added requestcounts endpoint

[jmensch1]

The endpoint provides data for both of the pie charts -- request types and how people contact 311.

[sellnat77]

Im thinking that we might want to adjust this to hit data service with countFields and filters and create an aggregateQuery method in dataservice....so functionality will look like

select [countFields] from table where [filters]

Then do a pandas groupby(countFields).size()
Should follow something like this https://stackoverflow.com/a/38933130

[sellnat77]

The reason behind this is to keep the database logic separate from any of the services which will make testing/validation much easier(when we get around to it)

[jmensch1]

Ok, that makes sense, I can move the database logic into the dataservice. We might want to get away from constructing SQL queries by string concatenation though, it's kind of error-prone and hard to read. The ORM in sqlalchemy basically generates (and executes) SQL queries in a cleaner way.

[sellnat77]

Yeah i agree, we can def tweak the existing query method to use the orm instead of creating a 💩 query string

[sellnat77]

I created #398 in the meantime so whoever takes it can use your code as an example

[jmensch1]

ok cool, yeah I'll take that one. I can work on it while I'm creating the aggregationQuery in the dataservice. I like the aggregationQuery concept, could be useful for other endpoints down the line.

[sellnat77]

Awesome, thanks!!
Im hoping that with query and aggregationQuery itll keep the backend to easy categorization of services. Either getting a single thing, a lot of stuff, or a summary of groupings
That way there wont be weird complex querying per endpoint

[jmensch1]

Ok, I gave the aggregationQuery a shot, also removed SQL string concatenation and used ORM instead in the pins service and request detail service.

[sellnat77]

👏 bravo!! I like what you've done,