<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <HTML> <HEAD> <TITLE>Tutorial</TITLE> <LINK REL="stylesheet" HREF="../../../../boost.css"> <LINK REL="stylesheet" HREF="../theme/iostreams.css"> </HEAD> <BODY> <!-- Begin Banner --> <H1 CLASS="title">Tutorial</H1> <HR CLASS="banner"> <!-- End Banner --> <!-- Begin Nav --> <DIV CLASS='nav'> <A HREF='unix2dos_filters.html'><IMG BORDER=0 WIDTH=19 HEIGHT=19 SRC='../../../../doc/html/images/prev.png'></A> <A HREF='tutorial.html'><IMG BORDER=0 WIDTH=19 HEIGHT=19 SRC='../../../../doc/html/images/up.png'></A> <A HREF='dual_use_filters.html'><IMG BORDER=0 WIDTH=19 HEIGHT=19 SRC='../../../../doc/html/images/next.png'></A> </DIV> <!-- End Nav --> <A NAME="multichar"></A> <H2>2.2.8. Multi-Character Filters</H2> <P> All the Filters you've seen so far — except for those that derive from <CODE>stdio_filter</CODE> — process characters one at a time. If you instead process several characters at once, you can often reduce the number of function calls it takes to filter a character sequence, resulting in more efficient code. This is what <A HREF="../concepts/multi_character.html">Multi-Character Filters</A> allow you to do. </P> <A NAME="multichar_input_filters"></A> <H4>Multi-Character InputFilters</H4> <P>A typical narrow-character <A HREF="../concepts/multi_character.html">Multi-Character</A> <A HREF="../concepts/input_filter.html">InputFilter</A> looks like this:<P> <PRE CLASS="broken_ie"><SPAN CLASS='preprocessor'>#include</SPAN> <SPAN CLASS="literal"><iosfwd></SPAN> <SPAN CLASS='comment'>// streamsize</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="HEADER" HREF="../../../../boost/iostreams/categories.hpp"><SPAN CLASS='literal'><boost/iostreams/categories.hpp></SPAN></A> <SPAN CLASS='comment'>// tags </SPAN> <SPAN CLASS='keyword'>class</SPAN> my_input_filter { <SPAN CLASS='keyword'>public</SPAN>: <SPAN CLASS='keyword'>typedef</SPAN> <SPAN CLASS='keyword'>char</SPAN> char_type; <SPAN CLASS='keyword'>struct</SPAN> category : input_filter_tag, multichar_tag { }; <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Source> std::streamsize read(Source& src, <SPAN CLASS='keyword'>char</SPAN>* s, std::streamsize n) { <SPAN CLASS='comment'>// Read up to n filtered characters into the buffer s,</SPAN> <SPAN CLASS='comment'>// returning the number of characters read or -1 for EOF.</SPAN> <SPAN CLASS='comment'>// Use src to access the unfiltered character sequence</SPAN> } <SPAN CLASS='comment'>/* Other members */</SPAN> };</PRE> <P> Notice that the member type category is a <CODE>struct</CODE> convertible to <CODE>input_filter_tag</CODE> and to <CODE>multichar_tag</CODE>. This tells the Iostream library that <CODE>my_input_filter</CODE> is a <A HREF="../concepts/multi_character.html">Multi-Character Filter</A> and an <A HREF="../concepts/input_filter.html">InputFilter</A>. You could have achieved the same effect several ways. <I>E.g.</I>, </P> <PRE CLASS="broken_ie"> <SPAN CLASS='keyword'>struct</SPAN> category : input, filter_tag, multichar_tag { }; <SPAN CLASS='comment'>/* or */</SPAN> <SPAN CLASS='keyword'>typedef</SPAN> multichar_input_filter_tag category;</PRE> <P>(For details, <I>see</I> <A HREF="../guide/modes.html#mode_tags">Mode Tags</A> and <A HREF="../guide/traits.html#category_tags">Category Tags</A>.) <P> You could also write the above example as follows:</P> <PRE CLASS="broken_ie"><SPAN CLASS='preprocessor'>#include</SPAN> <SPAN CLASS="literal"><iosfwd></SPAN> <SPAN CLASS='comment'>// streamsize</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="HEADER" HREF="../../../../boost/iostreams/concepts.hpp"><SPAN CLASS='literal'><boost/iostreams/concepts.hpp></SPAN></A> <SPAN CLASS='comment'>// multichar_input_filter</SPAN> <SPAN CLASS='keyword'>class</SPAN> my_input_filter : <SPAN