Re: More font-lock keywords for tcl.el

[Glenn Morris]

Stefan Monnier wrote:

> Just skimming through it, I notice "namespace". Is it really a
> builtin? I would've expected it to be a keyword.

It's a command, if that's what you're asking. (I got this list from
Some-Frighteningly-Thick-Tcl-Book.)

>> +          (list (concat "\\(\\s-\\|^\\|\\[\\)"
>> +                   (regexp-opt tcl-builtin-list t)
>> +                   "\\(\\s-\\|$\\|\\]\\)")
>
> How 'bout (concat "\\_<" (regexp-opt tcl-builtin-list t) "\\_>") ?

OK. I used the format I did because it's the same as the existing
stuff in tcl.el.

>> +          '("\\\\$" 0 font-lock-warning-face) ; escaped EOL
>
> Why warning?  Is it dangerous?  I don't think so.

Personally, I like it to stand out (that's no argument, of course!).
I think I was trying to be consistent with sh-script.el.
There, escaped EOLs used to be highlighted in string-face. But that
confused sh-get-indent-info, so I changed it to warning-face.

<http://lists.gnu.org/archive/html/emacs-pretest-bug/2003-09/msg00028.html>

> Also the regexp matched more than escaped EOL, it also matches an
> escaped backslash at the end of the line.

The regexp is also the same as the sh-script.el one.

What's a regexp for "/ at the end of a line, not preceded by an odd
number of other /"?

I tried this monstrosity:

"\\(?:[^\\\\]\\|\\(?:\\\\\\\\\\)+\\)\\(\\\\\\)$"

but it also matches \\\\ (I don't see why).

I can just leave this part out, I'm not especially bothered.
But sh-script should at least be corrected, I guess.