Python大十进制舍入问题

Question

I am trying to store

x=9999999999999.9999999999 

in a string variable. But it is getting round off.

c=str(9999999999999.9999999999)
print c
'1e+13'

Here I am trying to read a JSON file and converting it to CSV... But while reading 9999999999999.9999999999 it is automatically converted into 1e+13.

with open(inputfilename) as json_file:  
    data = json.load(json_file)
    selected_month=str(data['selected_month'])
    selected_year=str(data['selected_year'])
    record_count=str(data['record_count'])      
    for p in data['monthly_rates']:
        source_iso_currency_cd=p['source_iso_currency_cd']
        target_iso_currency_cd=p['target_iso_currency_cd']
        source_currency_cd=p['source_currency_cd']
        target_currency_cd=p['target_currency_cd']
        indirect_rate=p['indirect_rate']
        if indirect_rate == None:
            o_indirect_rate=""
        else:
            o_indirect_rate=repr(indirect_rate)

        direct_rate=p['direct_rate']
        if direct_rate == None:
            o_direct_rate=""
        else:
            o_direct_rate=repr(direct_rate)
        average_rate=p['average_rate']
        if average_rate == None:
            o_average_rate=""
        else:
            o_average_rate=repr(average_rate)

        o_str1=record_count+','+selected_month+','+selected_year+','+source_iso_currency_cd+','+target_iso_currency_cd+','+source_currency_cd+','+target_currency_cd+','+o_indirect_rate+','+o_direct_rate+','+o_average_rate

Answer 1

json normally parses float numbers into Python's float numbers, which cannot represent the precision you desire.

Use either the built-in decimal , or another library, eg mpmath package's arbitrary-precision floats. You will need to tell json to use a different function for parsing floating-point numbers.

from decimal import Decimal
json.load(json_file, parse_float=Decimal)