You can leverage the API built into the Casper Suite to do lots and lots of cool stuff, without interacting directly with the database. Here, I’ll use a simple curl command in a bash script that has myuser as the username for a server and mypassword as the password. The server is myserver.jamfcloud.com. Basically, we’re going to ask the computers and mobiledevices tables for all their datas. Once we have that, we’ll constrain the output to just the size attribute for each using sed:
curl -s -u myuser:mypassword https://myserver.jamfcloud.com/JSSResource/computers | sed -n -e 's/.*<size>\(.*\)<\/size>.*/\1/p'
curl -s -u myuser:mypassword https://myserver.jamfcloud.com/JSSResource/mobiledevices | sed -n -e 's/.*<size>\(.*\)<\/size>.*/\1/p'
This same logic can then be applied to any payload of XML data coming out of a REST API. Some API’s have different options to constrain output of a request, some don’t. But no matter whether there is or isn’t, you can loop through a bunch of statements like this. Why would you look to the API to constrain data, etc? Well, it comes down to a cost issue. Each time you run the above commands, you’re costing yourself runtime, you’re taxing the server with potentially a substantial query, and you’re potentially transferring a considerable amount of data over the wires between you and where the script is being run. So if the API is smart enough to give you less data, then you might as well do that. In this case, it isn’t, but if you apply this same sed logic in other scripts, it’s great to be cognizant of remaining as efficient as you can.