NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c-part2/14075

Internet Message Format | 1996-08-05 | 3.6 KB

Path: anvil.ugrad.cs.ubc.ca!not-for-mail
From: [e-mail] (Kazimir Kylheku)
Newsgroups: comp.lang.c,comp.unix.programmer
Subject: Re: Q: '\n' character
Date: 11 Apr 1996 10:40:00 -0700
Organization: Computer Science, University of B.C., Vancouver, B.C., Canada
Message-ID: <[e-mail]>
References: <[e-mail]> <4jtddt$[e-mail]> <[e-mail]> <[e-mail]>
NNTP-Posting-Host: anvil.ugrad.cs.ubc.ca

In article <[e-mail]>,
Bill Zissimopoulos  <[e-mail]> wrote:
>Stephen Baynes wrote:
>>
>> Spasmo ([e-mail]) wrote:
>> : Artur Wojdat ([e-mail]) wrote:
>> : : Hello everybody,
>> : :     Is there a function or some sort of way that I could remove '\n'
>> : : charecter form the end of the string. I'm reading from two files, want to
>> : Dunno if there are any functions available, but what I always do
>> : is just overwrite the '\n' with a '\0', which does the job nicely.
>>
>> : For example, let's say that the string that holds the data is called
>> : buf.  To get rid of the '\n' you'd merely do the following:
>>
>> :       buf[strlen(buf) - 1] = '\0';
>>
>> : And that takes care of that.
>>
>> Thats the best way, but do check that the character you are about to overwrite
>> is a newline. There are two occasions when fgets stops without reading a
>> newline, first if it has filled your buffer (because you have a very long line
>> in the file), second if there is no newline at the end of the last line of the
>> file.
>
>Which makes this the worst way for the exact reasons that you give.

In fact, if one can still trace back to the original article that started this
thread, any solutions to the problem involving buffers, string manipulationa nd
other buffoonery are completely wacky.

The original question, part of which is reveled in the quote above, calls for
a program that will read lines from two files and display them side by side on
the standard output---clearly a trivial problem.

This problem can be solved by scanning lines character by character alternating
between the the left and the right file, and putting them onto standard output
using just the getc() and putchar() macros. You read until a newline or EOF
from one file and print, but don't print the newline.  Then you do the same
thing from the second file,  and output a newline. If neither file is at EOF,
you repeat the process. If one file is at EOF but the other is not, you then
dump the rest of the remaining file to standard output.

Solutions to this problem involving strlen(), fgets() and other zoological
beasts show a lack of focus and clarity of thought.  I recommend a good dose of
meditation each day, a sound diet, and mandatory readings from a suitable bible
(a tome dedicated to some area of computer science or another, of course).

The fact is that the problem requires _zero_ input symbols of _lookahead_ in
either file, hence buffering is only needed for I/O efficiency, not for
isolating tokens (in this case lines). The efficiency buffering is already
done by the standard IO library. Further buffering in the application code (in
this case) is wasteful of memory and CPU cycles, and  results in an
implementation that is unnecessarily complex and hard to read.

For example, the check whether the result of fgets() has a newline or not is
completely wasteful, and is needed in order to hack what is a poor design in
the first place into working properly.   I agree with you completely about this
being a bad approach, if not the worst.
--