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
|
from gflanguages import languages_public_pb2
import enum
import re
class Udhr:
def __init__(
self, key, iso639_3, iso15924, bcp47, direction, ohchr, stage, loc, name
):
self.key = key
self.iso639_3 = iso639_3
self.iso15924 = iso15924
self.bcp47 = bcp47
self.direction = direction
self.ohchr = ohchr
self.stage = stage
self.loc = loc
self.name = name
self.title = None
self.preamble = None
self.articles = []
def Parse(self, translation_data):
if translation_data is None or self.stage < 2:
return
if translation_data.find("./{*}title") is not None:
self.title = translation_data.find("./{*}title").text
preamble_data = translation_data.find("./{*}preamble")
if preamble_data is not None:
if preamble_data.find("./{*}title") is not None:
self.preamble = {
"title": preamble_data.find("./{*}title").text,
"content": [
para.text for para in preamble_data.findall("./{*}para")
],
}
articles_data = translation_data.findall("./{*}article")
for article_data in articles_data:
title_data = article_data.find("./{*}title")
article = {
"id": int(article_data.get("number")),
"title": None if title_data is None else title_data.text,
"content": [para.text for para in article_data.findall("./{*}para")],
}
self.articles.append(article)
def LoadArticleOne(self, article_one):
self.articles.append({"id": 0, "title": None, "content": [article_one]})
def GetSampleTexts(self):
extractor = SampleTextExtractor(self)
return extractor.GetSampleTexts()
class SampleTextExtractor:
class TextType(enum.Enum):
GLYPHS = 1
WORD = 2
PHRASE = 3
SENTENCE = 4
PARAGRAPH = 5
PASSAGE = 6
def __init__(self, udhr):
self._udhr = udhr
self._glyphs = iter(self._GetGlyphs())
self._words = iter(self._GetWords())
self._paragraphs = iter(self._GetParagraphs())
self._phrase_history = set()
self._non_word_regex = re.compile(r"[^\w]+")
self._space_regex = re.compile(r"\s+")
self._non_space_regex = re.compile(r"[^\s]+")
self._non_word_space_regex = re.compile(r"[^\w\s]+")
self._any_regex = re.compile(r".")
def _DisplayLength(self, s):
"""Returns length of given string. Omits combining characters.
Some entire scripts will not be counted; in those cases, the raw length of
the string is returned.
"""
word_space_length = len(self._non_word_space_regex.sub("", s))
space_length = len(self._non_space_regex.sub("", s))
if word_space_length == space_length:
return len(s)
return word_space_length
def _GetGlyphs(self):
seen = set()
for article in self._udhr.articles:
for para in article["content"]:
for ch in self._non_word_regex.sub("", para) or self._space_regex.sub(
"", para
):
ch = ch.lower()
if ch not in seen:
seen.add(ch)
yield ch
def _GetWords(self):
if self._space_regex.search(self._udhr.articles[0]["content"][0]) is not None:
splitter = self._space_regex
else:
splitter = self._non_word_regex
seen = set()
for article in self._udhr.articles:
for para in article["content"]:
for s in splitter.split(para):
if s not in seen:
seen.add(s)
yield s
def _GetParagraphs(self):
if self._udhr.preamble is not None:
for para in self._udhr.preamble["content"]:
yield para
for article in self._udhr.articles:
for para in article["content"]:
yield para
def _ExtractGlyphs(self, min_chars, max_chars):
s = ""
for ch in self._glyphs:
s += ch.upper()
if len(s) >= min_chars:
break
if ch != ch.upper():
s += ch
if len(s) >= min_chars:
break
return s
def _ExtractWord(self, min_chars, max_chars):
for iterator in [self._words, self._GetWords()]:
for w in iterator:
if w is None:
continue
if min_chars <= self._DisplayLength(w) <= max_chars:
return w
# Fallback to using multiple words for languages with very small words
return self._ExtractPhrase(min_chars, max_chars)
def _ExtractPhrase(self, min_chars, max_chars):
for iterator in [self._paragraphs, self._GetParagraphs()]:
for para in iterator:
if para is None:
continue
for regex in [self._any_regex, self._space_regex, self._non_word_regex]:
breaks = [-1]
for match in regex.finditer(para, min_chars):
breaks.append(match.start())
phrase = para[breaks[0] + 1 : breaks[len(breaks) - 1]]
p_size = self._DisplayLength(phrase)
while p_size > max_chars and len(breaks) > 1:
breaks.pop()
phrase = para[breaks[0] + 1 : breaks[len(breaks) - 1]]
p_size = self._DisplayLength(phrase)
if min_chars <= p_size and phrase not in self._phrase_history:
self._phrase_history.add(phrase)
return phrase
return self._ExtractParagraph(min_chars, max_chars)
def _ExtractSentence(self, min_chars, max_chars):
