Write (and continuously update) documentation structured as a "how to" tutorial (or Problem/Solution).
Do this in a way that is testable so we assure that the docs are correct.
That is: Figure out how to test python from md docs, or generate them from testable code (see test2doc for example).
Figure out a way to mitigate imports visual noise. For example, using from mongodol import * and possibly any other needed imports at the top?
Here are questions to start with:
- How do I add full (i.e. containing
_id fields) docs to a collection, controlling for unicity (overwrites allowed or not, with or without warnings...)
- How do I add a doc that doesn't have a
_id?
- How can I control
_id creation, and possibly other fields created automatically at insertion (append?) time?
- How can I get interfaces such as
list.append, list.extend, set.add, or dict.update to mongo?
- tuple keys:
s[i,j]
- Slicing:
s[:, j], s[i_query, j]
Write (and continuously update) documentation structured as a "how to" tutorial (or Problem/Solution).
Do this in a way that is testable so we assure that the docs are correct.
That is: Figure out how to test python from md docs, or generate them from testable code (see test2doc for example).
Figure out a way to mitigate imports visual noise. For example, using
from mongodol import *and possibly any other needed imports at the top?Here are questions to start with:
_idfields) docs to a collection, controlling for unicity (overwrites allowed or not, with or without warnings...)_id?_idcreation, and possibly other fields created automatically at insertion (append?) time?list.append,list.extend,set.add, ordict.updateto mongo?s[i,j]s[:, j],s[i_query, j]