def add_mark_char(char, mark):
"""
Add mark to a single char.
"""
if char == u'':
return u''
case = char.isupper()
ac = accent.get_accent_char(char)
char = accent.add_accent_char(char.lower(), Accent.NONE)
new_char = char
if mark == Mark.HAT:
if char in FAMILY_A:
new_char = u"â"
elif char in FAMILY_O:
new_char = u"ô"
elif char in FAMILY_E:
new_char = u"ê"
elif mark == Mark.HORN:
if char in FAMILY_O:
new_char = u"ơ"
elif char in FAMILY_U:
new_char = u"ư"
elif mark == Mark.BREVE:
if char in FAMILY_A:
new_char = u"ă"
elif mark == Mark.BAR:
if char in FAMILY_D:
new_char = u"đ"
elif mark == Mark.NONE:
if char in FAMILY_A:
new_char = u"a"
elif char in FAMILY_E:
new_char = u"e"
elif char in FAMILY_O:
new_char = u"o"
elif char in FAMILY_U:
new_char = u"u"
elif char in FAMILY_D:
new_char = u"d"
new_char = accent.add_accent_char(new_char, ac)
return utils.change_case(new_char, case)
def add_mark_char(char, mark):
"""
Add mark to a single char.
"""
if char == u'':
return u''
case = char.isupper()
ac = accent.get_accent_char(char)
char = accent.add_accent_char(char.lower(), Accent.NONE)
new_char = char
if mark == Mark.HAT:
if char in FAMILY_A:
new_char = u"â"
elif char in FAMILY_O:
new_char = u"ô"
elif char in FAMILY_E:
new_char = u"ê"
elif mark == Mark.HORN:
if char in FAMILY_O:
new_char = u"ơ"
elif char in FAMILY_U:
new_char = u"ư"
elif mark == Mark.BREVE:
if char in FAMILY_A:
new_char = u"ă"
elif mark == Mark.BAR:
if char in FAMILY_D:
new_char = u"đ"
elif mark == Mark.NONE:
if char in FAMILY_A:
new_char = u"a"
elif char in FAMILY_E:
new_char = u"e"
elif char in FAMILY_O:
new_char = u"o"
elif char in FAMILY_U:
new_char = u"u"
elif char in FAMILY_D:
new_char = u"d"
new_char = accent.add_accent_char(new_char, ac)
return utils.change_case(new_char, case)
def is_valid_combination(components):
"""Check if a character combination complies to Vietnamese spelling.
Input:
components - a list of the form [u'c', u'a', u'm']
Output:
True if OK, False otherwise.
"""
comps = list(components)
# We only work with lower case
for i in range(len(comps)):
comps[i] = utils.change_case(comps[i], 0)
# Allow 'đ' to appear in abbreviations like 'đm', 'đc', 'kgcđ', etc.
#if comps[0] and not comps[1] and not comps[2] and \
#not comps[0] in ('gi', 'qu'):
#for c in comps[0]:
#if not c in CONSONANTS:
#return False
#return True
if comps[0] and not comps[1] and not comps[2]:
return True
# Check if our start sound is a proper consonant
if (comps[0] != u'') and (not (comps[0] in CONSONANTS)):
return False
# And if our ending sound is a proper ending consonant
if (comps[2] != u'') and (not (comps[2] in ENDING_CONSONANTS)):
return False
vowel = accent.remove_accent_string(comps[1])
if len(vowel) > 1:
if not (vowel in OPEN_COMPOUND_VOWELS or \
vowel in CLOSED_COMPOUND_VOWELS):
return False
if vowel in CLOSED_COMPOUND_VOWELS and \
not vowel in OPEN_COMPOUND_VOWELS and comps[2] != u'':
return False
# 'ăch'?
if comps[2] == u'ch' and ((vowel in u'ăâeôơuư') or \
(vowel in OPEN_COMPOUND_VOWELS and not vowel in CLOSED_COMPOUND_VOWELS)):
return False
# 'ương' is ok but 'ơng' ?
if comps[2] == u'ng' and vowel in (u'ơ'):
return False
# Sadly, this interferes with 'nhếch' :<
#if comps[2] == u'c' and vowel in u'ê':
# return False
# Get the first accent
ac = Accent.NONE
for i in range(len(comps[1])):
a = accent.get_accent_char(comps[1][i])
if a != Accent.NONE:
ac = a
break
# These consonants can only go with ACUTE, DOT or NONE accents
if comps[2] in [u'c', u'p', u't', u'ch'] and \
not ac in [Accent.NONE, Accent.ACUTE, Accent.DOT]:
return False
return True
def transform(comps, trans):
"""
Transform the given string with transform type trans
"""
components = list(comps)
# Special case for 'ư, ơ'
#if trans[0] == '<' and not trans[1] in (u'ư', u'ơ', u'Ư', u'Ơ'):
# trans = '+' + trans[1]
# (Not our job)
if trans[0] == u'<':
if not components[2]:
# Undo operation
if components[1][-1:] == trans[1]:
return components
# Only allow ư, ơ or ươ sitting alone in the middle part
elif not components[1] or \
(components[1].lower() == u'ư' and trans[1].lower() == u'ơ'):
components[1] += trans[1]
