gcc/libstdc++-v3/docs/27_io/howto.html
Benjamin Kosnik b2dad0e372 libstdc++-v3: New directory.
2000-04-21  Benjamin Kosnik  <bkoz@redhat.com>

	* libstdc++-v3: New directory.

From-SVN: r33317
2000-04-21 20:33:34 +00:00

347 lines
17 KiB
HTML

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="AUTHOR" CONTENT="pme@sourceware.cygnus.com (Phil Edwards)">
<META NAME="KEYWORDS" CONTENT="HOWTO, libstdc++, egcs, g++, libg++, STL">
<META NAME="DESCRIPTION" CONTENT="HOWTO for the libstdc++ chapter 27.">
<META NAME="GENERATOR" CONTENT="vi and eight fingers">
<TITLE>libstdc++-v3 HOWTO: Chapter 27</TITLE>
<LINK REL="home" HREF="http://sourceware.cygnus.com/libstdc++/docs/27_io/">
<LINK REL=StyleSheet HREF="../lib3styles.css">
<!-- $Id: howto.html,v 1.7 1999/12/15 16:57:06 pme Exp $ -->
</HEAD>
<BODY>
<H1 CLASS="centered"><A NAME="top">Chapter 27: Input/Output</A></H1>
<P>Chapter 27 deals with iostreams and all their subcomponents
and extensions. All <EM>kinds</EM> of fun stuff.
</P>
<!-- ####################################################### -->
<HR>
<H1>Contents</H1>
<UL>
<LI><A HREF="#1">Copying a file</A>
<LI><A HREF="#2">The buffering is screwing up my program!</A>
<LI><A HREF="#3">Binary I/O</A>
<LI><A HREF="#4">Iostreams class hierarchy diagram</A>
<LI><A HREF="#5">What is this &lt;sstream&gt;/stringstreams thing?</A>
</UL>
<HR>
<!-- ####################################################### -->
<H2><A NAME="1">Copying a file</A></H2>
<P>So you want to copy a file quickly and easily, and most important,
completely portably. And since this is C++, you have an open
ifstream (call it IN) and an open ofstream (call it OUT):
<PRE>
#include &lt;fstream&gt;
std::ifstream IN ("input_file");
std::ofstream OUT ("output_file"); </PRE>
</P>
<P>Here's the easiest way to get it completely wrong:
<PRE>
OUT &lt;&lt; IN;</PRE>
For those of you who don't already know why this doesn't work
(probably from having done it before), I invite you to quickly
create a simple text file called &quot;input_file&quot; containing
the sentence
<PRE>
The quick brown fox jumped over the lazy dog.</PRE>
surrounded by blank lines. Code it up and try it. The contents
of &quot;output_file&quot; may surprise you.
</P>
<P>Seriously, go do it. Get surprised, then come back. It's worth it.
</P>
<HR WIDTH="60%">
<P>The thing to remember is that the <TT>basic_[io]stream</TT> classes
handle formatting, nothing else. In particular, they break up on
whitespace. The actual reading, writing, and storing of data is
handled by the <TT>basic_streambuf</TT> family. Fortunately, the
<TT>operator&lt;&lt;</TT> is overloaded to take an ostream and
a pointer-to-streambuf, in order to help with just this kind of
&quot;dump the data verbatim&quot; situation.
</P>
<P>Why a <EM>pointer</EM> to streambuf and not just a streambuf? Well,
the [io]streams hold pointers (or references, depending on the
implementation) to their buffers, not the actual
buffers. This allows polymorphic behavior on the part of the buffers
as well as the streams themselves. The pointer is easily retrieved
using the <TT>rdbuf()</TT> member function. Therefore, the easiest
way to copy the file is:
<PRE>
OUT &lt;&lt; IN.rdbuf();</PRE>
</P>
<P>So what <EM>was</EM> happening with OUT&lt;&lt;IN? Undefined
behavior, since that particular &lt;&lt; isn't defined by the Standard.
