improve data download feature

[smcalilly]

Overview

For #816 and #807

Includes:

• additional make recipes for filtering data, based on the list in #816 comment
• updated the download template with text in #816 comment
• removed the old download buttons
• added nav link to download page
• blank readme for demo (per tom's instruction in an email)
• new import docket with country name
• updated the download data import to use country name

Notes

The data archive is a flat directory. The recipes write to <country_name>_<entity_type>.csv instead of creating a new directory with the country name. I couldn't figure out how to cleanly create a new directory based on the dynamic country name values. Let me know if you see a way to do this?

I had to create a custom processing script to blank out the values for two specific columns based on a discussion in our slack #computer-programming channel. For context: csvgrep can only remove the columns. csvsql can change the column values, but csvsql would completely remove the header row from a file that had no data rows, thus creating a blank file which isn't something we want.

Testing Instructions

• get aws s3 token from wwic-data-archive-staging token in last pass
• cp .env.s3.example .env and add your AWS access tokens
• for testing purposes, you might need to remove the myanmar row from the import docket, since that spreadsheet naming convention will break the import (this issue has been mentioned in an email, to be fixed)
• to create an archive locally, run docker-compose --env-file .env.s3 run --rm app make data_archive
• visit localhost:8000/en/download/ to download the data

[hancush]

Hey, @smcalilly, great refactor! Once you respond to my super minor comments, let's get this up on staging for review.

[smcalilly]

@hancush I made that python change, thanks for that.

When I added the individual sources file to the pipeline, it was sometimes missing a column and erroring the blank_columns.py script. So I did some explicit checks for that:

sfm-cms/data/processors/blank_columns.py

Lines 17 to 32 in 5a944f6

	for row in reader:
	comment_key = f'{args.entity}:comments:admin'
	comments = row.get(comment_key)
	if comments:
	row.update({
	comment_key: ''
	})
	owner_key = f'{args.entity}:owner:admin'
	owner = row.get(owner_key)
	if owner:
	row.update({
	owner_key: ''
	})

Not the cleanest but it works.