CSV : decimal separator about pyexcel

[jgonggrijp]

May I request that this ticket be reopened. pyexcel is actually in control here, in the sense that it is doing an extra number coercion step beyond what the builtin csv module is doing. That extra step is preventing people like me and @doudz, who need to work with decimal commas, from reading the numbers correctly. Consider the following CSV:

"2023-07-29";"0,02"

If I do this:

import csv

with open('the-above-file.csv') as my_file:
    print(next(csv.reader(my_file, delimiter=';')))

I get ('2023-07-29', '0,02') as output. The second field is not interpreted, I can do a string replacement to replace the decimal comma by a decimal point.

However, if I do this:

import pyexcel

print(pyexcel.get_array(file_name='the-above-file.csv', delimiter=';')[0])

I get ['2023-07-29', 2]. The second field has already been interpreted and the position of the decimal comma has been lost.

I can understand if you cannot or don't want to support decimal commas, but at least please give us a way to switch off automatic number coercion so we can address it ourselves. As it currently is, I have to revert to the standard csv module.

[jgonggrijp]

@CodeWithOz I didn't use iget_records. I'm fortunate enough that I only need to work with small spreadsheets.

[chfw]

that has to be with your data presentation. pyexcel or csv module are not in control.

[doudz]

Yes, I have converted float to string and then replace dot by comma

[CodeWithOz]

@jgonggrijp were you using pyexcel's iget_records method? If so, how did you replace its functionalty with the standard csv module? Asking because I'm using iget_records for processing large files without loading them all into memory, so I need a solution that won't blow out my memory if I switch away from pyexcel.