Re: [Boost-testing] Toolset Autoconfiguration [was: Martin Wille x86_64 V2 - python]

From: David Abrahams (dave_at_[hidden])
Date: 2007-03-26 15:32:55

on Mon Mar 26 2007, Roland Schwarz <roland.schwarz-AT-chello.at> wrote:

> David Abrahams wrote:
>> No, that's a different but related problem.  Martin's complaint, IIUC,
>> is that if you misspell a toolset version name on the command-line,
>> you can end up autoconfiguring a different toolset than the one you
>> intended to use, which is configured in your xxx-config.jam.
>
> I see. Thank you.
>
>> 
>>> But this of course is not a python problem (I am just mentioning
>>> this because we are in a python thread)
>> 
>> Yes.  And as nobody seems to be telling me in general how to parse
>> toolset descriptions like xxx-vvv-sss-ttt, or what the intended
>> semantics are, it probably won't be fixed, at least not by me.
>
> Umm, now I am lost :-(
> Isn't this code authored by you, that splits off these descriptions.

Yes, I wrote code that can parse a limited range of tooslet/version
strings, and pass the version to the correct toolset's "init" rule.
That doesn't tell me anything about how to handle the cases you're
complaining about.  Remember, users have, up till now, been given the
utmost flexibility in deciding how to represent their toolset and
version names.

Let me give you a few examples of questions that have not been answered:

1. Is it possible that the version or the toolset name contains a dash
   (it always used to be possible to make a toolset name containing a
   dash, c.f. intel-win)

2. If one substring is numeric, does that necessarily constitute the
   version number?

3. When invoking the toolset's "init" rule to configure it, what
   should I do with any additional dash-separated substrings?  

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com