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 | class TestLanguageDetector(XMLTester):
"""Test the functionality of LanguageDetector"""
def setUp(self):
self.root = etree.parse(
os.path.join(
HERE,
"converter_data/samediggi-article-48s-before-lang-"
"detection-with-multilingual-tag.xml",
)
).getroot()
def test_get_main_lang(self):
test_main_lang = "sme"
language_detector = languagedetector.LanguageDetector(
self.root, LANGUAGEGUESSER
)
self.assertEqual(test_main_lang, language_detector.mainlang)
def test_set_paragraph_language_preset_language(self):
orig_paragraph = '<p xml:lang="sme">I Orohagat</p>'
expected_paragraph = etree.fromstring('<p xml:lang="sme">I Orohagat</p>')
language_detector = languagedetector.LanguageDetector(
self.root, LANGUAGEGUESSER
)
got_paragraph = language_detector.set_paragraph_language(
etree.fromstring(orig_paragraph)
)
self.assertXmlEqual(got_paragraph, expected_paragraph)
def test_set_paragraph_language_mainlanguage(self):
orig_paragraph = (
"<p>Sámegiella lea 2004 čavčča rájes standárda giellaválga "
"Microsofta operatiivavuogádagas Windows XP. Dat mearkkaša ahte "
"sámegiel bustávaid ja hámiid sáhttá válljet buot prográmmain. "
"Buot leat dás dán fitnodaga Service Pack 2-páhkas, maid ferte "
"viežžat ja bidjat dihtorii. Boađus lea ahte buot boahttevaš "
"Microsoft prográmmat dorjot sámegiela. Dattetge sáhttet "
"deaividit váttisvuođat go čálát sámegiela Outlook-kaleandaris "
"dahje e-poastta namahussajis, ja go čálát sámegillii dakkár "
"prográmmain, maid Microsoft ii leat ráhkadan.</p>"
)
expected_paragraph = etree.fromstring(
"<p>Sámegiella lea 2004 čavčča rájes standárda giellaválga "
"Microsofta operatiivavuogádagas Windows XP. Dat mearkkaša ahte "
"sámegiel bustávaid ja hámiid sáhttá válljet buot prográmmain. "
"Buot leat dás dán fitnodaga Service Pack 2-páhkas, maid ferte "
"viežžat ja bidjat dihtorii. Boađus lea ahte buot boahttevaš "
"Microsoft prográmmat dorjot sámegiela. Dattetge sáhttet "
"deaividit váttisvuođat go čálát sámegiela Outlook-kaleandaris "
"dahje e-poastta namahussajis, ja go čálát sámegillii dakkár "
"prográmmain, maid Microsoft ii leat ráhkadan.</p>"
)
language_detector = languagedetector.LanguageDetector(
self.root, LANGUAGEGUESSER
)
got_paragraph = language_detector.set_paragraph_language(
etree.fromstring(orig_paragraph)
)
self.assertXmlEqual(got_paragraph, expected_paragraph)
def test_set_paragraph_language_mainlanguage_quote_mainlang(self):
orig_paragraph = (
"<p>Sámegiella lea 2004 čavčča rájes standárda giellaválga "
"Microsofta operatiivavuogádagas Windows XP. Dat mearkkaša ahte "
"sámegiel bustávaid ja hámiid sáhttá válljet buot prográmmain. "
'<span type="quote">«Buot leat dás dán fitnodaga Service Pack 2-'
"páhkas, maid ferte viežžat ja bidjat dihtorii»</span>. Boađus "
"lea ahte buot boahttevaš Microsoft prográmmat dorjot sámegiela. "
"Dattetge sáhttet deaividit váttisvuođat go čálát sámegiela "
"Outlook-kaleandaris dahje e-poastta namahussajis, ja go čálát "
"sámegillii dakkár prográmmain, maid Microsoft ii leat "
"ráhkadan.</p>"
)
expected_paragraph = etree.fromstring(
"<p>Sámegiella lea 2004 čavčča rájes standárda giellaválga "
"Microsofta operatiivavuogádagas Windows XP. Dat mearkkaša ahte "
"sámegiel bustávaid ja hámiid sáhttá válljet buot prográmmain. "
'<span type="quote">«Buot leat dás dán fitnodaga Service Pack 2-'
"páhkas, maid ferte viežžat ja bidjat dihtorii»</span>. Boađus "
"lea ahte buot boahttevaš Microsoft prográmmat dorjot sámegiela. "
"Dattetge sáhttet deaividit váttisvuođat go čálát sámegiela "
"Outlook-kaleandaris dahje e-poastta namahussajis, ja go čálát "
"sámegillii dakkár prográmmain, maid Microsoft ii leat "
"ráhkadan.</p>"
)
language_detector = languagedetector.LanguageDetector(
self.root, LANGUAGEGUESSER
)
got_paragraph = language_detector.set_paragraph_language(
etree.fromstring(orig_paragraph)
)
self.assertXmlEqual(got_paragraph, expected_paragraph)
def test_set_paragraph_language_mainlanguage_quote_not_mainlang(self):
orig_paragraph = (
"<p>Sámegiella lea 2004 čavčča rájes standárda giellaválga "
"Microsofta operatiivavuogádagas Windows XP. Dat mearkkaša ahte "
"sámegiel bustávaid ja hámiid sáhttá válljet buot prográmmain. "
'<span type="quote">«Alt finnes i den foreliggende Service Pack 2 '
"fra selskapet, som må lastes ned og installeres på din "
"datamaskin. Konsekvensen er at all framtidig programvare fra "
