» Home › Forums › For Coders › Locum ›

Millennium v2007?

» Submitted by ahojchris on Wed, 03/04/2009 - 8:38pm

I notice there is a Locum connector for Millennium v2006. However it looks like there might also be a v2007 as well. Can the 2006 connector be used for 2007, or is there a need for a different connector for v2007?

‹ Call to undefined method MDB2_Error::fetchAll() in /usr/local/lib/locum/locum-server.php Handling Special Characters ›
Submitted by brianegan on Thu, 03/05/2009 - 8:57pm.

Hello,

There is in a fact a 2007 version of Millennium, which we here at UNLV are running. For testing, I've modified the 2006 connector slightly to work with 2007.

The only thing I had to do was change locum_iii_2006.php. There is a line in there (Line 76), which says:

"case 'MAT TYPE':"

I changed this to:

"case 'BCODE2':"

without any of the quotes.

The XML files for Mil. 2006 and 2007 are slightly different, in that the material type is no longer taken from an XML node called "MAT TYPE" but now from a node called "BCODE2." Once this is changed, I was able to import all of my data without a problem (well, watch out how many child processes you start up, you might put some severe load on your III server!).

Hope this helps!
Brian Egan

Submitted by jblyberg on Thu, 03/12/2009 - 2:14pm.

Brian, Thank-you. We're about to upgrade to 2007 here in the next several weeks and I will be releasing the updated 2007 connector.

Submitted by smaskit on Wed, 07/22/2009 - 6:11pm.

I have committed a slightly upgraded version of the 2007 connector. As part of his upgrade from 2006, John modified how we gather on the fly information about an item's status. It now goes to a summary URL where it can find out whether any holds have been placed on the item. It also checks whether the summary URL contains all of the records for the item. If it does not, it goes to a second URL which contains all of the records, but does not indicate whether any holds it in place). However, it was gathering information about the holdings from the first URL, even when the second URL was needed. Thus any additional records added by the second URL were being missed. I have now corrected this by gathering the information after the second URL has been loaded.