Execution of distributive property using regex

user_13859899 0
Viewed 94

I am trying to change the order using regex, but running into a problem with alternative groups.

Example: 1a,1b,1c ABC
Output: ABC*1a,ABC*1b,ABC*1c

Example: 1a,2b ABC
Output: ABC*1a,ABC*2b

Example: 1a ABC
Output: ABC*1a

So far what I have is:

re.sub(r'((\d\W,\d\W,\d\W)|(\d\W,\d\W)|(\d\W))()(\d\W\d\W\d\W)',r'\3\*\1',string)

Any ideas on how to deal with the possible subgroups?

regex-group python regex
4 Answers

I see that the question is about regular expressions, but what as @Ronald stated using them to entirely swap the content would be rather difficult. This is an example how you could use them and combine them with python methods to achieve the results that you are looking for:

import re

def parse_string_1(str_):
    prog1 = re.compile('([\d\w]*)[\,\s]')
    prog2 = re.compile('\s(\w*)')

    constant = re.findall(prog2, str_)[0]
    multipliers = re.findall(prog1, str_)

    result = []
    for multiplier in multipliers: 
        result.append(f"{constant}*{multiplier}")
    return ','.join(result)

parse_string_1("1a,1b,1c ABC")
>>> ABC*1a,ABC*1b,ABC*1c
parse_string_1("1a,2b ABC")
>>> ABC*1a,ABC*2b
parse_string_1("1a ABC")
>>> ABC*1a

If you can leave them completely you could do this:

def parse_string_2(str_):
    result = []
    multipliers, constant = str_.split(' ')
    for multiplier in multipliers.split(','):
        result.append(f"{constant}*{multiplier}")
    return ','.join(result)

parse_string_2("1a,1b,1c ABC")
>>> ABC*1a,ABC*1b,ABC*1c
parse_string_2("1a,2b ABC")
>>> ABC*1a,ABC*2b
parse_string_2("1a ABC")
>>> ABC*1a
user_133901390

Without using regular expressions, you may consider to use itertools.product:

from itertools import product
data=['1a,1b,1c ABC', '1a,2b ABC', '1a ABC']

for s in data:
    lst = [e.split(',') for e in s.split()]
    for prod in product(*lst):
        print('{}*{}'.format(prod[1], prod[0]))
user_134925610

Maybe something like:

import re
str = '1a,1b,1c ABC'
print(re.sub(r'(?=\d)', str.split()[1]+'*', str.split()[0]))

Result:

ABC*1a,ABC*1b,ABC*1c
user_97581940

For that data structure, you might also use the regex PyPi module and the \G anchor

(?:\G(\d[a-zA-Z]+,?)| ([A-Z]+))

Explanation

  • (?: Non capture group
    • \G Assert the end of the previous match
    • (\d[a-zA-Z]+,?) Capture group 1, match 1+ times a char a-zA-Z and optional ,
    • | Or
    • ([A-Z]+) Capture group 1, match a space and 1+ chars A-Z
  • ) Close group

Regex demo | Python demo

import regex

strings = [
    "1a,1b,1c ABC",
    "1a,2b ABC",
    "1a ABC"
]

for s in strings:
    tuples = regex.findall(r"(?:\G(\d[a-zA-Z]+,?)| ([A-Z]+))", s)
    prefix = ''.join(tuples.pop())
    result = "".join(f"{prefix}*{t[0]}" for t in tuples)
    print(result)

Output

ABC*1a,ABC*1b,ABC*1c
ABC*1a,ABC*2b
ABC*1a
user_54249880
Related