Find word on given position in text

there is more elega开发者_如何学运维nt (pythonic + effective) way to find word on given position?

FIRST_WORD = re.compile(r'^(\w+)', re.UNICODE) 
LAST_WORD = re.compile(r'(\w+)$', re.UNICODE)

def _get_word(self, text, position):
    """
    Get word on given position
    """
    assert position >= 0
    assert position < len(text)

    # get second part of word
    # slice string and get first word
    match = FIRST_WORD.search(text[position:])
    assert match is not None
    postfix = match.group(1)

    # get first part of word, can be empty
    # slice text and get last word
    match2 = LAST_WORD.search(text[:position])
    if match2 : prefix = match2.group(1)
    else : prefix = ''

    return prefix + postfix


#                                  | 21.
>>> _get_word("Hello, my name is Earl.", 21)
Earl
>>> _get_word("Hello, my name is Earl.", 20)
Earl

Thanks

---

Here's how I'd do it:

s = "Hello, my name is Earl."
def get_word(text, position):
    words = text.split()
    characters = -1
    for word in words:
        characters += len(word)
        if characters > = position:
            return word
>>> get_word(s, 21)
Earl.

Stripping off the punctuation can be done with ''.strip() or regular expressions or something hacky like

for c in word:
    final += c if c.lower() in 'abcdefghijklmnopqrstuvwxyz'

---

import string

s = "Hello, my name is Earl."
def get_word(text, position):
    _, _, start = text[:position].rpartition(' ')
    word,_,_ = text[position:].partition(' ')
    return start+word

print get_word(s, 21).strip(string.punctuation)

---

The following solution is to get the alpha characters around the given position:

def get_word(text, position):
    if position < 0 or position >= len(text):
        return ''

    str_list = []

    i = position
    while text[i].isalpha():
        str_list.insert(0, text[i])
        i -= 1

    i = position + 1
    while text[i].isalpha():
        str_list.append(text[i])
        i += 1

    return ''.join(str_list)

The following is a test case:

get_word("Hello, my name is Earl.", 21)  # 'Earl'
get_word("Hello, my name is Earl.", 20)  # 'Earl'

I don't think it is a good idea to split the text into words with the split function here, because position is essential for this problem. If there are continuous blanks in a text, the split function may cause troubles.