` tags. If the text cannot be translated, return the original text as is. Do not add any additional text to the translation. Do not translate person's name. The HTML webpage to be translated is:
57 |
58 | ### 待补充内容：
59 | - [ ] token 的使用情况
60 | - [ ] 时间花费情况
61 | - [ ] 工作原理
62 |
63 |
--------------------------------------------------------------------------------
/BilingualHTML.py:
--------------------------------------------------------------------------------
1 | from bs4 import BeautifulSoup as bs
2 | import openai
3 | import time
4 | import json
5 | import logging
6 | import os
7 | from pathlib import Path
8 |
9 | logging.basicConfig(
10 | level=logging.INFO,
11 | filename="./log.txt",
12 | filemode="a",
13 | format="%(asctime)s - %(levelname)s - %(funcName)s: %(message)s",
14 | )
15 |
16 |
17 | def contains_only_code_tag(tag):
18 | """
19 | 检查标记是否只包含一个代码标记。
20 | """
21 | code_tag = tag.find('code')
22 | if code_tag is None:
23 | return False
24 | code_text = code_tag.text.replace('\n', '').replace(' ', '')
25 | p_text = tag.text.replace('\n', '').replace(' ', '')
26 | if code_text == p_text:
27 | return True
28 |
29 |
30 | def is_single_character_tag(tag):
31 | """
32 | 判断一个标签内的文本是否只有一个字符
33 | 只有一个字符的话，就不需要翻译了
34 | """
35 | text = tag.text.replace('\n', '').replace(' ', '')
36 | return len(text) == 1
37 |
38 |
39 | def is_jump_translate(tag):
40 | """
41 | 判断是否可以跳过当前标签，不用翻译。
42 | """
43 | return contains_only_code_tag(tag) or is_single_character_tag(tag)
44 |
45 |
46 | def get_translation(prompt, code):
47 | """
48 | 将 HTML 网页翻译为简体中文，返回翻译后的 HTML 代码，包裹在 标签中的文本不会被翻译。
49 | 如果无法翻译，则返回原始的 HTML 代码。
50 | """
51 | completion = openai.ChatCompletion.create(
52 | model="gpt-3.5-turbo",
53 | messages=[
54 | {"role": "user", "content": f"{prompt} {code}"}
55 | ]
56 | )
57 | return completion.choices[0].message.content
58 |
59 |
60 | def translate_tag(prompt, code):
61 | """
62 | 将给定的 HTML 网页翻译为简体中文，如果翻译失败则进行重试。
63 | """
64 | max_attempts = 5
65 | for i in range(max_attempts):
66 | try:
67 | t_text = get_translation(prompt, code)
68 | time.sleep(3)
69 | print(t_text)
70 | logging.info(t_text)
71 | return t_text
72 | except Exception as e:
73 | sleep_time = 60
74 | print(e)
75 | logging.error(e)
76 | print(f"请求失败，将等待 {sleep_time} 秒后重试")
77 | logging.info(f"请求失败，将等待 {sleep_time} 秒后重试")
78 | time.sleep(sleep_time)
79 | print(f"开始重试第 {i + 1}/{max_attempts}")
80 | logging.info(f"开始重试第 {i + 1}/{max_attempts}")
81 | print(f"请求失败，重试次数{max_attempts}/{max_attempts}，放弃请求")
82 | logging.error(f"请求失败，重试次数{max_attempts}/{max_attempts}，放弃请求")
83 |
84 |
85 | def read_api_key(path):
86 | with open(path, 'r') as f:
87 | api_key = f.read().strip()
88 | return api_key
89 |
90 |
91 | def read_chatGPT_prompt(path):
92 | with open(path, 'r') as f:
93 | prompt = f.read().strip()
94 | return prompt
95 |
96 |
97 | def read_page(path):
98 | """
99 | 打开 HTML 网页，返回 BeautifulSoup 对象。
100 | """
101 | with open(path, 'r') as f:
102 | soup = bs(f, "html.parser")
103 | return soup
104 |
105 |
106 | def read_json(path):
107 | """
108 | 打开 JSON 文件，返回 JSON 对象。
109 | """
110 | with open(path, 'r') as f:
111 | json_obj = json.load(f)
112 | return json_obj
113 |
114 |
115 | def write_json(path, code, mode='w'):
116 | """
117 | 将 JSON 对象写入 JSON 文件。
118 | """
119 | with open(path, mode) as f:
120 | json.dump(code, f, ensure_ascii=False)
121 |
122 |
123 | def write_page(path, soup):
124 | with open(path, 'w') as f:
125 | f.write(soup.prettify())
126 |
127 |
128 | def resume_translate():
129 | """
130 | 恢复之前的翻译进度
131 | """
132 | try:
133 | translated_texts = read_json('translated.json')
134 | start_index = len(translated_texts)
135 | print(f"从索引 {start_index + 1} 处继续翻译")
136 | logging.info(f"从索引 {start_index + 1} 处继续翻译")
137 | except:
138 | print("没有找到之前的翻译结果，从头开始翻译")
139 | logging.info("没有找到之前的翻译结果，从头开始翻译")
140 | start_index = 0
141 | translated_texts = []
142 | return translated_texts, start_index
143 |
144 |
