stdio.texi

@node I/O on Streams, Low-Level I/O, I/O Overview, Top
@c %MENU% High-level, portable I/O facilities
@chapter Input/Output on Streams
@c fix an overfull:
@tex
\hyphenation{which-ever}
@end tex

This chapter describes the functions for creating streams and performing
input and output operations on them.  As discussed in @ref{I/O
Overview}, a stream is a fairly abstract, high-level concept
representing a communications channel to a file, device, or process.

@menu
* Streams::                     About the data type representing a stream.
* Standard Streams::            Streams to the standard input and output
				 devices are created for you.
* Opening Streams::             How to create a stream to talk to a file.
* Closing Streams::             Close a stream when you are finished with it.
* Streams and Threads::         Issues with streams in threaded programs.
* Streams and I18N::            Streams in internationalized applications.
* Simple Output::               Unformatted output by characters and lines.
* Character Input::             Unformatted input by characters and words.
* Line Input::                  Reading a line or a record from a stream.
* Unreading::                   Peeking ahead/pushing back input just read.
* Block Input/Output::          Input and output operations on blocks of data.
* Formatted Output::            @code{printf} and related functions.
* Customizing Printf::          You can define new conversion specifiers for
				 @code{printf} and friends.
* Formatted Input::             @code{scanf} and related functions.
* EOF and Errors::              How you can tell if an I/O error happens.
* Error Recovery::		What you can do about errors.
* Binary Streams::              Some systems distinguish between text files
				 and binary files.
* File Positioning::            About random-access streams.
* Portable Positioning::        Random access on peculiar ISO C systems.
* Stream Buffering::            How to control buffering of streams.
* Other Kinds of Streams::      Streams that do not necessarily correspond
				 to an open file.
* Formatted Messages::          Print strictly formatted messages.
@end menu

@node Streams
@section Streams

For historical reasons, the type of the C data structure that represents
a stream is called @code{FILE} rather than ``stream''.  Since most of
the library functions deal with objects of type @code{FILE *}, sometimes
the term @dfn{file pointer} is also used to mean ``stream''.  This leads
to unfortunate confusion over terminology in many books on C.  This
manual, however, is careful to use the terms ``file'' and ``stream''
only in the technical sense.
@cindex file pointer

@pindex stdio.h
The @code{FILE} type is declared in the header file @file{stdio.h}.

@comment stdio.h
@comment ISO
@deftp {Data Type} FILE
This is the data type used to represent stream objects.  A @code{FILE}
object holds all of the internal state information about the connection
to the associated file, including such things as the file position
indicator and buffering information.  Each stream also has error and
end-of-file status indicators that can be tested with the @code{ferror}
and @code{feof} functions; see @ref{EOF and Errors}.
@end deftp

@code{FILE} objects are allocated and managed internally by the
input/output library functions.  Don't try to create your own objects of
type @code{FILE}; let the library do it.  Your programs should
deal only with pointers to these objects (that is, @code{FILE *} values)
rather than the objects themselves.
@c !!! should say that FILE's have "No user-serviceable parts inside."

@node Standard Streams
@section Standard Streams
@cindex standard streams
@cindex streams, standard

When the @code{main} function of your program is invoked, it already has
three predefined streams open and available for use.  These represent
the ``standard'' input and output channels that have been established
for the process.

These streams are declared in the header file @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypevar {FILE *} stdin
The @dfn{standard input} stream, which is the normal source of input for the
program.
@end deftypevar
@cindex standard input stream

@comment stdio.h
@comment ISO
@deftypevar {FILE *} stdout
The @dfn{standard output} stream, which is used for normal output from
the program.
@end deftypevar
@cindex standard output stream

@comment stdio.h
@comment ISO
@deftypevar {FILE *} stderr
The @dfn{standard error} stream, which is used for error messages and
diagnostics issued by the program.
@end deftypevar
@cindex standard error stream

On @gnusystems{}, you can specify what files or processes correspond to
these streams using the pipe and redirection facilities provided by the
shell.  (The primitives shells use to implement these facilities are
described in @ref{File System Interface}.)  Most other operating systems
provide similar mechanisms, but the details of how to use them can vary.

In @theglibc{}, @code{stdin}, @code{stdout}, and @code{stderr} are
normal variables which you can set just like any others.  For example,
to redirect the standard output to a file, you could do:

@smallexample
fclose (stdout);
stdout = fopen ("standard-output-file", "w");
@end smallexample

Note however, that in other systems @code{stdin}, @code{stdout}, and
@code{stderr} are macros that you cannot assign to in the normal way.
But you can use @code{freopen} to get the effect of closing one and
reopening it.  @xref{Opening Streams}.

The three streams @code{stdin}, @code{stdout}, and @code{stderr} are not
unoriented at program start (@pxref{Streams and I18N}).

@node Opening Streams
@section Opening Streams

@cindex opening a stream
Opening a file with the @code{fopen} function creates a new stream and
establishes a connection between the stream and a file.  This may
involve creating a new file.

@pindex stdio.h
Everything described in this section is declared in the header file
@file{stdio.h}.

@comment stdio.h
@comment ISO
@deftypefun {FILE *} fopen (const char *@var{filename}, const char *@var{opentype})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@acsmem{} @acsfd{} @aculock{}}}
@c fopen may leak the list lock if cancelled within _IO_link_in.
The @code{fopen} function opens a stream for I/O to the file
@var{filename}, and returns a pointer to the stream.

The @var{opentype} argument is a string that controls how the file is
opened and specifies attributes of the resulting stream.  It must begin
with one of the following sequences of characters:

@table @samp
@item r
Open an existing file for reading only.

@item w
Open the file for writing only.  If the file already exists, it is
truncated to zero length.  Otherwise a new file is created.

@item a
Open a file for append access; that is, writing at the end of file only.
If the file already exists, its initial contents are unchanged and
output to the stream is appended to the end of the file.
Otherwise, a new, empty file is created.

@item r+
Open an existing file for both reading and writing.  The initial contents
of the file are unchanged and the initial file position is at the
beginning of the file.

@item w+
Open a file for both reading and writing.  If the file already exists, it
is truncated to zero length.  Otherwise, a new file is created.

@item a+
Open or create file for both reading and appending.  If the file exists,
its initial contents are unchanged.  Otherwise, a new file is created.
The initial file position for reading is at the beginning of the file,
but output is always appended to the end of the file.
@end table

As you can see, @samp{+} requests a stream that can do both input and
output.  When using such a stream, you must call @code{fflush}
(@pxref{Stream Buffering}) or a file positioning function such as
@code{fseek} (@pxref{File Positioning}) when switching from reading
to writing or vice versa.  Otherwise, internal buffers might not be
emptied properly.

Additional characters may appear after these to specify flags for the
call.  Always put the mode (@samp{r}, @samp{w+}, etc.) first; that is
the only part you are guaranteed will be understood by all systems.

@Theglibc{} defines additional characters for use in @var{opentype}:

@table @samp
@item c
The file is opened with cancellation in the I/O functions disabled.

@item e
The underlying file descriptor will be closed if you use any of the
@code{exec@dots{}} functions (@pxref{Executing a File}).  (This is
equivalent to having set @code{FD_CLOEXEC} on that descriptor.
@xref{Descriptor Flags}.)

@item m
The file is opened and accessed using @code{mmap}.  This is only
supported with files opened for reading.

@item x
Insist on creating a new file---if a file @var{filename} already
exists, @code{fopen} fails rather than opening it.  If you use
@samp{x} you are guaranteed that you will not clobber an existing
file.  This is equivalent to the @code{O_EXCL} option to the
@code{open} function (@pxref{Opening and Closing Files}).

The @samp{x} modifier is part of @w{ISO C11}.
@end table

The character @samp{b} in @var{opentype} has a standard meaning; it
requests a binary stream rather than a text stream.  But this makes no
difference in POSIX systems (including @gnusystems{}).  If both
@samp{+} and @samp{b} are specified, they can appear in either order.
@xref{Binary Streams}.

@cindex stream orientation
@cindex orientation, stream
If the @var{opentype} string contains the sequence
@code{,ccs=@var{STRING}} then @var{STRING} is taken as the name of a
coded character set and @code{fopen} will mark the stream as
wide-oriented with appropriate conversion functions in place to convert
from and to the character set @var{STRING}.  Any other stream
is opened initially unoriented and the orientation is decided with the
first file operation.  If the first operation is a wide character
operation, the stream is not only marked as wide-oriented, also the
conversion functions to convert to the coded character set used for the
current locale are loaded.  This will not change anymore from this point
on even if the locale selected for the @code{LC_CTYPE} category is
changed.

Any other characters in @var{opentype} are simply ignored.  They may be
meaningful in other systems.

If the open fails, @code{fopen} returns a null pointer.

When the sources are compiling with @code{_FILE_OFFSET_BITS == 64} on a
32 bit machine this function is in fact @code{fopen64} since the LFS
interface replaces transparently the old interface.
@end deftypefun

You can have multiple streams (or file descriptors) pointing to the same
file open at the same time.  If you do only input, this works
straightforwardly, but you must be careful if any output streams are
included.  @xref{Stream/Descriptor Precautions}.  This is equally true
whether the streams are in one program (not usual) or in several
programs (which can easily happen).  It may be advantageous to use the
file locking facilities to avoid simultaneous access.  @xref{File
Locks}.

@comment stdio.h
@comment Unix98
@deftypefun {FILE *} fopen64 (const char *@var{filename}, const char *@var{opentype})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@acsmem{} @acsfd{} @aculock{}}}
This function is similar to @code{fopen} but the stream it returns a
pointer for is opened using @code{open64}.  Therefore this stream can be
used even on files larger than @twoexp{31} bytes on 32 bit machines.

Please note that the return type is still @code{FILE *}.  There is no
special @code{FILE} type for the LFS interface.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a 32
bits machine this function is available under the name @code{fopen}
and so transparently replaces the old interface.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypevr Macro int FOPEN_MAX
The value of this macro is an integer constant expression that
represents the minimum number of streams that the implementation
guarantees can be open simultaneously.  You might be able to open more
than this many streams, but that is not guaranteed.  The value of this
constant is at least eight, which includes the three standard streams
@code{stdin}, @code{stdout}, and @code{stderr}.  In POSIX.1 systems this
value is determined by the @code{OPEN_MAX} parameter; @pxref{General
Limits}.  In BSD and GNU, it is controlled by the @code{RLIMIT_NOFILE}
resource limit; @pxref{Limits on Resources}.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypefun {FILE *} freopen (const char *@var{filename}, const char *@var{opentype}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @acsfd{}}}
@c Like most I/O operations, this one is guarded by a recursive lock,
@c released even upon cancellation, but cancellation may leak file
@c descriptors and leave the stream in an inconsistent state (e.g.,
@c still bound to the closed descriptor).  Also, if the stream is
@c part-way through a significant update (say running freopen) when a
@c signal handler calls freopen again on the same stream, the result is
@c likely to be an inconsistent stream, and the possibility of closing
@c twice file descriptor number that the stream used to use, the second
@c time when it might have already been reused by another thread.
This function is like a combination of @code{fclose} and @code{fopen}.
It first closes the stream referred to by @var{stream}, ignoring any
errors that are detected in the process.  (Because errors are ignored,
you should not use @code{freopen} on an output stream if you have
actually done any output using the stream.)  Then the file named by
@var{filename} is opened with mode @var{opentype} as for @code{fopen},
and associated with the same stream object @var{stream}.

If the operation fails, a null pointer is returned; otherwise,
@code{freopen} returns @var{stream}.

@code{freopen} has traditionally been used to connect a standard stream
such as @code{stdin} with a file of your own choice.  This is useful in
programs in which use of a standard stream for certain purposes is
hard-coded.  In @theglibc{}, you can simply close the standard
streams and open new ones with @code{fopen}.  But other systems lack
this ability, so using @code{freopen} is more portable.

When the sources are compiling with @code{_FILE_OFFSET_BITS == 64} on a
32 bit machine this function is in fact @code{freopen64} since the LFS
interface replaces transparently the old interface.
@end deftypefun

@comment stdio.h
@comment Unix98
@deftypefun {FILE *} freopen64 (const char *@var{filename}, const char *@var{opentype}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @acsfd{}}}
This function is similar to @code{freopen}.  The only difference is that
on 32 bit machine the stream returned is able to read beyond the
@twoexp{31} bytes limits imposed by the normal interface.  It should be
noted that the stream pointed to by @var{stream} need not be opened
using @code{fopen64} or @code{freopen64} since its mode is not important
for this function.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a 32
bits machine this function is available under the name @code{freopen}
and so transparently replaces the old interface.
@end deftypefun

In some situations it is useful to know whether a given stream is
available for reading or writing.  This information is normally not
available and would have to be remembered separately.  Solaris
introduced a few functions to get this information from the stream
descriptor and these functions are also available in @theglibc{}.

@comment stdio_ext.h
@comment GNU
@deftypefun int __freadable (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{__freadable} function determines whether the stream
@var{stream} was opened to allow reading.  In this case the return value
is nonzero.  For write-only streams the function returns zero.

This function is declared in @file{stdio_ext.h}.
@end deftypefun

@comment stdio_ext.h
@comment GNU
@deftypefun int __fwritable (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{__fwritable} function determines whether the stream
@var{stream} was opened to allow writing.  In this case the return value
is nonzero.  For read-only streams the function returns zero.

This function is declared in @file{stdio_ext.h}.
@end deftypefun

For slightly different kind of problems there are two more functions.
They provide even finer-grained information.

@comment stdio_ext.h
@comment GNU
@deftypefun int __freading (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{__freading} function determines whether the stream
@var{stream} was last read from or whether it is opened read-only.  In
this case the return value is nonzero, otherwise it is zero.
Determining whether a stream opened for reading and writing was last
used for writing allows to draw conclusions about the content about the
buffer, among other things.

This function is declared in @file{stdio_ext.h}.
@end deftypefun

@comment stdio_ext.h
@comment GNU
@deftypefun int __fwriting (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{__fwriting} function determines whether the stream
@var{stream} was last written to or whether it is opened write-only.  In
this case the return value is nonzero, otherwise it is zero.

This function is declared in @file{stdio_ext.h}.
@end deftypefun


@node Closing Streams
@section Closing Streams

@cindex closing a stream
When a stream is closed with @code{fclose}, the connection between the
stream and the file is canceled.  After you have closed a stream, you
cannot perform any additional operations on it.

@comment stdio.h
@comment ISO
@deftypefun int fclose (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@aculock{} @acsmem{} @acsfd{}}}
@c After fclose, it is undefined behavior to use the stream it points
@c to.  Therefore, one must only call fclose when the stream is
@c otherwise unused.  Concurrent uses started before will complete
@c successfully because of the lock, which makes it MT-Safe.  Calling it
@c from a signal handler is perfectly safe if the stream is known to be
@c no longer used, which is a precondition for fclose to be safe in the
@c first place; since this is no further requirement, fclose is safe for
@c use in async signals too.  After calling fclose, you can no longer
@c use the stream, not even to fclose it again, so its memory and file
@c descriptor may leak if fclose is canceled before @c releasing them.
@c That the stream must be unused and it becomes unused after the call
@c is what would enable fclose to be AS- and AC-Safe while freopen
@c isn't.  However, because of the possibility of leaving __gconv_lock
@c taken upon cancellation, AC-Safety is lost.
This function causes @var{stream} to be closed and the connection to
the corresponding file to be broken.  Any buffered output is written
and any buffered input is discarded.  The @code{fclose} function returns
a value of @code{0} if the file was closed successfully, and @code{EOF}
if an error was detected.

It is important to check for errors when you call @code{fclose} to close
an output stream, because real, everyday errors can be detected at this
time.  For example, when @code{fclose} writes the remaining buffered
output, it might get an error because the disk is full.  Even if you
know the buffer is empty, errors can still occur when closing a file if
you are using NFS.

The function @code{fclose} is declared in @file{stdio.h}.
@end deftypefun

To close all streams currently available @theglibc{} provides
another function.

@comment stdio.h
@comment GNU
@deftypefun int fcloseall (void)
@safety{@prelim{}@mtunsafe{@mtasurace{:streams}}@asunsafe{}@acsafe{}}
@c Like fclose, using any previously-opened streams after fcloseall is
@c undefined.  However, the implementation of fcloseall isn't equivalent
@c to calling fclose for all streams: it just flushes and unbuffers all
@c streams, without any locking.  It's the flushing without locking that
@c makes it unsafe.
This function causes all open streams of the process to be closed and
the connection to corresponding files to be broken.  All buffered data
is written and any buffered input is discarded.  The @code{fcloseall}
function returns a value of @code{0} if all the files were closed
successfully, and @code{EOF} if an error was detected.

This function should be used only in special situations, e.g., when an
error occurred and the program must be aborted.  Normally each single
stream should be closed separately so that problems with individual
streams can be identified.  It is also problematic since the standard
streams (@pxref{Standard Streams}) will also be closed.

The function @code{fcloseall} is declared in @file{stdio.h}.
@end deftypefun

If the @code{main} function to your program returns, or if you call the
@code{exit} function (@pxref{Normal Termination}), all open streams are
automatically closed properly.  If your program terminates in any other
manner, such as by calling the @code{abort} function (@pxref{Aborting a
Program}) or from a fatal signal (@pxref{Signal Handling}), open streams
might not be closed properly.  Buffered output might not be flushed and
files may be incomplete.  For more information on buffering of streams,
see @ref{Stream Buffering}.

@node Streams and Threads
@section Streams and Threads

@cindex threads
@cindex multi-threaded application
Streams can be used in multi-threaded applications in the same way they
are used in single-threaded applications.  But the programmer must be
aware of the possible complications.  It is important to know about
these also if the program one writes never use threads since the design
and implementation of many stream functions is heavily influenced by the
requirements added by multi-threaded programming.

The POSIX standard requires that by default the stream operations are
atomic.  I.e., issuing two stream operations for the same stream in two
threads at the same time will cause the operations to be executed as if
they were issued sequentially.  The buffer operations performed while
reading or writing are protected from other uses of the same stream.  To
do this each stream has an internal lock object which has to be
(implicitly) acquired before any work can be done.

But there are situations where this is not enough and there are also
situations where this is not wanted.  The implicit locking is not enough
if the program requires more than one stream function call to happen
atomically.  One example would be if an output line a program wants to
generate is created by several function calls.  The functions by
themselves would ensure only atomicity of their own operation, but not
atomicity over all the function calls.  For this it is necessary to
perform the stream locking in the application code.

@comment stdio.h
@comment POSIX
@deftypefun void flockfile (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acunsafe{@aculock{}}}
@c There's no way to tell whether the lock was acquired before or after
@c cancellation so as to unlock only when appropriate.
The @code{flockfile} function acquires the internal locking object
associated with the stream @var{stream}.  This ensures that no other
thread can explicitly through @code{flockfile}/@code{ftrylockfile} or
implicit through a call of a stream function lock the stream.  The
thread will block until the lock is acquired.  An explicit call to
@code{funlockfile} has to be used to release the lock.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int ftrylockfile (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acunsafe{@aculock{}}}
The @code{ftrylockfile} function tries to acquire the internal locking
object associated with the stream @var{stream} just like
@code{flockfile}.  But unlike @code{flockfile} this function does not
block if the lock is not available.  @code{ftrylockfile} returns zero if
the lock was successfully acquired.  Otherwise the stream is locked by
another thread.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun void funlockfile (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acunsafe{@aculock{}}}
The @code{funlockfile} function releases the internal locking object of
the stream @var{stream}.  The stream must have been locked before by a
call to @code{flockfile} or a successful call of @code{ftrylockfile}.
The implicit locking performed by the stream operations do not count.
The @code{funlockfile} function does not return an error status and the
behavior of a call for a stream which is not locked by the current
thread is undefined.
@end deftypefun

The following example shows how the functions above can be used to
generate an output line atomically even in multi-threaded applications
(yes, the same job could be done with one @code{fprintf} call but it is
sometimes not possible):

@smallexample
FILE *fp;
@{
   @dots{}
   flockfile (fp);
   fputs ("This is test number ", fp);
   fprintf (fp, "%d\n", test);
   funlockfile (fp)
@}
@end smallexample

Without the explicit locking it would be possible for another thread to
use the stream @var{fp} after the @code{fputs} call return and before
@code{fprintf} was called with the result that the number does not
follow the word @samp{number}.

From this description it might already be clear that the locking objects
in streams are no simple mutexes.  Since locking the same stream twice
in the same thread is allowed the locking objects must be equivalent to
recursive mutexes.  These mutexes keep track of the owner and the number
of times the lock is acquired.  The same number of @code{funlockfile}
calls by the same threads is necessary to unlock the stream completely.
For instance:

@smallexample
void
foo (FILE *fp)
@{
  ftrylockfile (fp);
  fputs ("in foo\n", fp);
  /* @r{This is very wrong!!!}  */
  funlockfile (fp);
@}
@end smallexample

It is important here that the @code{funlockfile} function is only called
if the @code{ftrylockfile} function succeeded in locking the stream.  It
is therefore always wrong to ignore the result of @code{ftrylockfile}.
And it makes no sense since otherwise one would use @code{flockfile}.
The result of code like that above is that either @code{funlockfile}
tries to free a stream that hasn't been locked by the current thread or it
frees the stream prematurely.  The code should look like this:

@smallexample
void
foo (FILE *fp)
@{
  if (ftrylockfile (fp) == 0)
    @{
      fputs ("in foo\n", fp);
      funlockfile (fp);
    @}
@}
@end smallexample

Now that we covered why it is necessary to have these locking it is
necessary to talk about situations when locking is unwanted and what can
be done.  The locking operations (explicit or implicit) don't come for
free.  Even if a lock is not taken the cost is not zero.  The operations
which have to be performed require memory operations that are safe in
multi-processor environments.  With the many local caches involved in
such systems this is quite costly.  So it is best to avoid the locking
completely if it is not needed -- because the code in question is never
used in a context where two or more threads may use a stream at a time.
This can be determined most of the time for application code; for
library code which can be used in many contexts one should default to be
conservative and use locking.

There are two basic mechanisms to avoid locking.  The first is to use
the @code{_unlocked} variants of the stream operations.  The POSIX
standard defines quite a few of those and @theglibc{} adds a few
more.  These variants of the functions behave just like the functions
with the name without the suffix except that they do not lock the
stream.  Using these functions is very desirable since they are
potentially much faster.  This is not only because the locking
operation itself is avoided.  More importantly, functions like
@code{putc} and @code{getc} are very simple and traditionally (before the
introduction of threads) were implemented as macros which are very fast
if the buffer is not empty.  With the addition of locking requirements
these functions are no longer implemented as macros since they would
expand to too much code.
But these macros are still available with the same functionality under the new
names @code{putc_unlocked} and @code{getc_unlocked}.  This possibly huge
difference of speed also suggests the use of the @code{_unlocked}
functions even if locking is required.  The difference is that the
locking then has to be performed in the program:

@smallexample
void
foo (FILE *fp, char *buf)
@{
  flockfile (fp);
  while (*buf != '/')
    putc_unlocked (*buf++, fp);
  funlockfile (fp);
@}
@end smallexample

If in this example the @code{putc} function would be used and the
explicit locking would be missing the @code{putc} function would have to
acquire the lock in every call, potentially many times depending on when
the loop terminates.  Writing it the way illustrated above allows the
@code{putc_unlocked} macro to be used which means no locking and direct
manipulation of the buffer of the stream.

