-
Notifications
You must be signed in to change notification settings - Fork 0
/
find_mistakes.py
155 lines (126 loc) · 4.8 KB
/
find_mistakes.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
import subprocess
import os
import re
from string import punctuation
from tqdm import tqdm
import urllib.request as urr
import urllib.error as ure
from transliterate import translit
from collections import defaultdict
HEADERS = {'User-Agent':
'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_3) '
'AppleWebKit/537.36 (KHTML, like Gecko) '
'Chrome/35.0.1916.47 '
'Safari/537.36'}
KG_LETTERS = {"ң": "n",
"ү": "u",
"ө": "o",
"j": "y",
"'": ""}
def run_apertium_tagger(input, mode="text"):
"""
Analyse input text
"""
working_directory = os.path.dirname(os.path.abspath(__file__))
if mode != "text":
echo_word = "echo '" + input + "'"
cmd = echo_word + "| " \
"lt-proc -w transducer/kir.automorf.bin | " \
"cg-proc -1 transducer/kir.rlx.bin"
else:
print("Processing text...")
cmd = "lt-proc -w transducer/kir.automorf.bin < " \
+ working_directory + "/" + input + " | " \
"cg-proc -1 transducer/kir.rlx.bin"
ps = subprocess.Popen(cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, cwd=working_directory)
output = ps.communicate()[0]
result = output.decode("utf-8")
words_split = re.split(r'[\^\$]', result)
result_list = list(filter(lambda x: (x.strip() not in punctuation), words_split))
word_match_list = []
for word in tqdm(result_list):
word_match = re.match(r"(?:.+?)\/\*?([А-ЯӨҮҢJjа-яёөүң\s]+)(<.+>)?", word)
if word_match:
if not word_match[2] and ("Й" in word_match[1] or "й" in word_match[1]):
word_new = str(word_match[1]).replace("Й", "J")
word_new = word_new.replace("й", "j", 1)
word_match = run_apertium_tagger(word_new, mode="word")[0]
word_match_list.append(word_match)
return word_match_list
def read_analyzed(text_input):
"""
Convert results into a list
"""
word_results = []
match_list = run_apertium_tagger(text_input)
for word_match in match_list:
if not word_match:
continue
else:
word_match_0 = str(word_match[0]).replace("J", "Й")
word_match_0 = word_match_0.replace("j", "й")
word_results.append(re.split(r'[/]', word_match_0))
return word_results
def check_link(source):
"""
Check for 404 error
"""
request = urr.Request(source, headers=HEADERS)
try:
response = urr.urlopen(request)
return True
except ure.HTTPError:
return False
def check_dict(unk_list):
"""
Check whether dictionary entry for stem exists and count false stems
"""
stems_translit = defaultdict(str)
stems, false_stems = [], []
false_count = 0
for unk_word in unk_list:
stem_match = re.match(r'([а-яёөүң\s]+)(<.+>)?', unk_word)
stem = stem_match[1]
stems.append(stem)
stem_translit = translit(stem, 'ru', reversed=True)
substrings = sorted(KG_LETTERS, key=len, reverse=True)
regex = re.compile('|'.join(map(re.escape, substrings)))
stem_final = regex.sub(lambda match: KG_LETTERS[match.group(0)], stem_translit)
stems_translit[stem] = stem_final
print("\n\nChecking for false stems...")
for stem, lat_stem in tqdm(stems_translit.items()):
dict_link = "http://el-sozduk.kg/ru/" + lat_stem
stem_exists = check_link(dict_link)
if not stem_exists:
false_stems.append(stem)
for stem in stems:
if stem in false_stems:
false_count += 1
print("False stems found: {}".format(len(false_stems)))
print("False stems total: {}".format(false_count))
return false_count
if __name__ == "__main__":
text_input = "test_corpus.txt"
word_results = read_analyzed(text_input)
stop_words = []
with open("stop_words.txt", "r", encoding="utf-8") as stop_list:
for word in stop_list:
stop_words.append(word.strip("\n"))
total_count, unk_tag_count, no_tags_count = 0, 0, 0
unk_list = []
for word in word_results:
if any(x in word[1].lower() for x in stop_words):
continue
else:
total_count += 1
if "unk" in word[1]:
unk_tag_count += 1
unk_list.append(word[1].lower())
if "*" in word[1]:
no_tags_count += 1
false_count = check_dict(unk_list)
mistakes_before = unk_tag_count + no_tags_count
mistakes_after = false_count + no_tags_count
print("total: {}".format(total_count))
print("before changes: {} ({}%)".format(mistakes_before, (mistakes_before / total_count)*100))
print("after changes: {} ({}%)".format(mistakes_after, (mistakes_after / total_count)*100))