DataMgr 2.2.0.2 and com.sebtools Build 2
I just fixed a couple of bugs in DataMgr which caused an update in both DataMgr and in the com.sebtools pack. The com.sebtools package also includes an update to FileMgr among other components.
One bug fixed dealt with filtering by NULLs the other was with handling what I would call "voids" in a structure. That is, where a key value is essentially obliterated, but the key will still be found when looping over a structure. This happens when a key of a structure is set to the result of a function that doesn't return anything.
So, if you are using either, you might want to upgrade.
Both DataMgr and com.sebtools are open source and free for any use.
Probably not in 2.2.X, but I hope to have that in an upcoming development build.
I just dug into into DataMgr. I really like what you did with it. Excellent work. My main issue with Transfer is having to run setters in order to populate the transfer object. Why not throw it a struct or argument collection?
I have been following in your footsteps with out knowing it. I wrote a wrapper for the default CFCs that Illudium kicks out and the syntax is almost identical to DataMgr.
Really weird and very cool at the same time.
G! Out!
Glad to hear it!
I haven't used Transfer myself. It sounds like just a difference between an OO paradigm and the more data-centric approach taken my DataMgr.
Is this CFC Generator Lite you speak of, or something else? You have certainly piqued my curiosity.
i too do not like using Transfer as i think, among other, defining object in it i is too invasive, and couples model with persistence; in my blog i describe a method of using generic setters and getters so i can define a class only specifying its (cf)properties and introspecting them.
i'll go on qith any posts building a generic datalayer built on dataMgr
regards
>>>Is this CFC Generator Lite you speak of, or something else? You have certainly piqued my curiosity.
Yes and no... I have a bunch of code generators that I am trying to get under one hood for a long time. I tend to get distracted Between all of them they build an entire app.
I dug in over the weekend and started rolling your stuff into my stuff and creating an unholy marriage of the default Illudium CFCs and Datamgr... Just exploring, getting my learn on and having fun really. ;o)
I did not dig in too far into the code but does the "smart cache" option cache the auto generated crud functions or the introspection? The one thing I was thinking about as the cost of introspection for every DB operation. I did some "stress tests" on it and got some cfloop time outs when beating the crap out of it. I noticed that when I used the default init function that it was calling all of the db specific CFC's when doing many to many updates. Any suggestions on performance tuning?
This is great btw. I love this sort of stuff.
BTW I am writing a custom tag library for a lot of this.... just to make DB abstraction ridiculously simple.
EX:
<cf_query
dbaction = "Select"
table ="tblemp"
Orderby ="lname,fname" >
<cf_manytomany
tablename = "#mytable#"
keyfield = "ID"
keyvalue = "#Form.ID#"
multifield = "categoryid"
multilist = "#Form.categoryid#" >
Sorry if the site was down. I moved it to a new server. Depending on the results, I may blog about that later.
The smart cache option is actually for data:
http://www.bryantwebconsulting.com/docs/datamgr/sm...
I would be very interested in your stress test results. Maybe you could post some more information to the DataMgr Google Group?:
http://groups.google.com/group/datamgr
The custom tags look great!