5:47 a.m.
if a string ends with $, you are in deep trouble with font-lock.
this happens quite often with regexps, and there appears to be no way to rewrite
the regexps to avoid this.
any ideas?
is there a mailing list for cperl?
--
ben
I'm sometimes slow in getting around to reading my mail, so if you
want to reach me faster, call 520-661-6661.
See http://www.666.com/ben/chronic-pain/ for the hell I've been
through.
10:47 a.m.
Ben Wing <ben(a)666.com> writes:
if a string ends with $, you are in deep trouble with font-lock.
This is because cperl mode defines $ to be an "escape" character, just
like backslash.
It actually has to do that, so that stuff like
$[ = 10;
or
$' = whatever;
work out right. Perl is truly convoluted.
1:24 p.m.
yeah, but cperl tries to have a different syntax table inside of strings, where
$ is not a backslash character.
if you're feeling clever, perhaps you could fix font lock so it can allow for a
different syntax table in strings?
Hrvoje Niksic wrote:
Ben Wing <ben(a)666.com> writes:
> if a string ends with $, you are in deep trouble with font-lock.
This is because cperl mode defines $ to be an "escape" character, just
like backslash.
It actually has to do that, so that stuff like
$[ = 10;
or
$' = whatever;
work out right. Perl is truly convoluted.
--
ben
I'm sometimes slow in getting around to reading my mail, so if you
want to reach me faster, call 520-661-6661.
See http://www.666.com/ben/chronic-pain/ for the hell I've been
through.
1:25 p.m.
Ben Wing <ben(a)666.com> writes:
yeah, but cperl tries to have a different syntax table inside of
strings, where
$ is not a backslash character.
if you're feeling clever, perhaps you could fix font lock so it can allow for a
different syntax table in strings?
As far as I know, this should "just work" with Matt Tucker's syntax
improvements. In fact, cperl was advertised as the biggest user of
"syntax-table-as-text-property" feature.
But I could be wrong.
8:39 a.m.
Stefan Monnier wrote:
>>>>> "Hrvoje" == Hrvoje Niksic <hniksic(a)arsdigita.com>
writes:
> In fact, cperl was advertised as the biggest user of
> "syntax-table-as-text-property" feature.
I don't think anybody would dare to argue otherwise,
but ... ?
Stefan
--
ben
I'm sometimes slow in getting around to reading my mail, so if you
want to reach me faster, call 520-661-6661.
See http://www.666.com/ben/chronic-pain/ for the hell I've been
through.
4:25 p.m.
>>>> "Ben" == Ben Wing <ben(a)666.com>
writes:
Stefan Monnier wrote:
> >>>>> "Hrvoje" == Hrvoje Niksic
<hniksic(a)arsdigita.com> writes:
> > In fact, cperl was advertised as the biggest user of
> > "syntax-table-as-text-property" feature.
> I don't think anybody would dare to argue otherwise,
but ... ?
But the problem is also present on Emacs, so it's a CPerl limitation,
Stefan
5:59 p.m.
"Stefan Monnier" <monnier+lists.xemacs.beta/news/(a)RUM.cs.yale.edu>
writes:
>>>>> "Ben" == Ben Wing <ben(a)666.com>
writes:
> Stefan Monnier wrote:
>> >>>>> "Hrvoje" == Hrvoje Niksic
<hniksic(a)arsdigita.com> writes:
>> > In fact, cperl was advertised as the biggest user of
>> > "syntax-table-as-text-property" feature.
>> I don't think anybody would dare to argue otherwise,
> but ... ?
But the problem is also present on Emacs, so it's a CPerl
limitation,
Good to know; thanks.
6:12 a.m.
well, i should be running with these changes, right? i still see the problems,
though.
Hrvoje Niksic wrote:
Ben Wing <ben(a)666.com> writes:
> yeah, but cperl tries to have a different syntax table inside of strings, where
> $ is not a backslash character.
>
> if you're feeling clever, perhaps you could fix font lock so it can allow for a
> different syntax table in strings?
As far as I know, this should "just work" with Matt Tucker's syntax
improvements. In fact, cperl was advertised as the biggest user of
"syntax-table-as-text-property" feature.
