keep-the-receipts / procurement-portal-backend Goto Github PK

What would we enter if it's a period though? Not sure this is reporting period?

http://ocpo.treasury.gov.za/COVID19/Documents/FREE%20STATE%20OFFICE%20OF%20THE%20PREMIER.pdf
keep-the-receipts/data-extraction#13

See how OCDS models this, but probably something like implementation status or implementation notes?

http://ocpo.treasury.gov.za/COVID19/Documents/FREE%20STATE%20HEALTH.pdf

Provincial Government/Free State/Economic and small business development, tourism and environmental affairs/FREE STATE ECONOMIC and SMALL BUSINESS DEVELOPMENT, TOURISM and ENVIRONMENTAL AFFAIRS.csv

example data:

REPORTING PERIOD
--
14/04/2020
14/04/2020
26/03/2020
26/03/2020
26/03/2020
26/03/2020
26/03/2020
14/04/2020
May-20
May-20
May-20
May-20
May-20
May-20
May-20
May-20
May-20

We have a field for central_supplier_database_number for the supplier, and supplier_name. We need a field for all the other sorts of numbers available.

It should be included in the supplier_full_text index.

http://ocpo.treasury.gov.za/COVID19/Documents/FREE%20STATE%20SPORTS%2c%20ARTS%2c%20CULTURE%20AND%20RECREATION.pdf
keep-the-receipts/data-extraction#19

#31

It would be nice if someone could download the original uploaded file to get closer to the original source data shown in the site - in case something weird happened on import, and currently only the known columns are imported, while there may be some additional columns that don't get imported.

The idea with the dataset version field is to easily revert to an older version if a serious mistake is uploaded.

This field is just distracting when adding a new dataset

• it should be convenient to upload a dataset file with the right metadata
• you should not easily forget to do that
• uploaded dataset files should be imported, creating PurchaseRecord objects
• if uploading fails, it should be clear
  • that it failed - that the dataset is not up to date
  • why it failed so we can resolve the issue

As a data admin, I would like to know whether uploads were successful so that I can tell whether it is complete or whether I need to change something and try again.

Options for satisfying that

• error feedback on creating a new dataset or updating a dataset file
  • requires access to the request variable
  • only works until an import takes longer than is acceptable for a browser form submit
• queue with error feedback and clear overview of datasets, upload success, and current versions
• do the import sync in a signal for now and just save the error message on the DatasetVersion object as we might if it was done async.

e.g.

Admin error feedback: tablib.exceptions.InvalidDimensions

e.g.

django.core.exceptions.ValidationError: ['Missing field(s): buyer_name, supplier_name, order_amount_zar']