How can I replace a text delimited string list item with multiple list items in a python list?

Question

Given a list:

mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']

I'd like a one-liner to return a new list:

['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

Answer 1

Another trick is first to join the list with underscores and then re-split it:

"_".join(mylist).split('_')

Answer 2

Just use 2 for clauses in your comprehension, e.g.:

>>> mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']
>>> [animal for word in mylist for animal in word.split('_')]
['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

Answer 3

This is not a one liner, but is nevertheless a valid option to consider if you want to return a generator:

def yield_underscore_split(lst):
     for x in lst:
         yield from x.split('_')

>>> list(yield_underscore_split(mylist))
['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

---

Original answer valid only for versions Python 3.3-3.7, kept here for interested readers. Do not use!

>>> list([(yield from x.split('_')) for x in l]) 
['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

Answer 4

using the itertools recipe to flatten a list you could do this:

from itertools import chain

mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']

new_list = list(chain.from_iterable(item.split('_') for item in mylist))
print(new_list) 
# ['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

...or does the import statement violate your one-liner requirement?

Answer 5

Since so many answers here were posted (over ten), I thought it'd be beneficial to show some timing stats to compare the different methods posted:

-----------------------------------------
AChampion time: 2.6322
-----------------------------------------
hiro_protagonist time: 3.1724
-----------------------------------------
Eugene_Sh time: 1.0108
-----------------------------------------
cᴏʟᴅsᴘᴇᴇᴅ time: 3.5386
-----------------------------------------
jdehesa time: 2.9406
-----------------------------------------
mogga time: 3.1645
-----------------------------------------
Ajax1234 time: 2.4659
-----------------------------------------

Here's the script I used to test:

from timeit import timeit

setup = """
from itertools import chain
mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']
"""

methods = {
    'AChampion': """[animal for word in mylist for animal in word.split('_')]""",
    'hiro_protagonist': """list(chain.from_iterable(item.split('_') for item in mylist))""",
    'Eugene_Sh': """'_'.join(mylist).split('_')""",
    'cᴏʟᴅsᴘᴇᴇᴅ': """list([(yield from x.split('_')) for x in mylist])""",
    'jdehesa': """sum((s.split("_") for s in mylist), [])""",
    'mogga': """[i for sublist in [j.split('_') for j in mylist] for i in sublist]""",
    'Ajax1234': """list(chain(*[[i] if "_" not in i else i.split("_") for i in mylist]))"""
}

print('-----------------------------------------')
for author, method in methods.items():
    print('{} time: {}'.format(author, round(timeit(setup=setup, stmt=method), 4)))
    print('-----------------------------------------')

Each method is tested against the sample list given in the question about one million times. To keep things readable, each timing result was rounded to four decimal places.

---

Note: If you have a new, unique method that has not been posted here yet, contact me in the comments and I'll try to add a timing for it too.

Answer 6

Split each item into sublists and flatten them:

[item for sublist in mylist for item in sublist.split("_")]

Answer 7

One-liners are over-rated. Here's a solution using a "traditional" for loop.

mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']

out = []
for s in mylist:
    if '_' in s:
        out.extend(s.split('_'))
    else:
        out.append(s)

print(out)

output

['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

This also works:

out = []
for s in mylist:
    out.extend(s.split('_'))

It's shorter, but I think the previous version is clearer.

Answer 8

You can do:

mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']
result = sum((s.split("_") for s in mylist), [])
print(result)
>>> ['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

Answer 9

This works:

[i for sublist in [j.split('_') for j in mylist] for i in sublist]

Answer 10

mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']
animals = [a for item in mylist for a in item.split('_')]
print (animals)

Answer 11

You can try this:

from itertools import chain

mylist = ['dog', 'cat', 'mouse_bear', 'lion_tiger_rabbit', 'ant']

new_list = list(chain(*[[i] if "_" not in i else i.split("_") for i in mylist]))

Output:

['dog', 'cat', 'mouse', 'bear', 'lion', 'tiger', 'rabbit', 'ant']

Answer 12

what I would actually do:

newlist = []

for i in mylist:
    newlist += i.split('_')