def test_unidecode_bmp(self):
for n in range(0,0x10000):
# Just check that it doesn't throw an exception
try:
t = unichr(n)
unidecode(t)
except:
print("catch error at %02x"%n)
def test_unidecode_kana(self):
for n in range(0x3000,0x30ff):
# Just check that it doesn't throw an exception
try:
t = six.unichr(n)
unidecode(t)
except:
print("catch error at %02x"%n)
def test_unidecode_kana(self):
for n in range(0x3000, 0x30ff):
# Just check that it doesn't throw an exception
try:
t = six.unichr(n)
unidecode(t)
except:
print("catch error at %02x" % n)
def test_unidecode_compatibility_composite(self):
TESTS = [
("\ufb01", "fi"),
("\u0032\u2075", "25"),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_specific_bmp(self):
TESTS = [
("Hello, World!", "Hello, World!"),
("'\"\r\n", "'\"\r\n"),
("ČŽŠčžš", "CZSczs"),
("\u00a0\u00a1\u00a2\u00a3\u00a4\u00a5\u00a6\u00a7",
" !C/PS\u005c$?Y=|SS"),
("\u00a8\u00a9\u00aa\u00ab\u00ac\u00ad\u00ae\u00af",
"\u0022(c)a<<!(r)-"),
("ア", "a"),
("α", "a"),
("а", "a"),
('ch\xe2teau', "chateau"),
('vi\xf1edos', "vinedos"),
("\u5317\u4EB0", "Bei Jing "),
("Efficient", "Efficient"),
# Table that doesn't exist
('\ua500', ''),
# Table that has less than 256 entriees
('\u1eff', ''),
# Mark area
(
"\u210a", #gram mark
"g"),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_compatibility_composite(self):
TESTS = [
(u"\ufb01","fi"),
(u"\u0032\u2075", "25"),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_combining_chars(self):
TESTS = [
# roman number "1" wrapped with solid square
("\u0031\u20de", "1"),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def _standardized_fsn(self, fsn):
"""Convert FSN-Latin part into FSN-English to standardize FSN:input original FSN，output FSN_std."""
# Determine whether the FSN is (English or numeric).
def is_English(s):
ans = re.search(r"[0-9A-Z\']+$", s)
return True if ans else False
fsn['isEnglish'] = fsn['FSN'].astype(str).apply(is_English)
print("FSN_English&Num shape=", list(fsn['isEnglish']).count(True))
print("FSN_Latin shape=", list(fsn['isEnglish']).count(False))
fsn_en = fsn[fsn['isEnglish'] == True]
fsn_latin = fsn[fsn['isEnglish'] == False]
for df in [fsn_latin, fsn_en]:
df.drop([
'isEnglish',
], axis=1, inplace=True)
# Convert FSN_Latin into FSN_English
fsn_latin['FSN_std'] = fsn_latin['FSN'].apply(
lambda x: unihandecode.unidecode(x))
fsn_latin = fsn_latin[[
'FSN', 'FSN_std', 'continent_code', 'country_iso_code',
'subdivision_1_iso_code', 'city'
]]
for i in "()/'`!- _":
fsn_latin['FSN_std'] = fsn_latin['FSN_std'].str.replace(i, '')
# fsn_latin.to_csv(OUTPUT_DIR+'/'+'FSN-latin_part_stded.csv',index=None)
fsn_en['FSN_std'] = fsn_en['FSN']
fsn_std = pd.concat([fsn_latin, fsn_en],
axis=0,
ignore_index=True,
sort=False)
fsn_std.to_csv(self._fsn_path, index=None)
def test_unidecode_combining_chars(self):
TESTS = [
# roman number "1" wrapped with solid square
(u"\u0031\u20de", "1"),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_zh(self):
ZHTESTS = [
(u"\u660e\u5929\u660e\u5929\u7684\u98ce\u5439",
'Ming Tian Ming Tian De Feng Chui '),
(u"馮", "Feng "),
]
for input, output in ZHTESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_zh(self):
ZHTESTS = [
("\u660e\u5929\u660e\u5929\u7684\u98ce\u5439",
'Ming Tian Ming Tian De Feng Chui '),
("馮", "Feng "),
]
for input, output in ZHTESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_decomposed_form(self):
TESTS = [
("\u0041\u0301", "A"), # "A" with accent mark
("\u0061\u0323\u0302", "a"), # "a" with accent marks
("\u30AB\u3099", "ga"), # "ガ" coded by decomposed from as ' カ゛ '
("\u304B\u3099", "ga"), # "が" coded by decomposed from as ' か゛ '
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_decomposed_form(self):
