A purge holds function is needed

[Thomas Berezansky]

Similar to the purge circulations function, a purge holds function is needed to remove old holds from the database for patron privacy reasons.

The branch below has two commits. The first one adds said function, the second modifies user visible holds to pay attention to the fill/cancel date instead of request date. That should prevent holds from "vanishing" because the option was turned on while the hold was active. This is a second commit because there are varying opinions on it.

http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/tsbere/purge_holds

[Kathy Lussier (klussier)]

Adding needsreleasenote tag since this would be a new feature requiring a release notes entry.

[Jason Stephenson (jstephenson)]

When I run the upgrade script for this branch, I get an error about a missing column. (Forget which one at the moment.) I'll fix it when I next update my dev server unless tsbere beats me to it.

[Thomas Berezansky (tsbere)]

I force pushed a fix for the upgrade script (and seed data, actually). I will see about adding a release notes commit shortly.

[Jason Stephenson (jstephenson)]

I thought that I was going to get to this by this past weekend, but I have been sick and so I did not. I am removing myself as assignee so someone else can have a look if they want. I don't know when I'll be able to get to this as I have a number of other things coming up.

[Jason Stephenson (jstephenson)]

Release notes were added.

I have pushed a rebased branch to http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/collab/dyrcona/purge_holds_rebase

I'm still looking into this one.

[Jason Stephenson (jstephenson)]

One thing concerns me about this code: The holds are simply deleted from action.hold_request. As there is no trigger on action.hold_request to anonymize and copy the data to an aged_holds table (as is done with circulations), it looks like you could lose data that you might want for reporting with this branch.

Perhaps a similar trigger and table should be added for holds?

[Mike Rylander (mrylander)]

Agreed, Jason. I'd hate to loose useful (and, in some cases, necessary-for-funding) stats data.

[Jason Stephenson (jstephenson)]

My current WIP: http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/collab/dyrcona/purge_holds

I'll work on adding a table (action.aged_holds) and a trigger to move holds to that table on delete and unassign the bug when I think that is ready.

I have already signed off on tsbere's commits in the branch above as I've tested them and they work.

This should probably wait to go in until after 2.4 unless someone feels strongly that it is a bug fix. I see it as a new feature.

[Mike Rylander (mrylander)]

Your humble 2.4 RM agrees that this is all feature.

[Jason Stephenson (jstephenson)]

This one is ready for someone else's signoff. I've tested and added to the release notes about the new table: action.aged_hold_request. The new code is the top 8 commits in

http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/collab/dyrcona/purge_holds

I've already signed off tsbere's commits, so just the last three commits need another signoff. It would be kind of hard to test those without testing all of the others.

[Dan Wells (dbw2)]

I wasn't too familiar with the circulation equivalent feature, but once I understood the settings, this worked as advertised. Pushed to master.

Thanks, Jason and Thomas!