hack/tools/scripts/llm-key-hunter/results/_cache/content/2239923138_trash-bin.json

{"repo": "2239923138/trash-bin", "files": {"ai.py": {"b19775b8427a676ef661a49a304e559872cb417e": {"t": 1785745406.4539847, "n": 1722, "h": "13ecc8f8aeb24993", "c": "from flask import Flask, render_template, request\r\nfrom openai import OpenAI\r\nimport os\r\nfrom pathlib import Path\r\n\r\napi_key = os.environ[\"DEEPSEEK_API_KEY\"]\r\nclient = OpenAI(base_url=\"https://api.deepseek.com\", api_key=api_key)\r\n\r\ndef retrieval(query):\r\n    context = \"\"\r\n    path_list = list(Path(\"my_knowledge\").glob(\"*.txt\"))\r\n    for path in path_list:\r\n        if path.stem in query:\r\n            context += path.read_text(encoding=\"utf-8\") \r\n            context += \"\\n\\n\"\r\n    return context\r\n\r\ndef augmented(query, context=\"\"):\r\n    if not context:\r\n        return f\"请回答下面问题:{query}，如果问题与计算机编程算法无关，请让我问重新问与算法相关的问题,如果有关，请回答我，要求：生成一段话即可，尽量简洁，最后给出一个相关网站的网页让我深入学习。\"\r\n    else:\r\n        prompt = f\"\"\"请根据上下文信息回答问题，如果上下文信息不足以回答问题，请直接说“根据上下文的信息我无法回答”\r\n        上下文:\r\n        {context}\r\n\r\n        问题：{query}\"\"\"\r\n        return prompt\r\n\r\ndef generation(prompt):\r\n    completion = client.chat.completions.create(\r\n        model=\"deepseek-chat\",\r\n        messages=[\r\n            {\r\n                \"role\": \"user\",\r\n                \"content\": prompt\r\n            }\r\n        ]\r\n    )\r\n    return completion.choices[0].message.content\r\n\r\napp = Flask(__name__)\r\n\r\ndef get_answer(question):\r\n    context = retrieval(question)\r\n    prompt = augmented(question, context)\r\n    return generation(prompt)\r\n\r\n@app.route('/', methods=['GET', 'POST'])\r\ndef index():\r\n    query = None\r\n    answer = None\r\n    if request.method == 'POST':\r\n        query = request.form.get('query')  # 获取前端输入的问题\r\n        if query:\r\n            answer = get_answer(query)  # 调用回答逻辑生成回应\r\n    # 渲染模板，把问题和回答传给前端页面\r\n    return render_template('ai_.html', query=query, answer=answer)\r\n\r\nif __name__ == '__main__':\r\n    app.run(debug=True)  # 调试模式，方便开发时看报错，生产环境需调整\r\n"}, "365c73daa4d1f4b1097aca646cfe6fa8a41cafea": {"t": 1785752886.893991, "n": 1722, "h": "13ecc8f8aeb24993", "c": "from flask import Flask, render_template, request\r\nfrom openai import OpenAI\r\nimport os\r\nfrom pathlib import Path\r\n\r\napi_key = os.environ[\"DEEPSEEK_API_KEY\"]\r\nclient = OpenAI(base_url=\"https://api.deepseek.com\", api_key=api_key)\r\n\r\ndef retrieval(query):\r\n    context = \"\"\r\n    path_list = list(Path(\"my_knowledge\").glob(\"*.txt\"))\r\n    for path in path_list:\r\n        if path.stem in query:\r\n            context += path.read_text(encoding=\"utf-8\") \r\n            context += \"\\n\\n\"\r\n    return context\r\n\r\ndef augmented(query, context=\"\"):\r\n    if not context:\r\n        return f\"请回答下面问题:{query}，如果问题与计算机编程算法无关，请让我问重新问与算法相关的问题,如果有关，请回答我，要求：生成一段话即可，尽量简洁，最后给出一个相关网站的网页让我深入学习。