[pacman-dev] [PATCH v2] libalpm: parse {check, make}depends when reading database
Commit 0994893b0e6b627d45a63884ac01af7d0967eff2 added the
alpm_pkg_get_{make,check}depends functions but forgot to include
logic for parsing these fields from the database. As a result these
functions will always return an empty list.
This commit adds the parsing logic.
Signed-off-by: morganamilo
On 9/10/18 11:05 am, morganamilo wrote:
static struct pkg_operations local_pkg_ops = { - .get_base = _cache_get_base, - .get_desc = _cache_get_desc, - .get_url = _cache_get_url, - .get_builddate = _cache_get_builddate, - .get_installdate = _cache_get_installdate, - .get_packager = _cache_get_packager, <snip> + .get_base = _cache_get_base, + .get_desc = _cache_get_desc, + .get_url = _cache_get_url, + .get_builddate = _cache_get_builddate, + .get_installdate = _cache_get_installdate, ...
Is there any objection to killing this formatting rather than having a lot of uselessness in the patch. A
On 10/8/18 9:15 PM, Allan McRae wrote:
On 9/10/18 11:05 am, morganamilo wrote: <snip>
+ .get_base = _cache_get_base, + .get_desc = _cache_get_desc, + .get_url = _cache_get_url, + .get_builddate = _cache_get_builddate, + .get_installdate = _cache_get_installdate, ...
Is there any objection to killing this formatting rather than having a lot of uselessness in the patch.
I'm rather meh about keeping things aligned like this in the first place, so I'm okay with killing the formatting. I mean, indenting the beginning of the line makes sense. Making variable assignments line up is useless IMO. But if you will do it then you should leave space for it to grow... -- Eli Schwartz Bug Wrangler and Trusted User
On 09/10/2018 2:15 am, Allan McRae wrote:
On 9/10/18 11:05 am, morganamilo wrote:
static struct pkg_operations local_pkg_ops = { - .get_base = _cache_get_base, - .get_desc = _cache_get_desc, - .get_url = _cache_get_url, - .get_builddate = _cache_get_builddate, - .get_installdate = _cache_get_installdate, - .get_packager = _cache_get_packager, <snip> + .get_base = _cache_get_base, + .get_desc = _cache_get_desc, + .get_url = _cache_get_url, + .get_builddate = _cache_get_builddate, + .get_installdate = _cache_get_installdate, ...
Is there any objection to killing this formatting rather than having a lot of uselessness in the patch.
A
I could split adding the fields and the formatting to two patches if you prefer? Easier to read diffs and the struct stays pretty.
On 10/09/18 at 02:20am, Morgan Adamiec wrote:
On 09/10/2018 2:15 am, Allan McRae wrote:
On 9/10/18 11:05 am, morganamilo wrote:
static struct pkg_operations local_pkg_ops = { - .get_base = _cache_get_base, - .get_desc = _cache_get_desc, - .get_url = _cache_get_url, - .get_builddate = _cache_get_builddate, - .get_installdate = _cache_get_installdate, - .get_packager = _cache_get_packager, <snip> + .get_base = _cache_get_base, + .get_desc = _cache_get_desc, + .get_url = _cache_get_url, + .get_builddate = _cache_get_builddate, + .get_installdate = _cache_get_installdate, ...
Is there any objection to killing this formatting rather than having a lot of uselessness in the patch.
A
I could split adding the fields and the formatting to two patches if you prefer? Easier to read diffs and the struct stays pretty.
Kill the formatting. We've had this discussion before, for this very feature even: https://lists.archlinux.org/pipermail/pacman-dev/2017-January/021784.html
On 9/10/18 12:19 pm, Andrew Gregory wrote:
On 10/09/18 at 02:20am, Morgan Adamiec wrote:
On 09/10/2018 2:15 am, Allan McRae wrote:
On 9/10/18 11:05 am, morganamilo wrote:
static struct pkg_operations local_pkg_ops = { - .get_base = _cache_get_base, - .get_desc = _cache_get_desc, - .get_url = _cache_get_url, - .get_builddate = _cache_get_builddate, - .get_installdate = _cache_get_installdate, - .get_packager = _cache_get_packager, <snip> + .get_base = _cache_get_base, + .get_desc = _cache_get_desc, + .get_url = _cache_get_url, + .get_builddate = _cache_get_builddate, + .get_installdate = _cache_get_installdate, ...
Is there any objection to killing this formatting rather than having a lot of uselessness in the patch.
A
I could split adding the fields and the formatting to two patches if you prefer? Easier to read diffs and the struct stays pretty.
Kill the formatting. We've had this discussion before, for this very feature even: https://lists.archlinux.org/pipermail/pacman-dev/2017-January/021784.html .
OK - a patch to kill the formating, then this patch rebased on top. A
participants (5)
-
Allan McRae
-
Andrew Gregory
-
Eli Schwartz
-
Morgan Adamiec
-
morganamilo