's <i>operator&lt;&lt;</i> to handle the stream's
  width, alignment and <i>showpos</i> paramaters by itself. This is
  demonstrated in <a href=
  "../example/sample_userType.cpp">sample_userType.cpp</a>. <a name=
  "manipulators" id="manipulators"></a>
  <hr>

  <h3>Manipulators, and internal stream state</h3>

  <p>The internal stream state of <b>format</b> is saved before and restored
  after output of an argument; therefore, the modifiers are not sticky and
  affect only the argument they are applied to.<br>
  The default state for streams, as stated by the standard, is : precision 6,
  width 0, right alignment, and decimal flag set.</p>

  <p>The state of the internal <b>format</b> stream can be changed by
  manipulators passed along with the argument; via the <i>group</i> function,
  like that :</p>

  <blockquote>
    <pre>
cout &lt;&lt; format("%1% %2% %1%\n") % group(hex, showbase, 40) % 50; // prints "0x28 50 0x28\n"
</pre>
  </blockquote>

  <p><br>
  When passing N items inside a 'group' Boost.format needs to process
  manipulators diferently from regular argument, and thus using group is
  subject to the following constraints :</p>

  <ol>
    <li>the object to be printed must be passed as the last item in the
    group</li>

    <li>the first N-1 items are treated as manipulators, and if they do
    produce output, it is discarded</li>
  </ol>

  <p>Such manipulators are passed to the streams right before the following
  argument, at every occurrence. Note that formatting options specified within
  the format string are overridden by stream state modifiers passed this way.
  For instance in the following code, the <i>hex</i> manipulator has priority
  over the <i>d</i> type-specification in the format-string which would set
  decimal output :</p>

  <blockquote>
    <pre>
cout &lt;&lt; format("%1$d %2% %1%\n") % group(hex, showbase, 40) % 50;
// prints "0x28 50 0x28\n"
</pre>
  </blockquote><a name="alternatives" id="alternatives"></a>

  <h2>Alternatives</h2>

  <ul>
    <li><b>printf</b> is the classical alternative, that is not type safe and
    not extendable to user-defined types.</li>

    <li>ofrstream.cc by Karl Nelson's design was a big source of inspiration
    to this format class.</li>

    <li>James Kanze's library has a format class (in
    <i>srcode/Extended/format</i> ) which looks very well polished. Its
    design has in common with this class the use of internal stream for the
    actual conversions, as well as using operators to pass arguments. (but
    his class, as ofrstream, uses <i>operator&lt;&lt;</i> rather <i>than
    operator%</i> )</li>

    <li><a href="http://groups.yahoo.com/group/boost/files/format3/">Karl
    Nelson's library</a> was intented as demonstration of alternative
    solutions in discussions on Boost's list for the design of
    Boost.format.</li>

    <li><a href="http://fmtlib.net/latest/index.html">{fmt}</a> by Victor Zverovich.</li>
  </ul><a name="exceptions" id="exceptions"></a>
  <hr>

  <h2>Exceptions</h2>

  <p>Boost.format enforces a number of rules on the usage of format objects.
  The format-string must obeys the syntax described above, the user must
  supply exactly the right number of arguments before outputting to the final
  destination, and if using modify_item or bind_arg, items and arguments
  index must not be out of range.<br>
  When format detects that one of these rules is not satisfied, it raises a
  corresponding exception, so that the mistakes don't go unnoticed and
  unhandled.<br>
  But the user can change this behaviour to fit his needs, and select which
  types of errors may raise exceptions using the following functions :</p>

  <blockquote>
    <pre>

unsigned char exceptions(unsigned char newexcept); // query and set
unsigned char exceptions() const;                  // just query

</pre>
  </blockquote>

  <p>The user can compute the argument <i>newexcept</i> by combining the
  following atoms using binary arithmetic :</p>

  <ul>
    <li><b>boost::io::bad_format_string_bit</b> selects errors due to
    ill-formed format-strings.</li>