# Quite a hack. If you want to type gi[f = 'giờ', separate()
# will create ['g', 'i', '']. Therefore we have to allow
# components[1] == 'i'.
elif components[1].lower() == 'i' and components[0].lower() == 'g':
components[1] += trans[1]
components = separate(utils.join(components))
if trans[0] == u'+':
# See this and you'll understand:
# transform([u'nn', '', ''],'+n') = [u'nnn', '', '']
# transform([u'c', '', ''],'+o') = [u'c', 'o', '']
# transform([u'c', 'o', ''],'+o') = [u'c', 'oo', '']
# transform([u'c', 'o', ''],'+n') = [u'c', 'o', 'n']
if components[1] == u'':
if utils.is_vowel(trans[1]):
components[1] += trans[1]
else:
components[0] += trans[1]
else:
if components[2] == u'' and utils.is_vowel(trans[1]):
components[1] += trans[1]
else:
components[2] += trans[1]
# If there is any accent, remove and reapply it
# because it is likely to be misplaced in previous transformations
ac = accent.Accent.NONE
for c in components[1]:
ac = accent.get_accent_char(c)
if ac:
break
if ac != accent.Accent.NONE:
# Remove accent
components = accent.add_accent(components, Accent.NONE)
components = accent.add_accent(components, ac)
return components
action, factor = get_action(trans)
if action == Action.ADD_ACCENT:
components = accent.add_accent(components, factor)
elif action == Action.ADD_MARK:
if (mark.is_valid_mark(components, trans)):
components = mark.add_mark(components, factor)
return components
def transform(comps, trans):
"""
Transform the given string with transform type trans
"""
components = list(comps)
# Special case for 'ư, ơ'
#if trans[0] == '<' and not trans[1] in (u'ư', u'ơ', u'Ư', u'Ơ'):
# trans = '+' + trans[1]
# (Not our job)
if trans[0] == u'<':
if not components[2]:
# Undo operation
if components[1][-1:] == trans[1]:
return components
# Only allow ư, ơ or ươ sitting alone in the middle part
elif not components[1] or \
(components[1].lower() == u'ư' and trans[1].lower() == u'ơ'):
components[1] += trans[1]
# Quite a hack. If you want to type gi[f = 'giờ', separate()
# will create ['g', 'i', '']. Therefore we have to allow
# components[1] == 'i'.
elif components[1].lower() == 'i' and components[0].lower() == 'g':
components[1] += trans[1]
components = separate(utils.join(components))
if trans[0] == u'+':
# See this and you'll understand:
# transform([u'nn', '', ''],'+n') = [u'nnn', '', '']
# transform([u'c', '', ''],'+o') = [u'c', 'o', '']
# transform([u'c', 'o', ''],'+o') = [u'c', 'oo', '']
# transform([u'c', 'o', ''],'+n') = [u'c', 'o', 'n']
if components[1] == u'':
if utils.is_vowel(trans[1]):
components[1] += trans[1]
else:
components[0] += trans[1]
else:
if components[2] == u'' and utils.is_vowel(trans[1]):
components[1] += trans[1]
else:
components[2] += trans[1]
# If there is any accent, remove and reapply it
# because it is likely to be misplaced in previous transformations
ac = accent.Accent.NONE
for c in components[1]:
ac = accent.get_accent_char(c)
if ac:
break
if ac != accent.Accent.NONE:
# Remove accent
components = accent.add_accent(components, Accent.NONE)
components = accent.add_accent(components, ac)
return components
action, factor = get_action (trans)
if action == Action.ADD_ACCENT:
components = accent.add_accent(components, factor)
elif action == Action.ADD_MARK:
if (mark.is_valid_mark(components, trans)):
components = mark.add_mark(components, factor)
return components
def is_valid_combination(components):
"""Check if a character combination complies to Vietnamese spelling.
Input:
components - a list of the form [u'c', u'a', u'm']
Output:
True if OK, False otherwise.
"""
comps = list(components)
# We only work with lower case
for i in range(len(comps)):
comps[i] = utils.change_case(comps[i], 0)
# Allow 'đ' to appear in abbreviations like 'đm', 'đc', 'kgcđ', etc.
#if comps[0] and not comps[1] and not comps[2] and \
#not comps[0] in ('gi', 'qu'):
#for c in comps[0]:
#if not c in CONSONANTS:
#return False
#return True
if comps[0] and not comps[1] and not comps[2]:
return True
# Check if our start sound is a proper consonant
if (comps[0] != u'') and (not (comps[0] in CONSONANTS)):
return False
# And if our ending sound is a proper ending consonant
if (comps[2] != u'') and (not (comps[2] in ENDING_CONSONANTS)):
return False
vowel = accent.remove_accent_string(comps[1])
if len(vowel) > 1:
if not (vowel in OPEN_COMPOUND_VOWELS or \
vowel in CLOSED_COMPOUND_VOWELS):
return False
if vowel in CLOSED_COMPOUND_VOWELS and \
not vowel in OPEN_COMPOUND_VOWELS and comps[2] != u'':
return False
# 'ăch'?
if comps[2] == u'ch' and ((vowel in u'ăâeôơuư') or \
(vowel in OPEN_COMPOUND_VOWELS and not vowel in CLOSED_COMPOUND_VOWELS)):
return False
# 'ương' is ok but 'ơng' ?
if comps[2] == u'ng' and vowel in (u'ơ'):
return False
# Sadly, this interferes with 'nhếch' :<
#if comps[2] == u'c' and vowel in u'ê':
# return False
# Get the first accent
ac = Accent.NONE
for i in range(len(comps[1])):
a = accent.get_accent_char(comps[1][i])
if a != Accent.NONE:
ac = a
break
# These consonants can only go with ACUTE, DOT or NONE accents
if comps[2] in [u'c', u'p', u't', u'ch'] and \
not ac in [Accent.NONE, Accent.ACUTE, Accent.DOT]:
return False
return True
