def separate(string):
"""
Separates a valid Vietnamese word into 3 components:
the start sound, the middle sound and the end sound.
Eg: toán -> [u't', u'oá', u't']
Otherwise returns None (not a valid Vietnamese word).
"""
comps = [u'', u'', u'']
if string == u'':
return comps
# Search for the first vowel
for i in range(len(string)):
if utils.is_vowel(string[i]):
comps[0] = u'' + string[:i]
string = u'' + string[i:]
break
# No vowel?
if comps[0] == u'' and not utils.is_vowel(string[0]):
comps[0] = string
string = u''
# Search for the first consonant after the first vowel
for i in range(len(string)):
if not utils.is_vowel(string[i]):
comps[1] = string[:i]
comps[2] = string[i:]
break
# No ending consonant? Then the rest of the string must be the vowel part
if comps[1] == u'':
comps[1] = string
# 'gi' and 'qu' need some special treatments
# We want something like this:
# ['g', 'ia', ''] -> ['gi', 'a', '']
if (comps[0] != u'' and comps[1] != u'') and \
((comps[0] in u'gG' and comps[1][0] in 'iI' and len(comps[1]) > 1) or \
(comps[0] in u'qQ' and comps[1][0] in 'uU')):
comps[0] += comps[1][:1]
comps[1] = comps[1][1:]
if not is_valid_combination(comps):
return None
return comps
def separate(string):
"""
Separates a valid Vietnamese word into 3 components:
the start sound, the middle sound and the end sound.
Eg: toán -> [u't', u'oá', u't']
Otherwise returns None (not a valid Vietnamese word).
"""
comps = [u'', u'', u'']
if string == u'':
return comps
# Search for the first vowel
for i in range(len(string)):
if utils.is_vowel(string[i]):
comps[0] = u'' + string[:i]
string = u'' + string[i:]
break
# No vowel?
if comps[0] == u'' and not utils.is_vowel(string[0]):
comps[0] = string
string = u''
# Search for the first consonant after the first vowel
for i in range(len(string)):
if not utils.is_vowel(string[i]):
comps[1] = string[:i]
comps[2] = string[i:]
break
# No ending consonant? Then the rest of the string must be the vowel part
if comps[1] == u'':
comps[1] = string
# 'gi' and 'qu' need some special treatments
# We want something like this:
# ['g', 'ia', ''] -> ['gi', 'a', '']
if (comps[0] != u'' and comps[1] != u'') and \
((comps[0] in u'gG' and comps[1][0] in 'iI' and len(comps[1]) > 1) or \
(comps[0] in u'qQ' and comps[1][0] in 'uU')):
comps[0] += comps[1][:1]
comps[1] = comps[1][1:]
if not is_valid_combination(comps):
return None
return comps
#
# Note that when undo 'ư' with 'w', this function will always return
# 'uw' because of lack of raw string information. It is up to the
# user of this module to change the returned value to 'w' when necessary.
#
if new_comps == comps:
for trans in trans_list:
new_comps = reverse(new_comps, trans)
tmp = list(new_comps)
if tmp != comps:
new_comps = utils.append_comps(new_comps, unicode(key))
return garbage + utils.join(new_comps)
new_comps = utils.append_comps(new_comps, unicode(key))
# One last check to rule out cases like 'ảch' or 'chuyểnl'
if not is_valid_combination(new_comps):
return None
return garbage + utils.join(new_comps)
def get_transformation_list(key, im, case=0):
"""
Return list of transformations inferred from entered key. The
map between transform types and keys is given by module
bogo_config (if exists) or by variable simple_telex_im
if entered key is not in im, return u"+key", meaning appending
the entered key to current text
"""
if key in im:
lkey = key
def process_key(string, key, case = 0, config = DefaultConfig()):
"""
Process the given string and key based on the given input method and
config.
Args:
string -
key -
case (optional) - Force the output's case. Mostly to determine
the case of TELEX's [, ] keys. 0: lower, 1: upper. Default: 0.
im (optional) - one of 'telex', 'simple-telex', 'vni'.
Default: 'telex'.
config - a dictionary.
"""
## BEGIN TRICKS (scroll down please)
im = config.input_method
# People can sometimes be really mischievous :<
if im in IMs:
im = IMs[im]
else:
im = IMs['telex']
# Handle non-alpha string like 'tôi_là_ai' by putting 'tôi_là_' in
# the `garbage` variable, effectively skipping it then put it back
# later.
# TODO Should this be the ibus engine's job?
garbage = u''
for i in range(-1, -len(string)-1, -1): # Reverse indices [-1, -2, -3, ...]
if not string[i].isalpha():
garbage += string[:i] + string[i]
string = u'' + string[i+1:] if i != -1 else u''
break
# Handle process_key('â', '_')
if not key in im and not key.isalpha():
string += key
return garbage + string
## END TRICKS (here comes real code)
# Try to break the string down to 3 components
# separate('chuyen') = ['ch', 'uye', 'n']
comps = separate(string)
# Refuse to process things like process('zzam', 'f')
if comps == None:
return None
# Apply transformations
trans_list = get_transformation_list(key, im, case = case);
new_comps = comps
for trans in trans_list:
new_comps = transform(new_comps, trans)
# Double typing an IM key to undo.
# Eg: process_key(u'à', 'f')
# -> transform(['', u'à', ''], '\\') = ['', 'à', '']
# -> reverse(u'à', '\\') = 'a'
#
# Note that when undo 'ư' with 'w', this function will always return
# 'uw' because of lack of raw string information. It is up to the
# user of this module to change the returned value to 'w' when necessary.
#
if new_comps == comps:
for trans in trans_list:
new_comps = reverse(new_comps, trans)
tmp = list(new_comps)
if tmp != comps:
new_comps = utils.append_comps(new_comps, unicode(key))
return garbage + utils.join(new_comps)
new_comps = utils.append_comps(new_comps, unicode(key))
# One last check to rule out cases like 'ảch' or 'chuyểnl'
if not is_valid_combination(new_comps):
return None
return garbage + utils.join(new_comps)