    <li><b>boost::io::too_few_args_bit</b> selects errors due to asking for
    the srting result before all arguments are passed.</li>

    <li><b>boost::io::too_many_args_bit</b> selects errors due to passing too
    many arguments.</li>

    <li><b>boost::io::out_of_range_bit</b> select errors due to out of range
    index supplied by the user when calling <i>modify_item</i> or other
    functions taking an item index (or an argument index)</li>

    <li><b>boost::io::all_error_bits</b> selects all errors</li>

    <li><b>boost::io::no_error_bits</b> selects no error.</li>
  </ul>

  <p>For instance, if you don't want Boost.format to detect bad number of
  arguments, you can define a specific wrapper function for building format
  objects with the right exceptions settings :</p>

  <blockquote>
    <pre>

boost::format  my_fmt(const std::string &amp; f_string) {
    using namespace boost::io;
    format fmter(f_string);
    fmter.exceptions( all_error_bits ^ ( too_many_args_bit | too_few_args_bit )  );
    return fmter;
}

</pre>
  </blockquote>It is then allowed to give more arguments than needed (they
  are simply ignored) :

  <blockquote>
    <pre>

cout &lt;&lt; my_fmt(" %1% %2% \n") % 1 % 2 % 3 % 4 % 5;

</pre>
  </blockquote>And if we ask for the result before all arguments are
  supplied, the corresponding part of the result is simply empty

  <blockquote>
    <pre>

cout &lt;&lt; my_fmt(" _%2%_ _%1%_ \n") % 1 ;
// prints      " __ _1_ \n"

</pre>
  </blockquote><a name="performance" id="performance"></a>
  <hr>

  <h2>A Note about performance</h2>

  <p>The performance of boost::format for formatting a few builtin type
  arguments with reordering can be compared to that of Posix-printf, and of
  the equivalent stream manual operations to give a measure of the overhead
  incurred. The result may greatly depend on the compiler, standard library
  implementation, and the precise choice of format-string and arguments.</p>

  <p>Common stream implementations eventually call functions of the
  printf family for the actual formatting of numbers.  In general, printf
  will be noticeably faster than the direct stream operations due to the
  reordering overhead (allocations to store the pieces of string, stream
  initialisation at each item formatting, ..).  The direct stream operations
  would be faster than boost::format - one can expect a ratio ranging from 2
  to 5 or more.</p>

  <p>When iterated formattings are a performance bottleneck, performance can
  be slightly increased by parsing the format string into a format object,
  and copying it at each formatting, in the following way:</p>

  <blockquote>
    <pre>
    const boost::format fmter(fstring);
    dest &lt;&lt; boost::format(fmter) % arg1 % arg2 % arg3 ;
    </pre>
  </blockquote>

  <p>As an example of performance results, the author measured the time of
  execution of iterated formattings with 4 different methods:</p>

  <ol>
    <li>posix printf</li>

    <li>manual stream output (to a dummy <i>nullStream</i> stream sending the
    bytes into oblivion)</li>

    <li>boost::format copied from a const object as shown above</li>

    <li>the straight boost::format usage</li>
  </ol>

  <p>the test was compiled with g++-3.3.3 and the following timings were
  measured (in seconds, and ratios):</p>

  <blockquote>
    <pre>
string     fstring="%3$0#6x %1$20.10E %2$g %3$0+5d \n";
double     arg1=45.23;
double     arg2=12.34;
int        arg3=23;

- release mode :
printf                 : 2.13
nullStream             : 3.43,  = 1.61033 * printf
boost::format copied   : 6.77,  = 3.1784  * printf ,  = 1.97376 * nullStream
boost::format straight :10.67,  = 5.00939 * printf ,  = 3.11079 * nullStream

- debug mode :
printf                 : 2.12
nullStream             : 3.69,  = 1.74057 * printf
boost::format copied   :10.02,  = 4.72642 * printf ,  = 2.71545 * nullStream
boost::format straight :17.03,  = 8.03302 * printf ,  = 4.61518 * nullStream
</pre>
  </blockquote><a name="extract" id="extract"></a>
  <hr>

