CODEWITHSUNDEEP
pythonstring
Safirah
Safirah

Reputation: 375

Python: How to sort the letters of a string keeping some characters in place?

I've already asked this question, but it was for Ruby, now it's Python's turn! I want to sort the words of a string, keeping non-alphanumeric characters in place, for example:

"hello, sally! seen 10/dec/2016 => ehllo, allsy! eens 01/cde/0126"

Based in the answer I've previously received, I've tried to do:

def sortThisList(listWords):
    for word in listWords:
        print(re.sub('\W+', sortStr(word), word)) #Error

def sortStr(word):
    return "".join(sorted(list(word)))

But this error pops up:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 3, in sortItAll
  File ".../lib/python3.6/re.py", line 191, in sub
    return _compile(pattern, flags).sub(repl, string, count)
TypeError: expected string or bytes-like object

Not anymore, thank you ^^. But it's still not sorting properly.

Upvotes: 0

Views: 290

Answers (2)

Martijn Pieters
Martijn Pieters

Reputation: 1121524

You are trying to apply the regular expression to the whole list, not the individual word:

for word in textInaList:  # textInaList presumably is a list
    print(re.sub('\W+', sortStr(word), textInaList))
#    you pass that list into re.sub(): ^^^^^^^^^^^

Next, you want to pass in a your sortStr function if you want it to be used for each replacement, and have that function deal with a match object. You'll also want to replace \w+ (word characters), not non-word characters:

def sortStr(match):
    return "".join(sorted(match.group()))

print(re.sub('\w+', sortStr, sentence))

When you pass in a function as the second argument to re.sub(), it is called for every match found in the third argument, passing in a match object; calling match.group() returns the matched text (so a single word in this case). The return value is then used as the replacement.

Demo:

>>> import re
>>> def sortStr(match):
...     return "".join(sorted(match.group()))
...
>>> sentence = "hello, sally! seen 10/dec/2016"
>>> re.sub('\w+', sortStr, sentence)
'ehllo, allsy! eens 01/cde/0126'

Upvotes: 3

sirfz
sirfz

Reputation: 4277

It makes more sense to match consecutive alphanumeric characters, sort them and replace the original words while keeping all other characters untouched. That is:

In [25]: s = "hello, sally! seen 10/dec/2016"

In [26]: ns = s

In [27]: for w in re.findall(r'\w+', s):
    ...:     ns = ns.replace(w, "".join(sorted(w)))
    ...:     

In [28]: ns
Out[28]: 'ehllo, allsy! eens 01/cde/0126'

Upvotes: 1

Related Questions