feof usage

[Mantorok Redgormor]

My professor that teaches C says that using feof to test if EOF is
reached while using a function like fgets on a file is okay. but i've
been told elsewhere that it is not okay. what is the case here? is it
okay or not?


nethlek

[Mac]

On Fri, 19 Sep 2003 19:11:20 +0000, Mantorok Redgormor wrote:

My professor that teaches C says that using feof to test if EOF is
reached while using a function like fgets on a file is okay. but i've
been told elsewhere that it is not okay. what is the case here? is it
okay or not?

The issue is that feof() only tells you AFTER you have attempted to read
past the end of the file. If you understand that, and code accordingly, it
is OK to use it, AFAIK.

That is, if a file has n characters, you can read all of them with fgets
and feof() will still return zero (false). But if you try to read n+1,
then fgets will still succeed, but feof() will return non-zero (true).

HTH

Mac
--

[Artie Gold]

Mantorok Redgormor wrote:

My professor that teaches C says that using feof to test if EOF is
reached while using a function like fgets on a file is okay. but i've
been told elsewhere that it is not okay. what is the case here? is it
okay or not?

It's OK -- but feof() only returns a non-zero (`true') value *after*
an attempted `read' has failed.

HTH,
--ag
--
Artie Gold -- Austin, Texas

[Jack Klein]

On 19 Sep 2003 19:11:20 -0700, ne*****@tokyo.c om (Mantorok Redgormor)
wrote in comp.lang.c:

My professor that teaches C says that using feof to test if EOF is
reached while using a function like fgets on a file is okay. but i've
been told elsewhere that it is not okay. what is the case here? is it
okay or not?

This is covered in the FAQ, specifically at
http://www.eskimo.com/~scs/C-faq/q12.2.html

Generally it is not a good idea to use feof() this way. It is not so
bad if you test for EOF after an input operation and before trying to
use the input data. But it can still cause problems, because an error
does not cause feof() to return a non-zero value.

There are all kinds of things that can cause an error in a file
operation. Hard drives are pretty reliable these days, but they are
still wear parts. Or a user could remove the media containing a file,
such as a floppy disk or CD. Or the network connection could go down.

All functions that can read from a file return a value that indicates
whether they succeeded or failed. fgets() returns the pointer you
passed it if it succeeded, or a null pointer if it failed. So to use
fgets() in a loop:

while (NULL != fgets(buffer, sizeof buffer, file_ptr))
{
/* ... */
}

After your input function indicates a failure, you can use feof() and
ferror() to determine whether you successfully read the entire file or
whether an error occurred.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://www.eskimo.com/~scs/C-faq/top.html
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.l earn.c-c++ ftp://snurse-l.org/pub/acllc-c++/faq

[Martin Ambuhl]

Mantorok Redgormor wrote:

My professor that teaches C says that using feof to test if EOF is
reached while using a function like fgets on a file is okay. but i've
been told elsewhere that it is not okay. what is the case here? is it
okay or not?

Sure, it's OK. But it is useless until EOF has already been reached, by
which time whatever you are using to read the file (fgets, fgetc, fread,
even fscanf) has already given an error indication. feof can be used to
sort out what kind of input error it was. You don't reach EOF on a read
that completely succeeds, so feof is useless then.
--
Martin Ambuhl

[Glen Herrmannsfeldt]

"Mantorok Redgormor" <ne*****@tokyo. com> wrote in message
news:41******** *************** ***@posting.goo gle.com...

My professor that teaches C says that using feof to test if EOF is
reached while using a function like fgets on a file is okay. but i've
been told elsewhere that it is not okay. what is the case here? is it
okay or not?

Most users of feof() have previously used Pascal, where it is required to
use feof() before reading to avoid a fatal error.

C requires an attempt to read past the end of file before feof() will
indicate it, so direct translation of Pascal feof() to C's feof() tend to
fail.

All C input functions signal the failure to do what they were asked to do,
and that tends to be a better way to do the test.

