Python Bottle json_dumps - pass parameter to default handler function

Question

How can I pass arguments to a default handler function while converting data to JSON using Bottle's json_dumps?

EDIT: I missed a point where I convert the data from json_dumps, back to the previous format using json_loads. The reason I use json_dumps is because i need to convert the datetime format to string

I have the result of a MySQL query as list of tuples:

data = [(u'user1', u'Topic1', datetime.datetime(2015, 8, 3, 23, 55), 2.0, 5), (u'user2', u'Topic2', datetime.datetime(2015, 8, 4, 23, 55), 3.0, 5)]

It contains some data in datetime format. And I send this data as response to an AJAX call, in JSON format, because of which I perform json_dumps on it.

Now, I can't simply perform json_dumps(data) as it gives this error:

TypeError: datetime.datetime(2015, 8, 3, 23, 55) is not JSON serializable

So, I define this handler function and use it as follows:

def dataHandler(obj):
    if isinstance(obj, datetime):
        return obj.strftime('%Y-%m-%d %H:%M')

json_dumps(data, default=dataHandler)

This works fine and the output is:

'[["user1", "Topic1", "2015-08-03 23:55", 2.0, 5], ["user2", "Topic2", "2015-08-04 23:55", 3.0, 5]

Now at different points in my code, for different data, I need different formats of datetime. So, I redefined the function like this:

def dataHandler(obj, showTime='no'):
    if isinstance(obj, datetime):
        if str(showTime).lower() == 'no':
            return obj.strftime('%Y-%m-%d')
        elif str(showTime).lower() == 'yes':
            return obj.strftime('%Y-%m-%d %H:%M')

Now if I perform json_dumps(data, default=dataHandler), it works properly, considers showTime as no and gives output same as above.

The problem comes when I try to pass it an argument:

json_dumps(data, default=dataHandler('Yes'))

It gives this error:

TypeError: datetime.datetime(2015, 8, 10, 23, 55) is not JSON serializable

How can I have different such cases defined in the same function?

Thanks.

bufh · Accepted Answer · 2015-08-17 14:43:05Z

You could use a @decorator (PEP-318).

Example:

def json_handler(show_time=True):
    def decorated(obj):
        if isinstance(obj, datetime.datetime):
            return obj.strftime('%FT%T' if show_time else '%F')
        if isinstance(obj, datetime.date):
            return obj.strftime('%F')
        return repr(obj) # catch-all
    return decorated

Then:

>>> json.dumps(data, default=json_handler())
'[["user1", "Topic1", "2015-08-03T23:55:00", 2.0, 5], ["user2", "Topic2", "2015-08-04T23:55:00", 3.0, 5]]'
>>> json.dumps(data, default=json_handler(False))
'[["user1", "Topic1", "2015-08-03", 2.0, 5], ["user2", "Topic2", "2015-08-04", 3.0, 5]]'

Edit

The line return repr(obj) # catch-all will use the representation of the object instead of refusing to dump it (ie: "<object object at 0x7f934f6ef0c0>" instead of TypeError: <object object at 0x7f934f6ef0e0> is not JSON serializable); this might or might not be desired.

Do not forget to call the decorator!
Failing to do so (ie: json.dumps(data, default=json_handler) instead of json.dumps(data, default=json_handler())) will result in a RuntimeError: maximum recursion depth exceeded while calling a Python object or Segmentation Faults.

Collectives™ on Stack Overflow

Python Bottle json_dumps - pass parameter to default handler function

1 Answer 1

Edit

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Edit

Comments

Your Answer

Sign up or log in

Post as a guest

Related