But I could be wrong.
--
ben
I'm sometimes slow in getting around to reading my mail, so if you
want to reach me faster, call 520-661-6661.
See http://www.666.com/ben/chronic-pain/ for the hell I've been
through.
2:26 p.m.
Ben Wing <ben(a)666.com> writes:
well, i should be running with these changes, right?
I think so. Matt originally implemented them to depend on a configure
check *and* a Lisp variable, but I think I persuaded him to get rid of
both.
i still see the problems, though.
I know. :-( This could be for a number of reasons:
* I could be wrong -- maybe cperl mode doesn't use this feature to get
this effect -- perhaps it uses some other FSF Emacs feature.
* Perhaps Matt's code misbehaves for the things cperl mode uses it for
(Matt, have you ever tested it with cperl mode?)
* Perhaps cperl mode uses the syntax property in a weird way which is
actually incompatible with what we have implemented.
* Finally, perhaps it's something completely silly such as turning on
the feature being triggered with (not (string-match "XEmacs" ...)).
Matt, do you have time to investigate what's going on here?
5:10 a.m.
matt, are you there? i haven't seen any responses to my messages to you
concerning the syntax-table breakage that a number of people are seeing.
let me clarify about my previous comments. i made a "mea culpa" comment --
with
double-quoted strings, i do indeed need to backslash the $.
but in single-quoted strings, i don't, and i still see the font-lock problem: a
literal like
'^foo.*bar$'
will be font-locked as a runaway string, since the $ will be seen as an escape.
matt, can this be fixed?
Hrvoje Niksic wrote:
Ben Wing <ben(a)666.com> writes:
> well, i should be running with these changes, right?
I think so. Matt originally implemented them to depend on a configure
check *and* a Lisp variable, but I think I persuaded him to get rid of
both.
> i still see the problems, though.
I know. :-( This could be for a number of reasons:
* I could be wrong -- maybe cperl mode doesn't use this feature to get
this effect -- perhaps it uses some other FSF Emacs feature.
* Perhaps Matt's code misbehaves for the things cperl mode uses it for
(Matt, have you ever tested it with cperl mode?)
* Perhaps cperl mode uses the syntax property in a weird way which is
actually incompatible with what we have implemented.
* Finally, perhaps it's something completely silly such as turning on
the feature being triggered with (not (string-match "XEmacs" ...)).
Matt, do you have time to investigate what's going on here?
--
ben
I'm sometimes slow in getting around to reading my mail, so if you
want to reach me faster, call 520-661-6661.
See http://www.666.com/ben/chronic-pain/ for the hell I've been
through.
2:58 a.m.
-- Ben Wing <ben(a)666.com> spake thusly:
matt, are you there? i haven't seen any responses to my messages
to
you concerning the syntax-table breakage that a number of people are
seeing.
My apologies. I've been out of town for 5 weeks (in Thailand), and
totally forgot to inform the list before I left that I'd be gone (I may
have mentioned it offhand in an email somewhere, but I should have been
more formal about it). I just got back this morning and am still
decompressing. I'll try to address the more urgent font-lock issues
over the next few days.
8:01 a.m.
never mind.
i do need to backslash that $.
Hrvoje Niksic wrote:
Ben Wing <ben(a)666.com> writes:
> yeah, but cperl tries to have a different syntax table inside of strings, where
> $ is not a backslash character.
>
> if you're feeling clever, perhaps you could fix font lock so it can allow for a
> different syntax table in strings?
As far as I know, this should "just work" with Matt Tucker's syntax
improvements. In fact, cperl was advertised as the biggest user of
"syntax-table-as-text-property" feature.
But I could be wrong.
--
ben
I'm sometimes slow in getting around to reading my mail, so if you
want to reach me faster, call 520-661-6661.
See http://www.666.com/ben/chronic-pain/ for the hell I've been
through.
8343
days inactive
8357
days old
12 comments
4 participants
participants (4)
-
Ben Wing -
Hrvoje Niksic -
Matt Tucker -
Stefan Monnier