-- glen

[Mac]

On Sat, 20 Sep 2003 04:16:09 +0000, Martin Ambuhl wrote:

Mantorok Redgormor wrote:

My professor that teaches C says that using feof to test if EOF is
reached while using a function like fgets on a file is okay. but i've
been told elsewhere that it is not okay. what is the case here? is it
okay or not?

Sure, it's OK. But it is useless until EOF has already been reached, by
which time whatever you are using to read the file (fgets, fgetc, fread,
even fscanf) has already given an error indication. feof can be used to
sort out what kind of input error it was. You don't reach EOF on a read
that completely succeeds, so feof is useless then.

Actually, fgets doesn't provide error notification on reaching the end of
a file unless it wasn't able to read any characters, in which case it
returns NULL. It also returns NULL if there is an error reading from the
file.

Apart from that I agree with you. It seems that feof() is seldom useful,
except to distinguish between end of file and some other file error after
the fact.

Mac
--

[LibraryUser]

Mac wrote:

On Fri, 19 Sep 2003 19:11:20 +0000, Mantorok Redgormor wrote:

My professor that teaches C says that using feof to test if EOF
is reached while using a function like fgets on a file is okay.
but i've been told elsewhere that it is not okay. what is the
case here? is it okay or not?

The issue is that feof() only tells you AFTER you have attempted
to read past the end of the file. If you understand that, and
code accordingly, it is OK to use it, AFAIK.

That is, if a file has n characters, you can read all of them with
fgets and feof() will still return zero (false). But if you try to
read n+1, then fgets will still succeed, but feof() will return
non-zero (true).

Correction - that final fgets call will not succeed, it will
return NULL. This may describe either the end-of-file, or a read
error. feof() allows you to disambiguate them.

--
Replies should be to the newsgroup
Chuck Falconer, on vacation.

[LibraryUser]

Glen Herrmannsfeldt wrote:

"Mantorok Redgormor" <ne*****@tokyo. com> wrote in message

My professor that teaches C says that using feof to test if EOF
is reached while using a function like fgets on a file is okay.
but i've been told elsewhere that it is not okay. what is the
case here? is it okay or not?

Most users of feof() have previously used Pascal, where it is
required to use feof() before reading to avoid a fatal error.

C requires an attempt to read past the end of file before feof()
will indicate it, so direct translation of Pascal feof() to C's
feof() tend to fail.

All C input functions signal the failure to do what they were
asked to do, and that tends to be a better way to do the test.

The reason for the difference is that Pascal input is always
buffered, so that a look ahead can be implemented by examining
f^. In C, such lookahead involves the use of ungetc() and the
user needs to be highly aware of what is going on. In Pascal
OTOH use of the buffering can cause problems with interactive
input, leading to so-called lazy-io schemes, which in turn
require the user to be aware when using eof and eoln calls.

Neither scheme is perfect, but the trade-offs are different.

--
Replies should be to the newsgroup
Chuck Falconer, on vacation.

[Glen Herrmannsfeldt]

"LibraryUse r" <de**********@m ade.invalid> wrote in message
news:3F******** *******@made.in valid...

Glen Herrmannsfeldt wrote:

"Mantorok Redgormor" <ne*****@tokyo. com> wrote in message

My professor that teaches C says that using feof to test if EOF
is reached while using a function like fgets on a file is okay.
but i've been told elsewhere that it is not okay. what is the
case here? is it okay or not?
Most users of feof() have previously used Pascal, where it is
required to use feof() before reading to avoid a fatal error.

C requires an attempt to read past the end of file before feof()
will indicate it, so direct translation of Pascal feof() to C's
feof() tend to fail.

All C input functions signal the failure to do what they were
asked to do, and that tends to be a better way to do the test.

The reason for the difference is that Pascal input is always
buffered, so that a look ahead can be implemented by examining
f^. In C, such lookahead involves the use of ungetc() and the
user needs to be highly aware of what is going on. In Pascal
OTOH use of the buffering can cause problems with interactive
input, leading to so-called lazy-io schemes, which in turn
require the user to be aware when using eof and eoln calls.

Interactive input can be confusing in C, too.
Neither scheme is perfect, but the trade-offs are different.

Using the assumptions of one language in programming the other tends to show
those trade-offs.

-- glen