Wednesday, 31 August 2016

Planning Metadata CSV parser



Hyperion Planning is a cool tool built around a HOLAP framework. It does a lot of things to ease administrative tasks from a process perspective compared to Essbase which can be looked at as a data engine. But there are some things which are good in Essbase but not so in Planning. One of them is the hierarchy view for members…While EAS console allows me to see all the metadata in a hierarchy in a single window, Planning does it one step at a time… (This is again a case of tradeoff since communicating information of a hierarchy having potentially thousands of members from which only 10 to 15 would be queried on web is more expensive than on a thick client since a thick client can be optimized internally). However, sometimes you just need to see the hierarchy in its full form, particularly with shared members since I would like to see how they roll up, under which parent and so on…And with PBCS in the picture, I cannot go to the backend EAS console and see the metadata…

Well, I got tired of expanding nodes in the workspace and I thought of designing my own metadata parser… 
Starters, this is alpha so not much is available by way of error codes and so on…Still working on it…

This is the hello world for the jar file when you launch the metadata parser…



The file name field can be manually updated to enter the full file path of a CSV file or we can click on the Browse button that is shown below to access a file using a Navigation wizard… 


The Process button shown below is used to process the CSV extract that you have chosen…



The search button that is shown below can be used for doing a case insensitive search for members in the hierarchy…
 
If you click on the browse button, a file navigation wizard opens up that allows you to choose a CSV planning metadata file…Below is a snapshot of the browsing for the file and choosing one…



As soon as you hit the open button, the file would be loaded and parsed as shown in the below snapshot….







Note that in the above snapshot, the metadata was parsed for account dimension and it shows the metadata as a tree format…

If you manually update the full file name, you need to hit the process button to initiate the processing of the extract….

Now, to search for a node like “assets” you enter it in the search field and hit the search button… Snapshots are shown below…




Observe that the members are shown in in breadcrumb path since this was the main problem I wanted to solve, that is see the rollups in a hierarchy so that I can make logical inferences and correct errors if any...

Now, for the best part….

If you click on a node in the main hierarchy, it will pop up the list of associated properties of that member… below snapshot shows the properties of Balance sheet member parsed and in a simple easy to read format…

Its way faster than the same thing over the web since all of the web communication is done away with using the Planning CSV extract…



Now, I am still working on a few ifs and buts and adding a couple of JOptionPanes for error message display and stuff… However, if you would like to test it out let me know… I would upload it on GitHub and am looking for folks to test it out since I might have developer bias…

Will keep you posted…