I have seen instances where it is implemented, but the character
extraction process removes all the whitespace, leaving you with no
blank lines and only &quot;Thequickbrownfox...&quot;. With
libraries that do not define that operator, IN (or one of IN's
member pointers) sometimes gets converted to a void*, and the output
file then contains a perfect text representation of a hexidecimal
address (quite a big surprise). Others don't compile at all.
</P>
<P>Also note that none of this is specific to o<B>*f*</B>streams.
The operators shown above are all defined in the parent
basic_ostream class and are therefore available with all possible
descendents.
</P>
<P>Return <A HREF="#top">to top of page</A> or
<A HREF="../faq/index.html">to the FAQ</A>.
</P>
<HR>
<H2><A NAME="2">The buffering is screwing up my program!</A></H2>
<!--
This is not written very well. I need to redo this section.
-->
<P>First, are you sure that you understand buffering? Particularly
the fact that C++ may not, in fact, have anything to do with it?
</P>
<P>The rules for buffering can be a little odd, but they aren't any
different from those of C. (Maybe that's why they can be a bit
odd.) Many people think that writing a newline to an output
stream automatically flushes the output buffer. This is true only
when the output stream is, in fact, a terminal and not a file
or some other device -- and <EM>that</EM> may not even be true
since C++ says nothing about files nor terminals. All of that is
system-dependant. (The &quot;newline-buffer-flushing only occuring
on terminals&quot; thing is mostly true on Unix systems, though.)
</P>
<P>Some people also believe that sending <TT>endl</TT> down an
output stream only writes a newline. This is incorrect; after a
newline is written, the buffer is also flushed. Perhaps this
is the effect you want when writing to a screen -- get the text
out as soon as possible, etc -- but the buffering is largely
wasted when doing this to a file:
<PRE>
output &lt;&lt; &quot;a line of text&quot; &lt;&lt; endl;
output &lt;&lt; some_data_variable &lt;&lt; endl;
output &lt;&lt; &quot;another line of text&quot; &lt;&lt; endl; </PRE>
The proper thing to do in this case to just write the data out
and let the libraries and the system worry about the buffering.
If you need a newline, just write a newline:
<PRE>
output &lt;&lt; &quot;a line of text\n&quot;
&lt;&lt; some_data_variable &lt;&lt; '\n'
&lt;&lt; &quot;another line of text\n&quot;; </PRE>
I have also joined the output statements into a single statement.
You could make the code prettier by moving the single newline to
the start of the quoted text on the thing line, for example.
</P>
<P>If you do need to flush the buffer above, you can send an
<TT>endl</TT> if you also need a newline, or just flush the buffer
yourself:
<PRE>
output &lt;&lt; ...... &lt;&lt; flush; // can use std::flush manipulator
output.flush(); // or call a member fn </PRE>
</P>
<P>On the other hand, there are times when writing to a file should
be like writing to standard error; no buffering should be done
because the data needs to appear quickly (a prime example is a
log file for security-related information). The way to do this is
just to turn off the buffering <EM>before any I/O operations at
all</EM> have been done, i.e., as soon as possible after opening:
<PRE>
std::ofstream os (&quot;/foo/bar/baz&quot;);
std::ifstream is (&quot;/qux/quux/quuux&quot;);
int i;
os.rdbuf()-&gt;pubsetbuf(0,0);
is.rdbuf()-&gt;pubsetbuf(0,0);
...
os &lt;&lt; &quot;this data is written immediately\n&quot;;
is &gt;&gt; i; // and this will probably cause a disk read </PRE>
</P>
<P>Since all aspects of buffering are handled by a streambuf-derived
member, it is necessary to get at that member with <TT>rdbuf()</TT>.
Then the public version of <TT>setbuf</TT> can be called. The
arguments are the same as those for the Standard C I/O Library
function (a buffer area followed by its size).