\"\r\n    else:\r\n        prompt = f\"\"\"请根据上下文信息回答问题，如果上下文信息不足以回答问题，请直接说“根据上下文的信息我无法回答”\r\n        上下文:\r\n        {context}\r\n\r\n        问题：{query}\"\"\"\r\n        return prompt\r\n\r\ndef generation(prompt):\r\n    completion = client.chat.completions.create(\r\n        model=\"deepseek-chat\",\r\n        messages=[\r\n            {\r\n                \"role\": \"user\",\r\n                \"content\": prompt\r\n            }\r\n        ]\r\n    )\r\n    return completion.choices[0].message.content\r\n\r\napp = Flask(__name__)\r\n\r\ndef get_answer(question):\r\n    context = retrieval(question)\r\n    prompt = augmented(question, context)\r\n    return generation(prompt)\r\n\r\n@app.route('/', methods=['GET', 'POST'])\r\ndef index():\r\n    query = None\r\n    answer = None\r\n    if request.method == 'POST':\r\n        query = request.form.get('query')  # 获取前端输入的问题\r\n        if query:\r\n            answer = get_answer(query)  # 调用回答逻辑生成回应\r\n    # 渲染模板，把问题和回答传给前端页面\r\n    return render_template('ai_.html', query=query, answer=answer)\r\n\r\nif __name__ == '__main__':\r\n    app.run(debug=True)  # 调试模式，方便开发时看报错，生产环境需调整\r\n"}, "620adfdc93ff072608b5d5390b0016349dbcbf08": {"t": 1785752888.2041917, "n": 1757, "h": "fc4cd0c737dd4d52", "c": "from flask import Flask, render_template, request\r\nfrom openai import OpenAI\r\nimport os\r\nfrom pathlib import Path\r\n\r\napi_key = os.environ[\"DEEPSEEK_API_KEY\"]\r\nclient = OpenAI(base_url=\"https://api.deepseek.com\", api_key=api_key)\r\n\r\ndef retrieval(query):\r\n    context = \"\"\r\n    path_list = list(Path(\"my_knowledge\").glob(\"*.txt\"))\r\n    for path in path_list:\r\n        if path.stem in query:\r\n            context += path.read_text(encoding=\"utf-8\") \r\n            context += \"\\n\\n\"\r\n    return context\r\n\r\ndef augmented(query, context=\"\"):\r\n    if not context:\r\n        return f\"请回答下面问题:{query}，如果问题与计算机编程算法无关，请让我问重新问与算法相关的问题,如果有关，请回答我，要求：生成一段话即可，尽量简洁，最后给出一个相关网站的网页让我深入学习。\"\r\n    else:\r\n        prompt = f\"\"\"请根据上下文信息回答问题，如果上下文信息不足以回答问题，请直接说“根据上下文的信息我无法回答”\r\n        上下文:\r\n        {context}\r\n\r\n        问题：{query}\"\"\"\r\n        return prompt\r\n\r\ndef generation(prompt):\r\n    completion = client.chat.completions.create(\r\n        model=\"deepseek-chat\",\r\n        messages=[\r\n            {\r\n                \"role\": \"user\",\r\n                \"content\": prompt\r\n            }\r\n        ]\r\n    )\r\n    return completion.choices[0].message.content\r\n\r\napp = Flask(__name__)\r\n\r\ndef get_answer(question):\r\n    context = retrieval(question)\r\n    prompt = augmented(question, context)\r\n    return generation(prompt)\r\n\r\n@app.route('/', methods=['GET', 'POST'])\r\ndef index():\r\n    query = None\r\n    answer = None\r\n    # 如果是 POST 请求，说明用户点了提交，处理表单数据\r\n    if