1 |
>>>>> On Tue, 10 Apr 2012, Brian Harring wrote: |
2 |
|
3 |
> On Tue, Apr 10, 2012 at 08:43:55AM -0700, Zac Medico wrote: |
4 |
>> We could make the ([A-Za-z0-9+_.-]*) group "optional", but treat it |
5 |
>> as an error if the group is missing. That way, the package manager |
6 |
>> can detect invalid assignments earlier, and can skip sourcing the |
7 |
>> ebuild in this case. |
8 |
|
9 |
Currently there isn't any ebuild in the tree that assigns an empty |
10 |
EAPI, so I think we can neglect any performance issues. |
11 |
|
12 |
> Or disallow EAPI= . |
13 |
|
14 |
Probably it's simpler to keep the regexp as it is, but error out if |
15 |
the second group is empty. |
16 |
|
17 |
> There isn't any reason to reset it to an unknown EAPI manually like |
18 |
> that; if code does that, the code is wrong imo, thus skip trying to |
19 |
> parse that. |
20 |
|
21 |
I see no valid usage case why an ebuild should assign the empty string |
22 |
to the EAPI variable. So we could add a sentence like |
23 |
"It is an error for an ebuild to assign an empty EAPI." |
24 |
to the first paragraph of this subsection: |
25 |
<http://dev.gentoo.org/~ulm/pms/4/pms.html#x1-710008.3.1> |
26 |
|
27 |
Ulrich |