Prerequisites

Make sure you have git and Docker desktop installed.

Installation

Clone/git pull the repo into any local directory

git https://github.com/rcemper/Dataset-OEX-reviews.git

Run the IRIS container with your project:

docker-compose up -d --build

How to Test it

docker-compose exec iris iris session iris

or use Online Demo
The demo is loaded already. There is no need to rebuild / reload it.

There are 4 SQL procedures written in embedded Python to manipulate the table

- oex.CLEAR()           to erase the whole table   
- oex.LOAD(first,last)  to load directory pages first to last..to   
- oex.PAGE(pn)          to load an inidividual directory page   
- oex.DETAIL(id)        to fill all details for a specific package

You may just use any of them by CALL procedure() or SELECT procedure()
To load details SELECT id,oex.DETAIL(id) from oex.map where author is null
is the most elegant way.

Data Load

In practical tests it turned out that loading directory pages is no problen.
Differently, with the package details I experienced network timeouts every 30..50 packages.
Restarting the download is no problem and works OK. Though you have to watch it.
As this rather unattractive and took 40 minutes or more I created an additional SQL procedure

- CALL oex.TOTAL()

It traps all odd network or other incidents and restarts until completed
Due to the long run time it is no well suited to SMB or Webterminal
You better use it from the console with SQL shell to escape from timeouts

$ docker-compose exec iris iris session iris
USER>do $system.SQL.Shell()
[SQL]USER>>CALL oex.TOTAL()
1.      call oex.TOTAL()
2023-06-07 19:10:25 load directory
2023-06-07 19:10:27 directory page 1
2023-06-07 19:10:29 directory page 2
2023-06-07 19:10:31 directory page 3
2023-06-07 19:10:34 directory page 4
   -- -- - -

And in Management Portal
you may watch the table and the progress in loading

Unit Test

from terminal / console prompt run

zpm "oex-mapping test -v"

or

zwrite ##class(oex.test).Run()

and see the results:
http://localhost:52773/csp/sys/%25UnitTest.Portal.Indices.cls?$NAMESPACE=USER

Analytics

I have added a Cube and Pivot table for various exercises.
The Cube is built based onthe remainders from Unit test.
An initial Pivot ‘oex’ is prepared.

Quick Loading

If do not want to wait for a complete fresh load
there is a short cut based on a snapshot of OEX from 2023-06-12 available

- CALL oex.QUICK()

Finally also the Analytics Cube is rebuilt to reflect the results.

1st Article in DC
2nd Article in DC

Video

Demo Server SMP
Demo Server WebTerminal
Demo Server Unit-Tests