Open
Description
There are a few message blocks that pypolibox
currently can't turn into an HLDS structure, since there is/are no corresponding rules/vocabulary in the OpenCCG grammar fragment we're using.
Have a look at lexicalize_messageblocks.py
:
def lexicalize_lastbook_nomatch(lastbook_nomatch_message_block):
r"""
Im Gegensatz zum ersten / vorhergehenden / anderen Buch ____
"""
raise NotImplementedError, "The grammar fragment can't handle lastbook non-matches, yet."
def lexicalize_usermodel_match(usermodel_match_message_block):
r"""erfüllt Anforderungen / entspricht ihren Wünschen"""
raise NotImplementedError, "The grammar fragment can't handle usermodel matches, yet."
def lexicalize_usermodel_nomatch(usermodel_nomatch_message_block):
r"""erfüllt (leider) Anforderungen nicht / entspricht nicht ihren
Wünschen"""
raise NotImplementedError, "The grammar fragment can't handle usermodel non-matches, yet."
As an example, let's ask pypolibox
for a book about parsing:
pypolibox --keywords parsing
[...]
The id message block can be realized as follows:
„ Computational_Linguistics._An_Introduction. “ von Ralph Grishman ist im Jahr 1986 erschienen .
Ralph Grishmans „ Computational_Linguistics._An_Introduction. “ beinhaltet keine Code-Beispiele .
das Buch richtet sich an Anfänger .
Ralph Grishmans „ Computational_Linguistics._An_Introduction. “ beinhaltet Übungen .
das Buch von Ralph Grishman umfasst 193 Seiten .
The grammar fragment can't handle usermodel matches, yet.
The message block contains these messages:
[ *msgType* = 'usermodel_match' ]
[ *reference_authors* = (frozenset(['Ralph Grishman']), 'neutral') ]
[ *reference_title* = ('Computational Linguistics. An Introduction.', 'neutral') ]
[ keywords = (frozenset(['parsing']), 'positive') ]
In this case, pypolibox
will generate the ID
message block for a matching book, but cannot express the usermodel_match
message block (something like "das Buch behandelt das Thema Parsing und entspricht damit ihren Wünschen").