1
|
|
|
import os |
2
|
|
|
import re |
3
|
|
|
import string |
4
|
|
|
|
5
|
|
|
import astroid |
6
|
|
|
from pylint.checkers import BaseChecker, utils |
7
|
|
|
from pylint.interfaces import IAstroidChecker |
8
|
|
|
from textdistance import levenshtein |
9
|
|
|
|
10
|
|
|
|
11
|
|
|
class SimilarStringChecker(BaseChecker): |
12
|
|
|
__implements__ = (IAstroidChecker,) |
13
|
|
|
name = "similar-string-checker" |
14
|
|
|
|
15
|
|
|
msgs = { |
16
|
|
|
"R5011": ( |
17
|
|
|
'"%s" is %.2f%% similar to "%s". Try keeping them the same ' |
18
|
|
|
"to reduce work for translators!", |
19
|
|
|
"similar-string", |
20
|
|
|
"Similar strings should be avoided for single translation", |
21
|
|
|
) |
22
|
|
|
} |
23
|
|
|
|
24
|
|
|
_dict_of_strings = {} |
25
|
|
|
threshold = 0.8 |
26
|
|
|
error_messages = [] |
27
|
|
|
|
28
|
|
|
project_root = os.path.abspath( |
29
|
|
|
os.path.join(os.path.dirname(__file__), "..", "tcms") |
30
|
|
|
) |
31
|
|
|
|
32
|
|
|
# NOTE: this works against tcms/ directory and will not take into account |
33
|
|
|
# if we want to examine only a sub-dir or a few files |
34
|
|
|
# all files found by os.walk |
35
|
|
|
all_template_files = set() |
36
|
|
|
|
37
|
|
|
def open(self): |
38
|
|
|
for rootpath, _dirs, files in os.walk(self.project_root, topdown=False): |
39
|
|
|
for file_name in files: |
40
|
|
|
if file_name.endswith((".html", ".txt")): |
41
|
|
|
self.all_template_files.add( |
42
|
|
|
os.path.join(self.project_root, rootpath, file_name) |
43
|
|
|
) |
44
|
|
|
|
45
|
|
|
@staticmethod |
46
|
|
|
def clean_string(text): |
47
|
|
|
""" |
48
|
|
|
This method removes the operators and other punctuations |
49
|
|
|
used in the string to avoid unwanted data to add up to |
50
|
|
|
the similarity between the strings. |
51
|
|
|
""" |
52
|
|
|
cleaned = "" |
53
|
|
|
for char in text: |
54
|
|
|
if char not in string.punctuation: |
55
|
|
|
cleaned += char |
56
|
|
|
return cleaned |
57
|
|
|
|
58
|
|
|
def check_similar_string(self, translation_string): |
59
|
|
|
cleaned_translation_string = self.clean_string(translation_string) |
60
|
|
|
for key in self._dict_of_strings: |
61
|
|
|
similarity = levenshtein.normalized_similarity( |
62
|
|
|
cleaned_translation_string, self.clean_string(key) |
63
|
|
|
) |
64
|
|
|
if similarity >= self.threshold: |
65
|
|
|
return key, similarity |
66
|
|
|
return None, None |
67
|
|
|
|
68
|
|
|
@utils.check_messages("similar-string") |
69
|
|
|
def visit_call(self, node): |
70
|
|
|
if not ( |
71
|
|
|
isinstance(node.func, astroid.Name) |
72
|
|
|
and node.func.name in ("_", "gettext_lazy") |
73
|
|
|
): |
74
|
|
|
return |
75
|
|
|
if not isinstance(node.args[0], astroid.nodes.Const): |
76
|
|
|
return |
77
|
|
|
|
78
|
|
|
translation_string = node.args[0].value |
79
|
|
|
self.check_similar_and_add_error_message(node, translation_string) |
80
|
|
|
|
81
|
|
|
# check if similar string found and add it to error_messages |
82
|
|
|
def check_similar_and_add_error_message( |
83
|
|
|
