Special handling of "0:root" labels in deps column

[jbrry]

I recently upgraded from conllu 1.3.1 to 2.2 due to the latter version's ability to deal with elided tokens/copy nodes (e.g. token 8.1 below) which was addressed in https://github.com/EmilStenstrom/conllu/issues/27.

I am parsing the deps column and have a loop which iterates over the deps tuples to put the heads into a heads list and the relations into a relations list. The upgrade now includes the copy nodes which is good but now all 0:root labels are returned as a string and not a tuple which breaks my loop.

# sent_id = weblog-blogspot.com_healingiraq_20040409053012_ENG_20040409_053012-0022
# text = Over 300 Iraqis are reported dead and 500 wounded in Fallujah alone.
1   Over    over    ADV RB  _   2   advmod  2:advmod    _
2   300 300 NUM CD  NumType=Card    3   nummod  3:nummod    _
3   Iraqis  Iraqis  PROPN   NNPS    Number=Plur 5   nsubj:pass  5:nsubj:pass|6:nsubj:xsubj|8:nsubj:pass _
4   are be  AUX VBP Mood=Ind|Tense=Pres|VerbForm=Fin    5   aux:pass    5:aux:pass  _
5   reported    report  VERB    VBN Tense=Past|VerbForm=Part|Voice=Pass 0   root    0:root  _
6   dead    dead    ADJ JJ  Degree=Pos  5   xcomp   5:xcomp _
7   and and CCONJ   CC  _   8   cc  8:cc|8.1:cc _
8   500 500 NUM CD  NumType=Card    5   conj    5:conj:and|8.1:nsubj:pass|9:nsubj:xsubj _
8.1 reported    report  VERB    VBN Tense=Past|VerbForm=Part|Voice=Pass _   _   5:conj:and  CopyOf=5
9   wounded wounded ADJ JJ  Degree=Pos  8   orphan  8.1:xcomp   _
10  in  in  ADP IN  _   11  case    11:case _
11  Fallujah    Fallujah    PROPN   NNP Number=Sing 5   obl 5:obl:in    _
12  alone   alone   ADV RB  _   11  advmod  11:advmod   SpaceAfter=No
13  .   .   PUNCT   .   _   5   punct   5:punct _

I'm just wondering is this the desired behaviour? e.g. the output of deps looks like:

deps [[('advmod', 2)], [('nummod', 3)], [('nsubj:pass', 5), ('nsubj:xsubj', 6), ('nsubj:pass', 8)], [('aux:pass', 5)], '0:root', [('xcomp', 5)], [('cc', 8), ('cc', (8, '.', 1))], [('conj:and', 5), ('nsubj:pass', (8, '.', 1)), ('nsubj:xsubj', 9)], [('conj:and', 5)], [('xcomp', (8, '.', 1))], [('case', 11)], [('obl:in', 5)], [('advmod', 11)], [('punct', 5)]]

Is there any particular reason why '0:root' shouldn't be [('root', 0)]?

Thanks!

[jbrry]

I was able to override this behaviour by changing: https://github.com/EmilStenstrom/conllu/blob/68199a7afcbf66660aec09bab5b0a2b995937dd6/conllu/parser.py#L173

to ID_SINGLE = re.compile(r"[0-9][0-9]*")

This enables a match here: https://github.com/EmilStenstrom/conllu/blob/68199a7afcbf66660aec09bab5b0a2b995937dd6/conllu/parser.py#L201

allowing the value to be returned in the tuple format.

It seems to have solved the problem but let me know if you would advise against it, thanks!

[EmilStenstrom]

@Jbar-ry Excellent catch, thanks for reporting it! It's definitely a bug, I'll think of a good way to solve it and release a new version soon!

[jbrry]

Thank you very much @EmilStenstrom!

[EmilStenstrom]

@Jbar-ry Thank you! I just released 2.2.1 with fixes this bug! Install it with pip install -U conllu.

[jbrry]

Thanks a lot @EmilStenstrom!