remove pinyin dict

pull/637/head
Hui Zhang 4 years ago
parent 8983880c75
commit 3a5cf7ac0e

@ -3,59 +3,7 @@
文本转拼音的模块依赖python-pinyinjiebaphrase-pinyin-data模块
"""
import re
from pypinyin import lazy_pinyin, Style, load_phrases_dict, load_phrases_dict
# def parse_pinyin_txt(inpath):
# # U+4E2D: zhōng,zhòng # 中
# outs = []
# with open(inpath, encoding="utf8") as fin:
# for line in tqdm(fin, desc='load pinyin', ncols=80, mininterval=1):
# if line.startswith("#"):
# continue
# res = _ziyin_re.search(line)
# if res:
# zi = res.group(3).strip()
# if len(zi) == 1:
# outs.append([zi, res.group(2).strip().split(",")])
# else:
# print(line)
# elif line.strip():
# print(line)
# return {ord(z): ','.join(p) for z, p in outs}
# def parse_phrase_txt(inpath):
# # 一一对应: yī yī duì yìng
# outs = []
# with open(inpath, encoding="utf8") as fin:
# for line in tqdm(fin, desc='load phrase', ncols=80, mininterval=1):
# if line.startswith("#"):
# continue
# parts = line.split(":")
# zs = parts[0].strip()
# ps = parts[1].strip().split()
# if len(parts) == 2 and len(zs) == len(ps) and len(zs) >= 2:
# outs.append([zs, ps])
# elif line.strip():
# print(line)
# return {zs: [[p] for p in ps] for zs, ps in outs}
# def initialize():
# # 导入数据
# inpath = Path(__file__).absolute().parent.joinpath('phrase_pinyin.txt.py')
# _phrases_dict = parse_phrase_txt(inpath)
# load_phrases_dict(_phrases_dict) # big:398815 small:36776
# inpath = Path(__file__).absolute().parent.joinpath('single_pinyin.txt.py')
# _pinyin_dict = parse_pinyin_txt(inpath)
# load_single_dict(_pinyin_dict) # 41451
# jieba.initialize()
# # for word, _ in tqdm(_phrases_dict.items(), desc='jieba add word', ncols=80, mininterval=1):
# # jieba.add_word(word)
from pypinyin import lazy_pinyin, Style
# 兼容0.1.0之前的版本。
# 音调5为轻声

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff
Loading…
Cancel
Save