[ragel-users] Re: changes to ragel, feedback requested

[Adrian Thurston]

Ah, I like that. CS is always initialized unless it is explicitly
requested that it isn't. When the user wants to init CS it can happen
before write init. Also, CS isn't initialized twice. Seems like a winner
to me!

In regards to the question of initializing with the first vs last
instantiation, I am leaning towards the last since Ragel programs are
normally specified in a bottom-up nature (leaf machine definitions come
first). Instantiations can go in any order, but in the programs I write
the main machine always ends up last.

Adrian

Oliver Schonefeld wrote:
> Hello Adrian,
> 
> Eines schoenen Tages schrieb Adrian Thurston:
>> Defaulting to the first/last is a
>> possibility. I'm not sure I like the requirement that cs be initialized
>> after the write init. Could be a source of bugs more common than the one
>> we're trying to prevent.
> 
> How about an additional option to 'init', which prevents the initalization
> of cs. Something link '%write init nocsinit'. The makes the need for the to
> initalize cs more explicit.
> 
> Best Regards,
> Oliver

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 252 bytes
Desc: OpenPGP digital signature
URL: <http://www.colm.net/pipermail/ragel-users/attachments/20070320/6dd6bda9/attachment-0001.sig>