A second way to avoid locking is by using a non-standard function which
was introduced in Solaris and is available in @theglibc{} as well.

@comment stdio_ext.h
@comment GNU
@deftypefun int __fsetlocking (FILE *@var{stream}, int @var{type})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asulock{}}@acsafe{}}
@c Changing the implicit-locking status of a stream while it's in use by
@c another thread may cause a lock to be implicitly acquired and not
@c released, or vice-versa.  This function should probably hold the lock
@c while changing this setting, to make sure we don't change it while
@c there are any concurrent uses.  Meanwhile, callers should acquire the
@c lock themselves to be safe, and even concurrent uses with external
@c locking will be fine, as long as functions that require external
@c locking are not called without holding locks.

The @code{__fsetlocking} function can be used to select whether the
stream operations will implicitly acquire the locking object of the
stream @var{stream}.  By default this is done but it can be disabled and
reinstated using this function.  There are three values defined for the
@var{type} parameter.

@vtable @code
@item FSETLOCKING_INTERNAL
The stream @code{stream} will from now on use the default internal
locking.  Every stream operation with exception of the @code{_unlocked}
variants will implicitly lock the stream.

@item FSETLOCKING_BYCALLER
After the @code{__fsetlocking} function returns the user is responsible
for locking the stream.  None of the stream operations will implicitly
do this anymore until the state is set back to
@code{FSETLOCKING_INTERNAL}.

@item FSETLOCKING_QUERY
@code{__fsetlocking} only queries the current locking state of the
stream.  The return value will be @code{FSETLOCKING_INTERNAL} or
@code{FSETLOCKING_BYCALLER} depending on the state.
@end vtable

The return value of @code{__fsetlocking} is either
@code{FSETLOCKING_INTERNAL} or @code{FSETLOCKING_BYCALLER} depending on
the state of the stream before the call.

This function and the values for the @var{type} parameter are declared
in @file{stdio_ext.h}.
@end deftypefun

This function is especially useful when program code has to be used
which is written without knowledge about the @code{_unlocked} functions
(or if the programmer was too lazy to use them).

@node Streams and I18N
@section Streams in Internationalized Applications

@w{ISO C90} introduced the new type @code{wchar_t} to allow handling
larger character sets.  What was missing was a possibility to output
strings of @code{wchar_t} directly.  One had to convert them into
multibyte strings using @code{mbstowcs} (there was no @code{mbsrtowcs}
yet) and then use the normal stream functions.  While this is doable it
is very cumbersome since performing the conversions is not trivial and
greatly increases program complexity and size.

The Unix standard early on (I think in XPG4.2) introduced two additional
format specifiers for the @code{printf} and @code{scanf} families of
functions.  Printing and reading of single wide characters was made
possible using the @code{%C} specifier and wide character strings can be
handled with @code{%S}.  These modifiers behave just like @code{%c} and
@code{%s} only that they expect the corresponding argument to have the
wide character type and that the wide character and string are
transformed into/from multibyte strings before being used.

This was a beginning but it is still not good enough.  Not always is it
desirable to use @code{printf} and @code{scanf}.  The other, smaller and
faster functions cannot handle wide characters.  Second, it is not
possible to have a format string for @code{printf} and @code{scanf}
consisting of wide characters.  The result is that format strings would
have to be generated if they have to contain non-basic characters.

@cindex C++ streams
@cindex streams, C++
In the @w{Amendment 1} to @w{ISO C90} a whole new set of functions was
added to solve the problem.  Most of the stream functions got a
counterpart which take a wide character or wide character string instead
of a character or string respectively.  The new functions operate on the
same streams (like @code{stdout}).  This is different from the model of
the C++ runtime library where separate streams for wide and normal I/O
are used.

@cindex orientation, stream
@cindex stream orientation
Being able to use the same stream for wide and normal operations comes
with a restriction: a stream can be used either for wide operations or
for normal operations.  Once it is decided there is no way back.  Only a
call to @code{freopen} or @code{freopen64} can reset the
@dfn{orientation}.  The orientation can be decided in three ways:

@itemize @bullet
@item
If any of the normal character functions is used (this includes the
@code{fread} and @code{fwrite} functions) the stream is marked as not
wide oriented.

@item
If any of the wide character functions is used the stream is marked as
wide oriented.

@item
The @code{fwide} function can be used to set the orientation either way.
@end itemize

It is important to never mix the use of wide and not wide operations on
a stream.  There are no diagnostics issued.  The application behavior
will simply be strange or the application will simply crash.  The
@code{fwide} function can help avoiding this.

@comment wchar.h
@comment ISO
@deftypefun int fwide (FILE *@var{stream}, int @var{mode})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{}}}
@c Querying is always safe, but changing the stream when it's in use
@c upthread may be problematic.  Like most lock-acquiring functions,
@c this one may leak the lock if canceled.

The @code{fwide} function can be used to set and query the state of the
orientation of the stream @var{stream}.  If the @var{mode} parameter has
a positive value the streams get wide oriented, for negative values
narrow oriented.  It is not possible to overwrite previous orientations
with @code{fwide}.  I.e., if the stream @var{stream} was already
oriented before the call nothing is done.

If @var{mode} is zero the current orientation state is queried and
nothing is changed.

The @code{fwide} function returns a negative value, zero, or a positive
value if the stream is narrow, not at all, or wide oriented
respectively.

This function was introduced in @w{Amendment 1} to @w{ISO C90} and is
declared in @file{wchar.h}.
@end deftypefun

It is generally a good idea to orient a stream as early as possible.
This can prevent surprise especially for the standard streams
@code{stdin}, @code{stdout}, and @code{stderr}.  If some library
function in some situations uses one of these streams and this use
orients the stream in a different way the rest of the application
expects it one might end up with hard to reproduce errors.  Remember
that no errors are signal if the streams are used incorrectly.  Leaving
a stream unoriented after creation is normally only necessary for
library functions which create streams which can be used in different
contexts.

When writing code which uses streams and which can be used in different
contexts it is important to query the orientation of the stream before
using it (unless the rules of the library interface demand a specific
orientation).  The following little, silly function illustrates this.

@smallexample
void
print_f (FILE *fp)
@{
  if (fwide (fp, 0) > 0)
    /* @r{Positive return value means wide orientation.}  */
    fputwc (L'f', fp);
  else
    fputc ('f', fp);
@}
@end smallexample

Note that in this case the function @code{print_f} decides about the
orientation of the stream if it was unoriented before (will not happen
if the advise above is followed).

The encoding used for the @code{wchar_t} values is unspecified and the
user must not make any assumptions about it.  For I/O of @code{wchar_t}
values this means that it is impossible to write these values directly
to the stream.  This is not what follows from the @w{ISO C} locale model
either.  What happens instead is that the bytes read from or written to
the underlying media are first converted into the internal encoding
chosen by the implementation for @code{wchar_t}.  The external encoding
is determined by the @code{LC_CTYPE} category of the current locale or
by the @samp{ccs} part of the mode specification given to @code{fopen},
@code{fopen64}, @code{freopen}, or @code{freopen64}.  How and when the
conversion happens is unspecified and it happens invisible to the user.

Since a stream is created in the unoriented state it has at that point
no conversion associated with it.  The conversion which will be used is
determined by the @code{LC_CTYPE} category selected at the time the
stream is oriented.  If the locales are changed at the runtime this
might produce surprising results unless one pays attention.  This is
just another good reason to orient the stream explicitly as soon as
possible, perhaps with a call to @code{fwide}.

@node Simple Output
@section Simple Output by Characters or Lines

@cindex writing to a stream, by characters
This section describes functions for performing character- and
line-oriented output.

These narrow streams functions are declared in the header file
@file{stdio.h} and the wide stream functions in @file{wchar.h}.
@pindex stdio.h
@pindex wchar.h

@comment stdio.h
@comment ISO
@deftypefun int fputc (int @var{c}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
@c If the stream is in use when interrupted by a signal, the recursive
@c lock won't help ensure the stream is consistent; indeed, if fputc
@c gets a signal precisely before the post-incremented _IO_write_ptr
@c value is stored, we may overwrite the interrupted write.  Conversely,
@c depending on compiler optimizations, the incremented _IO_write_ptr
@c may be stored before the character is stored in the buffer,
@c corrupting the stream if async cancel hits between the two stores.
@c There may be other reasons for AS- and AC-unsafety in the overflow
@c cases.
The @code{fputc} function converts the character @var{c} to type
@code{unsigned char}, and writes it to the stream @var{stream}.
@code{EOF} is returned if a write error occurs; otherwise the
character @var{c} is returned.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun wint_t fputwc (wchar_t @var{wc}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
The @code{fputwc} function writes the wide character @var{wc} to the
stream @var{stream}.  @code{WEOF} is returned if a write error occurs;
otherwise the character @var{wc} is returned.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int fputc_unlocked (int @var{c}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
@c The unlocked functions can't possibly satisfy the MT-Safety
@c requirements on their own, because they require external locking for
@c safety.
The @code{fputc_unlocked} function is equivalent to the @code{fputc}
function except that it does not implicitly lock the stream.
@end deftypefun

@comment wchar.h
@comment POSIX
@deftypefun wint_t fputwc_unlocked (wchar_t @var{wc}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fputwc_unlocked} function is equivalent to the @code{fputwc}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int putc (int @var{c}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
This is just like @code{fputc}, except that most systems implement it as
a macro, making it faster.  One consequence is that it may evaluate the
@var{stream} argument more than once, which is an exception to the
general rule for macros.  @code{putc} is usually the best function to
use for writing a single character.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun wint_t putwc (wchar_t @var{wc}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
This is just like @code{fputwc}, except that it can be implement as
a macro, making it faster.  One consequence is that it may evaluate the
@var{stream} argument more than once, which is an exception to the
general rule for macros.  @code{putwc} is usually the best function to
use for writing a single wide character.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int putc_unlocked (int @var{c}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{putc_unlocked} function is equivalent to the @code{putc}
function except that it does not implicitly lock the stream.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun wint_t putwc_unlocked (wchar_t @var{wc}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{putwc_unlocked} function is equivalent to the @code{putwc}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int putchar (int @var{c})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
The @code{putchar} function is equivalent to @code{putc} with
@code{stdout} as the value of the @var{stream} argument.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun wint_t putwchar (wchar_t @var{wc})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
The @code{putwchar} function is equivalent to @code{putwc} with
@code{stdout} as the value of the @var{stream} argument.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int putchar_unlocked (int @var{c})
@safety{@prelim{}@mtunsafe{@mtasurace{:stdout}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{putchar_unlocked} function is equivalent to the @code{putchar}
function except that it does not implicitly lock the stream.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun wint_t putwchar_unlocked (wchar_t @var{wc})
@safety{@prelim{}@mtunsafe{@mtasurace{:stdout}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{putwchar_unlocked} function is equivalent to the @code{putwchar}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int fputs (const char *@var{s}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
The function @code{fputs} writes the string @var{s} to the stream
@var{stream}.  The terminating null character is not written.
This function does @emph{not} add a newline character, either.
It outputs only the characters in the string.

This function returns @code{EOF} if a write error occurs, and otherwise
a non-negative value.

For example:

@smallexample
fputs ("Are ", stdout);
fputs ("you ", stdout);
fputs ("hungry?\n", stdout);
@end smallexample

@noindent
outputs the text @samp{Are you hungry?} followed by a newline.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int fputws (const wchar_t *@var{ws}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{} @aculock{}}}
The function @code{fputws} writes the wide character string @var{ws} to
the stream @var{stream}.  The terminating null character is not written.
This function does @emph{not} add a newline character, either.  It
outputs only the characters in the string.

This function returns @code{WEOF} if a write error occurs, and otherwise
a non-negative value.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int fputs_unlocked (const char *@var{s}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fputs_unlocked} function is equivalent to the @code{fputs}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun int fputws_unlocked (const wchar_t *@var{ws}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fputws_unlocked} function is equivalent to the @code{fputws}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int puts (const char *@var{s})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{puts} function writes the string @var{s} to the stream
@code{stdout} followed by a newline.  The terminating null character of
the string is not written.  (Note that @code{fputs} does @emph{not}
write a newline as this function does.)

@code{puts} is the most convenient function for printing simple
messages.  For example:

@smallexample
puts ("This is a message.");
@end smallexample

@noindent
outputs the text @samp{This is a message.} followed by a newline.
@end deftypefun

@comment stdio.h
@comment SVID
@deftypefun int putw (int @var{w}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function writes the word @var{w} (that is, an @code{int}) to
@var{stream}.  It is provided for compatibility with SVID, but we
recommend you use @code{fwrite} instead (@pxref{Block Input/Output}).
@end deftypefun

@node Character Input
@section Character Input

@cindex reading from a stream, by characters
This section describes functions for performing character-oriented
input.  These narrow streams functions are declared in the header file
@file{stdio.h} and the wide character functions are declared in
@file{wchar.h}.
@pindex stdio.h
@pindex wchar.h

These functions return an @code{int} or @code{wint_t} value (for narrow
and wide stream functions respectively) that is either a character of
input, or the special value @code{EOF}/@code{WEOF} (usually -1).  For
the narrow stream functions it is important to store the result of these
functions in a variable of type @code{int} instead of @code{char}, even
when you plan to use it only as a character.  Storing @code{EOF} in a
@code{char} variable truncates its value to the size of a character, so
that it is no longer distinguishable from the valid character
@samp{(char) -1}.  So always use an @code{int} for the result of
@code{getc} and friends, and check for @code{EOF} after the call; once
you've verified that the result is not @code{EOF}, you can be sure that
it will fit in a @samp{char} variable without loss of information.

@comment stdio.h
@comment ISO
@deftypefun int fgetc (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
@c Same caveats as fputc, but instead of losing a write in case of async
@c signals, we may read the same character more than once, and the
@c stream may be left in odd states due to cancellation in the underflow
@c cases.
This function reads the next character as an @code{unsigned char} from
the stream @var{stream} and returns its value, converted to an
@code{int}.  If an end-of-file condition or read error occurs,
@code{EOF} is returned instead.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun wint_t fgetwc (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function reads the next wide character from the stream @var{stream}
and returns its value.  If an end-of-file condition or read error
occurs, @code{WEOF} is returned instead.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int fgetc_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fgetc_unlocked} function is equivalent to the @code{fgetc}
function except that it does not implicitly lock the stream.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun wint_t fgetwc_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fgetwc_unlocked} function is equivalent to the @code{fgetwc}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int getc (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This is just like @code{fgetc}, except that it is permissible (and
typical) for it to be implemented as a macro that evaluates the
@var{stream} argument more than once.  @code{getc} is often highly
optimized, so it is usually the best function to use to read a single
character.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun wint_t getwc (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This is just like @code{fgetwc}, except that it is permissible for it to
be implemented as a macro that evaluates the @var{stream} argument more
than once.  @code{getwc} can be highly optimized, so it is usually the
best function to use to read a single wide character.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int getc_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{getc_unlocked} function is equivalent to the @code{getc}
function except that it does not implicitly lock the stream.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun wint_t getwc_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{getwc_unlocked} function is equivalent to the @code{getwc}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int getchar (void)
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{getchar} function is equivalent to @code{getc} with @code{stdin}
as the value of the @var{stream} argument.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun wint_t getwchar (void)
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{getwchar} function is equivalent to @code{getwc} with @code{stdin}
as the value of the @var{stream} argument.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int getchar_unlocked (void)
@safety{@prelim{}@mtunsafe{@mtasurace{:stdin}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{getchar_unlocked} function is equivalent to the @code{getchar}
function except that it does not implicitly lock the stream.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun wint_t getwchar_unlocked (void)
@safety{@prelim{}@mtunsafe{@mtasurace{:stdin}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{getwchar_unlocked} function is equivalent to the @code{getwchar}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

Here is an example of a function that does input using @code{fgetc}.  It
would work just as well using @code{getc} instead, or using
@code{getchar ()} instead of @w{@code{fgetc (stdin)}}.  The code would
also work the same for the wide character stream functions.

@smallexample
int
y_or_n_p (const char *question)
@{
  fputs (question, stdout);
  while (1)
    @{
      int c, answer;
      /* @r{Write a space to separate answer from question.} */
      fputc (' ', stdout);
      /* @r{Read the first character of the line.}
	 @r{This should be the answer character, but might not be.} */
      c = tolower (fgetc (stdin));
      answer = c;
      /* @r{Discard rest of input line.} */
      while (c != '\n' && c != EOF)
	c = fgetc (stdin);
      /* @r{Obey the answer if it was valid.} */
      if (answer == 'y')
	return 1;
      if (answer == 'n')
	return 0;
      /* @r{Answer was invalid: ask for valid answer.} */
      fputs ("Please answer y or n:", stdout);
    @}
@}
@end smallexample

@comment stdio.h
@comment SVID
@deftypefun int getw (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function reads a word (that is, an @code{int}) from @var{stream}.
It's provided for compatibility with SVID.  We recommend you use
@code{fread} instead (@pxref{Block Input/Output}).  Unlike @code{getc},
any @code{int} value could be a valid result.  @code{getw} returns
@code{EOF} when it encounters end-of-file or an error, but there is no
way to distinguish this from an input word with value -1.
@end deftypefun

@node Line Input
@section Line-Oriented Input

Since many programs interpret input on the basis of lines, it is
convenient to have functions to read a line of text from a stream.

Standard C has functions to do this, but they aren't very safe: null
characters and even (for @code{gets}) long lines can confuse them.  So
@theglibc{} provides the nonstandard @code{getline} function that
makes it easy to read lines reliably.

Another GNU extension, @code{getdelim}, generalizes @code{getline}.  It
reads a delimited record, defined as everything through the next
occurrence of a specified delimiter character.

All these functions are declared in @file{stdio.h}.

@comment stdio.h
@comment GNU
@deftypefun ssize_t getline (char **@var{lineptr}, size_t *@var{n}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@aculock{} @acucorrupt{} @acsmem{}}}
@c Besides the usual possibility of getting an inconsistent stream in a
@c signal handler or leaving it inconsistent in case of cancellation,
@c the possibility of leaving a dangling pointer upon cancellation
@c between reallocing the buffer at *lineptr and updating the pointer
@c brings about another case of @acucorrupt.
This function reads an entire line from @var{stream}, storing the text
(including the newline and a terminating null character) in a buffer
and storing the buffer address in @code{*@var{lineptr}}.

Before calling @code{getline}, you should place in @code{*@var{lineptr}}
the address of a buffer @code{*@var{n}} bytes long, allocated with
@code{malloc}.  If this buffer is long enough to hold the line,
@code{getline} stores the line in this buffer.  Otherwise,
@code{getline} makes the buffer bigger using @code{realloc}, storing the
new buffer address back in @code{*@var{lineptr}} and the increased size
back in @code{*@var{n}}.
@xref{Unconstrained Allocation}.

If you set @code{*@var{lineptr}} to a null pointer, and @code{*@var{n}}
to zero, before the call, then @code{getline} allocates the initial
buffer for you by calling @code{malloc}.  This buffer remains allocated
even if @code{getline} encounters errors and is unable to read any bytes.

In either case, when @code{getline} returns,  @code{*@var{lineptr}} is
a @code{char *} which points to the text of the line.

When @code{getline} is successful, it returns the number of characters
read (including the newline, but not including the terminating null).
This value enables you to distinguish null characters that are part of
the line from the null character inserted as a terminator.

This function is a GNU extension, but it is the recommended way to read
lines from a stream.  The alternative standard functions are unreliable.

If an error occurs or end of file is reached without any bytes read,
@code{getline} returns @code{-1}.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun ssize_t getdelim (char **@var{lineptr}, size_t *@var{n}, int @var{delimiter}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@aculock{} @acucorrupt{} @acsmem{}}}
@c See the getline @acucorrupt note.
This function is like @code{getline} except that the character which
tells it to stop reading is not necessarily newline.  The argument
@var{delimiter} specifies the delimiter character; @code{getdelim} keeps
reading until it sees that character (or end of file).

The text is stored in @var{lineptr}, including the delimiter character
and a terminating null.  Like @code{getline}, @code{getdelim} makes
@var{lineptr} bigger if it isn't big enough.

@code{getline} is in fact implemented in terms of @code{getdelim}, just
like this:

@smallexample
ssize_t
getline (char **lineptr, size_t *n, FILE *stream)
@{
  return getdelim (lineptr, n, '\n', stream);
@}
@end smallexample
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun {char *} fgets (char *@var{s}, int @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{fgets} function reads characters from the stream @var{stream}
up to and including a newline character and stores them in the string
@var{s}, adding a null character to mark the end of the string.  You
must supply @var{count} characters worth of space in @var{s}, but the
number of characters read is at most @var{count} @minus{} 1.  The extra
character space is used to hold the null character at the end of the
string.

If the system is already at end of file when you call @code{fgets}, then
the contents of the array @var{s} are unchanged and a null pointer is
returned.  A null pointer is also returned if a read error occurs.
Otherwise, the return value is the pointer @var{s}.

@strong{Warning:}  If the input data has a null character, you can't tell.
So don't use @code{fgets} unless you know the data cannot contain a null.
Don't use it to read files edited by the user because, if the user inserts
a null character, you should either handle it properly or print a clear
error message.  We recommend using @code{getline} instead of @code{fgets}.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun {wchar_t *} fgetws (wchar_t *@var{ws}, int @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{fgetws} function reads wide characters from the stream
@var{stream} up to and including a newline character and stores them in
the string @var{ws}, adding a null wide character to mark the end of the
string.  You must supply @var{count} wide characters worth of space in
@var{ws}, but the number of characters read is at most @var{count}
@minus{} 1.  The extra character space is used to hold the null wide
character at the end of the string.

If the system is already at end of file when you call @code{fgetws}, then
the contents of the array @var{ws} are unchanged and a null pointer is
returned.  A null pointer is also returned if a read error occurs.
Otherwise, the return value is the pointer @var{ws}.

@strong{Warning:} If the input data has a null wide character (which are
null bytes in the input stream), you can't tell.  So don't use
@code{fgetws} unless you know the data cannot contain a null.  Don't use
it to read files edited by the user because, if the user inserts a null
character, you should either handle it properly or print a clear error
message.
@comment XXX We need getwline!!!
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun {char *} fgets_unlocked (char *@var{s}, int @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fgets_unlocked} function is equivalent to the @code{fgets}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun {wchar_t *} fgetws_unlocked (wchar_t *@var{ws}, int @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fgetws_unlocked} function is equivalent to the @code{fgetws}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefn {Deprecated function} {char *} gets (char *@var{s})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The function @code{gets} reads characters from the stream @code{stdin}
up to the next newline character, and stores them in the string @var{s}.
The newline character is discarded (note that this differs from the
behavior of @code{fgets}, which copies the newline character into the
string).  If @code{gets} encounters a read error or end-of-file, it
returns a null pointer; otherwise it returns @var{s}.

