Highlight known dictionary in transfer requests

[mushon]

To quickly understand the reasons or type for transfer highlight certain keywords:

• "עודפי תקציב" or "עודפים"
• "החלטת ממשלה"
• "הרשאה להתחייב" or "הרשאות"
• add more

[mushon]

@azk

[azk]

I'm leveraging this issue to start learning the openbudget codebase(s), so I apologize if this is a little long for a comment...

As far as I could find, there are a number of places in the various repos where transfer "explanations" can be manipulated as required, and I'm not sure where it would make the most sense to add the term highlighting.

Namely:

1. Data upload phase:

   • open-budget-data/changes/aggregate_changes.py - changes_total.{csv,json}.gz are created containing change objects. Not sure what then happens to them.
   • open-budget-data/change_explanation/rules.yaml - ChangeExplanation model objects are uploaded to DB (after being downloaded in "download_pending_changes")

   [ rss.py processor (not sure where it's being used) makes use of the ChangeExplanation model API to get the explanation ]

2. API:
   • Uploader: open-budget-api/open-budget-il/upload.py@ULChangeExplanation - Can override "preprocess_item" method
   • GET: open-budget-api/open-budget-il/main.py@ChangeExplApi - Can modify "get_query" method.
3. Views:
   • Email app: open-budget-frontend/email/app/scripts/main.js - Can modify explanation after getting it from API
   • Main web app: open-budget-frontend/open-budget/app/scripts/modelsHelpers/ChangeExplanation.coffee - Can modify explanation after getting it from API.

As you can tell I still have some holes regarding how things work together...

My thought is to modify the explanation field in the ChangeExplApi get_query method, as it would immediately reflect any changes to the term dictionary or highlighting method to all data.

And an unrelated question: how do you test the backend with Google App Engine?

Thank you!

[mushon]

@akariv please review and resond to @azk's questions either before the meetup tonight or at it.

[akariv]

@azk Some of the data already exists (e.g. the kind of the change "עודפים" etc. is a per-change parameter). This data needs to be aggregated from the Change level into the ChangeGroup level. This happens in processors/extract_change_groups.py (only code that runs is in processors/). Other than that, I would put the other known-phrase highlighting in the client side. A cool first project would be to just find gov decisions ("החלטת ממשלה מספר") and get their actual content using this API: https://ceci.yit.co.il/help

Adam