++

api_deepseek.py

@@ -8,8 +8,9 @@ deepseek-chat
 
 
 class DeepSeek(object):
-    def call_deepseek(self, message):
+    def call_deepseek(self, text, prompt_words):
         try:
+            message = text + '\n\n' + prompt_words
             print('call deepseek')
             client = OpenAI(
                 api_key="sk-20afb9967a124b63a6f67dcb69f17d74",
@@ -19,7 +20,7 @@ class DeepSeek(object):
             completion = client.chat.completions.create(
                 model="deepseek-chat",
                 messages=[
-                    {"role": "system", "content": "You are a helpful assistant"},
+                    {"role": "system", "content": "你是一个数据整理助手"},
                     {"role": "user", "content": f"{message}"},
                 ],
                 stream=False,

config.json

@@ -11,7 +11,7 @@
       "https://foresightnews.pro/news",
       "https://www.web3sj.com/news/"
     ],
-    "prompt_words": "给你一个或多个网页的源代码, 里面是未清洗的网页源代码,你可以无视网页源代码的部分,关注内容就行,重复的话就不用说了,帮我总结一下这些网站的内容, 请用中文回答"
+    "prompt_words": "给你一个或多个网页的源代码, 里面是未清洗的网页源代码,你可以无视网页源代码的部分,关注内容就行,重复的话就不用说了,中间不需要有空行 请用中文回答"
   },
   "A": {
     "target_url_list": [

main.py

@@ -8,20 +8,21 @@ from bs4 import BeautifulSoup
 from api_ollama import *
 from api_kimi import *
 from api_deepseek import *
-from send_to_email import *
+from send_matrix import *
 
 key = 'web3'
 
 
 class AINEWS:
-    def save_to_txt(self, text):
+    def save_to_txt(self, url_to_text):
         current_file_path = os.path.dirname(__file__)
         save_file_path = os.path.join(current_file_path, 'save_txt')
         if not os.path.exists(save_file_path):
             os.makedirs(save_file_path)
         file = os.path.join(save_file_path, str(int(time.time())) + '.txt')
         with open(file, 'w', encoding='utf-8') as file:
-            file.write(text)
+            file.write(str(url_to_text))
+        print(f'txt文件已保存, 路径为: {file}')
 
     def load_config(self, key):
         config = {}
@@ -49,7 +50,6 @@ class AINEWS:
         await page.evaluate("window.scrollTo({top: 0, behavior: 'smooth'})")
 
     async def get_htmls(self, urls):
-        htmls = []
         async with async_playwright() as p:
             # 启动浏览器
             browser = await p.chromium.launch(headless=True)
@@ -62,12 +62,6 @@ class AINEWS:
                     # 在上下文中打开新页面
                     page = await context.new_page()
 
-                    # 监听导航事件
-                    def handle_navigation(frame):
-                        print(f"导航发生: {frame.url}")
-
-                    page.on('framenavigated', handle_navigation)
-
                     # 导航到指定网址
                     await page.goto(url, wait_until='networkidle')  # 等待网络空闲
 
@@ -78,52 +72,63 @@ class AINEWS:
                     html = await page.content()
                     # 关闭页面
                     await page.close()
-                    return html
+
+                    # 使用 BeautifulSoup 格式化 HTML 内容
+                    soup = BeautifulSoup(html, 'html.parser')
+                    formatted_html = soup.get_text()
+                    cleaned_text = re.sub(r'[\n\t\r]+', ' ', formatted_html)
+                    cleaned_text = re.sub(r'\s+', ' ', cleaned_text).strip()
+
+                    return url, cleaned_text
                 except Exception as e:
                     print(f"Error fetching {url}: {e}")
-                    return ""
+                    return url, ""
 
             # 使用 asyncio.gather 同时获取所有网站的 HTML
             tasks = [get_html(url) for url in urls]
-            htmls_list = await asyncio.gather(*tasks)
+            results = await asyncio.gather(*tasks)
 
-            # 使用 BeautifulSoup 格式化每个 HTML 内容
-            formatted_htmls = []
-            for html in htmls_list:
-                soup = BeautifulSoup(html, 'html.parser')
-                formatted_html = soup.get_text()
-                cleaned_text = re.sub(r'[\n\t\r]+', ' ', formatted_html)
-                cleaned_text = re.sub(r'\s+', ' ', cleaned_text).strip()
-                formatted_htmls.append(cleaned_text)
-
-            # 将所有格式化后的 HTML 内容合并到一个字符串中
-            text = "\n".join(formatted_htmls)
+            # 将结果存储在字典中
+            url_to_text = {url: text for url, text in results}
 
             # 关闭上下文和浏览器
             await context.close()
             await browser.close()
 
-            return text
+            return url_to_text
 
     def main(self, target_url_list, prompt_words):
-        text = asyncio.run(self.get_htmls(target_url_list))
+        url_to_text = asyncio.run(self.get_htmls(target_url_list))
+
+        self.save_to_txt(url_to_text)
 
-        self.save_to_txt(text)
+        prepare_to_send = ''
 
-        prompt_words += text
+        prompt_words += '帮我总结一下内容, 并且分析一下今天数字币为什么会跌'
 
-        C = ChatBot('http://127.0.0.1:11434', prompt_words, 'llava:13b')
-        response_context = C.start_chat()
-        print(response_context)
+        C = ChatBot('http://127.0.0.1:11434', k, prompt_words, 'llava:13b')
+        for k, v in url_to_text.items():
+            response_context = C.call_deepseek(v, prompt_words)
+            message = f'{k}\n{response_context}\n'
+            prepare_to_send += message
+            print(message)
 
         # K = KIMI()
         # response_context = K.call_kimi(prompt_words)
         # print(response_context)
 
         # D = DeepSeek()
-        # response_context = D.call_deepseek(prompt_words)
-        # print(response_context)
+        # for k, v in url_to_text.items():
+        #     response_context = D.call_deepseek(v, prompt_words)
+
+        #     # 保存每一个字符串准备发送信息
+        #     message = f'{k}\n{response_context}\n'
+        #     prepare_to_send += message
+        #     print(message)
 
+        # 发送消息
+        bot = MatrixBot('message-bot', 'aaaAAA111!!!')
+        bot.send_message(message)
 
 if __name__ == "__main__":
     ainews = AINEWS()

send_matrix.py

@@ -11,7 +11,7 @@ class MatrixBot:
         self.password = password
         self.client = MatrixClient("https://matrix.erhe.top")
         self.token = self.login()
-        self.to = "!ddrrTpQmepfgivMxeW:chat.abeginner.cn"
+        self.to = "!CgWvWEnLbKYvhXLvil:chat.abeginner.cn"
 
     def login(self):
         self.token = self.client.login(username=self.user, password=self.password)
@@ -31,10 +31,10 @@ class MatrixBot:
             print("Bot is not logged in. Please login first.")
 
 # if __name__ == '__main__':
-# 测试调用
-# user = "bot1"
-# pw = "aaaAAA111!!!"
-# message = "123987456"
-#
-# bot = MatrixBot(user, pw)
-# bot.send_message(message)
+#     # 测试调用
+#     user = "message-bot"
+#     pw = "aaaAAA111!!!"
+#     message = "123987456"
+
+#     bot = MatrixBot(user, pw)
+#     bot.send_message(message)