Using multiple (similar) generator expressions

Question

In a csv file, I'm trying to replace certain characters with other characters.

My current code is similar to this:

import csv

set1 = set('abc')
set2 = set('def')
set3 = set('ghi')

with open(path, 'r') as input, open(path2, 'w') as output:
    reader = csv.reader(input)
    writer = csv.writer(output)

    for row in reader:
            newrow = row
            newrow = [''.join('x' if c in set1 else c for c in item) for item in newrow]
            newrow = [''.join('y' if c in set2 else c for c in item) for item in newrow]
            newrow = [''.join('z' if c in set3 else c for c in item) for item in newrow]

            writer.writerow(newrow)

In this example I am only using three generator expressions, but it could easily be more than that.

Does anyone know the right way to do this? My concern is that this structure might not be the fastest (and certainly doesn't look optimal).

Answer 1

str.translate might be appropriate; something along the lines of

replacements = [
    ('abc', 'x'),
    ('def', 'y'),
    ('ghi', 'z'),
]

trans = str.maketrans({ k: v for l, v in replacements for k in l })

and

new_row = [item.translate(trans) for item in row]

Answer 2

Here is something that combines both answers in some way and works great (multiple times faster than the code in the question):

replacements = [
('abc', 'x'),
('def', 'y'),
('ghi', 'z'),
]

mapping = {a: b for c, b in replacements for a in c}

for row in reader:
    newrow = [''.join(mapping.get(c, c) for c in item) for item in row]
    writer.writerow(newrow)

Answer 3

You can use a loop and parameterise the varying parts:

newrow = row
for v, s in (('x', set1), ('y', set2), ('z', set3)):
    newrow = [''.join(v if c in s else c for c in item) for item in newrow]

If you are replacing characters, don't use sets but a mapping:

mapping = dict.fromkeys(set1, 'x')
mapping.update(dict.fromkeys(set2, 'y'))
mapping.update(dict.fromkeys(set3, 'z'))
for row in reader:
    newrow = [''.join(mapping.get(c, c) for c in item) for item in newrow]