rc-list - mailing list for the rc(1) shell
 help / color / mirror / Atom feed
* Re: ifs substitution, how should it work?
@ 2000-05-18  6:32 Byron Rakitzis
  0 siblings, 0 replies; 10+ messages in thread
From: Byron Rakitzis @ 2000-05-18  6:32 UTC (permalink / raw)
  To: bengt, rc, tjg

> From: Bengt Kleberg <bengt@softwell.se>
> Subject: Re: ifs substitution, how should it work?
>
> Re: ifs substitution, how should it work?
>
> > What would you expect
>
> >     ; ifs=(' ' '<tab>' '
> '
>
> > to do (where `<tab>' represents a literal tab)?
>
>
> I had but seen code with lists as ifs arguments, and therefore thought
> that rc was using lists.
> Ie, I would have written it as:
>
> ; ifs=(' ' '<tab>' '
> ')
>
> Would it create much trouble to change rc :-)

I am not sure what the problem is...

At the moment rc scans ifs as a list and as a sequence of characters.
i.e., the implementation is as follows:

[from glom.c:bqinput]

        memzero(isifs, sizeof isifs);
        for (isifs['\0'] = TRUE; ifs != NULL; ifs = ifs->n)
                for (s = ifs->w; *s != '\0'; s++)
                        isifs[*(unsigned char *)s] = TRUE;

So,

	ifs='abc'

and

	ifs=(a b c)

should accomplish the same thing.

Isn't that what you want?

Byron.


^ permalink raw reply	[flat|nested] 10+ messages in thread
* Re: ifs substitution, how should it work?
@ 2000-05-22  7:55 Bengt Kleberg
  2000-05-22  8:53 ` Tim Goodwin
  0 siblings, 1 reply; 10+ messages in thread
From: Bengt Kleberg @ 2000-05-22  7:55 UTC (permalink / raw)
  To: rc, rsc, tjg

> From: "Russ Cox" <rsc@plan9.bell-labs.com>
> Introducing contexts where singleton lists
> and strings are not identical seems a very bad idea.

I agree about the context bit. ie, lists and strings should be (or not be) identical in all contexts.
it is just that I thought that they should not be identical.

i wanted
x = asd
and
y = (asd)
to mean that x and y where not identical.


Best Wishes, Bengt
===============================================================
Everything aforementioned should be regarded as totally private
opinions, and nothing else. bengt@softwell.se
``His great strength is that he is uncompromising. It would make
him physically ill to think of programming in C++.''


^ permalink raw reply	[flat|nested] 10+ messages in thread
* Re: ifs substitution, how should it work?
@ 2000-05-19 22:10 Russ Cox
  0 siblings, 0 replies; 10+ messages in thread
From: Russ Cox @ 2000-05-19 22:10 UTC (permalink / raw)
  To: rc, tjg

Introducing contexts where singleton lists
and strings are not identical seems a very bad idea.
Then you have to explain that in general they
are but ifs isn't; and what if save ifs away
to another variable and then restore it?
What does "x=(abc); ifs=$x" do?

If you want basename, why not use basename or sed or awk?

Russ


^ permalink raw reply	[flat|nested] 10+ messages in thread
* Re: ifs substitution, how should it work?
@ 2000-05-18 12:57 Bengt Kleberg
  0 siblings, 0 replies; 10+ messages in thread
From: Bengt Kleberg @ 2000-05-18 12:57 UTC (permalink / raw)
  To: bengt, tjg; +Cc: rc

> From: Tim Goodwin <tjg@star.le.ac.uk>
> > ; x = ``(abc) {cmd...}
> > ; echo $x => 1a2b3 4c5

> If you want awk, you know where to find it.  :-)

Yes, but I had this little problem and suddenly realised that the mysterious 
``(abc) {cmd...}
actually might be used for something! I have never used it before.

> I don't think anything more complicated than single character field
> seperators belongs in a shell: particularly not rc.

OK. Would you consider making ``(abc)
a parse time error? only allow ifs=asd, or ifs=( a s d ), not ifs=(asd) too.
Or, as with the confusing

if ( {}  && {} ) {cmd...}
vs
if ( () && () ) {cmd...} 

document it.

> What's the actual problem you're trying to solve?

split a filename into directory, filename and suffix. the suffix should be allowed to have more than 1 character.
the idea is to make the previously emailed base_name function (made by Carlo Strozzi <carlos@linux.it>)
more like the os utility basename.


