135 lines
4.1 KiB
Python
135 lines
4.1 KiB
Python
#!/usr/bin/env python3
|
|
# -*- encoding: utf-8 -*-
|
|
|
|
# Copyright (c) 2017 Adler Neves <adlerosn@gmail.com>
|
|
#
|
|
# MIT License
|
|
#
|
|
# Permission is hereby granted, free of charge, to any person obtaining
|
|
# a copy of this software and associated documentation files (the
|
|
# "Software"), to deal in the Software without restriction, including
|
|
# without limitation the rights to use, copy, modify, merge, publish,
|
|
# distribute, sublicense, and/or sell copies of the Software, and to
|
|
# permit persons to whom the Software is furnished to do so, subject to
|
|
# the following conditions:
|
|
#
|
|
# The above copyright notice and this permission notice shall be
|
|
# included in all copies or substantial portions of the Software.
|
|
#
|
|
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
|
# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
|
# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
|
# NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
|
# LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
|
# OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
|
# WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
|
|
|
import json
|
|
from utils import getListItemOr, sortedListBinarySearchNoRaise
|
|
|
|
def deepTupleConversion(lst):
|
|
if isinstance(lst, list):
|
|
for i, l in enumerate(lst):
|
|
if isinstance(l, list) or isinstance(l, dict):
|
|
lst[i] = deepTupleConversion(l)
|
|
return tuple(lst)
|
|
elif isinstance(lst, dict):
|
|
for k, v in lst.items():
|
|
if isinstance(v, list) or isinstance(v, dict):
|
|
lst[k] = deepTupleConversion(v)
|
|
return lst
|
|
|
|
with open('unitexable_test/corpus.txt.untagged.txt') as f:
|
|
corpus = [[wrd.strip().lower() for wrd in snt.split(' ') if len(wrd)>0] for snt in f.read().strip().splitlines() if len(snt)>0]
|
|
|
|
print('readed corpus')
|
|
|
|
with open('unitexable_test/traindata.json') as f:
|
|
taggerData = json.loads(f.read())
|
|
|
|
deepTupleConversion(taggerData)
|
|
|
|
tagset = taggerData['tagset']
|
|
tokset = taggerData['tokset']
|
|
|
|
print('loaded tagger data')
|
|
|
|
tokCorpus = list()
|
|
for i,snt in enumerate(corpus):
|
|
tokCorpus.append([sortedListBinarySearchNoRaise(tokset, wrd) for wrd in snt])
|
|
if i%25000 == 0:
|
|
print('%d of %d'%(i,len(corpus)))
|
|
|
|
tokCorpus = deepTupleConversion(tokCorpus)
|
|
|
|
print('loaded corpus - tagging may begin')
|
|
|
|
dictEmptyTriples = dict(taggerData['keypairs']['3empty'])
|
|
dictTriples = dict(taggerData['keypairs']['3middle'])
|
|
dictTuplesL = dict(taggerData['keypairs']['2left'])
|
|
dictTuplesR = dict(taggerData['keypairs']['2right'])
|
|
dictFallback = dict(taggerData['keypairs']['1fallback'])
|
|
|
|
def getTagFallback(triple):
|
|
return dictFallback.get[triple[1:2]]
|
|
|
|
def getTagTuplesL(triple):
|
|
return dictTuplesL[triple[0:2]]
|
|
|
|
def getTagTuplesR(triple):
|
|
return dictTuplesR[triple[1:3]]
|
|
|
|
def getTagTriples(triple):
|
|
return dictTuplesR[triple[0:3]]
|
|
|
|
def getTagTriplesEmpty(triple):
|
|
return dictEmptyTriples[triple[0:3:2]]
|
|
|
|
taggingStrategy = [
|
|
getTagTriples,
|
|
getTagTuplesR,
|
|
getTagTuplesL,
|
|
getTagTriplesEmpty,
|
|
getTagFallback
|
|
]
|
|
|
|
def getTagWithStategy(triple):
|
|
for stategy in taggingStrategy:
|
|
try:
|
|
return stategy(triple)
|
|
except:
|
|
pass
|
|
return -1
|
|
|
|
def processSentence(tokSentence):
|
|
l = list()
|
|
for i in range(len(tokSentence)):
|
|
triple = (getListItemOr(tokSentence, i-1, None), tokSentence[i], getListItemOr(tokSentence, i+1, None))
|
|
tag = getTagWithStategy(triple)
|
|
l.append((tokSentence[i], tag))
|
|
return l
|
|
|
|
tokTagged = list()
|
|
for i,snt in enumerate(tokCorpus):
|
|
tokTagged.append(processSentence(snt))
|
|
if i%25000 == 0:
|
|
print('%d of %d'%(i,len(tokCorpus)))
|
|
|
|
print('corpus tagged - preparing for serialization')
|
|
|
|
taggedCorpus = list()
|
|
for words in tokTagged:
|
|
sentence = list()
|
|
for word in words:
|
|
sentence.append(getListItemOr(tagset, word[1], "???"))
|
|
taggedCorpus.append(sentence)
|
|
|
|
stg = '\n\n'.join(['\n'.join(snt) for snt in taggedCorpus])
|
|
|
|
print('serialized')
|
|
|
|
with open('unitexable_test/corpus.txt.tagged.txt','w') as f:
|
|
f.write(stg)
|
|
|
|
print('done')
|