Add float precision variable to csv.writer / reader

[chaeynz]

Feature or enhancement

Proposal:

var_float_precision = 2

def format_item(item):
        if isinstance(item, float):
            return f"{item:.(var_float_precision)f}"

I just now have the situation, that I want to print a 0 as 0.00, because the ERP system only detects it that way.
Please add the ability to define a custom precision for float numbers. Just recently the QUOTE_STRINGS was introduced into python, but when I manually format it to a 0.00 using {:.2f}".format(var)) it will obviously become a string, therefore it will print the variable in Quotes, which the ERP system again, does not detect.

Thank you!

Has this already been discussed elsewhere?

This is a minor feature, which does not need previous discussion elsewhere

Links to previous discussion of this feature:

No response

[serhiy-storchaka]

gh-155097 (PR #155099) adds a formatter parameter to csv.writer which replaces the str() call. Quoting is still decided by the original value, so a formatted number stays unquoted — which is exactly what '{:.2f}'.format(v) in your own code cannot do:

>>> csv.writer(sys.stdout, quoting=csv.QUOTE_NONNUMERIC,
...            formatter=lambda i, v: format(v, '.2f')).writerow([0.0, 'text', 12.3456])
0.00,"text",12.35

The first argument is the 0-based column index, so the precision can differ per column. Strings and None never reach the formatter.