Re: [BUG] Regexp compiler, problem with character classes

[Johan Bockgård]

Richard Stallman <address@hidden> writes:

>     (with-temp-buffer
>       (string-match "x[[:space:]]" "x\n"))
>
>     => nil
>
>     (Fchar_syntax does
>       gl_state.current_syntax_table = current_buffer->syntax_table;)
>
> re_search_2 does it also, thru the macro
> SETUP_SYNTAX_TABLE_FOR_OBJECT.  I just verified that
> gl_state.current_buffer->syntax_table has the proper value after that
> line.

Yes, but that is too late. What matters is which syntax is seen by the
regexp *compiler* (this is why there is a problem in the first place)
when it transforms [[:space:]] into [\t\f etc]. This has already
happened before re_search_2 comes into play.

-- 
Johan Bockgård