# Sentence delimination may differ between scripts, so tokenizing on spaces
# would be unreliable. Prefer to use _ExtractPhrase.
return self._ExtractPhrase(min_chars, max_chars)
def _ExtractParagraph(self, min_chars, max_chars):
for iterator in [self._paragraphs, self._GetParagraphs()]:
for para in iterator:
if para is None:
continue
if min_chars <= self._DisplayLength(para) <= max_chars:
return para
# Paragraphs likely insufficient length; try combining into passages
return self._ExtractPassage(min_chars, max_chars)
def _ExtractPassage(self, min_chars, max_chars):
p = []
p_size = 0
while p_size < min_chars:
for iterator in [self._paragraphs, self._GetParagraphs()]:
for para in iterator:
if para is None:
continue
p.append(para)
p_size = self._DisplayLength(" ".join(p))
if max_chars < p_size:
p.pop()
elif min_chars <= p_size:
return "\n".join(p)
assert len(p) > 0, "Unable to extract passage: " + self._udhr.key
if len(p) == 0:
p.append([p for p in self._GetParagraphs()][0])
return "\n".join(p)
def _Get(self, text_type, **kwargs):
if "char_count" in kwargs:
min_chars = kwargs["char_count"]
max_chars = kwargs["char_count"]
else:
min_chars = kwargs["min_chars"]
max_chars = kwargs["max_chars"]
if text_type == self.TextType.GLYPHS:
return self._ExtractGlyphs(min_chars, max_chars)
if text_type == self.TextType.WORD:
return self._ExtractWord(min_chars, max_chars)
if text_type == self.TextType.PHRASE:
return self._ExtractPhrase(min_chars, max_chars)
if text_type == self.TextType.SENTENCE:
return self._ExtractSentence(min_chars, max_chars)
if text_type == self.TextType.PARAGRAPH:
return self._ExtractParagraph(min_chars, max_chars)
if text_type == self.TextType.PASSAGE:
return self._ExtractPassage(min_chars, max_chars)
raise Exception("Unsupported text type: " + text_type)
def GetSampleTexts(self):
sample_text = languages_public_pb2.SampleTextProto()
sample_text.masthead_full = self._Get(self.TextType.GLYPHS, char_count=4)
sample_text.masthead_partial = self._Get(self.TextType.GLYPHS, char_count=2)
sample_text.styles = self._Get(self.TextType.PHRASE, min_chars=40, max_chars=60)
sample_text.tester = self._Get(self.TextType.PHRASE, min_chars=60, max_chars=90)
sample_text.poster_sm = self._Get(
self.TextType.PHRASE, min_chars=10, max_chars=17
)
sample_text.poster_md = self._Get(
self.TextType.PHRASE, min_chars=6, max_chars=12
)
sample_text.poster_lg = self._Get(self.TextType.WORD, min_chars=3, max_chars=8)
sample_text.specimen_48 = self._Get(
self.TextType.SENTENCE, min_chars=50, max_chars=80
)
sample_text.specimen_36 = self._Get(
self.TextType.PARAGRAPH, min_chars=100, max_chars=120
)
sample_text.specimen_32 = self._Get(
self.TextType.PARAGRAPH, min_chars=140, max_chars=180
)
sample_text.specimen_21 = self._Get(
self.TextType.PASSAGE, min_chars=300, max_chars=500
)
sample_text.specimen_16 = self._Get(
self.TextType.PASSAGE, min_chars=550, max_chars=750
)
return sample_text
|
