Re: w3c-libwww-5.2.8-7 bug report

Hi,

I don't have a Compaq C compiler to hand to ensure this :)
But changing both instances of:

start = start ? ++start : line;
to
start = start ? (start + 1) : line;

will get rid of the first two warnings, and is a more 'correct' use of the
conditional operator.  Plus it reads clearer, IMO.  I haven't got the code
preceeding the pCSUser case in front of me so I can't help with that one.

--
Iain


 -----Original Message-----
From: www-lib-request@w3.org [mailto:www-lib-request@w3.org]On Behalf Of
David Binderman
Sent: Monday, January 28, 2002 1:40 PM
To: www-lib@w3.org
Subject: w3c-libwww-5.2.8-7 bug report


Hello there,

In compiling package w3c-libwww-5.2.8-7, from Redhat 7.1,
the Compaq C compiler produced the following messages
1.
cc: Warning: HTGopher.c, line 343: In this statement, the expression
"start=start?++start:line" modifies the variable "start" more than once
without an intervening sequence point. This behavior is undefined.
(undefvarmod)
start = start ? ++start : line;
--------^
2.
cc: Warning: HTGopher.c, line 349: In this statement, the expression
"start=start?++start:line" modifies the variable "start" more than once
without an intervening sequence point. This behavior is undefined.
(undefvarmod)
start = start ? ++start : line;
--------^
3.
cc: Warning: HTGopher.c, line 402: In this statement, the expression
"start=start?++start:line" modifies the variable "start" more than once
without an intervening sequence point. This behavior is undefined.
(undefvarmod)
start = start ? ++start : line;
--------^
4.
cc: Warning: CSApp.c, line 382: The scalar variable "pCSUser" is fetched but
not initialized. And there may be other such fetches of this variable that
have not been reported in this compilation. (uninit1)
HTTRACE(PICS_TRACE, "PICS: User \"%s\" not found.\n" _
CSUser_name(pCSUser));
--------^
I'm afraid I have no fixes for any of these problems I have found.
Regards

--

David Binderman

Received on Tuesday, 29 January 2002 18:40:18 UTC