mirror of
https://github.com/explosion/spaCy.git
synced 2025-04-29 05:13:42 +03:00
added hashtag rule and fixed currency rules
This commit is contained in:
parent
e9af79a803
commit
bb08d696f9
|
@ -11,7 +11,7 @@ _units = (_units + 's bit Gbps Mbps mbps Kbps kbps ƒ ppi px '
|
||||||
'ratus rb ribu ribuan '
|
'ratus rb ribu ribuan '
|
||||||
'juta jt jutaan mill?iar million bil[l]?iun bilyun billion '
|
'juta jt jutaan mill?iar million bil[l]?iun bilyun billion '
|
||||||
)
|
)
|
||||||
_currency = (_currency + r'Rp IDR RMB SGD S\$ ')
|
_currency = (_currency + r' Rp IDR RMB SGD S\$')
|
||||||
_months = ('Januari Februari Maret April Mei Juni Juli Agustus September '
|
_months = ('Januari Februari Maret April Mei Juni Juli Agustus September '
|
||||||
'Oktober November Desember January February March May June '
|
'Oktober November Desember January February March May June '
|
||||||
'July August October December Jan Feb Mar Jun Jul Aug Sept '
|
'July August October December Jan Feb Mar Jun Jul Aug Sept '
|
||||||
|
@ -25,7 +25,7 @@ HTML_SUFFIX = r'</(b|strong|i|em|p|span|div|a)>'
|
||||||
MONTHS = merge_chars(_months)
|
MONTHS = merge_chars(_months)
|
||||||
LIST_CURRENCY = split_chars(_currency)
|
LIST_CURRENCY = split_chars(_currency)
|
||||||
|
|
||||||
|
TOKENIZER_PREFIXES.remove('#') # hashtag
|
||||||
_prefixes = TOKENIZER_PREFIXES + LIST_CURRENCY + [HTML_PREFIX] + ['/', '—']
|
_prefixes = TOKENIZER_PREFIXES + LIST_CURRENCY + [HTML_PREFIX] + ['/', '—']
|
||||||
|
|
||||||
_suffixes = TOKENIZER_SUFFIXES + [r'\-[Nn]ya', '-[KkMm]u', '[—-]'] + [
|
_suffixes = TOKENIZER_SUFFIXES + [r'\-[Nn]ya', '-[KkMm]u', '[—-]'] + [
|
||||||
|
|
Loading…
Reference in New Issue
Block a user