request.method == 'POST':\r\n        query = request.form.get('query')  # 获取前端输入的问题\r\n        if query:\r\n            answer = get_answer(query)  # 调用回答逻辑生成回应\r\n    # 渲染模板，把问题和回答传给前端页面\r\n    return render_template('ai_.html', query=query, answer=answer)\r\n\r\nif __name__ == '__main__':\r\n    app.run(debug=True)  # 调试模式，方便开发时看报错，生产环境需调整\r\n"}, "807a1d97d3eb7f866be58f0c66de1e47b3f850fe": {"t": 1785752889.404161, "n": 1762, "h": "224ec5c173ca0a37", "c": "from flask import Flask, render_template, request\r\nfrom openai import OpenAI\r\nimport os\r\nfrom pathlib import Path\r\n\r\napi_key = os.environ[\"DEEPSEEK_API_KEY\"]\r\nclient = OpenAI(base_url=\"https://api.deepseek.com\", api_key=api_key)\r\n\r\ndef retrieval(query):\r\n    context = \"\"\r\n    path_list = list(Path(\"my_knowledge\").glob(\"*.txt\"))\r\n    for path in path_list:\r\n        if path.stem in query:\r\n            context += path.read_text(encoding=\"utf-8\")  # 读取\r\n            context += \"\\n\\n\"\r\n    return context\r\n\r\ndef augmented(query, context=\"\"):\r\n    if not context:\r\n        return f\"请回答下面问题:{query}，如果问题与计算机编程算法无关，请让我问重新问与算法相关的问题,如果有关，请回答我，要求：生成一段话即可，尽量简洁，最后给出一个相关网站的网页让我深入学习。\"\r\n    else:\r\n        prompt = f\"\"\"请根据上下文信息回答问题，如果上下文信息不足以回答问题，请直接说“根据上下文的信息我无法回答”\r\n        上下文:\r\n        {context}\r\n\r\n        问题：{query}\"\"\"\r\n        return prompt\r\n\r\ndef generation(prompt):\r\n    completion = client.chat.completions.create(\r\n        model=\"deepseek-chat\",\r\n        messages=[\r\n            {\r\n                \"role\": \"user\",\r\n                \"content\": prompt\r\n            }\r\n        ]\r\n    )\r\n    return completion.choices[0].message.content\r\n\r\napp = Flask(__name__)\r\n\r\ndef get_answer(question):\r\n    context = retrieval(question)\r\n    prompt = augmented(question, context)\r\n    return generation(prompt)\r\n\r\n@app.route('/', methods=['GET', 'POST'])\r\ndef index():\r\n    query = None\r\n    answer = None\r\n    # 如果是 POST 请求，说明用户点了提交，处理表单数据\r\n    if request.method == 'POST':\r\n        query = request.form.get('query')  # 获取前端输入的问题\r\n        if query:\r\n            answer = get_answer(query)  # 调用回答逻辑生成回应\r\n    # 渲染模板，把问题和回答传给前端页面\r\n    return render_template('ai_.html', query=query, answer=answer)\r\n\r\nif __name__ == '__main__':\r\n    app.run(debug=True)  # 调试模式，方便开发时看报错，生产环境需调整\r\n"}, "6ba95eaf4fe37d1eb36b90791a3902408a63a756": {"t": 1785752890.811533, "n": 1801, "h": "60eb8fd3b6e9ac27", "c": "from flask import Flask, render_template, request\r\nfrom openai import OpenAI\r\nimport os\r\nfrom pathlib import Path\r\n\r\n# 设置 API \r\napi_key = os.environ[\"DEEPSEEK_API_KEY\"]\r\nclient = OpenAI(base_url=\"https://api.deepseek.com\", api_key=api_key)\r\n\r\ndef retrieval(query):\r\n    context = \"\"\r\n    # 遍历所有文件\r\n    path_list = list(Path(\"my_knowledge\").glob(\"*.txt\"))\r\n    # 找到相关文件\r\n    for path in path_list:\r\n        if path.stem in query:\r\n            context += path.read_text(encoding=\"utf-8\")  # 读取\r\n            context += \"\\n\\n\"\r\n    return context\r\n\r\ndef augmented(query, context=\"\"):\r\n    if not context:\r\n        return f\"请回答下面问题:{query}，如果问题与计算机编程算法无关，请让我问重新问与算法相关的问题,如果有关，请回答我，要求：生成一段话即可，尽量简洁，最后给出一个相关网站的网页让我深入学习。