def get_middle_name(conn, n1, n2, saju, gender, n2_rsc_type, mode):
name_dict = {}
s = conn.cursor()
query = """
SELECT hanja,reading,strokes,add_strokes,rsc_type,pronunciations
FROM naming_hanja
WHERE is_naming_hanja=1 AND reading
NOT IN ('만', '병', '백', '장', '춘', '최', '충', '창', '치', '참', '천',
'택', '탁', '태', '외', '사', '매', '읍', '소', '종', '순', '요', '자',
'경', '옥', '해', '부', '효', '존', '난', '류', '홍', '처', '필')
"""
for n3 in s.execute(query):
if n1[HANJA] == n3[HANJA] or n2[HANJA] == n3[HANJA]: # 김주김, 김소소
continue
elif n1[READING] == n3[READING] or n2[READING] == n3[READING]: # 김주김, 김소소
continue
n2n3 = '%s%s' % (n2[READING], n3[READING])
if check_name(n2n3) is False:
continue
if check_last_name_gender(n3[READING], n2n3, gender) is False:
continue
if saju['c1'] != n3[RSC_TYPE] and saju['c2'] != n3[RSC_TYPE]:
continue
if n2_rsc_type == n3[RSC_TYPE]: # 중간이름에서 사용한 자원오행은 배제함
continue
if check_positive_negative(conn, n1, n2, n3) is False:
continue
# check positive negative hanja strokes as well
if check_total_stroke(conn, n1, n2, n3, mode) is False:
continue
s1 = split_syllable_char(n1[READING])
s2 = split_syllable_char(n2[READING])
s3 = split_syllable_char(n3[READING])
if balum_oheng(s1, s2, s3) is False:
continue
if check_all_name_hard_pronounce(s1, s2, s3) is False:
continue
temp_hanja = '%s%s%s' % (n1[HANJA], n2[HANJA], n3[HANJA])
# temp_hanja = '%s %s %s [%s/ %s]' % (n1[HANJA], n2[HANJA], n3[HANJA],
# n2[PRONUNCIATIONS], n3[PRONUNCIATIONS])
temp_name = '%s%s%s' % (n1[READING], n2[READING], n3[READING])
name_dict.update({temp_hanja: temp_name})
if len(name_dict) == 0:
return None
return name_dict
def check_two_words_hard_pronounce(n1, n2):
s1 = split_syllable_char(n1)
s2 = split_syllable_char(n2)
if len(s1) == 3 and len(s2) == 3:
if s1[1] == 'ㅕ' and s1[2] == 'ㅇ' and s2[1] == 'ㅡ' and s2[2] == 'ㅇ': # 경흥원
return False
elif s1[1] == 'ㅏ' and s1[2] == 'ㅇ' and s2[1] == 'ㅏ' and s2[2] == 'ㅇ': # 강항준
return False
elif s1[1] == 'ㅏ' and s1[2] == 'ㅇ' and s2[1] == 'ㅑ' and s2[2] == 'ㅇ': # 강향준
return False
return True
def get_label(index_data):
# len + 1 for one 'invalid' label
label_ko_cho = np.zeros([len(index_data), len(ko_chset_cho)+1])
label_ko_jung = np.zeros([len(index_data), len(ko_chset_jung)+1])
label_ko_jong = np.zeros([len(index_data), len(ko_chset_jong)+1])
label_en = np.zeros([len(index_data), len(en_chset)+1])
for i, member in enumerate(index_data):
target = member['target'] # Target Character
# Is Hangeul?
if (len(target) == 1 and check_syllable(target)):
splited = split_syllable_char(target)
label_ko_cho[i][ko_chset_cho.index(splited[0])] = 1
label_ko_jung[i][ko_chset_jung.index(splited[1])] = 1
if len(splited) < 3:
label_ko_jong[i][0] = 1
else:
label_ko_jong[i][ko_chset_jong.index(splited[2])] = 1
else :
label_ko_cho[i][len(ko_chset_cho)] = 1
label_ko_jung[i][len(ko_chset_jung)] = 1
label_ko_jong[i][len(ko_chset_jong)] = 1
# Is English?
if (target in en_chset):
label_en[i][en_chset.index(target)] = 1
else:
label_en[i][len(en_chset)] = 1
# Concatenate all labels
label = np.concatenate((label_ko_cho, label_ko_jung, label_ko_jong, label_en), axis=1)
print("label loaded")
return label
def set_five_type(conn):
query = 'SELECT hanja,is_naming_hanja,reading FROM naming_hanja'
ftype_select = conn.cursor()
for row in ftype_select.execute(query):
if row[1] == 0: # Can not use Korean name
continue
cs = split_syllable_char(row[2]) # 초성
# hlen = hangul_len(row[2])
f_type = get_five_type(cs[0]) # 음양'오행'
if f_type == -1:
continue
print(f_type, row[0])
update_hangul_strokes(f_type, row[0], conn)
def eval_tail(c, cand):
"""Analyze tail point which is the measure of too many splits.
