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
|
#!/usr/bin/python3
import os
import sqlite3
from argparse import ArgumentParser
import utils
from myconfig import MyConfig
from dirwalk import walkIndex
config = MyConfig()
#default pinyin total frequency
default = config.getDefaultPinyinTotalFrequency()
#minimum pinyin frequency
minimum = config.getMinimumPinyinFrequency()
#change cwd to the word recognizer directory
words_dir = config.getWordRecognizerDir()
os.chdir(words_dir)
#chdir done
atomic_words_dict = {}
merged_words_dict = {}
def load_atomic_words(filename):
wordsfile = open(filename, "r")
for oneline in wordsfile.readlines():
oneline = oneline.rstrip(os.linesep)
if len(oneline) == 0:
continue
(word, pinyin, freq) = oneline.split(None, 2)
freq = int(freq)
if word in atomic_words_dict:
atomic_words_dict[word].append((pinyin, freq))
else:
atomic_words_dict[word] = [(pinyin, freq)]
wordsfile.close()
def load_merged_words(filename):
wordsfile = open(filename, "r")
for oneline in wordsfile.readlines():
oneline = oneline.rstrip(os.linesep)
if len(oneline) == 0:
continue
(word, prefix, postfix, freq) = oneline.split(None, 3)
freq = int(freq)
if word in merged_words_dict:
merged_words_dict[word].append((prefix, postfix, freq))
else:
merged_words_dict[word] = [(prefix, postfix, freq)]
wordsfile.close()
def mergePinyin(pinyin_list):
print(pinyin_list)
pinyins = {}
for (pinyin, freq) in pinyin_list:
if pinyin in pinyins:
pinyins[pinyin] += freq
else:
pinyins[pinyin] = freq
pinyins = list(pinyins.items())
total_freq = sum([ freq for pinyin, freq in pinyins ])
results = []
for (pinyin, freq) in pinyins:
freq = default * freq / total_freq
freq = int(freq)
if freq < minimum:
continue
results.append((pinyin, freq))
print(results)
return results
def markAtomicWord(word):
assert word in atomic_words_dict
results = atomic_words_dict[word]
return mergePinyin(results)
def markMergedWord(word):
assert word in merged_words_dict
merged_list = merged_words_dict[word]
print(merged_list)
merged_sum = sum([ freq for prefix, postfix, freq in merged_list ])
results = []
for (prefix, postfix, freq) in merged_list:
prefix_list = markPinyin(prefix)
prefix_sum = sum([ freq for pinyin, freq in prefix_list ])
postfix_list = markPinyin(postfix)
postfix_sum = sum([ freq for pinyin, freq in postfix_list ])
for prefix_pinyin, prefix_freq in prefix_list:
for postfix_pinyin, postfix_freq in postfix_list:
merged_pinyin = prefix_pinyin + "'" + postfix_pinyin
merged_freq = default * freq * prefix_freq * postfix_freq / \
merged_sum / prefix_sum / postfix_sum
results.append((merged_pinyin, merged_freq))
return mergePinyin(results)
def markPinyin(word):
print(word)
if word in atomic_words_dict:
return markAtomicWord(word)
elif word in merged_words_dict:
return markMergedWord(word)
else:
assert False, "missed word.\n"
def markPinyins(workdir):
merged_words_dict = {}
filename = config.getPartialWordFileName()
filepath = workdir + os.sep + filename
load_merged_words(filepath)
filename = config.getNewWordFileName()
filepath = workdir + os.sep + filename
newwordfile = open(filepath, "r")
filename = config.getRecognizedWordFileName()
filepath = workdir + os.sep + filename
recordfile = open(filepath, "w")
for oneline in newwordfile.readlines():
oneline = oneline.rstrip(os.linesep)
if len(oneline) == 0:
continue
word = oneline
pinyin_list = markPinyin(word)
for pinyin, freq in pinyin_list:
freq = str(freq)
oneline = '\t'.join((word, pinyin, freq))
recordfile.writelines([oneline, os.linesep])
recordfile.close()
newwordfile.close()
def handleOneIndex(indexpath, subdir, indexname):
print(indexpath, subdir, indexname)
indexstatuspath = indexpath + config.getStatusPostfix()
indexstatus = utils.load_status(indexstatuspath)
if not utils.check_epoch(indexstatus, 'NewWord'):
raise utils.EpochError('Please new word first.\n')
if utils.check_epoch(indexstatus, 'MarkPinyin'):
return
workdir = config.getWordRecognizerDir() + os.sep + \
subdir + os.sep + indexname
print(workdir)
markPinyins(workdir)
#sign epoch
utils.sign_epoch(indexstatus, 'MarkPinyin')
utils.store_status(indexstatuspath, indexstatus)
#loading old words
load_atomic_words(config.getWordsWithPinyinFileName())
#print(atomic_words_dict)
if __name__ == '__main__':
parser = ArgumentParser(description='Mark pinyins.')
parser.add_argument('--indexdir', action='store', \
help='index directory', \
default=config.getTextIndexDir())
args = parser.parse_args()
print(args)
walkIndex(handleOneIndex, args.indexdir)
print('done')
|
