Re: Gelex problem

I made a completely cut-down version of the system with the 10 or so 
classes it needs from my environment, and then just referring to my 
current $GOBO and $ISE_EIFFEL (7.3 official GPL) locations.

I can show the problem occurring with a parser input file that is a cut 
down version of my usual files, where there are leading spaces on the 
lines that should match but don't. In my little test project, I made a 
copy of the parser input file, and made just one change - convert the 
leading spaces on those lines to TABs (four spaces = 1 TAB), and the 
file parses as it should, i.e. TABs are matched, but spaces are not.

I also tried changing the regex to be:

^[^\040\t] { -- match non-white space at start

from the original, which was:

^[^ \t] { -- match non-white space at start

So, in summary, I have a file containing lines that start with 0x20s, 
i.e. space characters (checked in VIM Hex mode), and the above regex is 
matching the first space on such a line with the regex pattern ^[^ \t]. 
But convert those spaces to TABs and everything appears to work.

Am I doing something dumb here?

If anyone cares to have a look, I can supply a small zip of about 20 
files that contains the entire system and test files.

- thomas

On 11/10/2013 22:48, Eric Bezault wrote:
>
> I wrote a simple example based on the code above, and it worked
> as expected. I got the following:
>
> ##### matched whitespace non-lf* lf #########
>         #####    toto
>  #########
> ##### matched non-whitespace at start #########
>         ##### 100 #########
>
>