  <h2>Class Interface Extract</h2>

  <blockquote>
    <pre>
namespace boost {

template&lt;class charT, class Traits=std::char_traits&lt;charT&gt; &gt;
class basic_format
{
public:
  typedef std::basic_string&lt;charT, Traits&gt; string_type;
  typedef typename string_type::size_type     size_type;
  basic_format(const charT* str);
  basic_format(const charT* str, const std::locale &amp; loc);
  basic_format(const string_type&amp; s);
  basic_format(const string_type&amp; s, const std::locale &amp; loc);
  basic_format&amp; operator= (const basic_format&amp; x);

  void clear(); // reset buffers
  basic_format&amp; parse(const string_type&amp;); // clears and parse a new format string

  string_type str() const;
  size_type size() const;

  // pass arguments through those operators :
  template&lt;class T&gt;  basic_format&amp;   operator%(T&amp; x);
  template&lt;class T&gt;  basic_format&amp;   operator%(const T&amp; x);

  // dump buffers to ostream :
  friend std::basic_ostream&lt;charT, Traits&gt;&amp;
  operator&lt;&lt; &lt;&gt; ( std::basic_ostream&lt;charT, Traits&gt;&amp; , basic_format&amp; );

   // Choosing which errors will throw exceptions :
   unsigned char exceptions() const;
   unsigned char exceptions(unsigned char newexcept);

// ............  this is just an extract .......
}; // basic_format

typedef basic_format&lt;char &gt;          format;
typedef basic_format&lt;wchar_t &gt;      wformat;


// free function for ease of use :
template&lt;class charT, class Traits&gt;
std::basic_string&lt;charT,Traits&gt;  str(const basic_format&lt;charT,Traits&gt;&amp; f) {
      return f.str();
}


} // namespace boost
</pre>
  </blockquote>
  <hr>
  <a name="rationale" id="rationale"></a>

  <h2>Rationale</h2>

  <p>This class's goal is to bring a better, C++, type-safe and
  type-extendable <i>printf</i> equivalent to be used with
  streams.</p>Precisely, <b>format</b> was designed to provide the following
  features :

  <ul>
    <li>support positional arguments (required for internationalisation)</li>

    <li>accept an unlimited number of arguments.</li>

    <li>make formatting commands visually natural.</li>

    <li>support the use of manipulators to modify the display of an argument.
    in addition to the format-string syntax.</li>

    <li>accept any types of variables, by relying on streams for the actual
    conversion to string. This specifically concerns user-defined types, for
    which the formatting options effects should be intuitively natural.</li>

    <li>provide printf-compatibility, as much as it makes sense in a
    type-safe and type-extendable context.</li>
  </ul>

  <p>In the process of the design, many issues were faced, and some choices
  were made, that might not be intuitively right. But in each case they were
  taken for <a href="choices.html">some reasons</a>.</p>
  <hr>

  <h2>Credits</h2>

  <p>The author of Boost format is Samuel Krempp. &nbsp; He used ideas from
  R&uuml;diger Loos' format.hpp and Karl Nelson's formatting classes.</p>
  <hr>

  <p><a href="http://validator.w3.org/check?uri=referer"><img border="0" src=
  "../../../doc/images/valid-html401.png" alt="Valid HTML 4.01 Transitional"
  height="31" width="88"></a></p>

  <p>Revised
  <!--webbot bot="Timestamp" s-type="EDITED" s-format="%d %B, %Y" startspan -->23 October, 2017<!--webbot bot="Timestamp" endspan i-checksum="38510" --></p>

  <p><i>Copyright &copy; 2002 Samuel Krempp</i></p>

  <p><i>Distributed under the Boost Software License, Version 1.0. (See
  accompanying file <a href="../../../LICENSE_1_0.txt">LICENSE_1_0.txt</a> or
  copy at <a href=
  "http://www.boost.org/LICENSE_1_0.txt">http://www.boost.org/LICENSE_1_0.txt</a>)</i></p>
</body>
</html>