</P>
<P>A great deal of this is implementation-dependant. For example,
<TT>streambuf</TT> does not specify any actions for its own
<TT>setbuf()</TT>-ish functions; the classes derived from
<TT>streambuf</TT> each define behavior that &quot;makes
sense&quot; for that class: an argument of (0,0) turns off
buffering for <TT>filebuf</TT> but has undefined behavior for
its sibling <TT>stringbuf</TT>, and specifying anything other
than (0,0) has varying effects. Other user-defined class derived
from streambuf can do whatever they want.
</P>
<P>A last reminder: there are usually more buffers involved than
just those at the language/library level. Kernel buffers, disk
buffers, and the like will also have an effect. Inspecting and
changing those are system-dependant.
</P>
<P>Return <A HREF="#top">to top of page</A> or
<A HREF="../faq/index.html">to the FAQ</A>.
</P>
<HR>
<H2><A NAME="3">Binary I/O</A></H2>
<P>The first and most important thing to remember about binary I/O is
that opening a file with <TT>ios::binary</TT> is not, repeat
<EM>not</EM>, the only thing you have to do. It is not a silver
bullet, and will not allow you to use the <TT>&lt;&lt;/&gt;&gt;</TT>
operators of the normal fstreams to do binary I/O.
</P>
<P>Sorry. Them's the breaks.
</P>
<P>This isn't going to try and be a complete tutorial on reading and
writing binary files (because &quot;binary&quot; covers a lot of
ground), but we will try and clear up a couple of misconceptions
and common errors.
</P>
<P>First, <TT>ios::binary</TT> has exactly one defined effect, no more
and no less. Normal text mode has to be concerned with the newline
characters, and the runtime system will translate between (for
example) '\n' and the appropriate end-of-line sequence (LF on Unix,
CRLF on DOS, CR on Macintosh, etc). (There are other things that
normal mode does, but that's the most obvious.) Opening a file in
binary mode disables this conversion, so reading a CRLF sequence
under Windows won't accidentally get mapped to a '\n' character, etc.
Binary mode is not supposed to suddenly give you a bitstream, and
if it is doing so in your program then you've discovered a bug in
your vendor's compiler (or some other part of the C++ implementation,
possibly the runtime system).
</P>
<P>Second, using <TT>&lt;&lt;</TT> to write and <TT>&gt;&gt;</TT> to
read isn't going to work with the standard file stream classes, even
if you use <TT>skipws</TT> during reading. Why not? Because
ifstream and ofstream exist for the purpose of <EM>formatting</EM>,
not reading and writing. Their job is to interpret the data into
text characters, and that's exactly what you don't want to happen
during binary I/O.
</P>
<P>Third, using the <TT>get()</TT> and <TT>put()/write()</TT> member
functions still aren't guaranteed to help you. These are
&quot;unformatted&quot; I/O functions, but still character-based.
(This may or may not be what you want.)
</P>
<P>Notice how all the problems here are due to the inappropriate use
of <EM>formatting</EM> functions and classes to perform something
which <EM>requires</EM> that formatting not be done? There are a
seemingly infinite number of solutions, and a few are listed here:
<UL>
<LI>&quot;Derive your own fstream-type classes and write your own
&lt;&lt;/&gt;&gt; operators to do binary I/O on whatever data
types you're using.&quot; This is a Bad Thing, because while
the compiler would probably be just fine with it, other humans
are going to be confused. The overloaded bitshift operators
have a well-defined meaning (formatting), and this breaks it.
<LI>&quot;Build the file structure in memory, then <TT>mmap()</TT>
the file and copy the structure.&quot; Well, this is easy to
make work, and easy to break, and is pretty equivalent to
using <TT>::read()</TT> and <TT>::write()</TT> directly, and
makes no use of the iostream library at all...
<LI>&quot;Use streambufs, that's what they're there for.&quot;
While not trivial for the beginner, this is the best of all
solutions. The streambuf/filebuf layer is the layer that is
responsible for actual I/O. If you want to use the C++
library for binary I/O, this is where you start.
</UL>
</P>
<P>How to go about using streambufs is a bit beyond the scope of this
document (at least for now), but while streambufs go a long way,
they still leave a couple of things up to you, the programmer.
As an example, byte ordering is completely between you and the
operating system, and you have to handle it yourself.
