How to split a string into chunks per number of characters and delimiter?

Question

Given:

s = 'foo, bar, baz, qudz'

I would like to split this string given 2 criteria:

a character count n (in this case, I'll use n = 10)
the delimiter: ", " (before the nth character)

Desired Result:

['foo, bar', 'baz, qudz']

I know I can split it by delimiter easily enough:

s.split(', ')
['foo', 'bar', 'baz', 'qudz']

I also know that I can split it into even chunks of n like this:

[s[i:i+n] for i in range(0, len(s), n)]

I've also seen where I can split by the nth delimiter here.

What do you expect with s='foo,bar,cow' and n=5? – Venkatesh-Prasad Ranganath Feb 14 '20 at 01:44 — Venkatesh-Prasad Ranganath, Feb 14 '20 at 01:44
What about in case of s='foobar,cow' and n=5? – Venkatesh-Prasad Ranganath Feb 14 '20 at 01:46 — Venkatesh-Prasad Ranganath, Feb 14 '20 at 01:46

Venkatesh-Prasad Ranganath · Accepted Answer · 2020-02-14T02:35:37.743

2

You could use functools.reduce to accomplish this.

import functools


def splitter(s, n):
    def helper(acc, v):
        tmp1 = acc[-1] 
        tmp2 = len(tmp1)
        if tmp2 >= n or tmp2 + len(v) >= n:
            acc.append(v)
        else:
            acc[-1] = tmp1 + ',' + v

        return acc

    tmp1 = s.split(',')
    if len(tmp1) == 1:
        return tmp1

    return list(functools.reduce(helper, tmp1[1:], [tmp1[0]]))

edited Feb 14 '20 at 02:35

answered Feb 14 '20 at 02:11

Venkatesh-Prasad Ranganath

1,776
11
19

I was writing a recursive generator thing, but your solution's better. – ShapeOfMatter Feb 14 '20 at 02:21
This is very close, but there's just one thing: s = 'foo, bar, baz, qudz', n = 10 returns ['foo bar', ' baz qudz'] but I need ['foo, bar', ' baz, qudz'] – Dance Party Feb 14 '20 at 02:26
1

Changing `acc[-1] = tmp1 + v` to `acc[-1] = tmp1 + ',' + v` will address this issue (as done in the updated solution). – Venkatesh-Prasad Ranganath Feb 14 '20 at 02:37
My man! I just split it on ', ' (with a space after the comma) instead and it works great. I wish I could give this 10 "correct answers". – Dance Party Feb 14 '20 at 02:44

Todd · Answer 2 · 2020-02-14T03:22:20.997

1

You can use a regular expression to get the substrings like so...

>>> re.findall(r"[\w,][\w, ]{1,8},?", "foo, bar, baz, qudz")
['foo, bar,', 'baz, qudz']

I find myself making edits to this after I posted. Regular expressions are tricky, but this is close to a solution if not exact. Maybe some fine tuning needed. There might be a space at the end of matches - that could be trimmed off.

edited Feb 14 '20 at 03:22

answered Feb 14 '20 at 02:58

Todd

4,669
1
22
30

If the commas count as one of the ten characters, then this here is probably the best regex: re.findall(r"\w[\w, ]{1,8}(?:,|\w$)", "foo, bar, baz, qudz") It assumes that we don't want the substring starting character to be a comma. – Todd Feb 14 '20 at 06:47

How to split a string into chunks per number of characters and delimiter?

2 Answers2