TESTS = [
(u"\u0041\u0301", "A"), # "A" with accent mark
(u"\u0061\u0323\u0302", "a"), # "a" with accent marks
(u"\u30AB\u3099", "ga"), # "ガ" coded by decomposed from as ' カ゛ '
(u"\u304B\u3099", "ga"), # "が" coded by decomposed from as ' か゛ '
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_mac_japanese_pua(self):
TESTS = [
(u"\uF862\u6709\u9650\u4F1A\u793E", #Adobe CID 8321
"You Xian Hui She "), # "yuugengaisha" in unihandecode(ja)
(u"\u5927\u20dd", "Da "), # "大" with circle "Dai " in unihandecode(ja)
(u"\u5c0f\u20dd", "Xiao "), # "小" with circle "Shou " in unihandecode(ja)
(u"\u63a7\u20dd", "Kong "), # "控" with circle "Hikae " in unihandecode(ja)
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_mathematical_digits(self):
if sys.maxunicode < 0x1d800:
print("skip test because of Narrow Python")
return
# 5 consecutive sequences of 0-9
for n in range(0x1d7ce, 0x1d800):
a = chr(ord('0') + (n - 0x1d7ce) % 10)
b = unidecode(six.unichr(n))
self.assertEqual(b, a)
def test_unidecode_mathematical_digits(self):
if sys.maxunicode < 0x1d800:
print("skip test because of Narrow Python")
return
# 5 consecutive sequences of 0-9
for n in range(0x1d7ce, 0x1d800):
a = chr(ord('0') + (n-0x1d7ce) % 10)
b = unidecode(unichr(n))
self.assertEqual(b, a)
def test_unidecode_specific_bmp(self):
TESTS = [
(u"Hello, World!",
"Hello, World!"),
(u"'\"\r\n",
"'\"\r\n"),
(u"ČŽŠčžš",
"CZSczs"),
(u"\u00a0\u00a1\u00a2\u00a3\u00a4\u00a5\u00a6\u00a7",
u" !C/PS\u005c$?Y=|SS"),
(u"\u00a8\u00a9\u00aa\u00ab\u00ac\u00ad\u00ae\u00af",
u"\u0022(c)a<<!(r)-"),
(u"ア",
"a"),
(u"α",
"a"),
(u"а",
"a"),
(u'ch\xe2teau',
"chateau"),
(u'vi\xf1edos',
"vinedos"),
(u"\u5317\u4EB0",
"Bei Jing "),
(u"Efficient",
"Efficient"),
# Table that doesn't exist
(u'\ua500',
''),
# Table that has less than 256 entriees
(u'\u1eff',
''),
# Mark area
(u"\u210a", #gram mark
"g"),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def createImage(postContent):
font = ImageFont.truetype("Andale Mono.ttf", 15)
im = Image.new(mode="RGB", size=(1280, 720), color=(153, 153, 255))
draw = ImageDraw.Draw(im)
# convert from latin-1 encoding to ascii
postContent = unihandecode.unidecode(postContent)
# lines are only 120 characters
text = textwrap.fill(postContent, 120)
draw.text((100, 100), text, fill=(255, 255, 0)) # draw lines on image
# im.show()
im.save("image.jpg")
def text_clean(x):
x = re.sub("<.*?>", "", x) #remove html tags
x = re.sub(r'([^\s\w]|_)+', '',
x) #retain only alphanumeric chars and space
x = unihandecode.unidecode(x) #Convert latin to ASCII characters
x = x.lower() #convert to lower case characters
x = x.translate(str.maketrans(
"", "", string.punctuation)) #removing punctuation marks
x = x.translate(str.maketrans("", "", digits)) #removing numbers
x = x.strip() #removing leading and trailing white spaces
x = PorterStemmer().stem(x) #stem to root word
#x = x.replace(" ", "") #remove white space
return x
def test_unidecode_specific_supplementary(self):
if sys.maxunicode < 0x1d6a4:
print("skip test because of Narrow Python")
return
TESTS = [
# Non-BMP character
('\U0001d5a0', 'A'),
# Mathematical
('\U0001d5c4\U0001d5c6/\U0001d5c1', 'km/h'),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_mac_japanese_pua(self):
TESTS = [
(
"\uF862\u6709\u9650\u4F1A\u793E", #Adobe CID 8321
"You Xian Hui She "), # "yuugengaisha" in unihandecode(ja)
("\u5927\u20dd",
"Da "), # "大" with circle "Dai " in unihandecode(ja)
("\u5c0f\u20dd",
"Xiao "), # "小" with circle "Shou " in unihandecode(ja)
("\u63a7\u20dd",
"Kong "), # "控" with circle "Hikae " in unihandecode(ja)
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_mathematical_latin():
