Handling of None / null fields in to_json

[JeffPaine]

It appears that micromodels does not (in my humble opinion) properly handle None / null fields when using to_json.

Modifying micromodels to properly return null when using to_json appears fairly straightforward (and I would happily submit a pull request). Is there some good reason for using the current approach that I'm not aware of? Thanks,

Example Code

import micromodels

class Everything(micromodels.Model):
    '''A test of every field type'''

    charfield = micromodels.CharField()
    integerfield = micromodels.IntegerField()
    floatfield = micromodels.FloatField()
    booleanfield = micromodels.BooleanField()
    datetimefield = micromodels.DateTimeField(format='%m')
    datefield = micromodels.DateField(format='%m')
    timefield = micromodels.TimeField(format='%m')

empty_data = {
    'charfield': None,
    'integerfield': None,
    'floatfield': None,
    'booleanfield': None,
    'datetimefield': None,
    'datefield': None,
    'timefield': None,
}

g = Everything.from_dict(empty_data)
print g.to_json()

Expected Result

{"datefield": null, "integerfield": null, "datetimefield": null, "floatfield": null, "timefield": null, "charfield": null, "booleanfield": null}

Actual Result

charfield returns ""
integerfield returns 0
floatfield returns 0.0
booleanfield returns False
datetimefield returns AttributeError: 'NoneType' object has no attribute 'isoformat'
datefield returns TypeError: NoneType could not be serialized by DateField
timefield returns TypeError: NoneType could not be serialized by TimeField

[JeffPaine]

@j4mie Is this project still active? Just a friendly, sincere question. Thanks very much for your time.

[j4mie]

Hi Jeff,

Many thanks for opening the issue. I agree that the case of null data vs missing keys in the data should probably be handled better.

Unfortunately I'm not actively maintaining micromodels at the moment. The project that I was using it for has now ended and so it's difficult for me to find the time to support it. I'd recommend looking at https://github.com/j2labs/dictshield or https://github.com/saymedia/remoteobjects as potential alternatives.

If you'd like to open a pull request I'll take a look at it when I can, but I can't promise when that will be. Alternatively you can fork the project and install your version with pip install git+git://github.com/JeffPaine/micromodels.git.

Cheers

[JeffPaine]

Hello J4mie,

Unfortunately I'm not actively maintaining micromodels at the moment. The project that I was using it for has now ended and so it's difficult for me to find the time to support it.

I totally understand. Thanks very much for filling me in. Surely I would do the same if I was in your position.

Alternatively you can fork the project and install your version with pip install git+git://github.com/JeffPaine/micromodels.git.

Ahh, I did not realize that, nice trick! I'll look into this as an appropriate route forwards, or the other projects you mentioned.

Thanks again for your time,

[ericmoritz]

Hi JeffPaine,

I am sure you have moved on by now but if you still want to work on this, or have solved this problem in a fork. I'd like to include your solution in the next release.

[JeffPaine]

Hello @ericmoritz ,

Thanks much for following up. I ultimately didn't end up attempting to fix this issue. I decided to go with a different solution. I appreciate you checking into it though!