如何建立文字中的单词词典

Question

How would I return a dictionary with the key being a word in the given text and the values being a list of previous words in the text?

eg

text = "hi my name is"    
get_previous_words_dict(text):

prints a dictionary:

>>> my_dict['hi']
[]
>>> my_dict['my']
['hi']    
>>> my_dict['name']
['hi', 'my']

Answer 1

This only makes sense if the words in the sentence are unique, as @cjds points out. Also, the value for the first word should surely be an empty list, not a list containing the empty string. The following will fit this specification:

def get_previous_words_dict(text):
    words = []
    dictionary = {}
    for word in text.split():
        dictionary[word] = words[:]
        words.append(word)
    return dictionary

The most important thing to understand is the assignment:

dictionary[word] = words[:]

The effect of this is to copy the words array. If it was a normal assignment:

dictionary[word] = words

Then that would just make each dictionary entry refer to the same words list, and so at the end of the loop every entry in the dictionary would have all of the words.

Answer 2

>>> t="hi my name is"
>>> li=t.split()

You can use a dict comprehension:

>>> {w:[li[si] for si in range(i-1,-1,-1)] for i, w in enumerate(li)}
{'is': ['name', 'my', 'hi'], 'hi': [], 'my': ['hi'], 'name': ['my', 'hi']}

Or, counting up:

>>> {w:[li[si] for si in range(0,i)] for i, w in enumerate(li)}
{'is': ['hi', 'my', 'name'], 'hi': [], 'my': ['hi'], 'name': ['hi', 'my']}

Or use a slice instead of the nested list comprehension:

>>> {w:li[0:i] for i, w in enumerate(li)}
{'is': ['hi', 'my', 'name'], 'hi': [], 'my': ['hi'], 'name': ['hi', 'my']}

Answer 3

If I were to implement from scratch:

Use a hash to store words, this used as dictionary. When insert into hash, insert as key => [previous keys in hash].

Answer 4

1. Split the sentence into words:

    sentence_words = sentence.split(' ') 

2. Create a dictionary where the key is the word, and the value is a slice of sentence_words from the beginning to the position of this word.

    d = {w: sentence_words[:i] for i, w in enumerate(sentence_words)} 

Sample Code

sentence = "Hi my name is John"
sentence_words = sentence.split(' ')
d = {w: sentence_words[:i] for i, w in enumerate(sentence_words)}