# 13 consecutive sequences of A-Z, a-z with some codepoints
# undefined. We just count the undefined ones and don't check
# positions.
empty = 0
for n in range(0x1d400, 0x1d6a4):
if n % 52 < 26:
a = chr(ord('A') + n % 26)
else:
a = chr(ord('a') + n % 26)
b = unihandecode.unidecode(chr(n))
if not b:
empty += 1
else:
assert b == a
assert empty == 24
def test_unidecode_specific_supplementary(self):
if sys.maxunicode < 0x1d6a4:
print("skip test because of Narrow Python")
return
TESTS = [
# Non-BMP character
(u'\U0001d5a0',
'A'),
# Mathematical
(u'\U0001d5c4\U0001d5c6/\U0001d5c1',
'km/h'),
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_mathematical_latin(self):
# 13 consecutive sequences of A-Z, a-z with some codepoints
# undefined. We just count the undefined ones and don't check
# positions.
if sys.maxunicode < 0x1d6a4:
print("skip test because of Narrow Python")
return
empty = 0
for n in range(0x1d400, 0x1d6a4):
if n % 52 < 26:
a = chr(ord('A') + n % 26)
else:
a = chr(ord('a') + n % 26)
b = unidecode(six.unichr(n))
if not b:
empty += 1
else:
self.assertEqual(b, a)
self.assertEqual(empty, 24)
def test_unidecode_mathematical_latin(self):
# 13 consecutive sequences of A-Z, a-z with some codepoints
# undefined. We just count the undefined ones and don't check
# positions.
if sys.maxunicode < 0x1d6a4:
print("skip test because of Narrow Python")
return
empty = 0
for n in range(0x1d400, 0x1d6a4):
if n % 52 < 26:
a = chr(ord('A') + n % 26)
else:
a = chr(ord('a') + n % 26)
b = unidecode(unichr(n))
if not b:
empty += 1
else:
self.assertEqual(b, a)
self.assertEqual(empty, 24)
def test_unidecode_squared_chars(self):
TESTS = [
(u"\u3301", "alpha"), # combined Alpha in Katakana
(u"\u3302", "ampere"), # combined Ampere in Katakana
(u"\u3304", "inning"),
(u"\u3306", "won"), # combined Won in Katakana
(u"\u3307", "escudo"),
(u"\u3308", "acre"), # combined Acre in Katakana
(u"\u3309", "ounce"), # combined ounce in Katakana
(u"\u330a", "ohm"), # combined Ohm in Katakana
(u"\u3349", "milli"), # milli in Katakana
(u"\u3314", "kilo"), # kilo in Katakana
(u"\u3315", "kilogram"), # kilo gram in Katakana
(u"\u3316", "kilometer"), # kilo metre in Katakana
(u"\u3322", "centi"), # centi in Katakana
(u"\u334d", "meter"), #metre in Katakana
(u"\u3318", "gram"), # gram in Katakana
(u"\u3327", "ton"), # ton in Katakana
(u"\u3303", "are"), # are in Katakana
(u"\u3336", "hectare"), # hect-are in Katakana
(u"\u337f", "Inc."), # kabusiki kaisha in Katakana
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_squared_chars(self):
TESTS = [
("\u3301", "alpha"), # combined Alpha in Katakana
("\u3302", "ampere"), # combined Ampere in Katakana
("\u3304", "inning"),
("\u3306", "won"), # combined Won in Katakana
("\u3307", "escudo"),
("\u3308", "acre"), # combined Acre in Katakana
("\u3309", "ounce"), # combined ounce in Katakana
("\u330a", "ohm"), # combined Ohm in Katakana
("\u3349", "milli"), # milli in Katakana
("\u3314", "kilo"), # kilo in Katakana
("\u3315", "kilogram"), # kilo gram in Katakana
("\u3316", "kilometer"), # kilo metre in Katakana
("\u3322", "centi"), # centi in Katakana
("\u334d", "meter"), #metre in Katakana
("\u3318", "gram"), # gram in Katakana
("\u3327", "ton"), # ton in Katakana
("\u3303", "are"), # are in Katakana
("\u3336", "hectare"), # hect-are in Katakana
("\u337f", "Inc."), # kabusiki kaisha in Katakana
]
for input, output in TESTS:
self.assertEqual(unidecode(input), output)
def test_unidecode_kana():
for n in range(0x3000, 0x30ff):
# Just check that it doesn't throw an exception
t = chr(n)
unihandecode.unidecode(t)
def test_unidecode_mathematical_digits():
# 5 consecutive sequences of 0-9
for n in range(0x1d7ce, 0x1d800):
a = chr(ord('0') + (n - 0x1d7ce) % 10)
b = unihandecode.unidecode(chr(n))
assert b == a
def test_unidecode_ascii():
for n in range(0, 128):
t = chr(n)
assert (n, unihandecode.unidecode(t)) == (n, t)
def test_unidecode_ascii(self):
for n in range(0,128):
t = chr(n)
self.assertEqual(unidecode(t), t)
def test_unidecode_ascii(self):
for n in range(0, 128):
t = chr(n)
self.assertEqual(unidecode(t), t)
def test_unidecode_bmp():
# Just check that it doesn't throw an exception
for n in range(0, 0x10000):
t = chr(n)
unihandecode.unidecode(t)
