CODEWITHSUNDEEP
pythonstringpython-3.xpython-2.7split
andrew
andrew

Reputation: 3141

Tricky way to split string by \t

String like that:

x = dir\tsubdir1\t\tfile1.ext\t\tsubsubdir1\tsubdir2\t\tsubsubdir2\t\t\tfile2.ext

How to 'pythonish/tricky' split above string into list?

dir
\tsubdir1
\t\tfile1.ext
\t\tsubsubdir1
\tsubdir2
\t\tsubsubdir2
\t\t\tfile2.ext

['dir', '\tsubdir1', '\t\tfile1.ext', '\t\tsubsubdir1', '\tsubdir2', '\t\tsubsubdir2', '\t\t\tfile2.ext']

Prove of concept:

x = r'dir\tsubdir1\t\tfile1.ext\t\tsubsubdir1\tsubdir2\t\tsubsubdir2\t\t\tfile2.ext'
y = x.split(r'\t')
print(y)

Upvotes: 1

Views: 282

Answers (4)

kantal
kantal

Reputation: 2407

Another regex solution with findall():

x = dir\tsubdir1\t\tfile1.ext\t\tsubsubdir1\tsubdir2\t\tsubsubdir2\t\t\tfile2.ext
re.findall(r"\t+[^\t]+|[^\t]+",x)                                                                                     
Out: 
['dir',
 '\tsubdir1',
 '\t\tfile1.ext',
 '\t\tsubsubdir1',
 '\tsubdir2',
 '\t\tsubsubdir2',
 '\t\t\tfile2.ext']

Upvotes: 0

Milo Bem
Milo Bem

Reputation: 1033

import re

x = 'dir\tsubdir1\t\tfile1.ext\t\tsubsubdir1\tsubdir2\t\tsubsubdir2\t\t\tfile2.ext'

s = re.sub('([^\t])\t', '\\1\n\t', x).split('\n')

print(s)

output:

['dir', '\tsubdir1', '\t\tfile1.ext', '\t\tsubsubdir1', '\tsubdir2', '\t\tsubsubdir2', '\t\t\tfile2.ext']

Upvotes: 0

Patrick Artner
Patrick Artner

Reputation: 51683

You can do this by touching each character of your path input once + some list comp:

path = "dir\tsubdir1\t\tfile1.ext\t\tsubsubdir1\tsubdir2\t\tsubsubdir2\t\t\tfile2.ext"

l = [[]]
for c in path:
    if c != "\t":              # append to last element of list if not a \t
        l[-1].append(c)
    elif l[-1][-1] == "\t":    # also append to last element of list if it's last is a \t
        l[-1].append(c)        # (you could 'or' it into the if before)
    else:
        l.append([])           # else create a new "word" and append the \t
        l[-1].append(c)

l = [''.join(elem) for elem in l]   # join the things back together
print(l)

Output:

['dir', 
 '\tsubdir1', 
 '\t\tfile1.ext', 
 '\t\tsubsubdir1', 
 '\tsubdir2',
 '\t\tsubsubdir2', 
 '\t\t\tfile2.ext']

Before the join-step the accumulated lists look like this:

[['d', 'i', 'r'], 
 ['\t', 's', 'u', 'b', 'd', 'i', 'r', '1'], 
 ['\t', '\t', 'f', 'i', 'l', 'e', '1', '.', 'e', 'x', 't'], 
 ['\t', '\t', 's', 'u', 'b', 's', 'u', 'b', 'd', 'i', 'r', '1'], 
 ['\t', 's', 'u', 'b', 'd', 'i', 'r', '2'], 
 ['\t', '\t', 's', 'u', 'b', 's', 'u', 'b', 'd', 'i', 'r', '2'], 
 ['\t', '\t', '\t', 'f', 'i', 'l', 'e', '2', '.', 'e', 'x', 't']]

You do not want to add to strings because it creates lots of intermediate "throw-away" string instances which slows it down - usings list is much faster and less strain.

Upvotes: 2

iBug
iBug

Reputation: 37297

Maybe use a regular expression?

>>> import regex
>>> L = regex.split(r"(?<!\t)\t", "dir\tsubdir1\t\tfile1.ext\t\tsubsubdir1\tsubdir2\t\tsubsubdir2\t\t\tfile2.ext")
>>> L
['dir', 'subdir1', '\tfile1.ext', '\tsubsubdir1', 'subdir2', '\tsubsubdir2', '\t\tfile2.ext']
>>> L[:1] + ['\t' + i for i in L[1:]]
['dir', '\tsubdir1', '\t\tfile1.ext', '\t\tsubsubdir1', '\tsubdir2', '\t\tsubsubdir2', '\t\t\tfile2.ext']

How does it work?

The regular expression is

(?<!\t)\t

which means "a tab that's not preceded by another tab", so every first tab in a sequence of tags is matched by the regex. It's then used as the splitting mark.

After splitting, one tab is stripped from every subsequent items, so the last line L[:1] + ['\t' + i for i in L[1:]] prepends the missing tab back.

Upvotes: 3

Related Questions