1 |
Jan Kundrát wrote: |
2 |
> If the user knows that keywords are set by the KEYWORDS variable, then |
3 |
> she must be familiar with the EAPI. The meaning of the KEYWORDS variable |
4 |
> is defined by the EAPI. |
5 |
|
6 |
But that's not really what I find objectionable. There's no need to |
7 |
make EAPI so special that it alters the file extension. If EAPI is |
8 |
accessible as part of the file or in metadata.xml, e.g., it still serves |
9 |
this purpose well, and that info is exposed at the right level. |
10 |
|
11 |
> Sure. If current EAPI specified that a sequence of four bytes starting |
12 |
> at offset 0x10 is a little-endian magic number that is used to identify |
13 |
> an EAPI, that'd be all we want. However, current format definition is |
14 |
> rather complex; there's nothing as simple as "read several bytes at some |
15 |
> offset and use them". |
16 |
|
17 |
I would not suggest byte-level methods, no - these files are not binary. |
18 |
There are simple ways to parse text files to find the EAPI. If it is |
19 |
in the ebuild, putting it in the header (out-of-band) works. If it is |
20 |
in metadata, simply parsing XML works. |
21 |
|
22 |
-Joe |
23 |
-- |
24 |
gentoo-dev@l.g.o mailing list |