NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c-part1/9933

Internet Message Format | 1996-08-05 | 4.1 KB

Path: newshost.lanl.gov!tanmoy
From: [e-mail] (Tanmoy Bhattacharya)
Newsgroups: comp.lang.c
Subject: Re: Help !!!!! on 'strstr"
Date: 14 Mar 1996 15:46:18 GMT
Organization: Los Alamos National Laboratory
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]>
    <[e-mail]> <[e-mail]>
NNTP-Posting-Host: qcd.lanl.gov
Mime-Version: 1.0
Content-Type: text
In-reply-to: Steve Fogoros's message of Wed, 13 Mar 1996 23:15:56 -0800

In article <[e-mail]>
Steve Fogoros <[e-mail]> writes:

SF: Tanmoy Bhattacharya wrote:
SF:
SF: > Any user function whose name begins with `is' followed by a lower case
SF: > letter violates the ANSI C standard.
SF:
SF: You mentioned the same kind of violation regarding user functions that start with
SF: 'str' followed by lower case letter in another post. I apologize that I question
SF: this without being intimately familiar with ANSI C Standard. I am having a hard time
SF: believing the standard would consider my use of other than reserved words a violation.

Belief depends on experience :-) As a fortran programmer, I had hard
time believing that a language _can_ have reserved words.

This is the name space issue: many implementations cannot link code
which defines the same names as in any standard library. The standard
makes a guarantee that the standard library shall not define or use
arbitrary names so that programs which follow the namespace rules in
the standard are portable. Thus, for example, if I name a function
`read', it cannot conflict with a library function name on any ANSI C
implementation where such conflict would cause a problem.

Some names are exempt from this guarantee (i.e. they do not belong to
the user name space). These typically start with underscores (the
detailed rules depend on exactly what context they are being used in),
but in addition, for symbols with external linkage, they include the
names of all library routines defined in the standard. Thus, for
example, you cannot name your routine `fprintf'. To allow for
extensions, some names are reserved in exactly the same way as
existing library functions (such as those beginning with `is' and a
lower case letter, or `to' and a lower case letter: presumably to
allow extensions whose names may be similar to the existing functions
declared in <ctype.h>, e.g. `islower' or `toupper').

By the way, we are only talking of symbols with external
linkage. Unless the appropriate header (<ctype.h> or <string.h> in the
cases under consideration) is included, the use of these names in
other contexts (e.g. static functions or variables, block scope
variables etc.) is perfectly okay.

Technically, the use of a name beiginning with `is' and a lower case
letter and with external linkage leads to a violation of a `shall'
condition in the standard, which is outside of any constraint section.

SF: (What kind of ticket would I get for an infraction? Is the implementation required to
SF: check for this and warn or error?)

It leads to undefined behaviour. So, the compiler is not required to
give a diagnostic: it may however fail to compile your code, compile
it and fail to run it to completion, compile and run it with wrong
results or display a variety of other behviours.

SF:
SF: Seriously, I will look this up because it seems strange to me and I didn't know
SF: anything about it. I quick scanned my K&R2 but did not find anything there.

ISO standard: 7.1.3 read in conjunction with 7.13.2. I do not have
K&R2 handy right now, and do not remember if they go into the
namespace issues.

Cheers
Tanmoy
--
[e-mail](128.165.23.46) DECNET: BETA::"[e-mail]"(1.218=1242)
Tanmoy Bhattacharya O:T-8(MS B285)LANL,NM87545 H:#9,3000,Trinity Drive,NM87544
Others see <gopher://yaleinfo.yale.edu:7700/00/Internet-People/internet-mail>,
<http://alpha.acast.nova.edu/cgi-bin/inmgq.pl>or<ftp://csd4.csd.uwm.edu/pub/
internetwork-mail-guide>. -- <http://nqcd.lanl.gov/people/tanmoy/tanmoy.html>
fax: [phone]   voice: 1 [phone]    [ Home: 1 [phone] ]