@strong{Warning:} The @code{gets} function is @strong{very dangerous}
because it provides no protection against overflowing the string
@var{s}.  @Theglibc{} includes it for compatibility only.  You
should @strong{always} use @code{fgets} or @code{getline} instead.  To
remind you of this, the linker (if using GNU @code{ld}) will issue a
warning whenever you use @code{gets}.
@end deftypefn

@node Unreading
@section Unreading
@cindex peeking at input
@cindex unreading characters
@cindex pushing input back

In parser programs it is often useful to examine the next character in
the input stream without removing it from the stream.  This is called
``peeking ahead'' at the input because your program gets a glimpse of
the input it will read next.

Using stream I/O, you can peek ahead at input by first reading it and
then @dfn{unreading} it (also called  @dfn{pushing it back} on the stream).
Unreading a character makes it available to be input again from the stream,
by  the next call to @code{fgetc} or other input function on that stream.

@menu
* Unreading Idea::              An explanation of unreading with pictures.
* How Unread::                  How to call @code{ungetc} to do unreading.
@end menu

@node Unreading Idea
@subsection What Unreading Means

Here is a pictorial explanation of unreading.  Suppose you have a
stream reading a file that contains just six characters, the letters
@samp{foobar}.  Suppose you have read three characters so far.  The
situation looks like this:

@smallexample
f  o  o  b  a  r
	 ^
@end smallexample

@noindent
so the next input character will be @samp{b}.

@c @group   Invalid outside @example
If instead of reading @samp{b} you unread the letter @samp{o}, you get a
situation like this:

@smallexample
f  o  o  b  a  r
	 |
      o--
      ^
@end smallexample

@noindent
so that the next input characters will be @samp{o} and @samp{b}.
@c @end group

@c @group
If you unread @samp{9} instead of @samp{o}, you get this situation:

@smallexample
f  o  o  b  a  r
	 |
      9--
      ^
@end smallexample

@noindent
so that the next input characters will be @samp{9} and @samp{b}.
@c @end group

@node How Unread
@subsection Using @code{ungetc} To Do Unreading

The function to unread a character is called @code{ungetc}, because it
reverses the action of @code{getc}.

@comment stdio.h
@comment ISO
@deftypefun int ungetc (int @var{c}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{ungetc} function pushes back the character @var{c} onto the
input stream @var{stream}.  So the next input from @var{stream} will
read @var{c} before anything else.

If @var{c} is @code{EOF}, @code{ungetc} does nothing and just returns
@code{EOF}.  This lets you call @code{ungetc} with the return value of
@code{getc} without needing to check for an error from @code{getc}.

The character that you push back doesn't have to be the same as the last
character that was actually read from the stream.  In fact, it isn't
necessary to actually read any characters from the stream before
unreading them with @code{ungetc}!  But that is a strange way to write a
program; usually @code{ungetc} is used only to unread a character that
was just read from the same stream.  @Theglibc{} supports this
even on files opened in binary mode, but other systems might not.

@Theglibc{} only supports one character of pushback---in other
words, it does not work to call @code{ungetc} twice without doing input
in between.  Other systems might let you push back multiple characters;
then reading from the stream retrieves the characters in the reverse
order that they were pushed.

Pushing back characters doesn't alter the file; only the internal
buffering for the stream is affected.  If a file positioning function
(such as @code{fseek}, @code{fseeko} or @code{rewind}; @pxref{File
Positioning}) is called, any pending pushed-back characters are
discarded.

Unreading a character on a stream that is at end of file clears the
end-of-file indicator for the stream, because it makes the character of
input available.  After you read that character, trying to read again
will encounter end of file.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun wint_t ungetwc (wint_t @var{wc}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{ungetwc} function behaves just like @code{ungetc} just that it
pushes back a wide character.
@end deftypefun

Here is an example showing the use of @code{getc} and @code{ungetc} to
skip over whitespace characters.  When this function reaches a
non-whitespace character, it unreads that character to be seen again on
the next read operation on the stream.

@smallexample
#include <stdio.h>
#include <ctype.h>

void
skip_whitespace (FILE *stream)
@{
  int c;
  do
    /* @r{No need to check for @code{EOF} because it is not}
       @r{@code{isspace}, and @code{ungetc} ignores @code{EOF}.}  */
    c = getc (stream);
  while (isspace (c));
  ungetc (c, stream);
@}
@end smallexample

@node Block Input/Output
@section Block Input/Output

This section describes how to do input and output operations on blocks
of data.  You can use these functions to read and write binary data, as
well as to read and write text in fixed-size blocks instead of by
characters or lines.
@cindex binary I/O to a stream
@cindex block I/O to a stream
@cindex reading from a stream, by blocks
@cindex writing to a stream, by blocks

Binary files are typically used to read and write blocks of data in the
same format as is used to represent the data in a running program.  In
other words, arbitrary blocks of memory---not just character or string
objects---can be written to a binary file, and meaningfully read in
again by the same program.

Storing data in binary form is often considerably more efficient than
using the formatted I/O functions.  Also, for floating-point numbers,
the binary form avoids possible loss of precision in the conversion
process.  On the other hand, binary files can't be examined or modified
easily using many standard file utilities (such as text editors), and
are not portable between different implementations of the language, or
different kinds of computers.

These functions are declared in @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypefun size_t fread (void *@var{data}, size_t @var{size}, size_t @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function reads up to @var{count} objects of size @var{size} into
the array @var{data}, from the stream @var{stream}.  It returns the
number of objects actually read, which might be less than @var{count} if
a read error occurs or the end of the file is reached.  This function
returns a value of zero (and doesn't read anything) if either @var{size}
or @var{count} is zero.

If @code{fread} encounters end of file in the middle of an object, it
returns the number of complete objects read, and discards the partial
object.  Therefore, the stream remains at the actual end of the file.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun size_t fread_unlocked (void *@var{data}, size_t @var{size}, size_t @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fread_unlocked} function is equivalent to the @code{fread}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun size_t fwrite (const void *@var{data}, size_t @var{size}, size_t @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function writes up to @var{count} objects of size @var{size} from
the array @var{data}, to the stream @var{stream}.  The return value is
normally @var{count}, if the call succeeds.  Any other value indicates
some sort of error, such as running out of space.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun size_t fwrite_unlocked (const void *@var{data}, size_t @var{size}, size_t @var{count}, FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fwrite_unlocked} function is equivalent to the @code{fwrite}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

@node Formatted Output
@section Formatted Output

@cindex format string, for @code{printf}
@cindex template, for @code{printf}
@cindex formatted output to a stream
@cindex writing to a stream, formatted
The functions described in this section (@code{printf} and related
functions) provide a convenient way to perform formatted output.  You
call @code{printf} with a @dfn{format string} or @dfn{template string}
that specifies how to format the values of the remaining arguments.

Unless your program is a filter that specifically performs line- or
character-oriented processing, using @code{printf} or one of the other
related functions described in this section is usually the easiest and
most concise way to perform output.  These functions are especially
useful for printing error messages, tables of data, and the like.

@menu
* Formatted Output Basics::     Some examples to get you started.
* Output Conversion Syntax::    General syntax of conversion
				 specifications.
* Table of Output Conversions:: Summary of output conversions and
				 what they do.
* Integer Conversions::         Details about formatting of integers.
* Floating-Point Conversions::  Details about formatting of
				 floating-point numbers.
* Other Output Conversions::    Details about formatting of strings,
				 characters, pointers, and the like.
* Formatted Output Functions::  Descriptions of the actual functions.
* Dynamic Output::		Functions that allocate memory for the output.
* Variable Arguments Output::   @code{vprintf} and friends.
* Parsing a Template String::   What kinds of args does a given template
				 call for?
* Example of Parsing::          Sample program using @code{parse_printf_format}.
@end menu

@node Formatted Output Basics
@subsection Formatted Output Basics

The @code{printf} function can be used to print any number of arguments.
The template string argument you supply in a call provides
information not only about the number of additional arguments, but also
about their types and what style should be used for printing them.

Ordinary characters in the template string are simply written to the
output stream as-is, while @dfn{conversion specifications} introduced by
a @samp{%} character in the template cause subsequent arguments to be
formatted and written to the output stream.  For example,
@cindex conversion specifications (@code{printf})

@smallexample
int pct = 37;
char filename[] = "foo.txt";
printf ("Processing of `%s' is %d%% finished.\nPlease be patient.\n",
	filename, pct);
@end smallexample

@noindent
produces output like

@smallexample
Processing of `foo.txt' is 37% finished.
Please be patient.
@end smallexample

This example shows the use of the @samp{%d} conversion to specify that
an @code{int} argument should be printed in decimal notation, the
@samp{%s} conversion to specify printing of a string argument, and
the @samp{%%} conversion to print a literal @samp{%} character.

There are also conversions for printing an integer argument as an
unsigned value in octal, decimal, or hexadecimal radix (@samp{%o},
@samp{%u}, or @samp{%x}, respectively); or as a character value
(@samp{%c}).

Floating-point numbers can be printed in normal, fixed-point notation
using the @samp{%f} conversion or in exponential notation using the
@samp{%e} conversion.  The @samp{%g} conversion uses either @samp{%e}
or @samp{%f} format, depending on what is more appropriate for the
magnitude of the particular number.

You can control formatting more precisely by writing @dfn{modifiers}
between the @samp{%} and the character that indicates which conversion
to apply.  These slightly alter the ordinary behavior of the conversion.
For example, most conversion specifications permit you to specify a
minimum field width and a flag indicating whether you want the result
left- or right-justified within the field.

The specific flags and modifiers that are permitted and their
interpretation vary depending on the particular conversion.  They're all
described in more detail in the following sections.  Don't worry if this
all seems excessively complicated at first; you can almost always get
reasonable free-format output without using any of the modifiers at all.
The modifiers are mostly used to make the output look ``prettier'' in
tables.

@node Output Conversion Syntax
@subsection Output Conversion Syntax

This section provides details about the precise syntax of conversion
specifications that can appear in a @code{printf} template
string.

Characters in the template string that are not part of a conversion
specification are printed as-is to the output stream.  Multibyte
character sequences (@pxref{Character Set Handling}) are permitted in a
template string.

The conversion specifications in a @code{printf} template string have
the general form:

@smallexample
% @r{[} @var{param-no} @r{$]} @var{flags} @var{width} @r{[} . @var{precision} @r{]} @var{type} @var{conversion}
@end smallexample

@noindent
or

@smallexample
% @r{[} @var{param-no} @r{$]} @var{flags} @var{width} . @r{*} @r{[} @var{param-no} @r{$]} @var{type} @var{conversion}
@end smallexample

For example, in the conversion specifier @samp{%-10.8ld}, the @samp{-}
is a flag, @samp{10} specifies the field width, the precision is
@samp{8}, the letter @samp{l} is a type modifier, and @samp{d} specifies
the conversion style.  (This particular type specifier says to
print a @code{long int} argument in decimal notation, with a minimum of
8 digits left-justified in a field at least 10 characters wide.)

In more detail, output conversion specifications consist of an
initial @samp{%} character followed in sequence by:

@itemize @bullet
@item
An optional specification of the parameter used for this format.
Normally the parameters to the @code{printf} function are assigned to the
formats in the order of appearance in the format string.  But in some
situations (such as message translation) this is not desirable and this
extension allows an explicit parameter to be specified.

The @var{param-no} parts of the format must be integers in the range of
1 to the maximum number of arguments present to the function call.  Some
implementations limit this number to a certainly upper bound.  The exact
limit can be retrieved by the following constant.

@defvr Macro NL_ARGMAX
The value of @code{NL_ARGMAX} is the maximum value allowed for the
specification of a positional parameter in a @code{printf} call.  The
actual value in effect at runtime can be retrieved by using
@code{sysconf} using the @code{_SC_NL_ARGMAX} parameter @pxref{Sysconf
Definition}.

Some system have a quite low limit such as @math{9} for @w{System V}
systems.  @Theglibc{} has no real limit.
@end defvr

If any of the formats has a specification for the parameter position all
of them in the format string shall have one.  Otherwise the behavior is
undefined.

@item
Zero or more @dfn{flag characters} that modify the normal behavior of
the conversion specification.
@cindex flag character (@code{printf})

@item
An optional decimal integer specifying the @dfn{minimum field width}.
If the normal conversion produces fewer characters than this, the field
is padded with spaces to the specified width.  This is a @emph{minimum}
value; if the normal conversion produces more characters than this, the
field is @emph{not} truncated.  Normally, the output is right-justified
within the field.
@cindex minimum field width (@code{printf})

You can also specify a field width of @samp{*}.  This means that the
next argument in the argument list (before the actual value to be
printed) is used as the field width.  The value must be an @code{int}.
If the value is negative, this means to set the @samp{-} flag (see
below) and to use the absolute value as the field width.

@item
An optional @dfn{precision} to specify the number of digits to be
written for the numeric conversions.  If the precision is specified, it
consists of a period (@samp{.}) followed optionally by a decimal integer
(which defaults to zero if omitted).
@cindex precision (@code{printf})

You can also specify a precision of @samp{*}.  This means that the next
argument in the argument list (before the actual value to be printed) is
used as the precision.  The value must be an @code{int}, and is ignored
if it is negative.  If you specify @samp{*} for both the field width and
precision, the field width argument precedes the precision argument.
Other C library versions may not recognize this syntax.

@item
An optional @dfn{type modifier character}, which is used to specify the
data type of the corresponding argument if it differs from the default
type.  (For example, the integer conversions assume a type of @code{int},
but you can specify @samp{h}, @samp{l}, or @samp{L} for other integer
types.)
@cindex type modifier character (@code{printf})

@item
A character that specifies the conversion to be applied.
@end itemize

The exact options that are permitted and how they are interpreted vary
between the different conversion specifiers.  See the descriptions of the
individual conversions for information about the particular options that
they use.

With the @samp{-Wformat} option, the GNU C compiler checks calls to
@code{printf} and related functions.  It examines the format string and
verifies that the correct number and types of arguments are supplied.
There is also a GNU C syntax to tell the compiler that a function you
write uses a @code{printf}-style format string.
@xref{Function Attributes, , Declaring Attributes of Functions,
gcc.info, Using GNU CC}, for more information.

@node Table of Output Conversions
@subsection Table of Output Conversions
@cindex output conversions, for @code{printf}

Here is a table summarizing what all the different conversions do:

@table @asis
@item @samp{%d}, @samp{%i}
Print an integer as a signed decimal number.  @xref{Integer
Conversions}, for details.  @samp{%d} and @samp{%i} are synonymous for
output, but are different when used with @code{scanf} for input
(@pxref{Table of Input Conversions}).

@item @samp{%o}
Print an integer as an unsigned octal number.  @xref{Integer
Conversions}, for details.

@item @samp{%u}
Print an integer as an unsigned decimal number.  @xref{Integer
Conversions}, for details.

@item @samp{%x}, @samp{%X}
Print an integer as an unsigned hexadecimal number.  @samp{%x} uses
lower-case letters and @samp{%X} uses upper-case.  @xref{Integer
Conversions}, for details.

@item @samp{%f}
Print a floating-point number in normal (fixed-point) notation.
@xref{Floating-Point Conversions}, for details.

@item @samp{%e}, @samp{%E}
Print a floating-point number in exponential notation.  @samp{%e} uses
lower-case letters and @samp{%E} uses upper-case.  @xref{Floating-Point
Conversions}, for details.

@item @samp{%g}, @samp{%G}
Print a floating-point number in either normal or exponential notation,
whichever is more appropriate for its magnitude.  @samp{%g} uses
lower-case letters and @samp{%G} uses upper-case.  @xref{Floating-Point
Conversions}, for details.

@item @samp{%a}, @samp{%A}
Print a floating-point number in a hexadecimal fractional notation which
the exponent to base 2 represented in decimal digits.  @samp{%a} uses
lower-case letters and @samp{%A} uses upper-case.  @xref{Floating-Point
Conversions}, for details.

@item @samp{%c}
Print a single character.  @xref{Other Output Conversions}.

@item @samp{%C}
This is an alias for @samp{%lc} which is supported for compatibility
with the Unix standard.

@item @samp{%s}
Print a string.  @xref{Other Output Conversions}.

@item @samp{%S}
This is an alias for @samp{%ls} which is supported for compatibility
with the Unix standard.

@item @samp{%p}
Print the value of a pointer.  @xref{Other Output Conversions}.

@item @samp{%n}
Get the number of characters printed so far.  @xref{Other Output Conversions}.
Note that this conversion specification never produces any output.

@item @samp{%m}
Print the string corresponding to the value of @code{errno}.
(This is a GNU extension.)
@xref{Other Output Conversions}.

@item @samp{%%}
Print a literal @samp{%} character.  @xref{Other Output Conversions}.
@end table

If the syntax of a conversion specification is invalid, unpredictable
things will happen, so don't do this.  If there aren't enough function
arguments provided to supply values for all the conversion
specifications in the template string, or if the arguments are not of
the correct types, the results are unpredictable.  If you supply more
arguments than conversion specifications, the extra argument values are
simply ignored; this is sometimes useful.

@node Integer Conversions
@subsection Integer Conversions

This section describes the options for the @samp{%d}, @samp{%i},
@samp{%o}, @samp{%u}, @samp{%x}, and @samp{%X} conversion
specifications.  These conversions print integers in various formats.

The @samp{%d} and @samp{%i} conversion specifications both print an
@code{int} argument as a signed decimal number; while @samp{%o},
@samp{%u}, and @samp{%x} print the argument as an unsigned octal,
decimal, or hexadecimal number (respectively).  The @samp{%X} conversion
specification is just like @samp{%x} except that it uses the characters
@samp{ABCDEF} as digits instead of @samp{abcdef}.

The following flags are meaningful:

@table @asis
@item @samp{-}
Left-justify the result in the field (instead of the normal
right-justification).

@item @samp{+}
For the signed @samp{%d} and @samp{%i} conversions, print a
plus sign if the value is positive.

@item @samp{ }
For the signed @samp{%d} and @samp{%i} conversions, if the result
doesn't start with a plus or minus sign, prefix it with a space
character instead.  Since the @samp{+} flag ensures that the result
includes a sign, this flag is ignored if you supply both of them.

@item @samp{#}
For the @samp{%o} conversion, this forces the leading digit to be
@samp{0}, as if by increasing the precision.  For @samp{%x} or
@samp{%X}, this prefixes a leading @samp{0x} or @samp{0X} (respectively)
to the result.  This doesn't do anything useful for the @samp{%d},
@samp{%i}, or @samp{%u} conversions.  Using this flag produces output
which can be parsed by the @code{strtoul} function (@pxref{Parsing of
Integers}) and @code{scanf} with the @samp{%i} conversion
(@pxref{Numeric Input Conversions}).

@item @samp{'}
Separate the digits into groups as specified by the locale specified for
the @code{LC_NUMERIC} category; @pxref{General Numeric}.  This flag is a
GNU extension.

@item @samp{0}
Pad the field with zeros instead of spaces.  The zeros are placed after
any indication of sign or base.  This flag is ignored if the @samp{-}
flag is also specified, or if a precision is specified.
@end table

If a precision is supplied, it specifies the minimum number of digits to
appear; leading zeros are produced if necessary.  If you don't specify a
precision, the number is printed with as many digits as it needs.  If
you convert a value of zero with an explicit precision of zero, then no
characters at all are produced.

Without a type modifier, the corresponding argument is treated as an
@code{int} (for the signed conversions @samp{%i} and @samp{%d}) or
@code{unsigned int} (for the unsigned conversions @samp{%o}, @samp{%u},
@samp{%x}, and @samp{%X}).  Recall that since @code{printf} and friends
are variadic, any @code{char} and @code{short} arguments are
automatically converted to @code{int} by the default argument
promotions.  For arguments of other integer types, you can use these
modifiers:

@table @samp
@item hh
Specifies that the argument is a @code{signed char} or @code{unsigned
char}, as appropriate.  A @code{char} argument is converted to an
@code{int} or @code{unsigned int} by the default argument promotions
anyway, but the @samp{h} modifier says to convert it back to a
@code{char} again.

This modifier was introduced in @w{ISO C99}.

@item h
Specifies that the argument is a @code{short int} or @code{unsigned
short int}, as appropriate.  A @code{short} argument is converted to an
@code{int} or @code{unsigned int} by the default argument promotions
anyway, but the @samp{h} modifier says to convert it back to a
@code{short} again.

@item j
Specifies that the argument is a @code{intmax_t} or @code{uintmax_t}, as
appropriate.

This modifier was introduced in @w{ISO C99}.

@item l
Specifies that the argument is a @code{long int} or @code{unsigned long
int}, as appropriate.  Two @samp{l} characters is like the @samp{L}
modifier, below.

If used with @samp{%c} or @samp{%s} the corresponding parameter is
considered as a wide character or wide character string respectively.
This use of @samp{l} was introduced in @w{Amendment 1} to @w{ISO C90}.

@item L
@itemx ll
@itemx q
Specifies that the argument is a @code{long long int}.  (This type is
an extension supported by the GNU C compiler.  On systems that don't
support extra-long integers, this is the same as @code{long int}.)

The @samp{q} modifier is another name for the same thing, which comes
from 4.4 BSD; a @w{@code{long long int}} is sometimes called a ``quad''
@code{int}.

@item t
Specifies that the argument is a @code{ptrdiff_t}.

This modifier was introduced in @w{ISO C99}.

@item z
@itemx Z
Specifies that the argument is a @code{size_t}.

@samp{z} was introduced in @w{ISO C99}.  @samp{Z} is a GNU extension
predating this addition and should not be used in new code.
@end table

Here is an example.  Using the template string:

@smallexample
"|%5d|%-5d|%+5d|%+-5d|% 5d|%05d|%5.0d|%5.2d|%d|\n"
@end smallexample

@noindent
to print numbers using the different options for the @samp{%d}
conversion gives results like:

@smallexample
|    0|0    |   +0|+0   |    0|00000|     |   00|0|
|    1|1    |   +1|+1   |    1|00001|    1|   01|1|
|   -1|-1   |   -1|-1   |   -1|-0001|   -1|  -01|-1|
|100000|100000|+100000|+100000| 100000|100000|100000|100000|100000|
@end smallexample

In particular, notice what happens in the last case where the number
is too large to fit in the minimum field width specified.

Here are some more examples showing how unsigned integers print under
various format options, using the template string:

@smallexample
"|%5u|%5o|%5x|%5X|%#5o|%#5x|%#5X|%#10.8x|\n"
@end smallexample

@smallexample
|    0|    0|    0|    0|    0|    0|    0|  00000000|
|    1|    1|    1|    1|   01|  0x1|  0X1|0x00000001|
|100000|303240|186a0|186A0|0303240|0x186a0|0X186A0|0x000186a0|
@end smallexample


@node Floating-Point Conversions
@subsection Floating-Point Conversions

This section discusses the conversion specifications for floating-point
numbers: the @samp{%f}, @samp{%e}, @samp{%E}, @samp{%g}, and @samp{%G}
conversions.

The @samp{%f} conversion prints its argument in fixed-point notation,
producing output of the form
@w{[@code{-}]@var{ddd}@code{.}@var{ddd}},
where the number of digits following the decimal point is controlled
by the precision you specify.

The @samp{%e} conversion prints its argument in exponential notation,
producing output of the form
@w{[@code{-}]@var{d}@code{.}@var{ddd}@code{e}[@code{+}|@code{-}]@var{dd}}.
Again, the number of digits following the decimal point is controlled by
the precision.  The exponent always contains at least two digits.  The
@samp{%E} conversion is similar but the exponent is marked with the letter
@samp{E} instead of @samp{e}.

