forked from nsantacruz/PSHAT
/
LangDatasetGenerator.py
301 lines (246 loc) · 10.9 KB
/
LangDatasetGenerator.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
# -*- coding: utf-8 -*-
import re,codecs,json,os,random
from collections import OrderedDict
from os import listdir
from os.path import isfile, join
import util
import local_settings
from research.talmud_pos_research.language_classifier import cal_tools
from sefaria.model import *
from sefaria.utils import hebrew
random.seed(2823274491)
def make_aramaic_training():
abbrev_dict = json.load(codecs.open('data/1_cal_input/abbreviations.json',encoding='utf8'))
for abbrev, defs in abbrev_dict.items():
sorted_defs = sorted(defs.items(),key=lambda x: x[1])
abbrev_dict[abbrev] = sorted_defs[0][0]
training = []
num_found = 0
num_missed = 0
with open('data/1_cal_input/caldbfull.txt','rb') as cal:
for line in cal:
line_obj = cal_tools.parseCalLine(line,True,withshinsin=False)
temp_word = line_obj['word']
words = []
if u"'" in temp_word:
temp_word = temp_word.replace(u"'",u'')
if temp_word in abbrev_dict:
words = re.split(ur'\s+',abbrev_dict[temp_word])
num_found += 1
else:
num_missed += 1
#print u'missed {}'.format(temp_word)
else:
words = [temp_word]
for w in words:
training.append({'word':w,'tag':'aramaic'})
print u'Num abbrevs replaced {}. Num Missed {}'.format(num_found,num_missed)
return training
def make_mishnaic_training():
training = []
num_mishnah_per_mesechta = 30000 # effectively all mishnah
mishnah_indexes = [library.get_index(ind) for ind in library.get_indexes_in_category("Mishnah")]
mishnah_indexes += [library.get_index(ind) for ind in library.get_indexes_in_category("Torah")]
mish_set = set()
num_removed = 0
for ind in mishnah_indexes:
mishna_segs = ind.all_section_refs()
if len(mishna_segs) >= num_mishnah_per_mesechta:
mishna_segs = mishna_segs[:num_mishnah_per_mesechta]
for seg in mishna_segs:
first_sec_str = hebrew.strip_cantillation(seg.text('he').as_string(), strip_vowels=True)
word_list = util.tokenize_words(first_sec_str)
for word in word_list:
if random.random() > 0.45 and word in mish_set:
num_removed += 1
continue
training.append({'word':word,'tag':'mishnaic'})
mish_set.add(word)
print "Num Mishna removed: {}".format(num_removed)
return training
def merge_sets(a,m):
a_set = {}
m_set = {}
let_set = set()
for w in a:
for c in w['word']:
let_set.add(c)
for w in m:
if w['word'] not in m_set:
m_set[w['word']] = 0
m_set[w['word']] += 1
num_in_m = 0
"""
with open('data/1_cal_input/jbaforms.txt','rb') as jba:
for line in jba:
line_obj = cal_tools.parseJBALine(line,True,withshinsin=False)
if 'word' in line_obj and line_obj['word'] not in m_set: # don't add ambiguous words from jba. there are too many
bad_char = False
for c in line_obj['word']:
if c not in let_set:
print 'continued'
bad_char = True
break
if not bad_char:
a.append({'word':line_obj['word'],'tag':'aramaic'})
elif 'word' in line_obj and line_obj['word'] in m_set:
num_in_m += 1
"""
for w in a:
if w['word'] not in a_set:
a_set[w['word']] = 0
a_set[w['word']] += 1
ambig_set = set()
for word,count in a_set.items():
if word in m_set and count < 10*m_set[word] and m_set[word] < 10*count:
ambig_set.add(word)
ambig = []
a_merge = []
m_merge = []
num_deleted = 0
for w in a:
if w['word'] in ambig_set:
num_deleted += 1
ambig.append({'word':w['word'],'tag':'ambiguous'})
else:
a_merge.append(w)
for w in m:
if w['word'] in ambig_set:
num_deleted += 1
ambig.append({'word':w['word'],'tag':'ambiguous'})
else:
m_merge.append(w)
print num_deleted
print "NUM IN M {}".format(num_in_m)
print 'YO A {} M {}'.format(len(a),len(m))
return a_merge + m_merge + ambig,len(a_merge),len(m_merge),len(ambig)
def print_tagged_corpus_to_html_table(lang_out):
str = u"<html><head><style>h1{text-align:center;background:grey}td{text-align:center}table{margin-top:20px;margin-bottom:20px;margin-right:auto;margin-left:auto;width:1200px}.aramaic{background-color:blue;color:white}.mishnaic{background-color:red;color:white}.ambiguous{background-color:yellow;color:black}</style><meta charset='utf-8'></head><body>"
for daf in lang_out:
str += u"<h1>DAF {}</h1>".format(daf)
str += u"<table>"
count = 0
while count < len(lang_out[daf]):
row_obj = lang_out[daf][count:count+10]
row = u"<tr>"
for w in reversed(row_obj):
row += u"<td class='{}'>{}</td>".format(w['lang'],w['word'])
row += u"</tr>"
#row_sef += u"<td>({}-{})</td></tr>".format(count,count+len(row_obj)-1)
str += row
count += 10
str += u"</table>"
str += u"</body></html>"
return str
def dilate_lang():