\"\r\n    else:\r\n        prompt = f\"\"\"请根据上下文信息回答问题，如果上下文信息不足以回答问题，请直接说“根据上下文的信息我无法回答”\r\n        上下文:\r\n        {context}\r\n\r\n        问题：{query}\"\"\"\r\n        return prompt\r\n\r\ndef generation(prompt):\r\n    completion = client.chat.completions.create(\r\n        model=\"deepseek-chat\",\r\n        messages=[\r\n            {\r\n                \"role\": \"user\",\r\n                \"content\": prompt\r\n            }\r\n        ]\r\n    )\r\n    return completion.choices[0].message.content\r\n\r\napp = Flask(__name__)\r\n\r\ndef get_answer(question):\r\n    context = retrieval(question)\r\n    prompt = augmented(question, context)\r\n    return generation(prompt)\r\n\r\n@app.route('/', methods=['GET', 'POST'])\r\ndef index():\r\n    query = None\r\n    answer = None\r\n    # 如果是 POST 请求，说明用户点了提交，处理表单数据\r\n    if request.method == 'POST':\r\n        query = request.form.get('query')  # 获取前端输入的问题\r\n        if query:\r\n            answer = get_answer(query)  # 调用回答逻辑生成回应\r\n    # 渲染模板，把问题和回答传给前端页面\r\n    return render_template('ai_.html', query=query, answer=answer)\r\n\r\nif __name__ == '__main__':\r\n    app.run(debug=True)  # 调试模式，方便开发时看报错，生产环境需调整\r\n"}, "88ce8cb8fb957a49ad30a94c24d01a5fbecb6998": {"t": 1785752892.6669967, "n": 1804, "h": "418aee9386576bb1", "c": "from flask import Flask, render_template, request\r\nfrom openai import OpenAI\r\nimport os\r\nfrom pathlib import Path\r\n\r\n# 设置 API 客户端\r\napi_key = os.environ[\"DEEPSEEK_API_KEY\"]\r\nclient = OpenAI(base_url=\"https://api.deepseek.com\", api_key=api_key)\r\n\r\ndef retrieval(query):\r\n    context = \"\"\r\n    # 遍历所有文件\r\n    path_list = list(Path(\"my_knowledge\").glob(\"*.txt\"))\r\n    # 找到相关文件\r\n    for path in path_list:\r\n        if path.stem in query:\r\n            context += path.read_text(encoding=\"utf-8\")  # 读取\r\n            context += \"\\n\\n\"\r\n    return context\r\n\r\ndef augmented(query, context=\"\"):\r\n    if not context:\r\n        return f\"请回答下面问题:{query}，如果问题与计算机编程算法无关，请让我问重新问与算法相关的问题,如果有关，请回答我，要求：生成一段话即可，尽量简洁，最后给出一个相关网站的网页让我深入学习。\"\r\n    else:\r\n        prompt = f\"\"\"请根据上下文信息回答问题，如果上下文信息不足以回答问题，请直接说“根据上下文的信息我无法回答”\r\n        上下文:\r\n        {context}\r\n\r\n        问题：{query}\"\"\"\r\n        return prompt\r\n\r\ndef generation(prompt):\r\n    completion = client.chat.completions.create(\r\n        model=\"deepseek-chat\",\r\n        messages=[\r\n            {\r\n                \"role\": \"user\",\r\n                \"content\": prompt\r\n            }\r\n        ]\r\n    )\r\n    return completion.choices[0].message.content\r\n\r\napp = Flask(__name__)\r\n\r\ndef get_answer(question):\r\n    context = retrieval(question)\r\n    prompt = augmented(question, context)\r\n    return generation(prompt)\r\n\r\n@app.route('/', methods=['GET', 'POST'])\r\ndef index():\r\n    query = None\r\n    answer = None\r\n    # 如果是 POST 请求，说明用户点了提交，处理表单数据\r\n    if request.method == 'POST':\r\n        query = request.form.get('query')  # 获取前端输入的问题\r\n        if query:\r\n            answer = get_answer(query)  # 调用回答逻辑生成回应\r\n    # 渲染模板，把问题和回答传给前端页面\r\n    return render_template('ai_.html', query=query, answer=answer)\r\n\r\nif __name__ == '__main__':\r\n    app.run(debug=True)  # 调试模式，方便开发时看报错，生产环境需调整\r\n"}, "608632e0534c1733f11920299f6ea6322a2bd2fa": {"t": 1785752894.978266, "n": 1876, "h": "2894bbe84ef5fc58", "c": "from flask import Flask, render_template, request\r\nfrom openai import OpenAI\r\nimport os\r\nfrom pathlib import Path\r\n\r\n# 设置 API 客户端\r\napi_key = os.environ[\"DEEPSEEK_API_KEY\"]\r\nclient = OpenAI(base_url=\"https://api.deepseek.com\", api_key=api_key)\r\n\r\n# 步骤1：实现一个简单的“关键词匹配”检索器\r\ndef retrieval(query):\r\n    context = \"\"\r\n    # 遍历所有文件\r\n    path_list = list(Path(\"my_knowledge\").glob(\"*.txt\"))\r\n    # 找到相关文件\r\n    for path in path_list:\r\n        if path.stem in query:\r\n            context += path.read_text(encoding=\"utf-8\")  # 读取\r\n            context += \"\\n\\n\"\r\n    return context\r\n\r\n# 步骤2：增强Query\r\ndef augmented(query, context=\"\"):\r\n    if not context:\r\n        return f\"请回答下面问题:{query}，如果问题与计算机编程算法无关，请让我问重新问与算法相关的问题,如果有关，请回答我，要求：生成一段话即可，尽量简洁，最后给出一个相关网站的网页让我深入学习。\"\r\n    else:\r\n        prompt = f\"\"\"请根据上下文信息回答问题，如果上下文信息不足以回答问题，请直接说“根据上下文的信息我无法回答”\r\n        上下文:\r\n        {context}\r\n\r\n        问题：{query}\"\"\"\r\n        return prompt\r\n\r\n# 步骤3：生成回答\r\ndef generation(prompt):\r\n    completion = client.chat.completions.create(\r\n        model=\"deepseek-chat\",\r\n        messages=[\r\n            {\r\n                \"role\": \"user\",\r\n                \"content\": prompt\r\n            }\r\n        ]\r\n    )\r\n    return completion.choices[0].message.content\r\n\r\napp = Flask(__name__)\r\n\r\n# 使用 RAG+LLM 逻辑的回答函数\r\ndef get_answer(question):\r\n    context = retrieval(question)\r\n    prompt = augmented(question, context)\r\n    return generation(prompt)\r\n\r\n@app.route('/', methods=['GET', 'POST'])\r\ndef index():\r\n    query = None\r\n    answer = None\r\n    # 如果是 POST 请求，说明用户点了提交，处理表单数据\r\n    if request.method == 'POST':\r\n        query = request.form.get('query')  # 获取前端输入的问题\r\n        if query:\r\n            answer = get_answer(query)  # 调用回答逻辑生成回应\r\n    # 渲染模板，把问题和回答传给前端页面\r\n    return render_template('ai_.html', query=query, answer=answer)\r\n\r\nif __name__ == '__main__':\r\n    app.run(debug=True)  # 调试模式，方便开发时看报错，生产环境需调整"}}}}