The @samp{%g} and @samp{%G} conversions print the argument in the style
of @samp{%e} or @samp{%E} (respectively) if the exponent would be less
than -4 or greater than or equal to the precision; otherwise they use
the @samp{%f} style.  A precision of @code{0}, is taken as 1.
Trailing zeros are removed from the fractional portion of the result and
a decimal-point character appears only if it is followed by a digit.

The @samp{%a} and @samp{%A} conversions are meant for representing
floating-point numbers exactly in textual form so that they can be
exchanged as texts between different programs and/or machines.  The
numbers are represented is the form
@w{[@code{-}]@code{0x}@var{h}@code{.}@var{hhh}@code{p}[@code{+}|@code{-}]@var{dd}}.
At the left of the decimal-point character exactly one digit is print.
This character is only @code{0} if the number is denormalized.
Otherwise the value is unspecified; it is implementation dependent how many
bits are used.  The number of hexadecimal digits on the right side of
the decimal-point character is equal to the precision.  If the precision
is zero it is determined to be large enough to provide an exact
representation of the number (or it is large enough to distinguish two
adjacent values if the @code{FLT_RADIX} is not a power of 2,
@pxref{Floating Point Parameters}).  For the @samp{%a} conversion
lower-case characters are used to represent the hexadecimal number and
the prefix and exponent sign are printed as @code{0x} and @code{p}
respectively.  Otherwise upper-case characters are used and @code{0X}
and @code{P} are used for the representation of prefix and exponent
string.  The exponent to the base of two is printed as a decimal number
using at least one digit but at most as many digits as necessary to
represent the value exactly.

If the value to be printed represents infinity or a NaN, the output is
@w{[@code{-}]@code{inf}} or @code{nan} respectively if the conversion
specifier is @samp{%a}, @samp{%e}, @samp{%f}, or @samp{%g} and it is
@w{[@code{-}]@code{INF}} or @code{NAN} respectively if the conversion is
@samp{%A}, @samp{%E}, or @samp{%G}.

The following flags can be used to modify the behavior:

@comment We use @asis instead of @samp so we can have ` ' as an item.
@table @asis
@item @samp{-}
Left-justify the result in the field.  Normally the result is
right-justified.

@item @samp{+}
Always include a plus or minus sign in the result.

@item @samp{ }
If the result doesn't start with a plus or minus sign, prefix it with a
space instead.  Since the @samp{+} flag ensures that the result includes
a sign, this flag is ignored if you supply both of them.

@item @samp{#}
Specifies that the result should always include a decimal point, even
if no digits follow it.  For the @samp{%g} and @samp{%G} conversions,
this also forces trailing zeros after the decimal point to be left
in place where they would otherwise be removed.

@item @samp{'}
Separate the digits of the integer part of the result into groups as
specified by the locale specified for the @code{LC_NUMERIC} category;
@pxref{General Numeric}.  This flag is a GNU extension.

@item @samp{0}
Pad the field with zeros instead of spaces; the zeros are placed
after any sign.  This flag is ignored if the @samp{-} flag is also
specified.
@end table

The precision specifies how many digits follow the decimal-point
character for the @samp{%f}, @samp{%e}, and @samp{%E} conversions.  For
these conversions, the default precision is @code{6}.  If the precision
is explicitly @code{0}, this suppresses the decimal point character
entirely.  For the @samp{%g} and @samp{%G} conversions, the precision
specifies how many significant digits to print.  Significant digits are
the first digit before the decimal point, and all the digits after it.
If the precision is @code{0} or not specified for @samp{%g} or @samp{%G},
it is treated like a value of @code{1}.  If the value being printed
cannot be expressed accurately in the specified number of digits, the
value is rounded to the nearest number that fits.

Without a type modifier, the floating-point conversions use an argument
of type @code{double}.  (By the default argument promotions, any
@code{float} arguments are automatically converted to @code{double}.)
The following type modifier is supported:

@table @samp
@item L
An uppercase @samp{L} specifies that the argument is a @code{long
double}.
@end table

Here are some examples showing how numbers print using the various
floating-point conversions.  All of the numbers were printed using
this template string:

@smallexample
"|%13.4a|%13.4f|%13.4e|%13.4g|\n"
@end smallexample

Here is the output:

@smallexample
|  0x0.0000p+0|       0.0000|   0.0000e+00|            0|
|  0x1.0000p-1|       0.5000|   5.0000e-01|          0.5|
|  0x1.0000p+0|       1.0000|   1.0000e+00|            1|
| -0x1.0000p+0|      -1.0000|  -1.0000e+00|           -1|
|  0x1.9000p+6|     100.0000|   1.0000e+02|          100|
|  0x1.f400p+9|    1000.0000|   1.0000e+03|         1000|
| 0x1.3880p+13|   10000.0000|   1.0000e+04|        1e+04|
| 0x1.81c8p+13|   12345.0000|   1.2345e+04|    1.234e+04|
| 0x1.86a0p+16|  100000.0000|   1.0000e+05|        1e+05|
| 0x1.e240p+16|  123456.0000|   1.2346e+05|    1.235e+05|
@end smallexample

Notice how the @samp{%g} conversion drops trailing zeros.

@node Other Output Conversions
@subsection Other Output Conversions

This section describes miscellaneous conversions for @code{printf}.

The @samp{%c} conversion prints a single character.  In case there is no
@samp{l} modifier the @code{int} argument is first converted to an
@code{unsigned char}.  Then, if used in a wide stream function, the
character is converted into the corresponding wide character.  The
@samp{-} flag can be used to specify left-justification in the field,
but no other flags are defined, and no precision or type modifier can be
given.  For example:

@smallexample
printf ("%c%c%c%c%c", 'h', 'e', 'l', 'l', 'o');
@end smallexample

@noindent
prints @samp{hello}.

If there is a @samp{l} modifier present the argument is expected to be
of type @code{wint_t}.  If used in a multibyte function the wide
character is converted into a multibyte character before being added to
the output.  In this case more than one output byte can be produced.

The @samp{%s} conversion prints a string.  If no @samp{l} modifier is
present the corresponding argument must be of type @code{char *} (or
@code{const char *}).  If used in a wide stream function the string is
first converted in a wide character string.  A precision can be
specified to indicate the maximum number of characters to write;
otherwise characters in the string up to but not including the
terminating null character are written to the output stream.  The
@samp{-} flag can be used to specify left-justification in the field,
but no other flags or type modifiers are defined for this conversion.
For example:

@smallexample
printf ("%3s%-6s", "no", "where");
@end smallexample

@noindent
prints @samp{ nowhere }.

If there is a @samp{l} modifier present the argument is expected to be of type @code{wchar_t} (or @code{const wchar_t *}).

If you accidentally pass a null pointer as the argument for a @samp{%s}
conversion, @theglibc{} prints it as @samp{(null)}.  We think this
is more useful than crashing.  But it's not good practice to pass a null
argument intentionally.

The @samp{%m} conversion prints the string corresponding to the error
code in @code{errno}.  @xref{Error Messages}.  Thus:

@smallexample
fprintf (stderr, "can't open `%s': %m\n", filename);
@end smallexample

@noindent
is equivalent to:

@smallexample
fprintf (stderr, "can't open `%s': %s\n", filename, strerror (errno));
@end smallexample

@noindent
The @samp{%m} conversion is a @glibcadj{} extension.

The @samp{%p} conversion prints a pointer value.  The corresponding
argument must be of type @code{void *}.  In practice, you can use any
type of pointer.

In @theglibc{}, non-null pointers are printed as unsigned integers,
as if a @samp{%#x} conversion were used.  Null pointers print as
@samp{(nil)}.  (Pointers might print differently in other systems.)

For example:

@smallexample
printf ("%p", "testing");
@end smallexample

@noindent
prints @samp{0x} followed by a hexadecimal number---the address of the
string constant @code{"testing"}.  It does not print the word
@samp{testing}.

You can supply the @samp{-} flag with the @samp{%p} conversion to
specify left-justification, but no other flags, precision, or type
modifiers are defined.

The @samp{%n} conversion is unlike any of the other output conversions.
It uses an argument which must be a pointer to an @code{int}, but
instead of printing anything it stores the number of characters printed
so far by this call at that location.  The @samp{h} and @samp{l} type
modifiers are permitted to specify that the argument is of type
@code{short int *} or @code{long int *} instead of @code{int *}, but no
flags, field width, or precision are permitted.

For example,

@smallexample
int nchar;
printf ("%d %s%n\n", 3, "bears", &nchar);
@end smallexample

@noindent
prints:

@smallexample
3 bears
@end smallexample

@noindent
and sets @code{nchar} to @code{7}, because @samp{3 bears} is seven
characters.


The @samp{%%} conversion prints a literal @samp{%} character.  This
conversion doesn't use an argument, and no flags, field width,
precision, or type modifiers are permitted.


@node Formatted Output Functions
@subsection Formatted Output Functions

This section describes how to call @code{printf} and related functions.
Prototypes for these functions are in the header file @file{stdio.h}.
Because these functions take a variable number of arguments, you
@emph{must} declare prototypes for them before using them.  Of course,
the easiest way to make sure you have all the right prototypes is to
just include @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypefun int printf (const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
The @code{printf} function prints the optional arguments under the
control of the template string @var{template} to the stream
@code{stdout}.  It returns the number of characters printed, or a
negative value if there was an output error.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int wprintf (const wchar_t *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
The @code{wprintf} function prints the optional arguments under the
control of the wide template string @var{template} to the stream
@code{stdout}.  It returns the number of wide characters printed, or a
negative value if there was an output error.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int fprintf (FILE *@var{stream}, const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is just like @code{printf}, except that the output is
written to the stream @var{stream} instead of @code{stdout}.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int fwprintf (FILE *@var{stream}, const wchar_t *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is just like @code{wprintf}, except that the output is
written to the stream @var{stream} instead of @code{stdout}.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int sprintf (char *@var{s}, const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is like @code{printf}, except that the output is stored in the character
array @var{s} instead of written to a stream.  A null character is written
to mark the end of the string.

The @code{sprintf} function returns the number of characters stored in
the array @var{s}, not including the terminating null character.

The behavior of this function is undefined if copying takes place
between objects that overlap---for example, if @var{s} is also given
as an argument to be printed under control of the @samp{%s} conversion.
@xref{Copying Strings and Arrays}.

@strong{Warning:} The @code{sprintf} function can be @strong{dangerous}
because it can potentially output more characters than can fit in the
allocation size of the string @var{s}.  Remember that the field width
given in a conversion specification is only a @emph{minimum} value.

To avoid this problem, you can use @code{snprintf} or @code{asprintf},
described below.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun int swprintf (wchar_t *@var{s}, size_t @var{size}, const wchar_t *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is like @code{wprintf}, except that the output is stored in the
wide character array @var{ws} instead of written to a stream.  A null
wide character is written to mark the end of the string.  The @var{size}
argument specifies the maximum number of characters to produce.  The
trailing null character is counted towards this limit, so you should
allocate at least @var{size} wide characters for the string @var{ws}.

The return value is the number of characters generated for the given
input, excluding the trailing null.  If not all output fits into the
provided buffer a negative value is returned.  You should try again with
a bigger output string.  @emph{Note:} this is different from how
@code{snprintf} handles this situation.

Note that the corresponding narrow stream function takes fewer
parameters.  @code{swprintf} in fact corresponds to the @code{snprintf}
function.  Since the @code{sprintf} function can be dangerous and should
be avoided the @w{ISO C} committee refused to make the same mistake
again and decided to not define a function exactly corresponding to
@code{sprintf}.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int snprintf (char *@var{s}, size_t @var{size}, const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
The @code{snprintf} function is similar to @code{sprintf}, except that
the @var{size} argument specifies the maximum number of characters to
produce.  The trailing null character is counted towards this limit, so
you should allocate at least @var{size} characters for the string @var{s}.
If @var{size} is zero, nothing, not even the null byte, shall be written and
@var{s} may be a null pointer.

The return value is the number of characters which would be generated
for the given input, excluding the trailing null.  If this value is
greater or equal to @var{size}, not all characters from the result have
been stored in @var{s}.  You should try again with a bigger output
string.  Here is an example of doing this:

@smallexample
@group
/* @r{Construct a message describing the value of a variable}
   @r{whose name is @var{name} and whose value is @var{value}.} */
char *
make_message (char *name, char *value)
@{
  /* @r{Guess we need no more than 100 chars of space.} */
  int size = 100;
  char *buffer = (char *) xmalloc (size);
  int nchars;
@end group
@group
  if (buffer == NULL)
    return NULL;

 /* @r{Try to print in the allocated space.} */
  nchars = snprintf (buffer, size, "value of %s is %s",
		     name, value);
@end group
@group
  if (nchars >= size)
    @{
      /* @r{Reallocate buffer now that we know
	 how much space is needed.} */
      size = nchars + 1;
      buffer = (char *) xrealloc (buffer, size);

      if (buffer != NULL)
	/* @r{Try again.} */
	snprintf (buffer, size, "value of %s is %s",
		  name, value);
    @}
  /* @r{The last call worked, return the string.} */
  return buffer;
@}
@end group
@end smallexample

In practice, it is often easier just to use @code{asprintf}, below.

@strong{Attention:} In versions of @theglibc{} prior to 2.1 the
return value is the number of characters stored, not including the
terminating null; unless there was not enough space in @var{s} to
store the result in which case @code{-1} is returned.  This was
changed in order to comply with the @w{ISO C99} standard.
@end deftypefun

@node Dynamic Output
@subsection Dynamically Allocating Formatted Output

The functions in this section do formatted output and place the results
in dynamically allocated memory.

@comment stdio.h
@comment GNU
@deftypefun int asprintf (char **@var{ptr}, const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This function is similar to @code{sprintf}, except that it dynamically
allocates a string (as with @code{malloc}; @pxref{Unconstrained
Allocation}) to hold the output, instead of putting the output in a
buffer you allocate in advance.  The @var{ptr} argument should be the
address of a @code{char *} object, and a successful call to
@code{asprintf} stores a pointer to the newly allocated string at that
location.

The return value is the number of characters allocated for the buffer, or
less than zero if an error occurred.  Usually this means that the buffer
could not be allocated.

Here is how to use @code{asprintf} to get the same result as the
@code{snprintf} example, but more easily:

@smallexample
/* @r{Construct a message describing the value of a variable}
   @r{whose name is @var{name} and whose value is @var{value}.} */
char *
make_message (char *name, char *value)
@{
  char *result;
  if (asprintf (&result, "value of %s is %s", name, value) < 0)
    return NULL;
  return result;
@}
@end smallexample
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int obstack_printf (struct obstack *@var{obstack}, const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtsrace{:obstack} @mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acucorrupt{} @acsmem{}}}
This function is similar to @code{asprintf}, except that it uses the
obstack @var{obstack} to allocate the space.  @xref{Obstacks}.

The characters are written onto the end of the current object.
To get at them, you must finish the object with @code{obstack_finish}
(@pxref{Growing Objects}).@refill
@end deftypefun

@node Variable Arguments Output
@subsection Variable Arguments Output Functions

The functions @code{vprintf} and friends are provided so that you can
define your own variadic @code{printf}-like functions that make use of
the same internals as the built-in formatted output functions.

The most natural way to define such functions would be to use a language
construct to say, ``Call @code{printf} and pass this template plus all
of my arguments after the first five.''  But there is no way to do this
in C, and it would be hard to provide a way, since at the C language
level there is no way to tell how many arguments your function received.

Since that method is impossible, we provide alternative functions, the
@code{vprintf} series, which lets you pass a @code{va_list} to describe
``all of my arguments after the first five.''

When it is sufficient to define a macro rather than a real function,
the GNU C compiler provides a way to do this much more easily with macros.
For example:

@smallexample
#define myprintf(a, b, c, d, e, rest...) \
	    printf (mytemplate , ## rest)
@end smallexample

@noindent
@xref{Variadic Macros,,, cpp, The C preprocessor}, for details.
But this is limited to macros, and does not apply to real functions at all.

Before calling @code{vprintf} or the other functions listed in this
section, you @emph{must} call @code{va_start} (@pxref{Variadic
Functions}) to initialize a pointer to the variable arguments.  Then you
can call @code{va_arg} to fetch the arguments that you want to handle
yourself.  This advances the pointer past those arguments.

Once your @code{va_list} pointer is pointing at the argument of your
choice, you are ready to call @code{vprintf}.  That argument and all
subsequent arguments that were passed to your function are used by
@code{vprintf} along with the template that you specified separately.

@strong{Portability Note:} The value of the @code{va_list} pointer is
undetermined after the call to @code{vprintf}, so you must not use
@code{va_arg} after you call @code{vprintf}.  Instead, you should call
@code{va_end} to retire the pointer from service.  You can call
@code{va_start} again and begin fetching the arguments from the start of
the variable argument list.  (Alternatively, you can use @code{va_copy}
to make a copy of the @code{va_list} pointer before calling
@code{vfprintf}.)  Calling @code{vprintf} does not destroy the argument
list of your function, merely the particular pointer that you passed to
it.

Prototypes for these functions are declared in @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypefun int vprintf (const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is similar to @code{printf} except that, instead of taking
a variable number of arguments directly, it takes an argument list
pointer @var{ap}.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int vwprintf (const wchar_t *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is similar to @code{wprintf} except that, instead of taking
a variable number of arguments directly, it takes an argument list
pointer @var{ap}.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int vfprintf (FILE *@var{stream}, const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
@c Although vfprintf sets up a cleanup region to release the lock on the
@c output stream, it doesn't use it to release args_value or string in
@c case of cancellation.  This doesn't make it unsafe, but cancelling it
@c may leak memory.  The unguarded use of __printf_function_table is
@c also of concern for all callers.
@c _itoa ok
@c   _udiv_qrnnd_preinv ok
@c group_number ok
@c _i18n_number_rewrite
@c   __wctrans ok
@c   __towctrans @mtslocale
@c   __wcrtomb ok? dup below
@c   outdigit_value ok
@c   outdigitwc_value ok
@c outchar ok
@c outstring ok
@c PAD ok
@c __printf_fp @mtslocale @ascuheap @acsmem
@c __printf_fphex @mtslocale
@c __readonly_area
@c   [GNU/Linux] fopen, strtoul, free
@c __strerror_r ok if no translation, check otherwise
@c __btowc ? gconv-modules
@c __wcrtomb ok (not using internal state) gconv-modules
@c ARGCHECK
@c UNBUFFERED_P (tested before taking the stream lock)
@c buffered_vfprintf ok
@c __find_spec(wc|mb)
@c read_int
@c __libc_use_alloca
@c process_arg
@c process_string_arg
@c extend_alloca
@c __parse_one_spec(wc|mb)
@c *__printf_arginfo_table unguarded
@c __printf_va_arg_table-> unguarded
@c *__printf_function_table unguarded
@c done_add
@c printf_unknown
@c   outchar
@c   _itoa_word
This is the equivalent of @code{fprintf} with the variable argument list
specified directly as for @code{vprintf}.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int vfwprintf (FILE *@var{stream}, const wchar_t *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This is the equivalent of @code{fwprintf} with the variable argument list
specified directly as for @code{vwprintf}.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int vsprintf (char *@var{s}, const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is the equivalent of @code{sprintf} with the variable argument list
specified directly as for @code{vprintf}.
@end deftypefun

@comment wchar.h
@comment GNU
@deftypefun int vswprintf (wchar_t *@var{s}, size_t @var{size}, const wchar_t *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is the equivalent of @code{swprintf} with the variable argument list
specified directly as for @code{vwprintf}.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int vsnprintf (char *@var{s}, size_t @var{size}, const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is the equivalent of @code{snprintf} with the variable argument list
specified directly as for @code{vprintf}.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int vasprintf (char **@var{ptr}, const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
The @code{vasprintf} function is the equivalent of @code{asprintf} with the
variable argument list specified directly as for @code{vprintf}.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int obstack_vprintf (struct obstack *@var{obstack}, const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtsrace{:obstack} @mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acucorrupt{} @acsmem{}}}
@c The obstack is not guarded by mutexes, it might be at an inconsistent
@c state within a signal handler, and it could be left at an
@c inconsistent state in case of cancellation.
The @code{obstack_vprintf} function is the equivalent of
@code{obstack_printf} with the variable argument list specified directly
as for @code{vprintf}.@refill
@end deftypefun

Here's an example showing how you might use @code{vfprintf}.  This is a
function that prints error messages to the stream @code{stderr}, along
with a prefix indicating the name of the program
(@pxref{Error Messages}, for a description of
@code{program_invocation_short_name}).

@smallexample
@group
#include <stdio.h>
#include <stdarg.h>

void
eprintf (const char *template, ...)
@{
  va_list ap;
  extern char *program_invocation_short_name;

  fprintf (stderr, "%s: ", program_invocation_short_name);
  va_start (ap, template);
  vfprintf (stderr, template, ap);
  va_end (ap);
@}
@end group
@end smallexample

@noindent
You could call @code{eprintf} like this:

@smallexample
eprintf ("file `%s' does not exist\n", filename);
@end smallexample

In GNU C, there is a special construct you can use to let the compiler
know that a function uses a @code{printf}-style format string.  Then it
can check the number and types of arguments in each call to the
function, and warn you when they do not match the format string.
For example, take this declaration of @code{eprintf}:

@smallexample
void eprintf (const char *template, ...)
	__attribute__ ((format (printf, 1, 2)));
@end smallexample

@noindent
This tells the compiler that @code{eprintf} uses a format string like
@code{printf} (as opposed to @code{scanf}; @pxref{Formatted Input});
the format string appears as the first argument;
and the arguments to satisfy the format begin with the second.
@xref{Function Attributes, , Declaring Attributes of Functions,
gcc.info, Using GNU CC}, for more information.

@node Parsing a Template String
@subsection Parsing a Template String
@cindex parsing a template string

You can use the function @code{parse_printf_format} to obtain
information about the number and types of arguments that are expected by
a given template string.  This function permits interpreters that
provide interfaces to @code{printf} to avoid passing along invalid
arguments from the user's program, which could cause a crash.

All the symbols described in this section are declared in the header
file @file{printf.h}.

@comment printf.h
@comment GNU
@deftypefun size_t parse_printf_format (const char *@var{template}, size_t @var{n}, int *@var{argtypes})
@safety{@prelim{}@mtsafe{@mtslocale{}}@assafe{}@acsafe{}}
This function returns information about the number and types of
arguments expected by the @code{printf} template string @var{template}.
The information is stored in the array @var{argtypes}; each element of
this array describes one argument.  This information is encoded using
the various @samp{PA_} macros, listed below.

The argument @var{n} specifies the number of elements in the array
@var{argtypes}.  This is the maximum number of elements that
@code{parse_printf_format} will try to write.

@code{parse_printf_format} returns the total number of arguments required
by @var{template}.  If this number is greater than @var{n}, then the
information returned describes only the first @var{n} arguments.  If you
want information about additional arguments, allocate a bigger
array and call @code{parse_printf_format} again.
@end deftypefun

The argument types are encoded as a combination of a basic type and
modifier flag bits.