CLASS='keyword'>public</SPAN> multichar_input_filter { <SPAN CLASS='keyword'>public</SPAN>: <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Source> std::streamsize read(<SPAN CLASS='keyword'>char</SPAN>* s, std::streamsize n); <SPAN CLASS='comment'>/* Other members */</SPAN> };</PRE> <P>Here <A HREF="../classes/filter.html#synopsis"><CODE>multichar_input_filter</CODE></A> is a convenience base class which provides the member types <CODE>char_type</CODE> and <CODE>category</CODE>, as well as no-op implementations of member functions <CODE>close</CODE> and <CODE>imbue</CODE>. <A NAME="shell_comments_multichar_input_filter"></A> <H4><CODE>shell_comments_multichar_input_filter</CODE></H4> <P>You can express a <A HREF="shell_comments_filters.html">shell comments Filter</A> as an <A HREF="../concepts/multi_character.html">Multi-Character</A> <A HREF="../concepts/input_filter.html">InputFilter</A> as follows:</P> <PRE CLASS="broken_ie"><SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="header" HREF="../../../../boost/iostreams/char_traits.hpp"><SPAN CLASS="literal"><boost/iostreams/char_traits.hpp></SPAN></A> <SPAN CLASS="comment">// EOF, WOULD_BLOCK</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="header" HREF="../../../../boost/iostreams/concepts.hpp"><SPAN CLASS="literal"><boost/iostreams/concepts.hpp></SPAN></A> <SPAN CLASS="comment">// multichar_input_filter</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="header" HREF="../../../../boost/iostreams/operations.hpp"><SPAN CLASS="literal"><boost/iostreams/operations.hpp></SPAN></A> <SPAN CLASS="comment">// get</SPAN> <SPAN CLASS='keyword'>namespace</SPAN> boost { <SPAN CLASS='keyword'>namespace</SPAN> iostreams { <SPAN CLASS='keyword'>namespace</SPAN> example { <SPAN CLASS="keyword">class</SPAN> shell_comments_multichar_input_filter : <SPAN CLASS="keyword">public</SPAN> multichar_input_filter { <SPAN CLASS="keyword">public</SPAN>: <SPAN CLASS="keyword">explicit</SPAN> shell_comments_multichar_input_filter(<SPAN CLASS="keyword">char</SPAN> comment_char = '#') : comment_char_(comment_char), skip_(<SPAN CLASS="keyword">false</SPAN>) { } <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Source> std::streamsize read(Source& src, <SPAN CLASS="keyword">char</SPAN>* s, std::streamsize n) { <SPAN CLASS="keyword">for</SPAN> (std::streamsize z = <SPAN CLASS='numeric_literal'>0</SPAN>; z < n; ++z) { <SPAN CLASS="keyword">int</SPAN> c; <SPAN CLASS="keyword">while</SPAN> (true) { <SPAN CLASS="keyword">if</SPAN> ((c = boost::iostreams::get(src)) == <SPAN CLASS='numeric_literal'>EOF</SPAN>) <SPAN CLASS="keyword">return</SPAN> z != <SPAN CLASS='numeric_literal'>0</SPAN> ? z : <SPAN CLASS='numeric_literal'>-1</SPAN>; <SPAN CLASS="keyword">else</SPAN> <SPAN CLASS="keyword">if</SPAN> (c == WOULD_BLOCK) <SPAN CLASS="keyword">return</SPAN> z; skip_ = c == comment_char_ ? <SPAN CLASS="keyword">true</SPAN> : c == '\n' ? <SPAN CLASS="keyword">false</SPAN> : skip_; <SPAN CLASS="keyword">if</SPAN> (!skip_) <SPAN CLASS="keyword">break</SPAN>; } s[z] = c; } <SPAN CLASS="keyword">return</SPAN> n; } <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Source> <SPAN CLASS="keyword">void</SPAN> close(Source&) { skip_ = false; } <SPAN CLASS="keyword">private</SPAN>: <SPAN CLASS="keyword">char</SPAN> comment_char_; <SPAN CLASS="keyword">bool</SPAN> skip_; }; } } } <SPAN CLASS="comment">// End namespace boost::iostreams:example</SPAN></PRE> <P> Note that the implementation of <CODE>read</CODE> is very similar to what you would get if you put the implementation of <A HREF="shell_comments_filters.html#shell_comments_input_filter"><CODE>shell_comments_input_filter::get</CODE></A> inside a <CODE>for</CODE> loop iterating from <CODE>0</CODE> to <CODE>n</CODE>. InputFilters which call themselves recursively, such as <A