GNU bug report logs - #22241
25.0.50; etags Ruby parser problems

Previous Next

Full log

Message #59 received at 22241 <at> debbugs.gnu.org (full text, mbox):

From: Dmitry Gutov <dgutov <at> yandex.ru>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: 22241 <at> debbugs.gnu.org
Subject: Re: bug#22241: 25.0.50; etags Ruby parser problems
Date: Mon, 1 Feb 2016 11:24:38 +0300

On 01/31/2016 09:01 PM, Eli Zaretskii wrote:

> But that was a trap, wasn't it?

Almost every feature is a trap, if one considers it long enough. :)

> What can legitimately follow the '+',
> in addition to whitespace?  (It's amazing, but among all the gazillion
> references to Ruby, I cannot easily find a formal description of its
> syntax.)

And there isn't one! Ruby is magical that way.

> According to this rare gem:
>
>    https://en.wikibooks.org/wiki/Ruby_Programming/Syntax
>
> (assuming I understand what it says), the RHS can be any literal, and
> also any constant expression, is that right?  If so, either (a) we
> recognize only '^[ \t]([A-Z][a-z0-9_])*[ \t]*=' and get potential
> false positives on the likes of
>
>     ABC == SOMETHING
>     ABC =< WHATEVER

=< is not a valid operator. You must be thinking of <=.

> etc. (are these possible?); or (b) you tell me which characters can
> potentially follow the '=' in an assignment of a constant.

Why not do it like this:

If 'ABC =' is followed by any character, except for '=' and '>', you 
record it as a tag "ABC".

>      " # % \' ( + - < ? [ {
>      0 1 2 3 4 5 6 7 8 9
>      A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
>
> And if we go the latter way, there are still multi-line expressions
> that I think are way too much.

What about them? Ideally, you'd skip over multi-line expressions, but 
you'd have to do that whether you record constants or not.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.