lang_tagged_path = 'data/3_lang_tagged'
lang_tagged_dilated_path = 'data/4_lang_tagged_dilated'
mesechtot_names = ['Berakhot','Shabbat','Eruvin','Pesachim','Bava Kamma','Bava Metzia','Bava Batra']
for mesechta in mesechtot_names:
util.make_folder_if_need_be('{}/json/{}'.format(lang_tagged_path, mesechta))
mesechta_path = '{}/json/{}'.format(lang_tagged_path, mesechta)
def sortdaf(fname):
daf = fname.split('/')[-1].split('.json')[0]
daf_int = int(daf[:-1])
amud_int = 1 if daf[-1] == 'b' else 0
return daf_int*2 + amud_int
files = [f for f in listdir(mesechta_path) if isfile(join(mesechta_path, f))]
files.sort(key=sortdaf)
html_out = OrderedDict()
for i_f,f_name in enumerate(files):
lang_out = []
lang_in = json.load(codecs.open('{}/{}'.format(mesechta_path,f_name), "rb", encoding="utf-8"))
for i_w,w in enumerate(lang_in):
if 1 < i_w < len(lang_in)-1:
neigh = [lang_in[i_w-1]['confidence'],lang_in[i_w+1]['confidence']]
elif i_w < len(lang_in) - 1:
neigh = [lang_in[i_w+1]['confidence']]
else:
neigh = [lang_in[i_w-1]['confidence']]
neigh_conf = [sum([c[0] for c in neigh])/2,sum([c[1] for c in neigh])/2]
weight = 1.1
new_conf = [sum([neigh_conf[0],weight*w['confidence'][0]]),sum([neigh_conf[1],weight*w['confidence'][1]])]
new_lang = 'aramaic' if new_conf[0] > new_conf[1] else 'mishnaic'
lang_out.append({'word':w['word'],'lang':new_lang,'confidence':new_conf})
util.make_folder_if_need_be("{}/json/{}".format(lang_tagged_dilated_path,mesechta))
fp = codecs.open("{}/json/{}/{}".format(lang_tagged_dilated_path,mesechta,f_name), "wb", encoding='utf-8')
json.dump(lang_out, fp, indent=4, encoding='utf-8', ensure_ascii=False)
fp.close()
daf = f_name.split('/')[-1].split('.json')[0]
html_out[daf] = lang_out
if i_f % 10 == 0:
print '{}/{}'.format(mesechta,f_name)
html = print_tagged_corpus_to_html_table(html_out)
util.make_folder_if_need_be("{}/html/{}".format(lang_tagged_dilated_path, mesechta))
fp = codecs.open("{}/html/{}/{}.html".format(lang_tagged_dilated_path, mesechta, daf), "wb",
encoding='utf-8')
fp.write(html)
fp.close()
html_out = OrderedDict()
#CONTEXT
def make_aramaic_training_context():
training = []
with open('data/1_cal_input/caldbfull.txt','rb') as cal:
temp_phrase = []
curr_line_num = None
curr_word_num = None
for line in cal:
try:
lineObj = cal_tools.parseCalLine(line,True,False)
except IndexError:
continue
if curr_line_num is None:
curr_line_num = lineObj['line_num']
if curr_word_num is None:
curr_word_num = lineObj['word_num'] - 1
if curr_line_num == lineObj['line_num'] and (curr_word_num + 1) == lineObj['word_num']:
temp_phrase.extend(lineObj['word'].split(' '))
curr_word_num = lineObj['word_num']
else:
training.append({'language': 'aramaic','phrase': temp_phrase[:]})
curr_line_num = lineObj['line_num']
curr_word_num = lineObj['word_num']
temp_phrase = lineObj['word'].split(' ')
total_words = 0
total_phrases = len(training)
for p in training:
total_words += len(p['phrase'])
print 'NUM PHRASES: {} AVG WORDS PER PHRASE: {}'.format(total_phrases,total_words/total_phrases)
return training
def make_mishnaic_training_context():
training = []
mishnah_indexes = [library.get_index(ind) for ind in library.get_indexes_in_category("Mishnah")]
mishnah_indexes += [library.get_index(ind) for ind in library.get_indexes_in_category("Torah")]
for ind in mishnah_indexes:
mishna_segs = ind.all_section_refs()
for seg in mishna_segs:
first_sec_str = hebrew.strip_cantillation(seg.text('he').as_string(), strip_vowels=True)
training += [{'language':'mishnaic', 'phrase': util.tokenize_words(p)} for p in first_sec_str.split(u'. ')]
total_words = 0
total_phrases = len(training)
for p in training:
total_words += len(p['phrase'])
print 'NUM PHRASES: {} AVG WORDS PER PHRASE: {}'.format(total_phrases,total_words/total_phrases)
return training
def merge_sets_context(a,m):
full = []
phrases = a + m
random.shuffle(phrases)
num_phrases_per_section = 100
i = 0
while i < len(phrases):
temp_phrases = phrases[i:i+num_phrases_per_section]
full.append([{'l': p['language'][0], 'w': w} for p in temp_phrases for w in p['phrase']])
i += num_phrases_per_section
return full
make_training = True:
if make_training:
a = make_aramaic_training()
m = make_mishnaic_training()
full,lena,lenm,lenambig = merge_sets(a,m)
print 'A {} M {} Ambig {}'.format(lena,lenm,lenambig)
print 'Full Length: {}'.format(len(full))
fp = codecs.open("data/3_lang_tagged/model/lstm_training.json", "wb", encoding='utf-8')
json.dump(full, fp, indent=4, encoding='utf-8', ensure_ascii=False)
else:
dilate_lang()