On Tue, Aug 2, 2011 at 7:41 AM, Lukas Fleischer <archlinux@cryptocrack.de> wrote:
On Tue, Aug 02, 2011 at 07:13:05AM -0500, Dan McGee wrote:
On Sun, Jul 31, 2011 at 12:05 PM, Lukas Fleischer <archlinux@cryptocrack.de> wrote:
This allows for merging comments and votes of deleted packages into another one which is useful if a package needs to be renamed.
Signed-off-by: Lukas Fleischer <archlinux@cryptocrack.de> --- web/lib/pkgfuncs.inc.php | 24 +++++++++++++++++++++++- 1 files changed, 23 insertions(+), 1 deletions(-)
diff --git a/web/lib/pkgfuncs.inc.php b/web/lib/pkgfuncs.inc.php index bb5a592..a81ee01 100644 --- a/web/lib/pkgfuncs.inc.php +++ b/web/lib/pkgfuncs.inc.php @@ -659,10 +659,11 @@ function pkg_flag ($atype, $ids, $action = True) { * * @param string $atype Account type, output of account_from_sid * @param array $ids Array of package IDs to delete + * @param int $mergepkgid Package to merge the deleted ones into * * @return string Translated error or success message */ -function pkg_delete ($atype, $ids) { +function pkg_delete ($atype, $ids, $mergepkgid) { if (!$atype) { return __("You must be logged in before you can delete packages."); } @@ -678,6 +679,27 @@ function pkg_delete ($atype, $ids) { }
$dbh = db_connect(); + + if ($mergepkgid) { + /* Merge comments */ + $q = "UPDATE IGNORE PackageComments "; Not a fan of this. This is not in any SQL standard so this would be another thing needing adjusted later if someone tried to run using non-MySQL. It is also extremely hacky IMO as you can end up suppressing real errors, and read this gem from the manual: "Rows for which columns are updated to values that would cause data conversion errors are updated to the closest valid values instead." Which just plain scares me.
I'm not even sure why you are doing IGNORE on the comments merge (these can never conflict?), but for the votes merge, simply do a * UPDATE where it doesn't already exist. * DELETE all remaining.
Well, I actually can't remember what I was thinking when I wrote the comments query. I probably just copy-pasted it. That's the best excuse I can come up with, yeah :)
I'll fix this and think of a better filter criteria for the votes merge query. I don't think we need the delete all remaining (duplicate) votes though, as we do remove the associated packages anyway and "ON DELETE" should do the rest.
Ahh, true on the auto-cascade DELETE thoughts. If you just add a little more logic to the WHERE clause you should be able to make the UPDATES selective enough to avoid the duplicates issue. -Dan