camelizeKeys only rather than the full dictionary (values and keys) about humps

[robinwkurtz]

Thanks for the comment @ificiana, my issue was in fact json.dumps which I was using within the camilize function.

>>> d = {"s_gv": "j_db", "ir_o": "op_m"}

>>> print('plain', d)
plain {'s_gv': 'j_db', 'ir_o': 'op_m'}

>>> print('humps', humps.camelize(d))
humps {'sGv': 'j_db', 'irO': 'op_m'}

>>> print('json', json.dumps(d))
json {"s_gv": "j_db", "ir_o": "op_m"}

>>> print('humps.json', humps.camelize(json.dumps(d)))
humps.json {"sGv":"jDb","irO":"opM"}

>>> print('json.humps', json.dumps(humps.camelize(d)))
json.humps {"sGv": "j_db", "irO": "op_m"}

I'm trying to camelize the keys of a large dataclass before sending it through a request, and I guess the combination of the two do not work well together.

The solution in my case (in the chance that someone else is doing the same), is to wrap my dataclass in an asdict first.

dict = asdict(dataclass)
data = humps.camelize(dict)
json.dumps(data, cls=DataclassEncoder)

Sorry for the issue!

[ificiana]

Right! I see, no problem. json.dumps returns a string which camelize processes as a string in place of a mapping. So, your "json.humps" is the way to go. camelize the mapping first, then json.dumps the output. Anyway, have a nice day!

[ificiana]

@robinwkurtz Hi, thanks for opening the issue. Sorry, but I couldn't replicate the problem. It'd be nice if you could list some use cases. I did this:

>>> d = {"s_gv": "j_db", "ir_o": "op_m"}
>>> humps.camelize(d)
{'sGv': 'j_db', 'irO': 'op_m'}

camelize does appear to be converting only the keys

However, it's worth noting that if the values include a mapping, it will convert their keys as well:

>>> d = {"s_gv": "j_db", "sub_dict": {"ir_o": "op_m"}}
>>> humps.camelize(d)
{'sGv': 'j_db', 'subDict': {'irO': 'op_m'}}

Currently, there's no way to limit this behaviour but there's an old issue requesting a similar functionality here: #233