I'm more of a database enthusiast than a developer but it actually sounds like a fun experiment for a test box. I need experience writing triggers so I'd be willing to put it on my to do list to play with though I couldn't promise a) when I'd get to it or b) that it will work. Someone else might chime in and point out a reason my idea wouldn't work.
On Thu, May 21, 2015 at 1:56 PM, Tony Bandy <[email protected]> wrote: > Hi all, Rogan, > > > > Appreciate the update….but not being a developer or SQL person, I was a > little lost after the second sentence. J If something like this would > get done, do you think this could be part of the documentation, etc? > > > > --Tony > > > > Tony Bandy > > [email protected] > > OHIONET > > 1500 West Lane Ave. > > Columbus, OH 43221-3975 > > 614-484-1074 (Direct) > > 614-486-2966 x19 > > > > *From:* Open-ils-general [mailto: > [email protected]] *On Behalf Of *Rogan > Hamby > *Sent:* Thursday, May 21, 2015 1:49 PM > *To:* Evergreen Discussion Group > *Subject:* Re: [OPEN-ILS-GENERAL] Authority Control Questions-Maybe a > Solution? > > > > Hi Tony, > > > > I've never done this but as I hope SCLENDS moves to authorities in the not > too distant future I'm interested in this topic a lot. > > > > If I remember correctly when a MARC record gets updated the old > metabib.real_full_rec entries get deleted and new ones are entered. I > would be tempted to create a conditional auditor trigger that only write > entries for the headings fields so that it's not a huge auditor table. > I've not done that but some quick googling showed up examples of it. It > would be a fun experiment if nothing else. Then it would be easy to write > a sql report to get distinct records from those whose inserted IDs (current > metabib) are different from the audited ones and then export just those > records. > > > > > > > > On Thu, May 21, 2015 at 1:04 PM, Tony Bandy <[email protected]> wrote: > > Hello all, > > > > At COOL, we’ve recently started authority control with BSLW. We have had > our entire DB cleaned up and are getting ready to start the quarterly > updates. In this, we will be sending them new bibs that have been added to > the system for headings cleanup. > > > > However, we are also considering sending edited bibs as well. This would > be bib records that were authorized and cleaned up, but have since been > overlaid or updated. > > > > The biggest issue we are having is how to do this. While the edit date in > the bib is a starting point, we really only want to send bibs whose > headings have changed, not other values in the record. > > > > The question is: How? > > > > We’ve looked at the auditing tables as a starting point, but am thinking > there might be a better solution: > > > > --------------------- > > > > Initial thoughts: > > > > 1. Using PGAdmin, we can extract a list of TCN’s that have been > modified (edited) since our last full DB dump. > > 2. Using this list, extract bibs in MARC format from the system. > > 3. Using MARC Compare (found in Marc Edit 6.x) I can compare these > bibs with the original cleaned up records from BSLW and get a list of > changed fields. (I’ve done a sample file this am) > > 4. This solution actually works to scan fields by hand for changes. > > 5. However, with MARC Compare the records have to be in order or > everything falls apart. (Plus if you have a ton-o-records, this can prove > time consuming). > > > > --------------------- > > > > Solution? > > > > So my proposed solution is to write a custom PERL script to take original > bibs and edited bibs, extract them down and compare headings in each of the > two files, sending output out (via TCN’s) of those bibs whose headings > fields have changed (such as the 100's, 400's, 600's, 700-740's as > examples). Once I have the TCN’s, it would be easy to dump these out of > the system for cleanup by BSLW. > > > > --------------------- > > > > My question is: Has this been done before? (I don’t want to reinvent the > wheel). Is there a better way that someone else has come up with to do > this? I’ve been digging around http://perldoc.perl.org/File/Compare.html > this am as well as http://marcpm.sourceforge.net/tutorial.pdf, and > https://metacpan.org/pod/MARC::Record and I think this would be do-able. > > > > I’m certainly not a developer, but have written perl scripts in the > past….so this would be fun to try. > > > > Any thoughts? > > > > --Tony > > > > Tony Bandy > > [email protected] > > OHIONET > > 1500 West Lane Ave. > > Columbus, OH 43221-3975 > > 614-484-1074 (Direct) > > 614-486-2966 x19 > > > > > > > > -- > > > > Rogan Hamby, MLS, CCNP, MIA > > Managers Headquarters Library and Reference Services, > > York County Library System > > > > “You can never get a cup of tea large enough or a book long enough to suit > me.” > ― C.S. Lewis <http://www.goodreads.com/author/show/1069006.C_S_Lewis> > -- Rogan Hamby, MLS, CCNP, MIA Managers Headquarters Library and Reference Services, York County Library System “You can never get a cup of tea large enough or a book long enough to suit me.” ― C.S. Lewis <http://www.goodreads.com/author/show/1069006.C_S_Lewis>