They are normally length - 1
"""
point = 0
# Please please split the . and , symbols...!!
if cand.value == '.' or cand.value == ',':
point += 10
if hasattr(c, "prob") and c.value is not None and len(c.value) > 0 and len(cand.value) > 0:
if c.value == '\'':
point += 2
if cand.prob > 0.99 and check_syllable(c.value) and split_syllable_char(c.value)[1] == 'ㅏ':
point += 2
return cand.tail + 1 - point
def get_hangul_list():
hanguls = []
with open('unicode_hangul') as f:
for line in f:
hangul = line.split()
for h in hangul:
if check_syllable(h) is False:
continue
split_h = split_syllable_char(h)
try:
if BLOCK_LIST1[split_h[0]] == 1:
continue
except:
try:
if BLOCK_LIST3[split_h[2]] == 1:
continue
except:
pass
hanguls.append(h)
return hanguls
def generate(flag):
cho_sung = {
'ㄱ': 'k',
'ㄴ': 'u',
'ㄷ': 'i',
'ㄹ': 'm',
'ㅁ': 'y',
'ㅂ': 'o',
'ㅅ': 'n',
'ㅇ': 'j',
'ㅈ': 'l',
'ㅊ': 'hl',
'ㅋ': 'hk',
'ㅌ': 'hi',
'ㅍ': 'ho',
'ㅎ': 'h',
'ㅆ': 'nn',
'ㄸ': 'ii',
'ㄲ': 'kk',
'ㅉ': 'll',
}
jung_sung = {
'ㅏ': 'f',
'ㅑ': '.g',
'ㅓ': 'r',
'ㅕ': 't',
'ㅗ': 'v',
'ㅛ': '.v',
'ㅜ': 'b',
'ㅠ': '.b',
'ㅡ': 'g',
'ㅣ': 'd',
'ㅐ': 'fd',
'ㅔ': 'c',
'ㅟ': 'bd',
'ㅘ': 'vf',
'ㅚ': 'vd',
'ㅢ': 'gd',
'ㅝ': 'vr',
}
jong_sung = {
'ㄱ': 'x',
'ㄴ': 's',
'ㄷ': ';z',
'ㄹ': 'e',
'ㅁ': 'z',
'ㅂ': 'w',
'ㅅ': 'q',
'ㅇ': 'a',
'ㅈ': ';e',
'ㅊ': ';q',
'ㅋ': ';x',
'ㅌ': 'sz',
'ㅍ': ';w',
'ㅎ': ';s',
'ㅆ': ';',
'ㄶ': 'sa',
'ㅄ': 'wq',
}
key_list = list()
for c in flag:
try:
jamo = split_syllable_char(c)
if len(jamo) == 2:
k1, k2 = cho_sung[jamo[0]], jung_sung[jamo[1]]
key_list += [k1, k2]
else:
k1, k2, k3 = cho_sung[jamo[0]], jung_sung[jamo[1]], jong_sung[jamo[2]]
key_list += [k1, k2, k3]
except ValueError:
if c == ' ' or c == '\n':
key_list.append(c)
else:
assert False, '"{}" is invalid'.format(c)
keyset = list(set(''.join(key_list)))
sound_map = {k: i for i, k in enumerate(keyset)}
msg = '-'.join(['{:02x}'.format(sound_map[c]) for c in ''.join(key_list)])
with open('keystroke.txt', 'wt') as f:
f.write(msg)
def generate(flag):
cho_sung = {
'ㄱ': 'k',
'ㄴ': 'u',
'ㄷ': 'i',
'ㄹ': 'm',
'ㅁ': 'y',
'ㅂ': 'o',
'ㅅ': 'n',
'ㅇ': 'j',
'ㅈ': 'l',
'ㅊ': 'hl',
'ㅋ': 'hk',
'ㅌ': 'hi',
'ㅍ': 'ho',
'ㅎ': 'h',
'ㅆ': 'nn',
'ㄸ': 'ii',
'ㄲ': 'kk',
'ㅉ': 'll',
}
jung_sung = {
'ㅏ': 'f',
'ㅑ': '.g',
'ㅓ': 'r',
'ㅕ': 't',
'ㅗ': 'v',
'ㅛ': '.v',
'ㅜ': 'b',
'ㅠ': '.b',
'ㅡ': 'g',
'ㅣ': 'd',
'ㅐ': 'fd',
'ㅔ': 'c',
'ㅟ': 'bd',
'ㅘ': 'vf',
'ㅚ': 'vd',
'ㅢ': 'gd',
'ㅝ': 'vr',
}
jong_sung = {
'ㄱ': 'x',
'ㄴ': 's',
'ㄷ': ';z',
'ㄹ': 'e',
'ㅁ': 'z',
'ㅂ': 'w',
'ㅅ': 'q',
'ㅇ': 'a',
'ㅈ': ';e',
'ㅊ': ';q',
'ㅋ': ';x',
'ㅌ': 'sz',
'ㅍ': ';w',
'ㅎ': ';s',
'ㅆ': ';',
'ㄶ': 'sa',
'ㅄ': 'wq',
}
key_list = list()
for c in flag:
try:
jamo = split_syllable_char(c)
if len(jamo) == 2:
k1, k2 = cho_sung[jamo[0]], jung_sung[jamo[1]]
key_list += [k1, k2]
else:
k1, k2, k3 = cho_sung[jamo[0]], jung_sung[jamo[1]], jong_sung[
jamo[2]]
key_list += [k1, k2, k3]
except ValueError:
if c == ' ' or c == '\n':
key_list.append(c)
else:
assert False, '"{}" is invalid'.format(c)
keyset = list(set(''.join(key_list)))
sound_map = {k: i for i, k in enumerate(keyset)}
msg = '-'.join(['{:02x}'.format(sound_map[c]) for c in ''.join(key_list)])
with open('keystroke.txt', 'wt') as f:
f.write(msg)
def split(c):
from hangul_utils import split_syllable_char
jm = split_syllable_char(c)
#if jm[0] == 'ㅇ': return jm[1:]
return jm