Allow create database in python

[flywire]

close #22

[pfalcon]

This PR is hosed by now and should be reworked.

It should be an option to create a database, default behavior of the application shouldn't change.

[flywire]

Agree. There are some conflicts with #24 so let's see your thoughts on that first.

Otherwise, prune it after e1b2948 (possibly)

[flywire]

It should be an option to create a database, default behaviour of the application shouldn't change.

Only two commits are relevant:

1. df5e2ec - Allow create database in python
2. e1b2948 - Enable create database in python

I understand you are supportive of the first commit but not the second. How would you enable the first?

[pfalcon]

Unused import, and please don't add any dependencies on pathlib (it doesn't do anything which Python couldn't do 20 years ago).

[flywire]

As you say, code was in fork I adapted.

[pfalcon]

What I meant is that there should be option like --createdb which would trigger execution of schema scripts. Implementation wise, it then probably should be a separate function like init_schema() in dbhelper.

I'm also not a fan of all the novelties like pathlib and f-strings. Whereas I'm a fan of minimalism, so prefer to rely on decades-known features vs recent novelties (where ones being replaced by others in rat-race).

[flywire]

I'm also not a fan of all the novelties like pathlib and f-strings.

Do you consider this PR worth accepting if I rework based on the two commits and a --createdb option leaving you to tweak to your preferred coding style?

[pfalcon]

@flywire

Do you consider this PR worth accepting

I remember it was one of your older concern that Makefile is used to create a database. It's not a concern to me, but I'm happy to listen to contributors who are willing to do legwork to change that (I myself wouldn't be willing to do legwork, because it already works and doesn't change much IMHO).

if I rework based on the two commits and a --createdb option leaving you to tweak to your preferred coding style?

Sorry, I unlikely would be able to tweak it any time soon, because I have hundreds of my own patches to tweak to make them suitable for upstreaming (e.g., for PP itself, but not only). I'm happy to provide review comments and suggestions, that's how open source contribution process usually works.

Thanks.

[flywire]

Only two commits are relevant:

1. df5e2ec - Allow create database in python
2. e1b2948 - Enable create database in python

I understand you are supportive of the first commit but not the second. How would you enable the first?

Your response is still not clear to me. Would you be supportive of the first (Allow create database in python) without the second (Enable create database in python)? It would be trivial for technical users to flip the boolean without stuffing around with options etc.