@comment printf.h
@comment GNU
@deftypevr Macro int PA_FLAG_MASK
This macro is a bitmask for the type modifier flag bits.  You can write
the expression @code{(argtypes[i] & PA_FLAG_MASK)} to extract just the
flag bits for an argument, or @code{(argtypes[i] & ~PA_FLAG_MASK)} to
extract just the basic type code.
@end deftypevr

Here are symbolic constants that represent the basic types; they stand
for integer values.

@vtable @code
@comment printf.h
@comment GNU
@item PA_INT
This specifies that the base type is @code{int}.

@comment printf.h
@comment GNU
@item PA_CHAR
This specifies that the base type is @code{int}, cast to @code{char}.

@comment printf.h
@comment GNU
@item PA_STRING
This specifies that the base type is @code{char *}, a null-terminated string.

@comment printf.h
@comment GNU
@item PA_POINTER
This specifies that the base type is @code{void *}, an arbitrary pointer.

@comment printf.h
@comment GNU
@item PA_FLOAT
This specifies that the base type is @code{float}.

@comment printf.h
@comment GNU
@item PA_DOUBLE
This specifies that the base type is @code{double}.

@comment printf.h
@comment GNU
@item PA_LAST
You can define additional base types for your own programs as offsets
from @code{PA_LAST}.  For example, if you have data types @samp{foo}
and @samp{bar} with their own specialized @code{printf} conversions,
you could define encodings for these types as:

@smallexample
#define PA_FOO  PA_LAST
#define PA_BAR  (PA_LAST + 1)
@end smallexample
@end vtable

Here are the flag bits that modify a basic type.  They are combined with
the code for the basic type using inclusive-or.

@vtable @code
@comment printf.h
@comment GNU
@item PA_FLAG_PTR
If this bit is set, it indicates that the encoded type is a pointer to
the base type, rather than an immediate value.
For example, @samp{PA_INT|PA_FLAG_PTR} represents the type @samp{int *}.

@comment printf.h
@comment GNU
@item PA_FLAG_SHORT
If this bit is set, it indicates that the base type is modified with
@code{short}.  (This corresponds to the @samp{h} type modifier.)

@comment printf.h
@comment GNU
@item PA_FLAG_LONG
If this bit is set, it indicates that the base type is modified with
@code{long}.  (This corresponds to the @samp{l} type modifier.)

@comment printf.h
@comment GNU
@item PA_FLAG_LONG_LONG
If this bit is set, it indicates that the base type is modified with
@code{long long}.  (This corresponds to the @samp{L} type modifier.)

@comment printf.h
@comment GNU
@item PA_FLAG_LONG_DOUBLE
This is a synonym for @code{PA_FLAG_LONG_LONG}, used by convention with
a base type of @code{PA_DOUBLE} to indicate a type of @code{long double}.
@end vtable

@ifinfo
For an example of using these facilities, see @ref{Example of Parsing}.
@end ifinfo

@node Example of Parsing
@subsection Example of Parsing a Template String

Here is an example of decoding argument types for a format string.  We
assume this is part of an interpreter which contains arguments of type
@code{NUMBER}, @code{CHAR}, @code{STRING} and @code{STRUCTURE} (and
perhaps others which are not valid here).

@smallexample
/* @r{Test whether the @var{nargs} specified objects}
   @r{in the vector @var{args} are valid}
   @r{for the format string @var{format}:}
   @r{if so, return 1.}
   @r{If not, return 0 after printing an error message.}  */

int
validate_args (char *format, int nargs, OBJECT *args)
@{
  int *argtypes;
  int nwanted;

  /* @r{Get the information about the arguments.}
     @r{Each conversion specification must be at least two characters}
     @r{long, so there cannot be more specifications than half the}
     @r{length of the string.}  */

  argtypes = (int *) alloca (strlen (format) / 2 * sizeof (int));
  nwanted = parse_printf_format (string, nelts, argtypes);

  /* @r{Check the number of arguments.}  */
  if (nwanted > nargs)
    @{
      error ("too few arguments (at least %d required)", nwanted);
      return 0;
    @}

  /* @r{Check the C type wanted for each argument}
     @r{and see if the object given is suitable.}  */
  for (i = 0; i < nwanted; i++)
    @{
      int wanted;

      if (argtypes[i] & PA_FLAG_PTR)
	wanted = STRUCTURE;
      else
	switch (argtypes[i] & ~PA_FLAG_MASK)
	  @{
	  case PA_INT:
	  case PA_FLOAT:
	  case PA_DOUBLE:
	    wanted = NUMBER;
	    break;
	  case PA_CHAR:
	    wanted = CHAR;
	    break;
	  case PA_STRING:
	    wanted = STRING;
	    break;
	  case PA_POINTER:
	    wanted = STRUCTURE;
	    break;
	  @}
      if (TYPE (args[i]) != wanted)
	@{
	  error ("type mismatch for arg number %d", i);
	  return 0;
	@}
    @}
  return 1;
@}
@end smallexample

@node Customizing Printf
@section Customizing @code{printf}
@cindex customizing @code{printf}
@cindex defining new @code{printf} conversions
@cindex extending @code{printf}

@Theglibc{} lets you define your own custom conversion specifiers
for @code{printf} template strings, to teach @code{printf} clever ways
to print the important data structures of your program.

The way you do this is by registering the conversion with the function
@code{register_printf_function}; see @ref{Registering New Conversions}.
One of the arguments you pass to this function is a pointer to a handler
function that produces the actual output; see @ref{Defining the Output
Handler}, for information on how to write this function.

You can also install a function that just returns information about the
number and type of arguments expected by the conversion specifier.
@xref{Parsing a Template String}, for information about this.

The facilities of this section are declared in the header file
@file{printf.h}.

@menu
* Registering New Conversions::         Using @code{register_printf_function}
					 to register a new output conversion.
* Conversion Specifier Options::        The handler must be able to get
					 the options specified in the
					 template when it is called.
* Defining the Output Handler::         Defining the handler and arginfo
					 functions that are passed as arguments
					 to @code{register_printf_function}.
* Printf Extension Example::            How to define a @code{printf}
					 handler function.
* Predefined Printf Handlers::          Predefined @code{printf} handlers.
@end menu

@strong{Portability Note:} The ability to extend the syntax of
@code{printf} template strings is a GNU extension.  ISO standard C has
nothing similar.

@node Registering New Conversions
@subsection Registering New Conversions

The function to register a new output conversion is
@code{register_printf_function}, declared in @file{printf.h}.
@pindex printf.h

@comment printf.h
@comment GNU
@deftypefun int register_printf_function (int @var{spec}, printf_function @var{handler-function}, printf_arginfo_function @var{arginfo-function})
@safety{@prelim{}@mtunsafe{@mtasuconst{:printfext}}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@acsmem{} @aculock{}}}
@c This function is guarded by the global non-recursive libc lock, but
@c users of the variables it sets aren't, and those should be MT-Safe,
@c so we're ruling out the use of this extension with threads.  Calling
@c it from a signal handler may self-deadlock, and cancellation may
@c leave the lock held, besides leaking allocated memory.
This function defines the conversion specifier character @var{spec}.
Thus, if @var{spec} is @code{'Y'}, it defines the conversion @samp{%Y}.
You can redefine the built-in conversions like @samp{%s}, but flag
characters like @samp{#} and type modifiers like @samp{l} can never be
used as conversions; calling @code{register_printf_function} for those
characters has no effect.  It is advisable not to use lowercase letters,
since the ISO C standard warns that additional lowercase letters may be
standardized in future editions of the standard.

The @var{handler-function} is the function called by @code{printf} and
friends when this conversion appears in a template string.
@xref{Defining the Output Handler}, for information about how to define
a function to pass as this argument.  If you specify a null pointer, any
existing handler function for @var{spec} is removed.

The @var{arginfo-function} is the function called by
@code{parse_printf_format} when this conversion appears in a
template string.  @xref{Parsing a Template String}, for information
about this.

@c The following is not true anymore.  The `parse_printf_format' function
@c is now also called from `vfprintf' via `parse_one_spec'.
@c --drepper@gnu, 1996/11/14
@c
@c Normally, you install both functions for a conversion at the same time,
@c but if you are never going to call @code{parse_printf_format}, you do
@c not need to define an arginfo function.

@strong{Attention:} In @theglibc{} versions before 2.0 the
@var{arginfo-function} function did not need to be installed unless
the user used the @code{parse_printf_format} function.  This has changed.
Now a call to any of the @code{printf} functions will call this
function when this format specifier appears in the format string.

The return value is @code{0} on success, and @code{-1} on failure
(which occurs if @var{spec} is out of range).

You can redefine the standard output conversions, but this is probably
not a good idea because of the potential for confusion.  Library routines
written by other people could break if you do this.
@end deftypefun

@node Conversion Specifier Options
@subsection Conversion Specifier Options

If you define a meaning for @samp{%A}, what if the template contains
@samp{%+23A} or @samp{%-#A}?  To implement a sensible meaning for these,
the handler when called needs to be able to get the options specified in
the template.

Both the @var{handler-function} and @var{arginfo-function} accept an
argument that points to a @code{struct printf_info}, which contains
information about the options appearing in an instance of the conversion
specifier.  This data type is declared in the header file
@file{printf.h}.
@pindex printf.h

@comment printf.h
@comment GNU
@deftp {Type} {struct printf_info}
This structure is used to pass information about the options appearing
in an instance of a conversion specifier in a @code{printf} template
string to the handler and arginfo functions for that specifier.  It
contains the following members:

@table @code
@item int prec
This is the precision specified.  The value is @code{-1} if no precision
was specified.  If the precision was given as @samp{*}, the
@code{printf_info} structure passed to the handler function contains the
actual value retrieved from the argument list.  But the structure passed
to the arginfo function contains a value of @code{INT_MIN}, since the
actual value is not known.

@item int width
This is the minimum field width specified.  The value is @code{0} if no
width was specified.  If the field width was given as @samp{*}, the
@code{printf_info} structure passed to the handler function contains the
actual value retrieved from the argument list.  But the structure passed
to the arginfo function contains a value of @code{INT_MIN}, since the
actual value is not known.

@item wchar_t spec
This is the conversion specifier character specified.  It's stored in
the structure so that you can register the same handler function for
multiple characters, but still have a way to tell them apart when the
handler function is called.

@item unsigned int is_long_double
This is a boolean that is true if the @samp{L}, @samp{ll}, or @samp{q}
type modifier was specified.  For integer conversions, this indicates
@code{long long int}, as opposed to @code{long double} for floating
point conversions.

@item unsigned int is_char
This is a boolean that is true if the @samp{hh} type modifier was specified.

@item unsigned int is_short
This is a boolean that is true if the @samp{h} type modifier was specified.

@item unsigned int is_long
This is a boolean that is true if the @samp{l} type modifier was specified.

@item unsigned int alt
This is a boolean that is true if the @samp{#} flag was specified.

@item unsigned int space
This is a boolean that is true if the @samp{ } flag was specified.

@item unsigned int left
This is a boolean that is true if the @samp{-} flag was specified.

@item unsigned int showsign
This is a boolean that is true if the @samp{+} flag was specified.

@item unsigned int group
This is a boolean that is true if the @samp{'} flag was specified.

@item unsigned int extra
This flag has a special meaning depending on the context.  It could
be used freely by the user-defined handlers but when called from
the @code{printf} function this variable always contains the value
@code{0}.

@item unsigned int wide
This flag is set if the stream is wide oriented.

@item wchar_t pad
This is the character to use for padding the output to the minimum field
width.  The value is @code{'0'} if the @samp{0} flag was specified, and
@code{' '} otherwise.
@end table
@end deftp


@node Defining the Output Handler
@subsection Defining the Output Handler

Now let's look at how to define the handler and arginfo functions
which are passed as arguments to @code{register_printf_function}.

@strong{Compatibility Note:} The interface changed in @theglibc{}
version 2.0.  Previously the third argument was of type
@code{va_list *}.

You should define your handler functions with a prototype like:

@smallexample
int @var{function} (FILE *stream, const struct printf_info *info,
		    const void *const *args)
@end smallexample

The @var{stream} argument passed to the handler function is the stream to
which it should write output.

The @var{info} argument is a pointer to a structure that contains
information about the various options that were included with the
conversion in the template string.  You should not modify this structure
inside your handler function.  @xref{Conversion Specifier Options}, for
a description of this data structure.

@c The following changes some time back.  --drepper@gnu, 1996/11/14
@c
@c The @code{ap_pointer} argument is used to pass the tail of the variable
@c argument list containing the values to be printed to your handler.
@c Unlike most other functions that can be passed an explicit variable
@c argument list, this is a @emph{pointer} to a @code{va_list}, rather than
@c the @code{va_list} itself.  Thus, you should fetch arguments by
@c means of @code{va_arg (*ap_pointer, @var{type})}.
@c
@c (Passing a pointer here allows the function that calls your handler
@c function to update its own @code{va_list} variable to account for the
@c arguments that your handler processes.  @xref{Variadic Functions}.)

The @var{args} is a vector of pointers to the arguments data.
The number of arguments was determined by calling the argument
information function provided by the user.

Your handler function should return a value just like @code{printf}
does: it should return the number of characters it has written, or a
negative value to indicate an error.

@comment printf.h
@comment GNU
@deftp {Data Type} printf_function
This is the data type that a handler function should have.
@end deftp

If you are going to use @w{@code{parse_printf_format}} in your
application, you must also define a function to pass as the
@var{arginfo-function} argument for each new conversion you install with
@code{register_printf_function}.

You have to define these functions with a prototype like:

@smallexample
int @var{function} (const struct printf_info *info,
		    size_t n, int *argtypes)
@end smallexample

The return value from the function should be the number of arguments the
conversion expects.  The function should also fill in no more than
@var{n} elements of the @var{argtypes} array with information about the
types of each of these arguments.  This information is encoded using the
various @samp{PA_} macros.  (You will notice that this is the same
calling convention @code{parse_printf_format} itself uses.)

@comment printf.h
@comment GNU
@deftp {Data Type} printf_arginfo_function
This type is used to describe functions that return information about
the number and type of arguments used by a conversion specifier.
@end deftp

@node Printf Extension Example
@subsection @code{printf} Extension Example

Here is an example showing how to define a @code{printf} handler function.
This program defines a data structure called a @code{Widget} and
defines the @samp{%W} conversion to print information about @w{@code{Widget *}}
arguments, including the pointer value and the name stored in the data
structure.  The @samp{%W} conversion supports the minimum field width and
left-justification options, but ignores everything else.

@smallexample
@include rprintf.c.texi
@end smallexample

The output produced by this program looks like:

@smallexample
|<Widget 0xffeffb7c: mywidget>|
|      <Widget 0xffeffb7c: mywidget>|
|<Widget 0xffeffb7c: mywidget>      |
@end smallexample

@node Predefined Printf Handlers
@subsection Predefined @code{printf} Handlers

@Theglibc{} also contains a concrete and useful application of the
@code{printf} handler extension.  There are two functions available
which implement a special way to print floating-point numbers.

@comment printf.h
@comment GNU
@deftypefun int printf_size (FILE *@var{fp}, const struct printf_info *@var{info}, const void *const *@var{args})
@safety{@prelim{}@mtsafe{@mtsrace{:fp} @mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @acucorrupt{}}}
@c This is meant to be called by vfprintf, that should hold the lock on
@c the stream, but if this function is called directly, output will be
@c racy, besides the uses of the global locale object while other
@c threads may be changing it and the possbility of leaving the stream
@c object in an inconsistent state in case of cancellation.
Print a given floating point number as for the format @code{%f} except
that there is a postfix character indicating the divisor for the
number to make this less than 1000.  There are two possible divisors:
powers of 1024 or powers of 1000.  Which one is used depends on the
format character specified while registered this handler.  If the
character is of lower case, 1024 is used.  For upper case characters,
1000 is used.

The postfix tag corresponds to bytes, kilobytes, megabytes, gigabytes,
etc.  The full table is:

@ifinfo
@multitable {' '} {2^10 (1024)} {zetta} {Upper} {10^24 (1000)}
@item low @tab Multiplier  @tab From  @tab Upper @tab Multiplier
@item ' ' @tab 1           @tab       @tab ' '   @tab 1
@item k   @tab 2^10 (1024) @tab kilo  @tab K     @tab 10^3 (1000)
@item m   @tab 2^20        @tab mega  @tab M     @tab 10^6
@item g   @tab 2^30        @tab giga  @tab G     @tab 10^9
@item t   @tab 2^40        @tab tera  @tab T     @tab 10^12
@item p   @tab 2^50        @tab peta  @tab P     @tab 10^15
@item e   @tab 2^60        @tab exa   @tab E     @tab 10^18
@item z   @tab 2^70        @tab zetta @tab Z     @tab 10^21
@item y   @tab 2^80        @tab yotta @tab Y     @tab 10^24
@end multitable
@end ifinfo
@iftex
@tex
\hbox to\hsize{\hfil\vbox{\offinterlineskip
\hrule
\halign{\strut#& \vrule#\tabskip=1em plus2em& {\tt#}\hfil& \vrule#& #\hfil& \vrule#& #\hfil& \vrule#& {\tt#}\hfil& \vrule#& #\hfil& \vrule#\tabskip=0pt\cr
\noalign{\hrule}
\omit&height2pt&\omit&&\omit&&\omit&&\omit&&\omit&\cr
&& \omit low && Multiplier && From && \omit Upper && Multiplier &\cr
\omit&height2pt&\omit&&\omit&&\omit&&\omit&&\omit&\cr
\noalign{\hrule}
&& {\tt\char32} &&  1 && && {\tt\char32} && 1 &\cr
&& k && $2^{10} = 1024$ && kilo && K && $10^3 = 1000$ &\cr
&& m && $2^{20}$ && mega && M && $10^6$ &\cr
&& g && $2^{30}$ && giga && G && $10^9$ &\cr
&& t && $2^{40}$ && tera && T && $10^{12}$ &\cr
&& p && $2^{50}$ && peta && P && $10^{15}$ &\cr
&& e && $2^{60}$ && exa && E && $10^{18}$ &\cr
&& z && $2^{70}$ && zetta && Z && $10^{21}$ &\cr
&& y && $2^{80}$ && yotta && Y && $10^{24}$ &\cr
\noalign{\hrule}}}\hfil}
@end tex
@end iftex

The default precision is 3, i.e., 1024 is printed with a lower-case
format character as if it were @code{%.3fk} and will yield @code{1.000k}.
@end deftypefun

Due to the requirements of @code{register_printf_function} we must also
provide the function which returns information about the arguments.

@comment printf.h
@comment GNU
@deftypefun int printf_size_info (const struct printf_info *@var{info}, size_t @var{n}, int *@var{argtypes})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
This function will return in @var{argtypes} the information about the
used parameters in the way the @code{vfprintf} implementation expects
it.  The format always takes one argument.
@end deftypefun

To use these functions both functions must be registered with a call like

@smallexample
register_printf_function ('B', printf_size, printf_size_info);
@end smallexample

Here we register the functions to print numbers as powers of 1000 since
the format character @code{'B'} is an upper-case character.  If we
would additionally use @code{'b'} in a line like

@smallexample
register_printf_function ('b', printf_size, printf_size_info);
@end smallexample

@noindent
we could also print using a power of 1024.  Please note that all that is
different in these two lines is the format specifier.  The
@code{printf_size} function knows about the difference between lower and upper
case format specifiers.

The use of @code{'B'} and @code{'b'} is no coincidence.  Rather it is
the preferred way to use this functionality since it is available on
some other systems which also use format specifiers.

@node Formatted Input
@section Formatted Input

@cindex formatted input from a stream
@cindex reading from a stream, formatted
@cindex format string, for @code{scanf}
@cindex template, for @code{scanf}
The functions described in this section (@code{scanf} and related
functions) provide facilities for formatted input analogous to the
formatted output facilities.  These functions provide a mechanism for
reading arbitrary values under the control of a @dfn{format string} or
@dfn{template string}.

@menu
* Formatted Input Basics::      Some basics to get you started.
* Input Conversion Syntax::     Syntax of conversion specifications.
* Table of Input Conversions::  Summary of input conversions and what they do.
* Numeric Input Conversions::   Details of conversions for reading numbers.
* String Input Conversions::    Details of conversions for reading strings.
* Dynamic String Input::	String conversions that @code{malloc} the buffer.
* Other Input Conversions::     Details of miscellaneous other conversions.
* Formatted Input Functions::   Descriptions of the actual functions.
* Variable Arguments Input::    @code{vscanf} and friends.
@end menu

@node Formatted Input Basics
@subsection Formatted Input Basics

Calls to @code{scanf} are superficially similar to calls to
@code{printf} in that arbitrary arguments are read under the control of
a template string.  While the syntax of the conversion specifications in
the template is very similar to that for @code{printf}, the
interpretation of the template is oriented more towards free-format
input and simple pattern matching, rather than fixed-field formatting.
For example, most @code{scanf} conversions skip over any amount of
``white space'' (including spaces, tabs, and newlines) in the input
file, and there is no concept of precision for the numeric input
conversions as there is for the corresponding output conversions.
Ordinarily, non-whitespace characters in the template are expected to
match characters in the input stream exactly, but a matching failure is
distinct from an input error on the stream.
@cindex conversion specifications (@code{scanf})

Another area of difference between @code{scanf} and @code{printf} is
that you must remember to supply pointers rather than immediate values
as the optional arguments to @code{scanf}; the values that are read are
stored in the objects that the pointers point to.  Even experienced
programmers tend to forget this occasionally, so if your program is
getting strange errors that seem to be related to @code{scanf}, you
might want to double-check this.

When a @dfn{matching failure} occurs, @code{scanf} returns immediately,
leaving the first non-matching character as the next character to be
read from the stream.  The normal return value from @code{scanf} is the
number of values that were assigned, so you can use this to determine if
a matching error happened before all the expected values were read.
@cindex matching failure, in @code{scanf}

The @code{scanf} function is typically used for things like reading in
the contents of tables.  For example, here is a function that uses
@code{scanf} to initialize an array of @code{double}:

@smallexample
void
readarray (double *array, int n)
@{
  int i;
  for (i=0; i<n; i++)
    if (scanf (" %lf", &(array[i])) != 1)
      invalid_input_error ();
@}
@end smallexample

The formatted input functions are not used as frequently as the
formatted output functions.  Partly, this is because it takes some care
to use them properly.  Another reason is that it is difficult to recover
from a matching error.

If you are trying to read input that doesn't match a single, fixed
pattern, you may be better off using a tool such as Flex to generate a
lexical scanner, or Bison to generate a parser, rather than using
@code{scanf}.  For more information about these tools, see @ref{Top, , ,
flex.info, Flex: The Lexical Scanner Generator}, and @ref{Top, , ,
bison.info, The Bison Reference Manual}.

@node Input Conversion Syntax
@subsection Input Conversion Syntax

A @code{scanf} template string is a string that contains ordinary
multibyte characters interspersed with conversion specifications that
start with @samp{%}.

Any whitespace character (as defined by the @code{isspace} function;
@pxref{Classification of Characters}) in the template causes any number
of whitespace characters in the input stream to be read and discarded.
The whitespace characters that are matched need not be exactly the same
whitespace characters that appear in the template string.  For example,
write @samp{ , } in the template to recognize a comma with optional
whitespace before and after.

Other characters in the template string that are not part of conversion
specifications must match characters in the input stream exactly; if
this is not the case, a matching failure occurs.