Best Wishes, Bengt
===============================================================
Everything aforementioned should be regarded as totally private
opinions, and nothing else. bengt@softwell.se
``His great strength is that he is uncompromising. It would make
him physically ill to think of programming in C++.''


^ permalink raw reply	[flat|nested] 10+ messages in thread
* Re: ifs substitution, how should it work?
@ 2000-05-18  8:12 Bengt Kleberg
  2000-05-18 12:12 ` Tim Goodwin
  0 siblings, 1 reply; 10+ messages in thread
From: Bengt Kleberg @ 2000-05-18  8:12 UTC (permalink / raw)
  To: bengt, byron, rc, tjg

> From: Byron Rakitzis <byron@rakitzis.com>
> So,
> 	ifs='abc'
> and
>	ifs=(a b c)
> should accomplish the same thing.
> Isn't that what you want?

No. I wanted ifs = (abc)
Ie, 
; x = ``(abc) {echo -n 1a2b3abc4c5}
; echo $x => 1a2b3 4c5

This is (IMHO) better, but I presume that there is too much code out there
that depends upon ifs='abc' beeing equal to ifs=(abc).
What if I volunteered to fix everybody elses code for them (provided that they send it to me :-)


Best Wishes, Bengt
===============================================================
Everything aforementioned should be regarded as totally private
opinions, and nothing else. bengt@softwell.se
``His great strength is that he is uncompromising. It would make
him physically ill to think of programming in C++.''


^ permalink raw reply	[flat|nested] 10+ messages in thread
* Re: ifs substitution, how should it work?
@ 2000-05-16 11:22 Bengt Kleberg
  0 siblings, 0 replies; 10+ messages in thread
From: Bengt Kleberg @ 2000-05-16 11:22 UTC (permalink / raw)
  To: rc, tjg

Re: ifs substitution, how should it work?

> What would you expect

>     ; ifs=(' ' '<tab>' '
'

> to do (where `<tab>' represents a literal tab)?


I had but seen code with lists as ifs arguments, and therefore thought
that rc was using lists.
Ie, I would have written it as:

; ifs=(' ' '<tab>' '
')

Would it create much trouble to change rc :-)


Best Wishes, Bengt
===============================================================
Everything aforementioned should be regarded as totally private
opinions, and nothing else. bengt@softwell.se
``His great strength is that he is uncompromising. It would make
him physically ill to think of programming in C++.''


^ permalink raw reply	[flat|nested] 10+ messages in thread
* ifs substitution, how should it work?
@ 2000-05-12 10:14 Bengt Kleberg
  2000-05-15  9:27 ` Tim Goodwin
  0 siblings, 1 reply; 10+ messages in thread
From: Bengt Kleberg @ 2000-05-12 10:14 UTC (permalink / raw)
  To: rc

Greetings,

I thought that these two expression would produce different output
; x = ``( a b c ) {echo -n 1a2b3abc4}
; echo $x
1 2 3 4

and then
x = ``( abc ) {echo -n 1a2b3abc4}
; echo $x
1 2 3 4

Obviously they do not, but since I was expecting
1a2b3 4
in the last case I wonder if this similarity is 'the real thing'?

(i have failed utterly to understand the simple rc syntax before, so presumably I only need
a friendly explanation as to why they must be the same :-)


Best Wishes, Bengt
===============================================================
Everything aforementioned should be regarded as totally private
opinions, and nothing else. bengt@softwell.se
``His great strength is that he is uncompromising. It would make
him physically ill to think of programming in C++.''


^ permalink raw reply	[flat|nested] 10+ messages in thread

end of thread, other threads:[~2000-05-23 23:01 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2000-05-18  6:32 ifs substitution, how should it work? Byron Rakitzis
  -- strict thread matches above, loose matches on Subject: below --
2000-05-22  7:55 Bengt Kleberg
2000-05-22  8:53 ` Tim Goodwin
2000-05-19 22:10 Russ Cox
2000-05-18 12:57 Bengt Kleberg
2000-05-18  8:12 Bengt Kleberg
2000-05-18 12:12 ` Tim Goodwin
2000-05-16 11:22 Bengt Kleberg
2000-05-12 10:14 Bengt Kleberg
2000-05-15  9:27 ` Tim Goodwin

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).