HREF="tab_expanding_filters.html#tab_expanding_input_filter"><CODE>tab_expanding_input_filter</CODE></A>, are much harder to transform into Multi-Character filters. </P> <A NAME="multichar_output_filters"></A> <H4>Multi-Character OutputFilters</H4> <P>A typical narrow-character <A HREF="../concepts/multi_character.html">Multi-Character</A> <A HREF="../concepts/output_filter.html">OutputFilter</A> looks like this:<P> <PRE CLASS="broken_ie"><SPAN CLASS='preprocessor'>#include</SPAN> <SPAN CLASS="literal"><iosfwd></SPAN> <SPAN CLASS='comment'>// streamsize</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="HEADER" HREF="../../../../boost/iostreams/categories.hpp"><SPAN CLASS='literal'><boost/iostreams/categories.hpp></SPAN></A> <SPAN CLASS='comment'>// tags </SPAN> <SPAN CLASS='keyword'>class</SPAN> my_output_filter { <SPAN CLASS='keyword'>public</SPAN>: <SPAN CLASS='keyword'>typedef</SPAN> <SPAN CLASS='keyword'>char</SPAN> char_type; <SPAN CLASS='keyword'>struct</SPAN> category : output_filter_tag, multichar_tag { }; <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Sink> std::streamsize write(Sink& dest, <SPAN CLASS='keyword'>const</SPAN> <SPAN CLASS='keyword'>char</SPAN>* s, std::streamsize n) { <SPAN CLASS='comment'>// Consume up to n filtered characters from the buffer s,</SPAN> <SPAN CLASS='comment'>// writing filtered characters to dest. Return the number</SPAN> <SPAN CLASS='comment'>// of characters consumed.</SPAN> } <SPAN CLASS='comment'>/* Other members */</SPAN> };</PRE> <P> Notice that the member type category is a <CODE>struct</CODE> convertible to <CODE>keyword</CODE> and to <CODE>multichar_tag</CODE>. This tells the Iostream library that <CODE>my_output_filter</CODE> is a <A HREF="../concepts/multi_character.html">Multi-Character Filter</A> and an <A HREF="../concepts/output_filter.html">OutputFilter</A>. As with <A HREF="../concepts/multi_character.html">Multi-Character</A> <A HREF="../concepts/input_filter.html">InputFilters</A>, you could have achieved the same effect several different ways. <I>E.g.</I>, </P> <PRE CLASS="broken_ie"> <SPAN CLASS='keyword'>struct</SPAN> category : output, filter_tag, multichar_tag { }; <SPAN CLASS='comment'>/* or */</SPAN> <SPAN CLASS='keyword'>typedef</SPAN> multichar_output_filter_tag category;</PRE> <P>(For details, <I>see</I> <A HREF="../guide/modes.html#mode_tags">Mode Tags</A> and <A HREF="../guide/traits.html#category_tags">Category Tags</A>.) <P> You could also write the above example as follows:</P> <PRE CLASS="broken_ie"><SPAN CLASS='preprocessor'>#include</SPAN> <SPAN CLASS="literal"><iosfwd></SPAN> <SPAN CLASS='comment'>// streamsize</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="HEADER" HREF="../../../../boost/iostreams/concepts.hpp"><SPAN CLASS='literal'><boost/iostreams/concepts.hpp></SPAN></A> <SPAN CLASS='comment'>// multichar_output_filter</SPAN> <SPAN CLASS='keyword'>class</SPAN> my_output_filter : <SPAN CLASS='keyword'>public</SPAN> multichar_output_filter { <SPAN CLASS='keyword'>public</SPAN>: <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Sink> std::streamsize write(Sink& dest, <SPAN CLASS='keyword'>const</SPAN> <SPAN CLASS='keyword'>char</SPAN>* s, std::streamsize n); <SPAN CLASS='comment'>/* Other members */</SPAN> };</PRE> <P>Here <A HREF="../classes/filter.html#synopsis"><CODE>multichar_output_filter</CODE></A> is a convenience base class which provides the member types <CODE>char_type</CODE> and <CODE>category</CODE>, as well as no-op implementations of member functions <CODE>close</CODE> and <CODE>imbue</CODE>. <A NAME="shell_comments_multichar_ouput_filter"></A> <H4><CODE>shell_comments_multichar_output_filter</CODE></H4> <P>You can express a <A HREF="shell_comments_filters.html">shell comments Filter</A> as an <A HREF="../concepts/multi_character.html">Multi-Character</A> <A