def get_bcmi_corpus(line, left_symbol='（（', right_symbol='））'):
"""
转换原始文本为encoder-decoder列表
:param line: 王老师心（（性））格温和，态度和爱（（蔼）），教学有方，得到了许多人的好平（（评））。
:param left_symbol:
:param right_symbol:
:return: ["王老师心格温和，态度和爱，教学有方，得到了许多人的好平。" , "王老师性格温和，态度和蔼，教学有方，得到了许多人的好评。"]
"""
error_sentence = ''
correct_sentence = ''
details = []
if left_symbol not in line or right_symbol not in line:
return error_sentence, correct_sentence, details
left_ids = find_all_idx(line, left_symbol)
right_ids = find_all_idx(line, right_symbol)
if len(left_ids) != len(right_ids):
return error_sentence, correct_sentence, details
begin = 0
for left, right in zip(left_ids, right_ids):
correct_len = right - left - len(left_symbol)
correct_word = line[(left + len(left_symbol)):right]
error_sentence += line[begin:left]
correct_sentence += line[begin:(left - correct_len)] + correct_word
begin = right + len(right_symbol)
details.append(correct_word)
error_sentence += line[begin:]
correct_sentence += line[begin:]
n_details = []
for i in details:
idx = correct_sentence.find(i)
end_idx = idx + len(i)
error_item = error_sentence[idx:end_idx]
n_details.append([error_item, i, idx, end_idx])
return error_sentence, correct_sentence, n_details
def get_bcmi_corpus(line, left_symbol='（（', right_symbol='））'):
"""
转换原始文本为encoder-decoder列表
:param line: 王老师心（（性））格温和，态度和爱（（蔼）），教学有方，得到了许多人的好平（（评））。
:param left_symbol:
:param right_symbol:
:return: ["王老师心格温和，态度和爱，教学有方，得到了许多人的好平。" , "王老师性格温和，态度和蔼，教学有方，得到了许多人的好评。"]
"""
error_sentence, correct_sentence = '', ''
if left_symbol not in line or right_symbol not in line:
return error_sentence, correct_sentence
left_ids = find_all_idx(line, left_symbol)
right_ids = find_all_idx(line, right_symbol)
if len(left_ids) != len(right_ids):
return error_sentence, correct_sentence
begin = 0
for left, right in zip(left_ids, right_ids):
correct_len = right - left - len(left_symbol)
correct_word = line[(left + len(left_symbol)):right]
error_sentence += line[begin:left]
correct_sentence += line[begin:(left - correct_len)] + correct_word
begin = right + len(right_symbol)
error_sentence += line[begin:]
correct_sentence += line[begin:]
return error_sentence, correct_sentence
