如何从字符串中删除标点符号，然后稍后再将其添加回相同的索引？

Question

I want the program to let's say if my word_str is "This is 'Cambridge University' for example." It will keep the first and last letter of the word and scramble up the inside of the word if the length of the word is greater than 3 chars long. My problem is that it shuffles words with punctuation at the beginning or end of word incorrectly. I need it to shuffle so that the punctuation remains in the correct index and then keeps the first and last letter of the word and shuffles the inside of the word adding punctuation at the end if there is one. Any ideas?

def scramble_word(word_str):
char = ".,!?';:"
import random
if len(word_str) <= 3:
    return word_str + ' '
else:
    word_str = word_str.strip(char)
    word_str = list(word_str)
    scramble = word_str[1:-1]
    random.shuffle(scramble)
    scramble = ''.join(scramble)
    word_str = ''.join(word_str)
    new_word = word_str[0] + scramble + word_str[-1]
    return new_word + ' '

Answer 1

Using regular expressions:

import random
import re

random.seed(1234) #remove this in production, just for replication of my results

def shuffle_word(m):
    word = m.group()
    inner = ''.join(random.sample(word[1:-1], len(word) - 2))
    return '%s%s%s' % (word[0], inner, word[-1])

s = """This is 'Cambridge University' for example."""

print re.sub(r'\b\w{3}\w+\b', shuffle_word, s)

Which prints

Tihs is 'Cadibrgme Uinrtvsiey' for exlampe.

re.sub allows you to pass it a function (which accepts an regex match object) instead of a replacement string.

EDIT - without regex

from StringIO import StringIO

def shuffle_word(m):
    inner = ''.join(random.sample(m[1:-1], len(m) - 2))
    return '%s%s%s' % (m[0], inner, m[-1])

def scramble(text)
    sio = StringIO(text)
    accum = []
    start = None
    while sio.tell() < sio.len:
        char = sio.read(1)
        if start is None:
            if char.isalnum():
                start = sio.tell() - 1
            else:
                accum.append(char)
        elif not char.isalnum():
            end = sio.tell() - 1
            sio.seek(start)
            accum.append(shuffle_word(sio.read(end - start)))
            print accum[-1]
            start = None
    else:
        if start is not None:
            sio.seek(start)
            word = sio.read()
            if len(word) > 3:
                accum.append(shuffle_word(sio.read()))
            else:
                accum.append(word)

    return ''.join(accum)

s = """This is 'Cambridge University' for example."""
print scramble(s)

Answer 2

Extremly easy with a regex:

import re
import random

s = ('Pitcairn Islands, Saint Helena, '
     'Ascension and Tristan da Cunha, '
     'Saint Kitts and Nevis, '
     'Saint Vincent and the Grenadines, Singapore')

reg = re.compile('(?<=[a-zA-Z])[a-zA-Z]{2,}(?=[a-zA-Z])')

def ripl(m):
    g = list(m.group())
    random.shuffle(g)
    return ''.join(g)

print reg.sub(ripl,s)

result

Piictran Islands, Sanit Heelna, Asnioecsn and Tiastrn da Cunha, Sniat Ktits and Neivs, Snait Vnnceit and the Giearndens, Snoiaprge