HREF="../concepts/output_filter.html">OutputFilter</A> as follows:</P> <PRE CLASS="broken_ie"><SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="header" HREF="../../../../boost/iostreams/char_traits.hpp"><SPAN CLASS="literal"><boost/iostreams/char_traits.hpp></SPAN></A> <SPAN CLASS="comment">// EOF, WOULD_BLOCK</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="header" HREF="../../../../boost/iostreams/concepts.hpp"><SPAN CLASS="literal"><boost/iostreams/concepts.hpp></SPAN></A> <SPAN CLASS="comment">// multichar_output_filter</SPAN> <SPAN CLASS='preprocessor'>#include</SPAN> <A CLASS="header" HREF="../../../../boost/iostreams/operations.hpp"><SPAN CLASS="literal"><boost/iostreams/operations.hpp></SPAN></A> <SPAN CLASS="comment">// get</SPAN> <SPAN CLASS='keyword'>namespace</SPAN> boost { <SPAN CLASS='keyword'>namespace</SPAN> iostreams { <SPAN CLASS='keyword'>namespace</SPAN> example { <SPAN CLASS="keyword">class</SPAN> shell_comments_multichar_output_filter : <SPAN CLASS="keyword">public</SPAN> multichar_output_filter { <SPAN CLASS="keyword">public</SPAN>: <SPAN CLASS="keyword">explicit</SPAN> shell_comments_multichar_output_filter(<SPAN CLASS="keyword">char</SPAN> comment_char = '#') : comment_char_(comment_char), skip_(<SPAN CLASS="keyword">false</SPAN>) { } <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Sink> std::streamsize write(Sink& dest, <SPAN CLASS="keyword">const</SPAN> char* s, std::streamsize n) { std::streamsize z; <SPAN CLASS="keyword">for</SPAN> (z = <SPAN CLASS='numeric_literal'>0</SPAN>; z < n; ++z) { <SPAN CLASS="keyword">int</SPAN> c = s[z]; skip_ = c == comment_char_ ? <SPAN CLASS="keyword">true</SPAN> : c == <SPAN CLASS='literal'>'\n'</SPAN> ? <SPAN CLASS="keyword">false</SPAN> : skip_; <SPAN CLASS="keyword">if</SPAN> (skip_) <SPAN CLASS="keyword">continue</SPAN>; <SPAN CLASS="keyword">if</SPAN> (!iostreams::put(dest, c)) <SPAN CLASS="keyword">break</SPAN>; } <SPAN CLASS="keyword">return</SPAN> z; } <SPAN CLASS="keyword">template</SPAN><<SPAN CLASS="keyword">typename</SPAN> Source> <SPAN CLASS="keyword">void</SPAN> close(Source&) { skip_ = <SPAN CLASS="keyword">false</SPAN>; } <SPAN CLASS="keyword">private</SPAN>: <SPAN CLASS="keyword">char</SPAN> comment_char_; <SPAN CLASS="keyword">bool</SPAN> skip_; }; } } } <SPAN CLASS="comment">// End namespace boost::iostreams:example</SPAN></PRE> <P> Note that the implementation of <CODE>write</CODE> is very similar to what you would get if you put the implementation of <A HREF="shell_comments_filters.html#shell_comments_output_filter"><CODE>shell_comments_output_filter::put</CODE></A> inside a <CODE>for</CODE> loop iterating from <CODE>0</CODE> to <CODE>n</CODE>. OutputFilters which call themselves recursively, such as <A HREF="unix2dos_filters.html#unix2dos_output_filter"><CODE>unix2dos_output_filter</CODE></A>, are much harder to transform into Multi-Character filters. </P> <!-- Begin Nav --> <DIV CLASS='nav'> <A HREF='unix2dos_filters.html'><IMG BORDER=0 WIDTH=19 HEIGHT=19 SRC='../../../../doc/html/images/prev.png'></A> <A HREF='tutorial.html'><IMG BORDER=0 WIDTH=19 HEIGHT=19 SRC='../../../../doc/html/images/up.png'></A> <A HREF='dual_use_filters.html'><IMG BORDER=0 WIDTH=19 HEIGHT=19 SRC='../../../../doc/html/images/next.png'></A> </DIV> <!-- End Nav --> <!-- Begin Footer --> <HR> <P CLASS="copyright">Revised 02 Feb 2008</P> <P CLASS="copyright">© Copyright 2008 <a href="http://www.coderage.com/" target="_top">CodeRage, LLC</a><br/>© Copyright 2004-2007 <a href="http://www.coderage.com/turkanis/" target="_top">Jonathan Turkanis</a></P> <P CLASS="copyright"> Use, modification, and distribution are subject to the Boost Software License, Version 2.0. (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>) </P> <!-- End Footer --> </BODY>