self, node, translation_string, **error_message |
84
|
|
|
): |
85
|
|
|
if translation_string in self._dict_of_strings: |
86
|
|
|
return |
87
|
|
|
|
88
|
|
|
similar_string, similarity = self.check_similar_string(translation_string) |
89
|
|
|
|
90
|
|
|
if similar_string: |
91
|
|
|
if isinstance(node, str): |
92
|
|
|
error_message["node"] = astroid.Module(node, file=node, doc=None) |
93
|
|
|
else: |
94
|
|
|
error_message["node"] = node |
95
|
|
|
error_message["args"] = ( |
96
|
|
|
translation_string, |
97
|
|
|
similarity * 100, |
98
|
|
|
similar_string, |
99
|
|
|
) |
100
|
|
|
self.error_messages.append(error_message) |
101
|
|
|
return |
102
|
|
|
self._dict_of_strings[translation_string] = True |
103
|
|
|
|
104
|
|
|
# checks each line and find trans or blocktrans tags |
105
|
|
|
def parse_translation_string(self, filename, lines): |
106
|
|
|
startline = 0 |
107
|
|
|
startcol = 0 |
108
|
|
|
blocktrans_string = "" |
109
|
|
|
|
110
|
|
|
for lineno, line in enumerate(lines): |
111
|
|
|
# if pylint disable comment is found ignore and continue with next line |
112
|
|
|
if re.search(r"<!--\s*pylint\s*:\s*disable\s*-->", line): |
113
|
|
|
continue |
114
|
|
|
|
115
|
|
|
# if blocktrans starting tag is found |
116
|
|
|
match_blocktrans_in_line = re.search(r"{% blocktrans[^%}]*%}(.+)", line) |
117
|
|
|
if match_blocktrans_in_line: |
118
|
|
|
startline = lineno |
119
|
|
|
startcol = match_blocktrans_in_line.start(1) |
120
|
|
|
blocktrans_string = match_blocktrans_in_line.group(1) |
121
|
|
|
|
122
|
|
|
# if line after blocktrans is found |
123
|
|
|
elif blocktrans_string != "": |
124
|
|
|
blocktrans_string += line |
125
|
|
|
|
126
|
|
|
# if blocktrans ending tag is found |
127
|
|
|
endblocktrans_line = re.search( |
128
|
|
|
r"((.|\n|)*){% endblocktrans %}", blocktrans_string |
129
|
|
|
) |
130
|
|
|
if endblocktrans_line: |
131
|
|
|
blocktrans_string = endblocktrans_line.group(1) |
132
|
|
|
self.check_similar_and_add_error_message( |
133
|
|
|
filename, |
134
|
|
|
blocktrans_string, |
135
|
|
|
line=startline + 1, |
136
|
|
|
col_offset=startcol, |
137
|
|
|
) |
138
|
|
|
blocktrans_string = "" |
139
|
|
|
|
140
|
|
|
# if trans tag is found |
141
|
|
|
# Note: trans tag could be more than one in same |
142
|
|
|
# line, hence re.finditer rather than re.search) |
143
|
|
|
match_in_line = re.finditer(r'{% trans "([\w ]+)" %}', line) |
144
|
|
|
for match in match_in_line: |
145
|
|
|
translation_string = match.group(1) |
146
|
|
|
self.check_similar_and_add_error_message( |
147
|
|
|
filename, |
148
|
|
|
translation_string, |
149
|
|
|
line=lineno + 1, |
150
|
|
|
col_offset=match.start(1), |
151
|
|
|
) |
152
|
|
|
|
153
|
|
|
def close(self): |
154
|
|
|
for filepath in self.all_template_files: |
155
|
|
|
with open(filepath, "r") as file: |
156
|
|
|
self.parse_translation_string(filepath, file.readlines()) |
157
|
|
|
|
158
|
|
|
for error_message in self.error_messages: |
159
|
|
|
self.add_message("similar-string", **error_message) |
160
|
|
|
|