1 |
On Sun, Mar 11, 2012 at 10:03 PM, Brian Harring <ferringb@×××××.com> wrote: |
2 |
> Pragmatic reality, the eapi function actually would work fine. As |
3 |
> pointed out elsewhere, bash parses as it goes- which isn't going to |
4 |
> change. |
5 |
|
6 |
Unless the ebuild isn't written in bash... |
7 |
|
8 |
How do you source the ebuild if you don't know what to use to source |
9 |
it? How do you know what to use to source it if you don't know the |
10 |
EAPI? Right now all the existing EAPIs use bash, but there is no |
11 |
reason the file couldn't be xml, or python, or just about anything |
12 |
else. |
13 |
|
14 |
If we want to allow for that kind of flexibility, then it might make |
15 |
sense to go ahead and state that our convention is to stick EAPI=5 in |
16 |
one of the first few lines of the ebuild, or inside a comment, but |
17 |
also go a step further and state that the text "EAPI=" cannot appear |
18 |
elsewhere in the ebuild (or perhaps within the first 10 lines). Just |
19 |
about any file format we might use would allow us to make "EAPI=" |
20 |
appear in it, but not all could guarantee that it would occur at the |
21 |
start of a line, or at the start of a line immediately after a #. |
22 |
|
23 |
In any case, I can really see the KISS value in a very rigid syntax |
24 |
that is trivial to parse. Stuff like this almost makes me wish our |
25 |
ebuilds already were xml files or such, with bash embedded inside |
26 |
sections. Finding a particular tag in an xml file is trivial as the |
27 |
fundamentals haven't changed in 15 years. |
28 |
|
29 |
Rich |