rcd@opus.UUCP (Dick Dunn) (09/20/85)
I'd like to poll you folks on a small matter of style. I've seen it
mentioned from time to time in lang.c. PLEASE email (DON'T POST) answers
to the questions below; I'll post responses and comments in a bit.
The issue is this: People tend to stumble over the difference between ==
and =. It's not that they don't know the difference; they just screw up,
typically writing something like
if (x=0) ...
when they mean
if (x==0) ...
Some people have suggested reversing the order of comparisons, viz:
if (0==x) ...
so that the compiler will complain if = is substituted for ==.
One argument against this form is that it is an unconventional order which
is harder to read.
- - - - - - - - - - - - - - - - - - - - - - - - - -
The questions:
1. Do you use this form of comparison?
2. Regardless of whether you use it, how do you feel about it (e.g., other
people using it in code that you will maintain)?
- dislike it
- neutral (OK if you want to do it)
- like it
3. If you use or like this way of writing comparisons, how do you feel
about also writing ordering tests with the constant first, as
if (0<x) ...
- dislike
- neutral
- like
Opinions welcome in addition to answers to the above; I'll summarize.
Again, please MAIL responses rather than posting them.
--
Dick Dunn {hao,ucbvax,allegra}!nbires!rcd (303)444-5710 x3086
...Lately it occurs to me what a long, strange trip it's been.jpm@BNL44.ARPA (John McNamee) (09/25/85)
[I tried to send this via mail to rcd%opus.uucp@BRL.ARPA, but it was rejected, so I'm posting it instead. The original article was asking about placing the constant on the left side of the equals sign in if tests (i.e. "if (0==strcmp(a,b))").] >1. Do you use this form of comparison? No. >2. Regardless of whether you use it, how do you feel about it (e.g., other > people using it in code that you will maintain)? > - dislike it > - neutral (OK if you want to do it) > - like it I refuse to work on code that is written that way. I once spent a whole day undoing that style from a program I had to make changes to. >3. If you use or like this way of writing comparisons, how do you feel > about also writing ordering tests with the constant first, as > if (0<x) ... > - dislike > - neutral > - like People who do that should be taken out and shot. No, on second thought that would be too quick. They should be tortured before being taken out and shot. Comment: What you do in private with consenting compilers is your own business, but if I'm going to have to maintain your code then you don't have the right to inflict your style on me. K&R style is readable, and is the standard. Other than special permission from God, there is no excuse for anything but K&R style. -- John McNamee ..!decvax!philabs!sbcs!bnl44!jpm jpm@BNL44.ARPA "MS-DOS is a communist plot"
alexis@reed.UUCP (Alexis Dimitriadis) (09/29/85)
This is in response to an article asking about the proposed style of comparisons where the constant is placed on the left, to eliminate typos like if (x=NULL) > >1. Do you use this form of comparison? > > No. As a matter of fact, I tried. I was rather intrigued when the subject came up, so I thought I'd give it a go. One pleasant side effect is things like if (NULL == (fp = fopen("foo", "r"))) scream(); The above looks, to me at least, more readable than the other leading product. The approach seems to work OK when testing for equality, but I felt that consistency required that I use it in all tests. However, to reverse if (foo > 0) I have to write if (0 < foo) which is more than a little twisted. It may be my conditioning, but I _do_not immediately perceive the two as equivalent. I stuck with it, but it required extra effort to reverse the operators. Another problem was, not all comparisons involve a hard constant. Do we reverse if (i == some.external->variable) to convey the fact that we are testing the value of i? And so on. A few weeks later I had to modify the program, and since I did not at the time feel like going through the same, I put everything back the way God meant it to be. Too bad, but mixing the two styles is a complete no-no. If there weren't all that existing code that we all need to look at, this style might stand a chance. It may have a chance if used for equality tests only, but that is not as orthogonal. I hope someone cared about my experience with this silly, but interesting style. -- Alexis Dimitriadis -- _______________________________________________ As soon as I get a full time job, the opinions expressed above will attach themselves to my employer, who will never be rid of them again. alexis @ reed {decvax,ihnp4,ucbcad,uw-beaver}!tektronix!reed.UUCP