The conversion specifications in a @code{scanf} template string
have the general form:

@smallexample
% @var{flags} @var{width} @var{type} @var{conversion}
@end smallexample

In more detail, an input conversion specification consists of an initial
@samp{%} character followed in sequence by:

@itemize @bullet
@item
An optional @dfn{flag character} @samp{*}, which says to ignore the text
read for this specification.  When @code{scanf} finds a conversion
specification that uses this flag, it reads input as directed by the
rest of the conversion specification, but it discards this input, does
not use a pointer argument, and does not increment the count of
successful assignments.
@cindex flag character (@code{scanf})

@item
An optional flag character @samp{a} (valid with string conversions only)
which requests allocation of a buffer long enough to store the string in.
(This is a GNU extension.)
@xref{Dynamic String Input}.

@item
An optional decimal integer that specifies the @dfn{maximum field
width}.  Reading of characters from the input stream stops either when
this maximum is reached or when a non-matching character is found,
whichever happens first.  Most conversions discard initial whitespace
characters (those that don't are explicitly documented), and these
discarded characters don't count towards the maximum field width.
String input conversions store a null character to mark the end of the
input; the maximum field width does not include this terminator.
@cindex maximum field width (@code{scanf})

@item
An optional @dfn{type modifier character}.  For example, you can
specify a type modifier of @samp{l} with integer conversions such as
@samp{%d} to specify that the argument is a pointer to a @code{long int}
rather than a pointer to an @code{int}.
@cindex type modifier character (@code{scanf})

@item
A character that specifies the conversion to be applied.
@end itemize

The exact options that are permitted and how they are interpreted vary
between the different conversion specifiers.  See the descriptions of the
individual conversions for information about the particular options that
they allow.

With the @samp{-Wformat} option, the GNU C compiler checks calls to
@code{scanf} and related functions.  It examines the format string and
verifies that the correct number and types of arguments are supplied.
There is also a GNU C syntax to tell the compiler that a function you
write uses a @code{scanf}-style format string.
@xref{Function Attributes, , Declaring Attributes of Functions,
gcc.info, Using GNU CC}, for more information.

@node Table of Input Conversions
@subsection Table of Input Conversions
@cindex input conversions, for @code{scanf}

Here is a table that summarizes the various conversion specifications:

@table @asis
@item @samp{%d}
Matches an optionally signed integer written in decimal.  @xref{Numeric
Input Conversions}.

@item @samp{%i}
Matches an optionally signed integer in any of the formats that the C
language defines for specifying an integer constant.  @xref{Numeric
Input Conversions}.

@item @samp{%o}
Matches an unsigned integer written in octal radix.
@xref{Numeric Input Conversions}.

@item @samp{%u}
Matches an unsigned integer written in decimal radix.
@xref{Numeric Input Conversions}.

@item @samp{%x}, @samp{%X}
Matches an unsigned integer written in hexadecimal radix.
@xref{Numeric Input Conversions}.

@item @samp{%e}, @samp{%f}, @samp{%g}, @samp{%E}, @samp{%G}
Matches an optionally signed floating-point number.  @xref{Numeric Input
Conversions}.

@item @samp{%s}

Matches a string containing only non-whitespace characters.
@xref{String Input Conversions}.  The presence of the @samp{l} modifier
determines whether the output is stored as a wide character string or a
multibyte string.  If @samp{%s} is used in a wide character function the
string is converted as with multiple calls to @code{wcrtomb} into a
multibyte string.  This means that the buffer must provide room for
@code{MB_CUR_MAX} bytes for each wide character read.  In case
@samp{%ls} is used in a multibyte function the result is converted into
wide characters as with multiple calls of @code{mbrtowc} before being
stored in the user provided buffer.

@item @samp{%S}
This is an alias for @samp{%ls} which is supported for compatibility
with the Unix standard.

@item @samp{%[}
Matches a string of characters that belong to a specified set.
@xref{String Input Conversions}.  The presence of the @samp{l} modifier
determines whether the output is stored as a wide character string or a
multibyte string.  If @samp{%[} is used in a wide character function the
string is converted as with multiple calls to @code{wcrtomb} into a
multibyte string.  This means that the buffer must provide room for
@code{MB_CUR_MAX} bytes for each wide character read.  In case
@samp{%l[} is used in a multibyte function the result is converted into
wide characters as with multiple calls of @code{mbrtowc} before being
stored in the user provided buffer.

@item @samp{%c}
Matches a string of one or more characters; the number of characters
read is controlled by the maximum field width given for the conversion.
@xref{String Input Conversions}.

If the @samp{%c} is used in a wide stream function the read value is
converted from a wide character to the corresponding multibyte character
before storing it.  Note that this conversion can produce more than one
byte of output and therefore the provided buffer be large enough for up
to @code{MB_CUR_MAX} bytes for each character.  If @samp{%lc} is used in
a multibyte function the input is treated as a multibyte sequence (and
not bytes) and the result is converted as with calls to @code{mbrtowc}.

@item @samp{%C}
This is an alias for @samp{%lc} which is supported for compatibility
with the Unix standard.

@item @samp{%p}
Matches a pointer value in the same implementation-defined format used
by the @samp{%p} output conversion for @code{printf}.  @xref{Other Input
Conversions}.

@item @samp{%n}
This conversion doesn't read any characters; it records the number of
characters read so far by this call.  @xref{Other Input Conversions}.

@item @samp{%%}
This matches a literal @samp{%} character in the input stream.  No
corresponding argument is used.  @xref{Other Input Conversions}.
@end table

If the syntax of a conversion specification is invalid, the behavior is
undefined.  If there aren't enough function arguments provided to supply
addresses for all the conversion specifications in the template strings
that perform assignments, or if the arguments are not of the correct
types, the behavior is also undefined.  On the other hand, extra
arguments are simply ignored.

@node Numeric Input Conversions
@subsection Numeric Input Conversions

This section describes the @code{scanf} conversions for reading numeric
values.

The @samp{%d} conversion matches an optionally signed integer in decimal
radix.  The syntax that is recognized is the same as that for the
@code{strtol} function (@pxref{Parsing of Integers}) with the value
@code{10} for the @var{base} argument.

The @samp{%i} conversion matches an optionally signed integer in any of
the formats that the C language defines for specifying an integer
constant.  The syntax that is recognized is the same as that for the
@code{strtol} function (@pxref{Parsing of Integers}) with the value
@code{0} for the @var{base} argument.  (You can print integers in this
syntax with @code{printf} by using the @samp{#} flag character with the
@samp{%x}, @samp{%o}, or @samp{%d} conversion.  @xref{Integer Conversions}.)

For example, any of the strings @samp{10}, @samp{0xa}, or @samp{012}
could be read in as integers under the @samp{%i} conversion.  Each of
these specifies a number with decimal value @code{10}.

The @samp{%o}, @samp{%u}, and @samp{%x} conversions match unsigned
integers in octal, decimal, and hexadecimal radices, respectively.  The
syntax that is recognized is the same as that for the @code{strtoul}
function (@pxref{Parsing of Integers}) with the appropriate value
(@code{8}, @code{10}, or @code{16}) for the @var{base} argument.

The @samp{%X} conversion is identical to the @samp{%x} conversion.  They
both permit either uppercase or lowercase letters to be used as digits.

The default type of the corresponding argument for the @code{%d} and
@code{%i} conversions is @code{int *}, and @code{unsigned int *} for the
other integer conversions.  You can use the following type modifiers to
specify other sizes of integer:

@table @samp
@item hh
Specifies that the argument is a @code{signed char *} or @code{unsigned
char *}.

This modifier was introduced in @w{ISO C99}.

@item h
Specifies that the argument is a @code{short int *} or @code{unsigned
short int *}.

@item j
Specifies that the argument is a @code{intmax_t *} or @code{uintmax_t *}.

This modifier was introduced in @w{ISO C99}.

@item l
Specifies that the argument is a @code{long int *} or @code{unsigned
long int *}.  Two @samp{l} characters is like the @samp{L} modifier, below.

If used with @samp{%c} or @samp{%s} the corresponding parameter is
considered as a pointer to a wide character or wide character string
respectively.  This use of @samp{l} was introduced in @w{Amendment 1} to
@w{ISO C90}.

@need 100
@item ll
@itemx L
@itemx q
Specifies that the argument is a @code{long long int *} or @code{unsigned long long int *}.  (The @code{long long} type is an extension supported by the
GNU C compiler.  For systems that don't provide extra-long integers, this
is the same as @code{long int}.)

The @samp{q} modifier is another name for the same thing, which comes
from 4.4 BSD; a @w{@code{long long int}} is sometimes called a ``quad''
@code{int}.

@item t
Specifies that the argument is a @code{ptrdiff_t *}.

This modifier was introduced in @w{ISO C99}.

@item z
Specifies that the argument is a @code{size_t *}.

This modifier was introduced in @w{ISO C99}.
@end table

All of the @samp{%e}, @samp{%f}, @samp{%g}, @samp{%E}, and @samp{%G}
input conversions are interchangeable.  They all match an optionally
signed floating point number, in the same syntax as for the
@code{strtod} function (@pxref{Parsing of Floats}).

For the floating-point input conversions, the default argument type is
@code{float *}.  (This is different from the corresponding output
conversions, where the default type is @code{double}; remember that
@code{float} arguments to @code{printf} are converted to @code{double}
by the default argument promotions, but @code{float *} arguments are
not promoted to @code{double *}.)  You can specify other sizes of float
using these type modifiers:

@table @samp
@item l
Specifies that the argument is of type @code{double *}.

@item L
Specifies that the argument is of type @code{long double *}.
@end table

For all the above number parsing formats there is an additional optional
flag @samp{'}.  When this flag is given the @code{scanf} function
expects the number represented in the input string to be formatted
according to the grouping rules of the currently selected locale
(@pxref{General Numeric}).

If the @code{"C"} or @code{"POSIX"} locale is selected there is no
difference.  But for a locale which specifies values for the appropriate
fields in the locale the input must have the correct form in the input.
Otherwise the longest prefix with a correct form is processed.

@node String Input Conversions
@subsection String Input Conversions

This section describes the @code{scanf} input conversions for reading
string and character values: @samp{%s}, @samp{%S}, @samp{%[}, @samp{%c},
and @samp{%C}.

You have two options for how to receive the input from these
conversions:

@itemize @bullet
@item
Provide a buffer to store it in.  This is the default.  You should
provide an argument of type @code{char *} or @code{wchar_t *} (the
latter of the @samp{l} modifier is present).

@strong{Warning:} To make a robust program, you must make sure that the
input (plus its terminating null) cannot possibly exceed the size of the
buffer you provide.  In general, the only way to do this is to specify a
maximum field width one less than the buffer size.  @strong{If you
provide the buffer, always specify a maximum field width to prevent
overflow.}

@item
Ask @code{scanf} to allocate a big enough buffer, by specifying the
@samp{a} flag character.  This is a GNU extension.  You should provide
an argument of type @code{char **} for the buffer address to be stored
in.  @xref{Dynamic String Input}.
@end itemize

The @samp{%c} conversion is the simplest: it matches a fixed number of
characters, always.  The maximum field width says how many characters to
read; if you don't specify the maximum, the default is 1.  This
conversion doesn't append a null character to the end of the text it
reads.  It also does not skip over initial whitespace characters.  It
reads precisely the next @var{n} characters, and fails if it cannot get
that many.  Since there is always a maximum field width with @samp{%c}
(whether specified, or 1 by default), you can always prevent overflow by
making the buffer long enough.
@comment Is character == byte here???  --drepper

If the format is @samp{%lc} or @samp{%C} the function stores wide
characters which are converted using the conversion determined at the
time the stream was opened from the external byte stream.  The number of
bytes read from the medium is limited by @code{MB_CUR_LEN * @var{n}} but
at most @var{n} wide character get stored in the output string.

The @samp{%s} conversion matches a string of non-whitespace characters.
It skips and discards initial whitespace, but stops when it encounters
more whitespace after having read something.  It stores a null character
at the end of the text that it reads.

For example, reading the input:

@smallexample
 hello, world
@end smallexample

@noindent
with the conversion @samp{%10c} produces @code{" hello, wo"}, but
reading the same input with the conversion @samp{%10s} produces
@code{"hello,"}.

@strong{Warning:} If you do not specify a field width for @samp{%s},
then the number of characters read is limited only by where the next
whitespace character appears.  This almost certainly means that invalid
input can make your program crash---which is a bug.

The @samp{%ls} and @samp{%S} format are handled just like @samp{%s}
except that the external byte sequence is converted using the conversion
associated with the stream to wide characters with their own encoding.
A width or precision specified with the format do not directly determine
how many bytes are read from the stream since they measure wide
characters.  But an upper limit can be computed by multiplying the value
of the width or precision by @code{MB_CUR_MAX}.

To read in characters that belong to an arbitrary set of your choice,
use the @samp{%[} conversion.  You specify the set between the @samp{[}
character and a following @samp{]} character, using the same syntax used
in regular expressions for explicit sets of characters.  As special cases:

@itemize @bullet
@item
A literal @samp{]} character can be specified as the first character
of the set.

@item
An embedded @samp{-} character (that is, one that is not the first or
last character of the set) is used to specify a range of characters.

@item
If a caret character @samp{^} immediately follows the initial @samp{[},
then the set of allowed input characters is the everything @emph{except}
the characters listed.
@end itemize

The @samp{%[} conversion does not skip over initial whitespace
characters.

Note that the @dfn{character class} syntax available in character sets
that appear inside regular expressions (such as @samp{[:alpha:]}) is
@emph{not} available in the @samp{%[} conversion.

Here are some examples of @samp{%[} conversions and what they mean:

@table @samp
@item %25[1234567890]
Matches a string of up to 25 digits.

@item %25[][]
Matches a string of up to 25 square brackets.

@item %25[^ \f\n\r\t\v]
Matches a string up to 25 characters long that doesn't contain any of
the standard whitespace characters.  This is slightly different from
@samp{%s}, because if the input begins with a whitespace character,
@samp{%[} reports a matching failure while @samp{%s} simply discards the
initial whitespace.

@item %25[a-z]
Matches up to 25 lowercase characters.
@end table

As for @samp{%c} and @samp{%s} the @samp{%[} format is also modified to
produce wide characters if the @samp{l} modifier is present.  All what
is said about @samp{%ls} above is true for @samp{%l[}.

One more reminder: the @samp{%s} and @samp{%[} conversions are
@strong{dangerous} if you don't specify a maximum width or use the
@samp{a} flag, because input too long would overflow whatever buffer you
have provided for it.  No matter how long your buffer is, a user could
supply input that is longer.  A well-written program reports invalid
input with a comprehensible error message, not with a crash.

@node Dynamic String Input
@subsection Dynamically Allocating String Conversions

A GNU extension to formatted input lets you safely read a string with no
maximum size.  Using this feature, you don't supply a buffer; instead,
@code{scanf} allocates a buffer big enough to hold the data and gives
you its address.  To use this feature, write @samp{a} as a flag
character, as in @samp{%as} or @samp{%a[0-9a-z]}.

The pointer argument you supply for where to store the input should have
type @code{char **}.  The @code{scanf} function allocates a buffer and
stores its address in the word that the argument points to.  You should
free the buffer with @code{free} when you no longer need it.

Here is an example of using the @samp{a} flag with the @samp{%[@dots{}]}
conversion specification to read a ``variable assignment'' of the form
@samp{@var{variable} = @var{value}}.

@smallexample
@{
  char *variable, *value;

  if (2 > scanf ("%a[a-zA-Z0-9] = %a[^\n]\n",
		 &variable, &value))
    @{
      invalid_input_error ();
      return 0;
    @}

  @dots{}
@}
@end smallexample

@node Other Input Conversions
@subsection Other Input Conversions

This section describes the miscellaneous input conversions.

The @samp{%p} conversion is used to read a pointer value.  It recognizes
the same syntax used by the @samp{%p} output conversion for
@code{printf} (@pxref{Other Output Conversions}); that is, a hexadecimal
number just as the @samp{%x} conversion accepts.  The corresponding
argument should be of type @code{void **}; that is, the address of a
place to store a pointer.

The resulting pointer value is not guaranteed to be valid if it was not
originally written during the same program execution that reads it in.

The @samp{%n} conversion produces the number of characters read so far
by this call.  The corresponding argument should be of type @code{int *}.
This conversion works in the same way as the @samp{%n} conversion for
@code{printf}; see @ref{Other Output Conversions}, for an example.

The @samp{%n} conversion is the only mechanism for determining the
success of literal matches or conversions with suppressed assignments.
If the @samp{%n} follows the locus of a matching failure, then no value
is stored for it since @code{scanf} returns before processing the
@samp{%n}.  If you store @code{-1} in that argument slot before calling
@code{scanf}, the presence of @code{-1} after @code{scanf} indicates an
error occurred before the @samp{%n} was reached.

Finally, the @samp{%%} conversion matches a literal @samp{%} character
in the input stream, without using an argument.  This conversion does
not permit any flags, field width, or type modifier to be specified.

@node Formatted Input Functions
@subsection Formatted Input Functions

Here are the descriptions of the functions for performing formatted
input.
Prototypes for these functions are in the header file @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypefun int scanf (const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
The @code{scanf} function reads formatted input from the stream
@code{stdin} under the control of the template string @var{template}.
The optional arguments are pointers to the places which receive the
resulting values.

The return value is normally the number of successful assignments.  If
an end-of-file condition is detected before any matches are performed,
including matches against whitespace and literal characters in the
template, then @code{EOF} is returned.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int wscanf (const wchar_t *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
The @code{wscanf} function reads formatted input from the stream
@code{stdin} under the control of the template string @var{template}.
The optional arguments are pointers to the places which receive the
resulting values.

The return value is normally the number of successful assignments.  If
an end-of-file condition is detected before any matches are performed,
including matches against whitespace and literal characters in the
template, then @code{WEOF} is returned.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int fscanf (FILE *@var{stream}, const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is just like @code{scanf}, except that the input is read
from the stream @var{stream} instead of @code{stdin}.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int fwscanf (FILE *@var{stream}, const wchar_t *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is just like @code{wscanf}, except that the input is read
from the stream @var{stream} instead of @code{stdin}.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int sscanf (const char *@var{s}, const char *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is like @code{scanf}, except that the characters are taken from the
null-terminated string @var{s} instead of from a stream.  Reaching the
end of the string is treated as an end-of-file condition.

The behavior of this function is undefined if copying takes place
between objects that overlap---for example, if @var{s} is also given
as an argument to receive a string read under control of the @samp{%s},
@samp{%S}, or @samp{%[} conversion.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int swscanf (const wchar_t *@var{ws}, const wchar_t *@var{template}, @dots{})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is like @code{wscanf}, except that the characters are taken from the
null-terminated string @var{ws} instead of from a stream.  Reaching the
end of the string is treated as an end-of-file condition.

The behavior of this function is undefined if copying takes place
between objects that overlap---for example, if @var{ws} is also given as
an argument to receive a string read under control of the @samp{%s},
@samp{%S}, or @samp{%[} conversion.
@end deftypefun

@node Variable Arguments Input
@subsection Variable Arguments Input Functions

The functions @code{vscanf} and friends are provided so that you can
define your own variadic @code{scanf}-like functions that make use of
the same internals as the built-in formatted output functions.
These functions are analogous to the @code{vprintf} series of output
functions.  @xref{Variable Arguments Output}, for important
information on how to use them.

@strong{Portability Note:} The functions listed in this section were
introduced in @w{ISO C99} and were before available as GNU extensions.

@comment stdio.h
@comment ISO
@deftypefun int vscanf (const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is similar to @code{scanf}, but instead of taking
a variable number of arguments directly, it takes an argument list
pointer @var{ap} of type @code{va_list} (@pxref{Variadic Functions}).
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int vwscanf (const wchar_t *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This function is similar to @code{wscanf}, but instead of taking
a variable number of arguments directly, it takes an argument list
pointer @var{ap} of type @code{va_list} (@pxref{Variadic Functions}).
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int vfscanf (FILE *@var{stream}, const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This is the equivalent of @code{fscanf} with the variable argument list
specified directly as for @code{vscanf}.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int vfwscanf (FILE *@var{stream}, const wchar_t *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@asucorrupt{} @ascuheap{}}@acunsafe{@acsmem{} @aculock{} @acucorrupt{}}}
This is the equivalent of @code{fwscanf} with the variable argument list
specified directly as for @code{vwscanf}.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int vsscanf (const char *@var{s}, const char *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is the equivalent of @code{sscanf} with the variable argument list
specified directly as for @code{vscanf}.
@end deftypefun

@comment wchar.h
@comment ISO
@deftypefun int vswscanf (const wchar_t *@var{s}, const wchar_t *@var{template}, va_list @var{ap})
@safety{@prelim{}@mtsafe{@mtslocale{}}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This is the equivalent of @code{swscanf} with the variable argument list
specified directly as for @code{vwscanf}.
@end deftypefun

In GNU C, there is a special construct you can use to let the compiler
know that a function uses a @code{scanf}-style format string.  Then it
can check the number and types of arguments in each call to the
function, and warn you when they do not match the format string.
For details, see @ref{Function Attributes, , Declaring Attributes of Functions,
gcc.info, Using GNU CC}.

@node EOF and Errors
@section End-Of-File and Errors

@cindex end of file, on a stream
Many of the functions described in this chapter return the value of the
macro @code{EOF} to indicate unsuccessful completion of the operation.
Since @code{EOF} is used to report both end of file and random errors,
it's often better to use the @code{feof} function to check explicitly
for end of file and @code{ferror} to check for errors.  These functions
check indicators that are part of the internal state of the stream
object, indicators set if the appropriate condition was detected by a
previous I/O operation on that stream.

@comment stdio.h
@comment ISO
@deftypevr Macro int EOF
This macro is an integer value that is returned by a number of narrow
stream functions to indicate an end-of-file condition, or some other
error situation.  With @theglibc{}, @code{EOF} is @code{-1}.  In
other libraries, its value may be some other negative number.

This symbol is declared in @file{stdio.h}.
@end deftypevr

@comment wchar.h
@comment ISO
@deftypevr Macro int WEOF
This macro is an integer value that is returned by a number of wide
stream functions to indicate an end-of-file condition, or some other
error situation.  With @theglibc{}, @code{WEOF} is @code{-1}.  In
other libraries, its value may be some other negative number.

This symbol is declared in @file{wchar.h}.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypefun int feof (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acunsafe{@aculock{}}}
The @code{feof} function returns nonzero if and only if the end-of-file
indicator for the stream @var{stream} is set.

This symbol is declared in @file{stdio.h}.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int feof_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
@c There isn't much of a thread unsafety risk in reading a flag word and
@c testing a bit in it.
The @code{feof_unlocked} function is equivalent to the @code{feof}
function except that it does not implicitly lock the stream.

This function is a GNU extension.

This symbol is declared in @file{stdio.h}.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int ferror (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acunsafe{@aculock{}}}
The @code{ferror} function returns nonzero if and only if the error
indicator for the stream @var{stream} is set, indicating that an error
has occurred on a previous operation on the stream.

This symbol is declared in @file{stdio.h}.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun int ferror_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{ferror_unlocked} function is equivalent to the @code{ferror}
function except that it does not implicitly lock the stream.