145 | def translate_page(start_time, page_num, page_count, page, prompt):
146 | """
147 | 读取 HTML 页面并翻译其中的所有 标签内容为中文。
148 | 如果翻译结果已经存在，则从上次翻译的位置继续翻译。
149 | 翻译结果会保存在 'translated.json' 文件中。
150 | """
151 | soup = read_page(page)
152 | p_list = soup.find_all('p')
153 | count = len(p_list)
154 | translated_texts, start_index = resume_translate()
155 | page_start_time = time.time()
156 | for i, p in enumerate(p_list[start_index:]):
157 | print("↓ " * 10 + "开始翻译 " + "↓ " * 10)
158 | logging.info("↓ " * 10 + "开始翻译 " + "↓ " * 10)
159 | p_start_time = time.time()
160 | p_code = p.prettify()
161 | print(p_code)
162 | logging.info(p_code)
163 | if p_code:
164 | if is_jump_translate(p):
165 | translated_texts.append(p_code)
166 | else:
167 | translated_texts.append(translate_tag(prompt, p_code))
168 | write_json('translated.json', translated_texts)
169 | p_end_time = time.time()
170 | print(f"已经翻译 {i + start_index + 1}/{count}", end='，')
171 | logging.info(f"已经翻译 {i + start_index + 1}/{count}")
172 | p_elapsed_time = int(p_end_time - p_start_time)
173 | print(f"本段耗时 {p_elapsed_time} 秒", end='，')
174 | logging.info(f"本段耗时 {p_elapsed_time} 秒")
175 | every_p_time = int((p_end_time - page_start_time) / (i + start_index + 1))
176 | print(f"平均每段耗时 {every_p_time} 秒")
177 | logging.info(f"平均每段耗时 {every_p_time} 秒")
178 | page_elapsed_time = int((p_end_time - page_start_time) / 60)
179 | print(f"本页耗时 {page_elapsed_time} 分钟", end='，')
180 | logging.info(f"本页耗时 {page_elapsed_time} 分钟")
181 | remaining_time = int(((count - i - start_index - 1) * every_p_time) / 60)
182 | print(f"本页预计剩余时间 {remaining_time} 分钟")
183 | logging.info(f"本页预计剩余时间 {remaining_time} 分钟")
184 | elapsed_time = int((p_end_time - start_time) / 60)
185 | print(f"总耗时 {elapsed_time} 分钟", end='，')
186 | logging.info(f"总耗时 {elapsed_time} 分钟")
187 | print(f"正在翻译第 {page_num + 1}/{page_count} 页")
188 | logging.info(f"正在翻译第 {page_num + 1}/{page_count} 页")
189 | print("↑ " * 10 + "翻译完成 " + "↑ " * 10)
190 | logging.info("↑ " * 10 + "翻译完成 " + "↑ " * 10)
191 | return count
192 |
193 |
194 | def save_translated_page(page, json):
195 | """
196 | 在 HTML 文件中添加翻译后的 p 标签
197 | """
198 | translated_texts = read_json(json)
199 | soup = read_page(page)
200 | p_list = soup.find_all('p')
201 | for i, p in enumerate(p_list):
202 | text = p.prettify()
203 | if text:
204 | translated_p = bs(translated_texts[i], 'html.parser')
205 | p.insert_after(translated_p)
206 | page_cn = page.parent / (page.stem + '_cn' + page.suffix)
207 | write_page(page_cn, soup)
208 |
209 |
210 | def get_translated_page(path):
211 | path = Path(path)
212 | return list(path.glob('**/*.html'))
213 |
214 |
215 | def resume_translate_page():
216 | """
217 | 恢复之前的翻译进度
218 | """
219 | try:
220 | translated_index = read_json('index.json')
221 | start_index = len(translated_index)
222 | print(f"从索引 {start_index + 1} 处继续翻译")
223 | logging.info(f"从索引 {start_index + 1} 处继续翻译")
224 | except:
225 | print("没有找到之前的翻译索引，从第一个文件开始翻译")
226 | logging.info("没有找到之前的翻译索引，从第一个文件开始翻译")
227 | start_index = 0
228 | return start_index
229 |
230 |
231 | openai.api_key = read_api_key('api_key.txt')
232 | prompt = read_chatGPT_prompt('chatGPT_prompt.txt')
233 | path = Path('translatable')
234 | pages = get_translated_page('translatable')
235 | pages_count = len(list(pages))
236 | start_time = time.time()
237 | start_index = resume_translate_page()
238 | for i, page in enumerate(pages[start_index:]):
239 | count = translate_page(start_time, i, pages_count, page, prompt)
240 | translated_texts = read_json('translated.json')
241 | if count == len(translated_texts):
242 | save_translated_page(page, 'translated.json')
243 | print(f"{page.stem} 翻译完成, 进度{i + 1}/{pages_count}")
244 | logging.info(f"{page.stem} 翻译完成, 进度{i + 1}/{pages_count}")
245 | os.remove('translated.json')
246 | write_json('index.json', str(page), 'a')
247 | os.remove('index.json')
248 | print("全部文件翻译完成")
249 |
--------------------------------------------------------------------------------