Add 'head token index' attribute to words in nlp(doc)

When navigating a spaCy-processed document, e.g –

from spacy.en import English
nlp = English()
doc = nlp('The old man shook hands with the young man.')
for word in doc:
    print(word, word.tag_, word.head)

– which outputs:

The DT man old JJ man man NN shook shook VBD shook hands NNS shook with IN shook the DT man young JJ man man NN with . . shook

It would be great to have a token number stored with the head (i.e. word.headi or something like that). This would allow you to unambiguously identify the head (e.g. man_3 versus man_9 in this sentence).

[honnibal]

There’s no immediate plan to use the Edit transition. There aren’t enough speech users, and people don’t know they want disfluency detection :p.

Try training a bidirectional LSTM to pre-process. It’ll probably work really well.

[honnibal]

Each token has a .i attribute. So token.head.i should give you exactly what you want.