Bump to 0.2.0
- Add Unweighted mode to command line - Download nltk data if needed.
This commit is contained in:
parent
6556164879
commit
d6e329ff39
Binary file not shown.
After Width: | Height: | Size: 392 KiB |
File diff suppressed because one or more lines are too long
After Width: | Height: | Size: 64 KiB |
|
@ -4,4 +4,4 @@
|
|||
# This is free software, see the included LICENSE for terms and conditions.
|
||||
#
|
||||
|
||||
version = '0.1.2'
|
||||
version = '0.2.0'
|
||||
|
|
|
@ -61,17 +61,18 @@ def parse_arguments() -> argparse.Namespace:
|
|||
|
||||
chain_sub = subparsers.add_parser('chain', help="Output from a chainer")
|
||||
chain_sub.add_argument('input', help="The chain file to load", type=pathlib.Path)
|
||||
chain_sub.add_argument('-u', '--unweighted', help="Don't weight token selection by statistical frequency.", action='store_true')
|
||||
chain_sub.add_argument('-c', '--count', help="Number of chain outputs to output", type=int, default=10)
|
||||
chain_sub.add_argument('-m', '--maxlen', help="Maximum length in tokens of output (0 is unlimited)", type=int, default=0)
|
||||
|
||||
return parser.parse_args()
|
||||
|
||||
|
||||
def print_chainer_output(chainer: Chain, random_state: Random):
|
||||
def print_chainer_output(chainer: Chain, random_state: Random, weighted: bool = False):
|
||||
if chainer.analyzer_class in JOINERS:
|
||||
print(JOINERS[cast(str, chainer.analyzer_class)](chainer.walk(random_state, True)))
|
||||
print(JOINERS[cast(str, chainer.analyzer_class)](chainer.walk(random_state, weighted)))
|
||||
else:
|
||||
print(chainer.walk(random_state, True))
|
||||
print(chainer.walk(random_state, weighted))
|
||||
|
||||
|
||||
def command_analyze(args: argparse.Namespace) -> int:
|
||||
|
@ -145,7 +146,7 @@ def command_chain(args: argparse.Namespace) -> int:
|
|||
args.count = 1
|
||||
|
||||
for _ in range(0, args.count):
|
||||
print_chainer_output(chainer, r)
|
||||
print_chainer_output(chainer, r, not args.unweighted)
|
||||
|
||||
return 0
|
||||
|
||||
|
|
|
@ -9,6 +9,12 @@ import nltk
|
|||
from .abstract import AbstractAnalyzer
|
||||
|
||||
|
||||
try:
|
||||
nltk.sent_tokenize("foo bar")
|
||||
except LookupError:
|
||||
nltk.download('punkt')
|
||||
|
||||
|
||||
class English(AbstractAnalyzer):
|
||||
def __init__(self, order, filters=None):
|
||||
if filters is None:
|
||||
|
|
|
@ -1,3 +0,0 @@
|
|||
[build-system]
|
||||
requires = ["setuptools", "wheel"]
|
||||
build-backend = "setuptools.build_meta"
|
Loading…
Reference in New Issue