</P>
<P>Deriving a streambuf or filebuf
class from the standard ones, one that is specific to your data
types (or an abstraction thereof) is probably a good idea, and
lots of examples exist in journals and on Usenet. Using the
standard filebufs directly (either by declaring your own or by
using the pointer returned from an fstream's <TT>rdbuf()</TT>)
is certainly feasible as well.
</P>
<P>One area that causes problems is trying to do bit-by-bit operations
with filebufs. C++ is no different from C in this respect: I/O
must be done at the byte level. If you're trying to read or write
a few bits at a time, you're going about it the wrong way. You
must read/write an integral number of bytes and then process the
bytes. (For example, the streambuf functions take and return
variables of type <TT>int_type</TT>.)
</P>
<P>Another area of problems is opening text files in binary mode.
Generally, binary mode is intended for binary files, and opening
text files in binary mode means that you now have to deal with all of
those end-of-line and end-of-file problems that we mentioned before.
An instructive thread from comp.lang.c++.moderated delved off into
this topic starting more or less at
<A HREF="http://www.deja.com/getdoc.xp?AN=436187505">this</A>
article and continuing to the end of the thread. (You'll have to
sort through some flames every couple of paragraphs, but the points
made are good ones.)
</P>
<HR>
<H2><A NAME="4">Iostreams class hierarchy diagram</A></H2>
<P>The <A HREF="iostreams_hierarchy.pdf">diagram</A> is in PDF. Rumor
has it that once Benjamin Kosnik has been dead for a few decades,
this work of his will be hung next to the Mona Lisa in the
<A HREF="http://www.louvre.fr/">Musee du Louvre</A>.
</P>
<HR>
<H2><A NAME="5">What is this &lt;sstream&gt;/stringstreams thing?</A></H2>
<P>Stringstreams (defined in the header <TT>&lt;sstream&gt;</TT>)
are in this author's opinion one of the coolest things since
sliced time. An example of their use is in the Received Wisdom
section for Chapter 21 (Strings),
<A HREF="../21_strings/howto.html#1.1internal"> describing how to
format strings</A>.
</P>
<P>The quick definition is: they are siblings of ifstream and ofstream,
and they do for <TT>std::string</TT> what their siblings do for
files. All that work you put into writing <TT>&lt;&lt;</TT> and
<TT>&gt;&gt;</TT> functions for your classes now pays off
<EM>again!</EM> Need to format a string before passing the string
to a function? Send your stuff via <TT>&lt;&lt;</TT> to an
ostringstream. You've read a string as input and need to parse it?
Initialize an istringstream with that string, and then pull pieces
out of it with <TT>&gt;&gt;</TT>. Have a stringstream and need to
get a copy of the string inside? Just call the <TT>str()</TT>
member function.
</P>
<P>This only works if you've written your
<TT>&lt;&lt;</TT>/<TT>&gt;&gt;</TT> functions correctly, though,
and correctly means that they take istreams and ostreams as
parameters, not i<B>f</B>streams and o<B>f</B>streams. If they
take the latter, then your I/O operators will work fine with
file streams, but with nothing else -- including stringstreams.
</P>
<P>If you are a user of the strstream classes, you need to update
your code. You don't have to explicitly append <TT>ends</TT> to
terminate the C-style character array, you don't have to mess with
&quot;freezing&quot; functions, and you don't have to manage the
memory yourself. The strstreams have been officially deprecated,
which means that 1) future revisions of the C++ Standard won't
support them, and 2) if you use them, people will laugh at you.
</P>
<!-- ####################################################### -->
<HR>
<P CLASS="fineprint"><EM>
Comments and suggestions are welcome, and may be sent to
<A HREF="mailto:pme@sourceware.cygnus.com">Phil Edwards</A> or
<A HREF="mailto:gdr@egcs.cygnus.com">Gabriel Dos Reis</A>.
<BR> $Id: howto.html,v 1.7 1999/12/15 16:57:06 pme Exp $
</EM></P>
</BODY>
</HTML>