This function is a GNU extension.

This symbol is declared in @file{stdio.h}.
@end deftypefun

In addition to setting the error indicator associated with the stream,
the functions that operate on streams also set @code{errno} in the same
way as the corresponding low-level functions that operate on file
descriptors.  For example, all of the functions that perform output to a
stream---such as @code{fputc}, @code{printf}, and @code{fflush}---are
implemented in terms of @code{write}, and all of the @code{errno} error
conditions defined for @code{write} are meaningful for these functions.
For more information about the descriptor-level I/O functions, see
@ref{Low-Level I/O}.

@node Error Recovery
@section Recovering from errors

You may explicitly clear the error and EOF flags with the @code{clearerr}
function.

@comment stdio.h
@comment ISO
@deftypefun void clearerr (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acunsafe{@aculock{}}}
This function clears the end-of-file and error indicators for the
stream @var{stream}.

The file positioning functions (@pxref{File Positioning}) also clear the
end-of-file indicator for the stream.
@end deftypefun

@comment stdio.h
@comment GNU
@deftypefun void clearerr_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@assafe{}@acsafe{}}
The @code{clearerr_unlocked} function is equivalent to the @code{clearerr}
function except that it does not implicitly lock the stream.

This function is a GNU extension.
@end deftypefun

Note that it is @emph{not} correct to just clear the error flag and retry
a failed stream operation.  After a failed write, any number of
characters since the last buffer flush may have been committed to the
file, while some buffered data may have been discarded.  Merely retrying
can thus cause lost or repeated data.

A failed read may leave the file pointer in an inappropriate position for
a second try.  In both cases, you should seek to a known position before
retrying.

Most errors that can happen are not recoverable --- a second try will
always fail again in the same way.  So usually it is best to give up and
report the error to the user, rather than install complicated recovery
logic.

One important exception is @code{EINTR} (@pxref{Interrupted Primitives}).
Many stream I/O implementations will treat it as an ordinary error, which
can be quite inconvenient.  You can avoid this hassle by installing all
signals with the @code{SA_RESTART} flag.

For similar reasons, setting nonblocking I/O on a stream's file
descriptor is not usually advisable.

@node Binary Streams
@section Text and Binary Streams

@gnusystems{} and other POSIX-compatible operating systems organize all
files as uniform sequences of characters.  However, some other systems
make a distinction between files containing text and files containing
binary data, and the input and output facilities of @w{ISO C} provide for
this distinction.  This section tells you how to write programs portable
to such systems.

@cindex text stream
@cindex binary stream
When you open a stream, you can specify either a @dfn{text stream} or a
@dfn{binary stream}.  You indicate that you want a binary stream by
specifying the @samp{b} modifier in the @var{opentype} argument to
@code{fopen}; see @ref{Opening Streams}.  Without this
option, @code{fopen} opens the file as a text stream.

Text and binary streams differ in several ways:

@itemize @bullet
@item
The data read from a text stream is divided into @dfn{lines} which are
terminated by newline (@code{'\n'}) characters, while a binary stream is
simply a long series of characters.  A text stream might on some systems
fail to handle lines more than 254 characters long (including the
terminating newline character).
@cindex lines (in a text file)

@item
On some systems, text files can contain only printing characters,
horizontal tab characters, and newlines, and so text streams may not
support other characters.  However, binary streams can handle any
character value.

@item
Space characters that are written immediately preceding a newline
character in a text stream may disappear when the file is read in again.

@item
More generally, there need not be a one-to-one mapping between
characters that are read from or written to a text stream, and the
characters in the actual file.
@end itemize

Since a binary stream is always more capable and more predictable than a
text stream, you might wonder what purpose text streams serve.  Why not
simply always use binary streams?  The answer is that on these operating
systems, text and binary streams use different file formats, and the
only way to read or write ``an ordinary file of text'' that can work
with other text-oriented programs is through a text stream.

In @theglibc{}, and on all POSIX systems, there is no difference
between text streams and binary streams.  When you open a stream, you
get the same kind of stream regardless of whether you ask for binary.
This stream can handle any file content, and has none of the
restrictions that text streams sometimes have.

@node File Positioning
@section File Positioning
@cindex file positioning on a stream
@cindex positioning a stream
@cindex seeking on a stream

The @dfn{file position} of a stream describes where in the file the
stream is currently reading or writing.  I/O on the stream advances the
file position through the file.  On @gnusystems{}, the file position is
represented as an integer, which counts the number of bytes from the
beginning of the file.  @xref{File Position}.

During I/O to an ordinary disk file, you can change the file position
whenever you wish, so as to read or write any portion of the file.  Some
other kinds of files may also permit this.  Files which support changing
the file position are sometimes referred to as @dfn{random-access}
files.

You can use the functions in this section to examine or modify the file
position indicator associated with a stream.  The symbols listed below
are declared in the header file @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypefun {long int} ftell (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function returns the current file position of the stream
@var{stream}.

This function can fail if the stream doesn't support file positioning,
or if the file position can't be represented in a @code{long int}, and
possibly for other reasons as well.  If a failure occurs, a value of
@code{-1} is returned.
@end deftypefun

@comment stdio.h
@comment Unix98
@deftypefun off_t ftello (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{ftello} function is similar to @code{ftell}, except that it
returns a value of type @code{off_t}.  Systems which support this type
use it to describe all file positions, unlike the POSIX specification
which uses a long int.  The two are not necessarily the same size.
Therefore, using ftell can lead to problems if the implementation is
written on top of a POSIX compliant low-level I/O implementation, and using
@code{ftello} is preferable whenever it is available.

If this function fails it returns @code{(off_t) -1}.  This can happen due
to missing support for file positioning or internal errors.  Otherwise
the return value is the current file position.

The function is an extension defined in the Unix Single Specification
version 2.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32 bit system this function is in fact @code{ftello64}.  I.e., the
LFS interface transparently replaces the old interface.
@end deftypefun

@comment stdio.h
@comment Unix98
@deftypefun off64_t ftello64 (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function is similar to @code{ftello} with the only difference that
the return value is of type @code{off64_t}.  This also requires that the
stream @var{stream} was opened using either @code{fopen64},
@code{freopen64}, or @code{tmpfile64} since otherwise the underlying
file operations to position the file pointer beyond the @twoexp{31}
bytes limit might fail.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a 32
bits machine this function is available under the name @code{ftello}
and so transparently replaces the old interface.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int fseek (FILE *@var{stream}, long int @var{offset}, int @var{whence})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{fseek} function is used to change the file position of the
stream @var{stream}.  The value of @var{whence} must be one of the
constants @code{SEEK_SET}, @code{SEEK_CUR}, or @code{SEEK_END}, to
indicate whether the @var{offset} is relative to the beginning of the
file, the current file position, or the end of the file, respectively.

This function returns a value of zero if the operation was successful,
and a nonzero value to indicate failure.  A successful call also clears
the end-of-file indicator of @var{stream} and discards any characters
that were ``pushed back'' by the use of @code{ungetc}.

@code{fseek} either flushes any buffered output before setting the file
position or else remembers it so it will be written later in its proper
place in the file.
@end deftypefun

@comment stdio.h
@comment Unix98
@deftypefun int fseeko (FILE *@var{stream}, off_t @var{offset}, int @var{whence})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function is similar to @code{fseek} but it corrects a problem with
@code{fseek} in a system with POSIX types.  Using a value of type
@code{long int} for the offset is not compatible with POSIX.
@code{fseeko} uses the correct type @code{off_t} for the @var{offset}
parameter.

For this reason it is a good idea to prefer @code{ftello} whenever it is
available since its functionality is (if different at all) closer the
underlying definition.

The functionality and return value is the same as for @code{fseek}.

The function is an extension defined in the Unix Single Specification
version 2.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32 bit system this function is in fact @code{fseeko64}.  I.e., the
LFS interface transparently replaces the old interface.
@end deftypefun

@comment stdio.h
@comment Unix98
@deftypefun int fseeko64 (FILE *@var{stream}, off64_t @var{offset}, int @var{whence})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function is similar to @code{fseeko} with the only difference that
the @var{offset} parameter is of type @code{off64_t}.  This also
requires that the stream @var{stream} was opened using either
@code{fopen64}, @code{freopen64}, or @code{tmpfile64} since otherwise
the underlying file operations to position the file pointer beyond the
@twoexp{31} bytes limit might fail.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a 32
bits machine this function is available under the name @code{fseeko}
and so transparently replaces the old interface.
@end deftypefun

@strong{Portability Note:} In non-POSIX systems, @code{ftell},
@code{ftello}, @code{fseek} and @code{fseeko} might work reliably only
on binary streams.  @xref{Binary Streams}.

The following symbolic constants are defined for use as the @var{whence}
argument to @code{fseek}.  They are also used with the @code{lseek}
function (@pxref{I/O Primitives}) and to specify offsets for file locks
(@pxref{Control Operations}).

@comment stdio.h
@comment ISO
@deftypevr Macro int SEEK_SET
This is an integer constant which, when used as the @var{whence}
argument to the @code{fseek} or @code{fseeko} function, specifies that
the offset provided is relative to the beginning of the file.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypevr Macro int SEEK_CUR
This is an integer constant which, when used as the @var{whence}
argument to the @code{fseek} or @code{fseeko} function, specifies that
the offset provided is relative to the current file position.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypevr Macro int SEEK_END
This is an integer constant which, when used as the @var{whence}
argument to the @code{fseek} or @code{fseeko} function, specifies that
the offset provided is relative to the end of the file.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypefun void rewind (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{rewind} function positions the stream @var{stream} at the
beginning of the file.  It is equivalent to calling @code{fseek} or
@code{fseeko} on the @var{stream} with an @var{offset} argument of
@code{0L} and a @var{whence} argument of @code{SEEK_SET}, except that
the return value is discarded and the error indicator for the stream is
reset.
@end deftypefun

These three aliases for the @samp{SEEK_@dots{}} constants exist for the
sake of compatibility with older BSD systems.  They are defined in two
different header files: @file{fcntl.h} and @file{sys/file.h}.

@table @code
@comment sys/file.h
@comment BSD
@item L_SET
@vindex L_SET
An alias for @code{SEEK_SET}.

@comment sys/file.h
@comment BSD
@item L_INCR
@vindex L_INCR
An alias for @code{SEEK_CUR}.

@comment sys/file.h
@comment BSD
@item L_XTND
@vindex L_XTND
An alias for @code{SEEK_END}.
@end table

@node Portable Positioning
@section Portable File-Position Functions

On @gnusystems{}, the file position is truly a character count.  You
can specify any character count value as an argument to @code{fseek} or
@code{fseeko} and get reliable results for any random access file.
However, some @w{ISO C} systems do not represent file positions in this
way.

On some systems where text streams truly differ from binary streams, it
is impossible to represent the file position of a text stream as a count
of characters from the beginning of the file.  For example, the file
position on some systems must encode both a record offset within the
file, and a character offset within the record.

As a consequence, if you want your programs to be portable to these
systems, you must observe certain rules:

@itemize @bullet
@item
The value returned from @code{ftell} on a text stream has no predictable
relationship to the number of characters you have read so far.  The only
thing you can rely on is that you can use it subsequently as the
@var{offset} argument to @code{fseek} or @code{fseeko} to move back to
the same file position.

@item
In a call to @code{fseek} or @code{fseeko} on a text stream, either the
@var{offset} must be zero, or @var{whence} must be @code{SEEK_SET} and
the @var{offset} must be the result of an earlier call to @code{ftell}
on the same stream.

@item
The value of the file position indicator of a text stream is undefined
while there are characters that have been pushed back with @code{ungetc}
that haven't been read or discarded.  @xref{Unreading}.
@end itemize

But even if you observe these rules, you may still have trouble for long
files, because @code{ftell} and @code{fseek} use a @code{long int} value
to represent the file position.  This type may not have room to encode
all the file positions in a large file.  Using the @code{ftello} and
@code{fseeko} functions might help here since the @code{off_t} type is
expected to be able to hold all file position values but this still does
not help to handle additional information which must be associated with
a file position.

So if you do want to support systems with peculiar encodings for the
file positions, it is better to use the functions @code{fgetpos} and
@code{fsetpos} instead.  These functions represent the file position
using the data type @code{fpos_t}, whose internal representation varies
from system to system.

These symbols are declared in the header file @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftp {Data Type} fpos_t
This is the type of an object that can encode information about the
file position of a stream, for use by the functions @code{fgetpos} and
@code{fsetpos}.

In @theglibc{}, @code{fpos_t} is an opaque data structure that
contains internal data to represent file offset and conversion state
information.  In other systems, it might have a different internal
representation.

When compiling with @code{_FILE_OFFSET_BITS == 64} on a 32 bit machine
this type is in fact equivalent to @code{fpos64_t} since the LFS
interface transparently replaces the old interface.
@end deftp

@comment stdio.h
@comment Unix98
@deftp {Data Type} fpos64_t
This is the type of an object that can encode information about the
file position of a stream, for use by the functions @code{fgetpos64} and
@code{fsetpos64}.

In @theglibc{}, @code{fpos64_t} is an opaque data structure that
contains internal data to represent file offset and conversion state
information.  In other systems, it might have a different internal
representation.
@end deftp

@comment stdio.h
@comment ISO
@deftypefun int fgetpos (FILE *@var{stream}, fpos_t *@var{position})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function stores the value of the file position indicator for the
stream @var{stream} in the @code{fpos_t} object pointed to by
@var{position}.  If successful, @code{fgetpos} returns zero; otherwise
it returns a nonzero value and stores an implementation-defined positive
value in @code{errno}.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32 bit system the function is in fact @code{fgetpos64}.  I.e., the LFS
interface transparently replaces the old interface.
@end deftypefun

@comment stdio.h
@comment Unix98
@deftypefun int fgetpos64 (FILE *@var{stream}, fpos64_t *@var{position})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function is similar to @code{fgetpos} but the file position is
returned in a variable of type @code{fpos64_t} to which @var{position}
points.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a 32
bits machine this function is available under the name @code{fgetpos}
and so transparently replaces the old interface.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypefun int fsetpos (FILE *@var{stream}, const fpos_t *@var{position})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function sets the file position indicator for the stream @var{stream}
to the position @var{position}, which must have been set by a previous
call to @code{fgetpos} on the same stream.  If successful, @code{fsetpos}
clears the end-of-file indicator on the stream, discards any characters
that were ``pushed back'' by the use of @code{ungetc}, and returns a value
of zero.  Otherwise, @code{fsetpos} returns a nonzero value and stores
an implementation-defined positive value in @code{errno}.

When the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a
32 bit system the function is in fact @code{fsetpos64}.  I.e., the LFS
interface transparently replaces the old interface.
@end deftypefun

@comment stdio.h
@comment Unix98
@deftypefun int fsetpos64 (FILE *@var{stream}, const fpos64_t *@var{position})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function is similar to @code{fsetpos} but the file position used
for positioning is provided in a variable of type @code{fpos64_t} to
which @var{position} points.

If the sources are compiled with @code{_FILE_OFFSET_BITS == 64} on a 32
bits machine this function is available under the name @code{fsetpos}
and so transparently replaces the old interface.
@end deftypefun

@node Stream Buffering
@section Stream Buffering

@cindex buffering of streams
Characters that are written to a stream are normally accumulated and
transmitted asynchronously to the file in a block, instead of appearing
as soon as they are output by the application program.  Similarly,
streams often retrieve input from the host environment in blocks rather
than on a character-by-character basis.  This is called @dfn{buffering}.

If you are writing programs that do interactive input and output using
streams, you need to understand how buffering works when you design the
user interface to your program.  Otherwise, you might find that output
(such as progress or prompt messages) doesn't appear when you intended
it to, or displays some other unexpected behavior.

This section deals only with controlling when characters are transmitted
between the stream and the file or device, and @emph{not} with how
things like echoing, flow control, and the like are handled on specific
classes of devices.  For information on common control operations on
terminal devices, see @ref{Low-Level Terminal Interface}.

You can bypass the stream buffering facilities altogether by using the
low-level input and output functions that operate on file descriptors
instead.  @xref{Low-Level I/O}.

@menu
* Buffering Concepts::          Terminology is defined here.
* Flushing Buffers::            How to ensure that output buffers are flushed.
* Controlling Buffering::       How to specify what kind of buffering to use.
@end menu

@node Buffering Concepts
@subsection Buffering Concepts

There are three different kinds of buffering strategies:

@itemize @bullet
@item
Characters written to or read from an @dfn{unbuffered} stream are
transmitted individually to or from the file as soon as possible.
@cindex unbuffered stream

@item
Characters written to a @dfn{line buffered} stream are transmitted to
the file in blocks when a newline character is encountered.
@cindex line buffered stream

@item
Characters written to or read from a @dfn{fully buffered} stream are
transmitted to or from the file in blocks of arbitrary size.
@cindex fully buffered stream
@end itemize

Newly opened streams are normally fully buffered, with one exception: a
stream connected to an interactive device such as a terminal is
initially line buffered.  @xref{Controlling Buffering}, for information
on how to select a different kind of buffering.  Usually the automatic
selection gives you the most convenient kind of buffering for the file
or device you open.

The use of line buffering for interactive devices implies that output
messages ending in a newline will appear immediately---which is usually
what you want.  Output that doesn't end in a newline might or might not
show up immediately, so if you want them to appear immediately, you
should flush buffered output explicitly with @code{fflush}, as described
in @ref{Flushing Buffers}.

@node Flushing Buffers
@subsection Flushing Buffers

@cindex flushing a stream
@dfn{Flushing} output on a buffered stream means transmitting all
accumulated characters to the file.  There are many circumstances when
buffered output on a stream is flushed automatically:

@itemize @bullet
@item
When you try to do output and the output buffer is full.

@item
When the stream is closed.  @xref{Closing Streams}.

@item
When the program terminates by calling @code{exit}.
@xref{Normal Termination}.

@item
When a newline is written, if the stream is line buffered.

@item
Whenever an input operation on @emph{any} stream actually reads data
from its file.
@end itemize

If you want to flush the buffered output at another time, call
@code{fflush}, which is declared in the header file @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypefun int fflush (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function causes any buffered output on @var{stream} to be delivered
to the file.  If @var{stream} is a null pointer, then
@code{fflush} causes buffered output on @emph{all} open output streams
to be flushed.

This function returns @code{EOF} if a write error occurs, or zero
otherwise.
@end deftypefun

@comment stdio.h
@comment POSIX
@deftypefun int fflush_unlocked (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{fflush_unlocked} function is equivalent to the @code{fflush}
function except that it does not implicitly lock the stream.
@end deftypefun

The @code{fflush} function can be used to flush all streams currently
opened.  While this is useful in some situations it does often more than
necessary since it might be done in situations when terminal input is
required and the program wants to be sure that all output is visible on
the terminal.  But this means that only line buffered streams have to be
flushed.  Solaris introduced a function especially for this.  It was
always available in @theglibc{} in some form but never officially
exported.

@comment stdio_ext.h
@comment GNU
@deftypefun void _flushlbf (void)
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
The @code{_flushlbf} function flushes all line buffered streams
currently opened.

This function is declared in the @file{stdio_ext.h} header.
@end deftypefun

@strong{Compatibility Note:} Some brain-damaged operating systems have
been known to be so thoroughly fixated on line-oriented input and output
that flushing a line buffered stream causes a newline to be written!
Fortunately, this ``feature'' seems to be becoming less common.  You do
not need to worry about this with @theglibc{}.

In some situations it might be useful to not flush the output pending
for a stream but instead simply forget it.  If transmission is costly
and the output is not needed anymore this is valid reasoning.  In this
situation a non-standard function introduced in Solaris and available in
@theglibc{} can be used.

@comment stdio_ext.h
@comment GNU
@deftypefun void __fpurge (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acunsafe{@acucorrupt{}}}
The @code{__fpurge} function causes the buffer of the stream
@var{stream} to be emptied.  If the stream is currently in read mode all
input in the buffer is lost.  If the stream is in output mode the
buffered output is not written to the device (or whatever other
underlying storage) and the buffer the cleared.

This function is declared in @file{stdio_ext.h}.
@end deftypefun

@node Controlling Buffering
@subsection Controlling Which Kind of Buffering

After opening a stream (but before any other operations have been
performed on it), you can explicitly specify what kind of buffering you
want it to have using the @code{setvbuf} function.
@cindex buffering, controlling

The facilities listed in this section are declared in the header
file @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment ISO
@deftypefun int setvbuf (FILE *@var{stream}, char *@var{buf}, int @var{mode}, size_t @var{size})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function is used to specify that the stream @var{stream} should
have the buffering mode @var{mode}, which can be either @code{_IOFBF}
(for full buffering), @code{_IOLBF} (for line buffering), or
@code{_IONBF} (for unbuffered input/output).

If you specify a null pointer as the @var{buf} argument, then @code{setvbuf}
allocates a buffer itself using @code{malloc}.  This buffer will be freed
when you close the stream.

Otherwise, @var{buf} should be a character array that can hold at least
@var{size} characters.  You should not free the space for this array as
long as the stream remains open and this array remains its buffer.  You
should usually either allocate it statically, or @code{malloc}
(@pxref{Unconstrained Allocation}) the buffer.  Using an automatic array
is not a good idea unless you close the file before exiting the block
that declares the array.

While the array remains a stream buffer, the stream I/O functions will
use the buffer for their internal purposes.  You shouldn't try to access
the values in the array directly while the stream is using it for
buffering.

The @code{setvbuf} function returns zero on success, or a nonzero value
if the value of @var{mode} is not valid or if the request could not
be honored.
@end deftypefun

@comment stdio.h
@comment ISO
@deftypevr Macro int _IOFBF
The value of this macro is an integer constant expression that can be
used as the @var{mode} argument to the @code{setvbuf} function to
specify that the stream should be fully buffered.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypevr Macro int _IOLBF
The value of this macro is an integer constant expression that can be
used as the @var{mode} argument to the @code{setvbuf} function to
specify that the stream should be line buffered.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypevr Macro int _IONBF
The value of this macro is an integer constant expression that can be
used as the @var{mode} argument to the @code{setvbuf} function to
specify that the stream should be unbuffered.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypevr Macro int BUFSIZ
The value of this macro is an integer constant expression that is good
to use for the @var{size} argument to @code{setvbuf}.  This value is
guaranteed to be at least @code{256}.

The value of @code{BUFSIZ} is chosen on each system so as to make stream
I/O efficient.  So it is a good idea to use @code{BUFSIZ} as the size
for the buffer when you call @code{setvbuf}.

Actually, you can get an even better value to use for the buffer size
by means of the @code{fstat} system call: it is found in the
@code{st_blksize} field of the file attributes.  @xref{Attribute Meanings}.

Sometimes people also use @code{BUFSIZ} as the allocation size of
buffers used for related purposes, such as strings used to receive a
line of input with @code{fgets} (@pxref{Character Input}).  There is no
particular reason to use @code{BUFSIZ} for this instead of any other
integer, except that it might lead to doing I/O in chunks of an
efficient size.
@end deftypevr

@comment stdio.h
@comment ISO
@deftypefun void setbuf (FILE *@var{stream}, char *@var{buf})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
If @var{buf} is a null pointer, the effect of this function is
equivalent to calling @code{setvbuf} with a @var{mode} argument of
@code{_IONBF}.  Otherwise, it is equivalent to calling @code{setvbuf}
with @var{buf}, and a @var{mode} of @code{_IOFBF} and a @var{size}
argument of @code{BUFSIZ}.

