将浮点数与逗号匹配，然后将逗号替换为点号？

Question

I have a text file which contains some different types of numbers: integers, binary and float. I want to match only float numbers and replace comma with dot.

example of my text file (there order is casual):

1000101 33434,34 1992 [3,41,4,5] 

after conversion :

1000101 33434.34 1992 [3,41,4,5] 

my code is :

lines = []
in_file = open("input.txt", "r")
for line in in_file:
    line = line.split(" ")
    for x in line:
        try:
            if isinstance(float(x.replace(',', '.')), float):
                line[line.index(x)] = float(x.replace(',', '.'))
        except:
            pass
    lines.append(line)
in_file.close()

but that will convert all other data to float, so what is the best way to resolve this?
I thought to use regex but I don't know how to do it in python.

Answer 1

Another approach, using regular expressions as well:

import re
with open('input.txt', 'r+') as f:
    newf = re.sub(r'(\s+[+-]?[0-9]+),([0-9]+\s+)',r'\1.\2', f.read())
    f.seek(0)
    f.write(newf)

test file:

1000101 33434,34 1992 [3,41,4,5] 
12,43 129012 91 [1,2]
1000101 33434,34 1992 [3, 41,4,5] 

result:

1000101 33434.34 1992 [3,41,4,5] 
12.43 129012 91 [1,2]
1000101 33434.34 1992 [3, 41,4,5] 

Answer 2

Try this:

import re
from ast import literal_eval

FLOAT_RE = re.compile('^\d+,\d+$')

lines = []
with open("input.txt", "r") as in_file:
    for line in in_file:
        line = line.strip().split(" ")
        for x in line:
            i = line.index(x)
            if FLOAT_RE.match(x):
                x = x.replace(',', '.')
            line[i] = literal_eval(x)
        lines.append(line)

print lines

Answer 3

this should work for you:

lines = []
in_file = open("input.txt", "r")
for line in in_file:
    line = line.split(" ")

    for x in line:
        tmp = eval(x)
        if isinstance(tmp, tuple):
            line[line.index(x)] = float(float(x.replace(',', '.')))
        else:
            line[line.index(x)] = tmp

    lines.append(line)
in_file.close()

it will convert everything to the right type

Answer 4

if all your strings are in the same format you can sub only the first occurrence of , :

s = "1000101 33434,34 1992 [3,41,4,5]"

print re.sub(",",".",s,1)
1000101 33434.34 1992 [3,41,4,5]