To me, and this is personal preference I think, I don't see a big difference between writing something once outside the db versus in it. Both approaches are fine in my mind. It does make me think of a point I didn't mention though, which is I wouldn't store a function in any existing Evergreen schema. If you do make your own functions for tasks like this create a distinct schema for them along with any tables. I'm not a fan of mixing project stuff into the stock schemas.
On Tue, Apr 7, 2020 at 8:45 AM Jason Stephenson <ja...@sigio.com> wrote: > Glen & Rogan, > > I prefer to do these things outside of the database, i.e. pull the > record out with DBI, modify it, and then update it with DBI. > > I'm not a fan of writing database functions for something that I'll use > only once or twice. > > I have some utility scripts that I've used and shared with the world > available on Github: > > https://github.com/Dyrcona/evergreen_utilities/tree/master/perl > > A couple of those might prove useful as examples, possibly loaderecords.pl > . > > My pre-conference presentation last year covered getting started writing > utility programs in Perl using the Evergreen back end. The files for it > are also available on Github: > > https://github.com/Dyrcona/evergreen2019-preconference > > I hope that you find the above to be useful, and if you have more > technical questions, the conversation should probably move to the > development list. > > Cheers, > Jason > >