"Microsoft vil inneholde støtte for samisk»</span>. Boađus lea "
"ahte buot boahttevaš Microsoft prográmmat dorjot sámegiela. "
"Dattetge sáhttet deaividit váttisvuođat go čálát sámegiela "
"Outlook-kaleandaris dahje e-poastta namahussajis, ja go čálát "
"sámegillii dakkár prográmmain, maid Microsoft ii leat "
"ráhkadan.</p>"
)
expected_paragraph = etree.fromstring(
"<p>Sámegiella lea 2004 čavčča rájes standárda giellaválga "
"Microsofta operatiivavuogádagas Windows XP. Dat mearkkaša ahte "
"sámegiel bustávaid ja hámiid sáhttá válljet buot prográmmain. "
'<span type="quote" xml:lang="nob">«Alt finnes i den foreliggende '
"Service Pack 2 fra selskapet, som må lastes ned og installeres "
"på din datamaskin. Konsekvensen er at all framtidig programvare "
"fra Microsoft vil inneholde støtte for samisk»</span>. Boađus "
"lea ahte buot boahttevaš Microsoft prográmmat dorjot sámegiela. "
"Dattetge sáhttet deaividit váttisvuođat go čálát sámegiela "
"Outlook-kaleandaris dahje e-poastta namahussajis, ja go čálát "
"sámegillii dakkár prográmmain, maid Microsoft ii leat "
"ráhkadan.</p>"
)
language_detector = languagedetector.LanguageDetector(
self.root, LANGUAGEGUESSER
)
got_paragraph = language_detector.set_paragraph_language(
etree.fromstring(orig_paragraph)
)
self.assertXmlEqual(got_paragraph, expected_paragraph)
def test_set_paragraph_language_not_mainlanguage(self):
orig_paragraph = (
"<p>Samisk er fra høsten 2004 et standard språkvalg Microsofts "
"operativsystem Windows XP. I praksis betyr det at samiske "
"bokstaver og formater kan velges i alle programmer. Alt finnes "
"i den foreliggende Service Pack 2 fra selskapet, som må lastes "
"ned og installeres på din datamaskin. Konsekvensen er at all "
"framtidig programvare fra Microsoft vil inneholde støtte for "
"samisk. Du vil imidlertid fremdeles kunne oppleve problemer med "
"å skrive samisk i Outlook-kalenderen eller i tittel-feltet i "
"e-post, og med å skrive samisk i programmer levert av andre enn "
"Microsoft.</p>"
)
language_detector = languagedetector.LanguageDetector(
self.root, LANGUAGEGUESSER
)
got_paragraph = language_detector.set_paragraph_language(
etree.fromstring(orig_paragraph)
)
self.assertEqual(
got_paragraph.get("{http://www.w3.org/XML/1998/namespace}lang"), "nob"
)
def test_remove_quote(self):
orig_paragraph = (
'<p>bla bla <span type="quote">bla1 bla</span> ble ble '
'<span type="quote">bla2 bla</span> <b>bli</b> bli '
'<span type="quote">bla3 bla</span> blo blo</p>'
)
expected_paragraph = "bla bla ble ble bli bli blo blo"
language_detector = languagedetector.LanguageDetector(
self.root, LANGUAGEGUESSER
)
got_paragraph = language_detector.remove_quote(etree.fromstring(orig_paragraph))
self.assertEqual(got_paragraph, expected_paragraph)
def test_detect_language_with_multilingualtag(self):
root = etree.parse(
os.path.join(
HERE,
"converter_data/samediggi-article-48s-before-"
"lang-detection-with-multilingual-tag.xml",
)
).getroot()
language_detector = languagedetector.LanguageDetector(root, LANGUAGEGUESSER)
language_detector.detect_language()
got_document = language_detector.document
expected_document = etree.parse(
os.path.join(
HERE,
"converter_data/samediggi-article-48s-after-lang-"
"detection-with-multilingual-tag.xml",
)
)
self.assertXmlEqual(got_document, expected_document)
def test_detect_language_without_multilingualtag(self):
root = etree.parse(
os.path.join(
HERE,
"converter_data/samediggi-article-48s-before-lang-"
"detection-without-multilingual-tag.xml",
)
).getroot()
language_detector = languagedetector.LanguageDetector(root, LANGUAGEGUESSER)
language_detector.detect_language()
got_document = language_detector.document
expected_document = etree.parse(
os.path.join(
HERE,
"converter_data/samediggi-article-48s-after-lang-"
"detection-without-multilingual-tag.xml",
)
)
self.assertXmlEqual(got_document, expected_document)
def test_no_lang_guessing_without_models(self):
test_document = """
<document xml:lang="non_existing_mainlang">
<header>
<title>title</title>
<multilingual>
<language xml:lang="non_existing_optional_lang"/>
</multilingual>
</header>
<body>
<p>content</p>
</body>
</document>
"""
root = etree.fromstring(test_document)
language_detector = languagedetector.LanguageDetector(root, LANGUAGEGUESSER)
language_detector = languagedetector.LanguageDetector(root, LANGUAGEGUESSER)
language_detector.detect_language()
got_document = language_detector.document
self.assertXmlEqual(got_document, root)
|