The @code{setbuf} function is provided for compatibility with old code;
use @code{setvbuf} in all new programs.
@end deftypefun

@comment stdio.h
@comment BSD
@deftypefun void setbuffer (FILE *@var{stream}, char *@var{buf}, size_t @var{size})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
If @var{buf} is a null pointer, this function makes @var{stream} unbuffered.
Otherwise, it makes @var{stream} fully buffered using @var{buf} as the
buffer.  The @var{size} argument specifies the length of @var{buf}.

This function is provided for compatibility with old BSD code.  Use
@code{setvbuf} instead.
@end deftypefun

@comment stdio.h
@comment BSD
@deftypefun void setlinebuf (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@asunsafe{@asucorrupt{}}@acunsafe{@aculock{} @acucorrupt{}}}
This function makes @var{stream} be line buffered, and allocates the
buffer for you.

This function is provided for compatibility with old BSD code.  Use
@code{setvbuf} instead.
@end deftypefun

It is possible to query whether a given stream is line buffered or not
using a non-standard function introduced in Solaris and available in
@theglibc{}.

@comment stdio_ext.h
@comment GNU
@deftypefun int __flbf (FILE *@var{stream})
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
The @code{__flbf} function will return a nonzero value in case the
stream @var{stream} is line buffered.  Otherwise the return value is
zero.

This function is declared in the @file{stdio_ext.h} header.
@end deftypefun

Two more extensions allow to determine the size of the buffer and how
much of it is used.  These functions were also introduced in Solaris.

@comment stdio_ext.h
@comment GNU
@deftypefun size_t __fbufsize (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acsafe{}}
The @code{__fbufsize} function return the size of the buffer in the
stream @var{stream}.  This value can be used to optimize the use of the
stream.

This function is declared in the @file{stdio_ext.h} header.
@end deftypefun

@comment stdio_ext.h
@comment GNU
@deftypefun size_t __fpending (FILE *@var{stream})
@safety{@prelim{}@mtsafe{@mtsrace{:stream}}@asunsafe{@asucorrupt{}}@acsafe{}}
The @code{__fpending}
function returns the number of bytes currently in the output buffer.
For wide-oriented stream the measuring unit is wide characters.  This
function should not be used on buffers in read mode or opened read-only.

This function is declared in the @file{stdio_ext.h} header.
@end deftypefun

@node Other Kinds of Streams
@section Other Kinds of Streams

@Theglibc{} provides ways for you to define additional kinds of
streams that do not necessarily correspond to an open file.

One such type of stream takes input from or writes output to a string.
These kinds of streams are used internally to implement the
@code{sprintf} and @code{sscanf} functions.  You can also create such a
stream explicitly, using the functions described in @ref{String Streams}.

More generally, you can define streams that do input/output to arbitrary
objects using functions supplied by your program.  This protocol is
discussed in @ref{Custom Streams}.

@strong{Portability Note:} The facilities described in this section are
specific to GNU.  Other systems or C implementations might or might not
provide equivalent functionality.

@menu
* String Streams::              Streams that get data from or put data in
				 a string or memory buffer.
* Custom Streams::              Defining your own streams with an arbitrary
				 input data source and/or output data sink.
@end menu

@node String Streams
@subsection String Streams

@cindex stream, for I/O to a string
@cindex string stream
The @code{fmemopen} and @code{open_memstream} functions allow you to do
I/O to a string or memory buffer.  These facilities are declared in
@file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment GNU
@deftypefun {FILE *} fmemopen (void *@var{buf}, size_t @var{size}, const char *@var{opentype})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@acsmem{} @aculock{}}}
@c Unlike open_memstream, fmemopen does (indirectly) call _IO_link_in,
@c bringing with it additional potential for async trouble with
@c list_all_lock.
This function opens a stream that allows the access specified by the
@var{opentype} argument, that reads from or writes to the buffer specified
by the argument @var{buf}.  This array must be at least @var{size} bytes long.

If you specify a null pointer as the @var{buf} argument, @code{fmemopen}
dynamically allocates an array @var{size} bytes long (as with @code{malloc};
@pxref{Unconstrained Allocation}).  This is really only useful
if you are going to write things to the buffer and then read them back
in again, because you have no way of actually getting a pointer to the
buffer (for this, try @code{open_memstream}, below).  The buffer is
freed when the stream is closed.

The argument @var{opentype} is the same as in @code{fopen}
(@pxref{Opening Streams}).  If the @var{opentype} specifies
append mode, then the initial file position is set to the first null
character in the buffer.  Otherwise the initial file position is at the
beginning of the buffer.

When a stream open for writing is flushed or closed, a null character
(zero byte) is written at the end of the buffer if it fits.  You
should add an extra byte to the @var{size} argument to account for this.
Attempts to write more than @var{size} bytes to the buffer result
in an error.

For a stream open for reading, null characters (zero bytes) in the
buffer do not count as ``end of file''.  Read operations indicate end of
file only when the file position advances past @var{size} bytes.  So, if
you want to read characters from a null-terminated string, you should
supply the length of the string as the @var{size} argument.
@end deftypefun

Here is an example of using @code{fmemopen} to create a stream for
reading from a string:

@smallexample
@include memopen.c.texi
@end smallexample

This program produces the following output:

@smallexample
Got f
Got o
Got o
Got b
Got a
Got r
@end smallexample

@comment stdio.h
@comment GNU
@deftypefun {FILE *} open_memstream (char **@var{ptr}, size_t *@var{sizeloc})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{}}@acunsafe{@acsmem{}}}
This function opens a stream for writing to a buffer.  The buffer is
allocated dynamically and grown as necessary, using @code{malloc}.
After you've closed the stream, this buffer is your responsibility to
clean up using @code{free} or @code{realloc}.  @xref{Unconstrained Allocation}.

When the stream is closed with @code{fclose} or flushed with
@code{fflush}, the locations @var{ptr} and @var{sizeloc} are updated to
contain the pointer to the buffer and its size.  The values thus stored
remain valid only as long as no further output on the stream takes
place.  If you do more output, you must flush the stream again to store
new values before you use them again.

A null character is written at the end of the buffer.  This null character
is @emph{not} included in the size value stored at @var{sizeloc}.

You can move the stream's file position with @code{fseek} or
@code{fseeko} (@pxref{File Positioning}).  Moving the file position past
the end of the data already written fills the intervening space with
zeroes.
@end deftypefun

Here is an example of using @code{open_memstream}:

@smallexample
@include memstrm.c.texi
@end smallexample

This program produces the following output:

@smallexample
buf = `hello', size = 5
buf = `hello, world', size = 12
@end smallexample

@node Custom Streams
@subsection Programming Your Own Custom Streams
@cindex custom streams
@cindex programming your own streams

This section describes how you can make a stream that gets input from an
arbitrary data source or writes output to an arbitrary data sink
programmed by you.  We call these @dfn{custom streams}.  The functions
and types described here are all GNU extensions.

@c !!! this does not talk at all about the higher-level hooks

@menu
* Streams and Cookies::         The @dfn{cookie} records where to fetch or
				 store data that is read or written.
* Hook Functions::              How you should define the four @dfn{hook
				 functions} that a custom stream needs.
@end menu

@node Streams and Cookies
@subsubsection Custom Streams and Cookies
@cindex cookie, for custom stream

Inside every custom stream is a special object called the @dfn{cookie}.
This is an object supplied by you which records where to fetch or store
the data read or written.  It is up to you to define a data type to use
for the cookie.  The stream functions in the library never refer
directly to its contents, and they don't even know what the type is;
they record its address with type @code{void *}.

To implement a custom stream, you must specify @emph{how} to fetch or
store the data in the specified place.  You do this by defining
@dfn{hook functions} to read, write, change ``file position'', and close
the stream.  All four of these functions will be passed the stream's
cookie so they can tell where to fetch or store the data.  The library
functions don't know what's inside the cookie, but your functions will
know.

When you create a custom stream, you must specify the cookie pointer,
and also the four hook functions stored in a structure of type
@code{cookie_io_functions_t}.

These facilities are declared in @file{stdio.h}.
@pindex stdio.h

@comment stdio.h
@comment GNU
@deftp {Data Type} {cookie_io_functions_t}
This is a structure type that holds the functions that define the
communications protocol between the stream and its cookie.  It has
the following members:

@table @code
@item cookie_read_function_t *read
This is the function that reads data from the cookie.  If the value is a
null pointer instead of a function, then read operations on this stream
always return @code{EOF}.

@item cookie_write_function_t *write
This is the function that writes data to the cookie.  If the value is a
null pointer instead of a function, then data written to the stream is
discarded.

@item cookie_seek_function_t *seek
This is the function that performs the equivalent of file positioning on
the cookie.  If the value is a null pointer instead of a function, calls
to @code{fseek} or @code{fseeko} on this stream can only seek to
locations within the buffer; any attempt to seek outside the buffer will
return an @code{ESPIPE} error.

@item cookie_close_function_t *close
This function performs any appropriate cleanup on the cookie when
closing the stream.  If the value is a null pointer instead of a
function, nothing special is done to close the cookie when the stream is
closed.
@end table
@end deftp

@comment stdio.h
@comment GNU
@deftypefun {FILE *} fopencookie (void *@var{cookie}, const char *@var{opentype}, cookie_io_functions_t @var{io-functions})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@acsmem{} @aculock{}}}
This function actually creates the stream for communicating with the
@var{cookie} using the functions in the @var{io-functions} argument.
The @var{opentype} argument is interpreted as for @code{fopen};
see @ref{Opening Streams}.  (But note that the ``truncate on
open'' option is ignored.)  The new stream is fully buffered.

The @code{fopencookie} function returns the newly created stream, or a null
pointer in case of an error.
@end deftypefun

@node Hook Functions
@subsubsection Custom Stream Hook Functions
@cindex hook functions (of custom streams)

Here are more details on how you should define the four hook functions
that a custom stream needs.

You should define the function to read data from the cookie as:

@smallexample
ssize_t @var{reader} (void *@var{cookie}, char *@var{buffer}, size_t @var{size})
@end smallexample

This is very similar to the @code{read} function; see @ref{I/O
Primitives}.  Your function should transfer up to @var{size} bytes into
the @var{buffer}, and return the number of bytes read, or zero to
indicate end-of-file.  You can return a value of @code{-1} to indicate
an error.

You should define the function to write data to the cookie as:

@smallexample
ssize_t @var{writer} (void *@var{cookie}, const char *@var{buffer}, size_t @var{size})
@end smallexample

This is very similar to the @code{write} function; see @ref{I/O
Primitives}.  Your function should transfer up to @var{size} bytes from
the buffer, and return the number of bytes written.  You can return a
value of @code{0} to indicate an error.  You must not return any
negative value.

You should define the function to perform seek operations on the cookie
as:

@smallexample
int @var{seeker} (void *@var{cookie}, off64_t *@var{position}, int @var{whence})
@end smallexample

For this function, the @var{position} and @var{whence} arguments are
interpreted as for @code{fgetpos}; see @ref{Portable Positioning}.

After doing the seek operation, your function should store the resulting
file position relative to the beginning of the file in @var{position}.
Your function should return a value of @code{0} on success and @code{-1}
to indicate an error.

You should define the function to do cleanup operations on the cookie
appropriate for closing the stream as:

@smallexample
int @var{cleaner} (void *@var{cookie})
@end smallexample

Your function should return @code{-1} to indicate an error, and @code{0}
otherwise.

@comment stdio.h
@comment GNU
@deftp {Data Type} cookie_read_function_t
This is the data type that the read function for a custom stream should have.
If you declare the function as shown above, this is the type it will have.
@end deftp

@comment stdio.h
@comment GNU
@deftp {Data Type} cookie_write_function_t
The data type of the write function for a custom stream.
@end deftp

@comment stdio.h
@comment GNU
@deftp {Data Type} cookie_seek_function_t
The data type of the seek function for a custom stream.
@end deftp

@comment stdio.h
@comment GNU
@deftp {Data Type} cookie_close_function_t
The data type of the close function for a custom stream.
@end deftp

@ignore
Roland says:

@quotation
There is another set of functions one can give a stream, the
input-room and output-room functions.  These functions must
understand stdio internals.  To describe how to use these
functions, you also need to document lots of how stdio works
internally (which isn't relevant for other uses of stdio).
Perhaps I can write an interface spec from which you can write
good documentation.  But it's pretty complex and deals with lots
of nitty-gritty details.  I think it might be better to let this
wait until the rest of the manual is more done and polished.
@end quotation
@end ignore

@c ??? This section could use an example.


@node Formatted Messages
@section Formatted Messages
@cindex formatted messages

On systems which are based on System V messages of programs (especially
the system tools) are printed in a strict form using the @code{fmtmsg}
function.  The uniformity sometimes helps the user to interpret messages
and the strictness tests of the @code{fmtmsg} function ensure that the
programmer follows some minimal requirements.

@menu
* Printing Formatted Messages::   The @code{fmtmsg} function.
* Adding Severity Classes::       Add more severity classes.
* Example::                       How to use @code{fmtmsg} and @code{addseverity}.
@end menu


@node Printing Formatted Messages
@subsection Printing Formatted Messages

Messages can be printed to standard error and/or to the console.  To
select the destination the programmer can use the following two values,
bitwise OR combined if wanted, for the @var{classification} parameter of
@code{fmtmsg}:

@vtable @code
@item MM_PRINT
Display the message in standard error.
@item MM_CONSOLE
Display the message on the system console.
@end vtable

The erroneous piece of the system can be signalled by exactly one of the
following values which also is bitwise ORed with the
@var{classification} parameter to @code{fmtmsg}:

@vtable @code
@item MM_HARD
The source of the condition is some hardware.
@item MM_SOFT
The source of the condition is some software.
@item MM_FIRM
The source of the condition is some firmware.
@end vtable

A third component of the @var{classification} parameter to @code{fmtmsg}
can describe the part of the system which detects the problem.  This is
done by using exactly one of the following values:

@vtable @code
@item MM_APPL
The erroneous condition is detected by the application.
@item MM_UTIL
The erroneous condition is detected by a utility.
@item MM_OPSYS
The erroneous condition is detected by the operating system.
@end vtable

A last component of @var{classification} can signal the results of this
message.  Exactly one of the following values can be used:

@vtable @code
@item MM_RECOVER
It is a recoverable error.
@item MM_NRECOV
It is a non-recoverable error.
@end vtable

@comment fmtmsg.h
@comment XPG
@deftypefun int fmtmsg (long int @var{classification}, const char *@var{label}, int @var{severity}, const char *@var{text}, const char *@var{action}, const char *@var{tag})
@safety{@prelim{}@mtsafe{}@asunsafe{@asulock{}}@acsafe{}}
Display a message described by its parameters on the device(s) specified
in the @var{classification} parameter.  The @var{label} parameter
identifies the source of the message.  The string should consist of two
colon separated parts where the first part has not more than 10 and the
second part not more than 14 characters.  The @var{text} parameter
describes the condition of the error, the @var{action} parameter possible
steps to recover from the error and the @var{tag} parameter is a
reference to the online documentation where more information can be
found.  It should contain the @var{label} value and a unique
identification number.

Each of the parameters can be a special value which means this value
is to be omitted.  The symbolic names for these values are:

@vtable @code
@item MM_NULLLBL
Ignore @var{label} parameter.
@item MM_NULLSEV
Ignore @var{severity} parameter.
@item MM_NULLMC
Ignore @var{classification} parameter.  This implies that nothing is
actually printed.
@item MM_NULLTXT
Ignore @var{text} parameter.
@item MM_NULLACT
Ignore @var{action} parameter.
@item MM_NULLTAG
Ignore @var{tag} parameter.
@end vtable

There is another way certain fields can be omitted from the output to
standard error.  This is described below in the description of
environment variables influencing the behavior.

The @var{severity} parameter can have one of the values in the following
table:
@cindex severity class

@vtable @code
@item MM_NOSEV
Nothing is printed, this value is the same as @code{MM_NULLSEV}.
@item MM_HALT
This value is printed as @code{HALT}.
@item MM_ERROR
This value is printed as @code{ERROR}.
@item MM_WARNING
This value is printed as @code{WARNING}.
@item MM_INFO
This value is printed as @code{INFO}.
@end vtable

The numeric value of these five macros are between @code{0} and
@code{4}.  Using the environment variable @code{SEV_LEVEL} or using the
@code{addseverity} function one can add more severity levels with their
corresponding string to print.  This is described below
(@pxref{Adding Severity Classes}).

@noindent
If no parameter is ignored the output looks like this:

@smallexample
@var{label}: @var{severity-string}: @var{text}
TO FIX: @var{action} @var{tag}
@end smallexample

The colons, new line characters and the @code{TO FIX} string are
inserted if necessary, i.e., if the corresponding parameter is not
ignored.

This function is specified in the X/Open Portability Guide.  It is also
available on all systems derived from System V.

The function returns the value @code{MM_OK} if no error occurred.  If
only the printing to standard error failed, it returns @code{MM_NOMSG}.
If printing to the console fails, it returns @code{MM_NOCON}.  If
nothing is printed @code{MM_NOTOK} is returned.  Among situations where
all outputs fail this last value is also returned if a parameter value
is incorrect.
@end deftypefun

There are two environment variables which influence the behavior of
@code{fmtmsg}.  The first is @code{MSGVERB}.  It is used to control the
output actually happening on standard error (@emph{not} the console
output).  Each of the five fields can explicitly be enabled.  To do
this the user has to put the @code{MSGVERB} variable with a format like
the following in the environment before calling the @code{fmtmsg} function
the first time:

@smallexample
MSGVERB=@var{keyword}[:@var{keyword}[:@dots{}]]
@end smallexample

Valid @var{keyword}s are @code{label}, @code{severity}, @code{text},
@code{action}, and @code{tag}.  If the environment variable is not given
or is the empty string, a not supported keyword is given or the value is
somehow else invalid, no part of the message is masked out.

The second environment variable which influences the behavior of
@code{fmtmsg} is @code{SEV_LEVEL}.  This variable and the change in the
behavior of @code{fmtmsg} is not specified in the X/Open Portability
Guide.  It is available in System V systems, though.  It can be used to
introduce new severity levels.  By default, only the five severity levels
described above are available.  Any other numeric value would make
@code{fmtmsg} print nothing.

If the user puts @code{SEV_LEVEL} with a format like

@smallexample
SEV_LEVEL=[@var{description}[:@var{description}[:@dots{}]]]
@end smallexample

@noindent
in the environment of the process before the first call to
@code{fmtmsg}, where @var{description} has a value of the form

@smallexample
@var{severity-keyword},@var{level},@var{printstring}
@end smallexample

The @var{severity-keyword} part is not used by @code{fmtmsg} but it has
to be present.  The @var{level} part is a string representation of a
number.  The numeric value must be a number greater than 4.  This value
must be used in the @var{severity} parameter of @code{fmtmsg} to select
this class.  It is not possible to overwrite any of the predefined
classes.  The @var{printstring} is the string printed when a message of
this class is processed by @code{fmtmsg} (see above, @code{fmtsmg} does
not print the numeric value but instead the string representation).


@node Adding Severity Classes
@subsection Adding Severity Classes
@cindex severity class

There is another possibility to introduce severity classes besides using
the environment variable @code{SEV_LEVEL}.  This simplifies the task of
introducing new classes in a running program.  One could use the
@code{setenv} or @code{putenv} function to set the environment variable,
but this is toilsome.

@deftypefun int addseverity (int @var{severity}, const char *@var{string})
@safety{@prelim{}@mtsafe{}@asunsafe{@ascuheap{} @asulock{}}@acunsafe{@aculock{} @acsmem{}}}
This function allows the introduction of new severity classes which can be
addressed by the @var{severity} parameter of the @code{fmtmsg} function.
The @var{severity} parameter of @code{addseverity} must match the value
for the parameter with the same name of @code{fmtmsg}, and @var{string}
is the string printed in the actual messages instead of the numeric
value.

If @var{string} is @code{NULL} the severity class with the numeric value
according to @var{severity} is removed.

It is not possible to overwrite or remove one of the default severity
classes.  All calls to @code{addseverity} with @var{severity} set to one
of the values for the default classes will fail.

The return value is @code{MM_OK} if the task was successfully performed.
If the return value is @code{MM_NOTOK} something went wrong.  This could
mean that no more memory is available or a class is not available when
it has to be removed.

This function is not specified in the X/Open Portability Guide although
the @code{fmtsmg} function is.  It is available on System V systems.
@end deftypefun


@node Example
@subsection How to use @code{fmtmsg} and @code{addseverity}

Here is a simple example program to illustrate the use of the both
functions described in this section.

@smallexample
@include fmtmsgexpl.c.texi
@end smallexample

The second call to @code{fmtmsg} illustrates a use of this function as
it usually occurs on System V systems, which heavily use this function.
It seems worthwhile to give a short explanation here of how this system
works on System V.  The value of the
@var{label} field (@code{UX:cat}) says that the error occurred in the
Unix program @code{cat}.  The explanation of the error follows and the
value for the @var{action} parameter is @code{"refer to manual"}.  One
could be more specific here, if necessary.  The @var{tag} field contains,
as proposed above, the value of the string given for the @var{label}
parameter, and additionally a unique ID (@code{001} in this case).  For
a GNU environment this string could contain a reference to the
corresponding node in the Info page for the program.

@noindent
Running this program without specifying the @code{MSGVERB} and
@code{SEV_LEVEL} function produces the following output:

@smallexample
UX:cat: NOTE2: invalid syntax
TO FIX: refer to manual UX:cat:001
@end smallexample

We see the different fields of the message and how the extra glue (the
colons and the @code{TO FIX} string) are printed.  But only one of the
three calls to @code{fmtmsg} produced output.  The first call does not
print anything because the @var{label} parameter is not in the correct
form.  The string must contain two fields, separated by a colon
(@pxref{Printing Formatted Messages}).  The third @code{fmtmsg} call
produced no output since the class with the numeric value @code{6} is
not defined.  Although a class with numeric value @code{5} is also not
defined by default, the call to @code{addseverity} introduces it and
the second call to @code{fmtmsg} produces the above output.

When we change the environment of the program to contain
@code{SEV_LEVEL=XXX,6,NOTE} when running it we get a different result:

@smallexample
UX:cat: NOTE2: invalid syntax
TO FIX: refer to manual UX:cat:001
label:foo: NOTE: text
TO FIX: action tag
@end smallexample

Now the third call to @code{fmtmsg} produced some output and we see how
the string @code{NOTE} from the environment variable appears in the
message.

Now we can reduce the output by specifying which fields we are
interested in.  If we additionally set the environment variable
@code{MSGVERB} to the value @code{severity:label:action} we get the
following output:

@smallexample
UX:cat: NOTE2
TO FIX: refer to manual
label:foo: NOTE
TO FIX: action
@end smallexample

@noindent
I.e., the output produced by the @var{text} and the @var{tag} parameters
to @code{fmtmsg} vanished.  Please also note that now there is no colon
after the @code{NOTE} and @code{NOTE2} strings in the output.  This is
not necessary since there is no more output on this line because the text
is missing.