PyGeoModel 1.0.0__py3-none-any.whl

scripts.py

@@ -0,0 +1,128 @@
+import requests
+import tenacity
+from openai import OpenAI
+
+
+class AcademicQueryService:
+    def __init__(self, api_key: str = "sk-4bp5a1DcdLSHCiw1401270055f47424b9eA58cAd587266A3",
+                 modelingContext: str = "",
+                 base_url: str = "https://aihubmix.com/v1"):
+        self.client = OpenAI(api_key=api_key, base_url=base_url)
+        self.modelingContext = modelingContext
+
+    async def get_academic_question_answer(self, query: str) -> dict:
+        """
+        获取学术问题的答案（这是对外暴露的主要接口）
+        """
+        paper_list = await self._get_academic_question_answer_list(query)
+        question_answer = await self._get_openai_summary(query, paper_list)
+        return {
+            "question": query,
+            "answer": question_answer,
+            "paperList": paper_list
+        }
+
+    @tenacity.retry(wait=tenacity.wait_fixed(1), stop=tenacity.stop_after_attempt(5))
+    async def _get_academic_question_answer_list(self, query: str, page: int = 1, size: int = 10) -> list:
+        """内部方法：获取论文列表 - 使用新的两步调用流程"""
+        try:
+            # 第一步：创建搜索线程，获取 search_id
+            search_url = "https://consensus.app/api/threads/"
+            search_payload = {
+                "user_message": query,
+                "is_pro_enabled": False,
+                "is_incognito": False,
+                "size": size,
+                "filters": {},
+                "search_mode": "SUMMARY"
+            }
+            search_headers = {
+                'Referer': 'https://consensus.app',
+                'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36',
+                'Content-Type': 'application/json',
+                'Accept': 'application/json, text/plain, */*'
+            }
+
+            search_response = requests.post(
+                search_url, json=search_payload, headers=search_headers, timeout=15)
+
+            if search_response.status_code not in [200, 201]:
+                # 降级到直接查询方式
+                return await self._fallback_direct_query(query, page, size)
+
+            search_data = search_response.json()
+            search_id = None
+
+            # 尝试从不同位置获取search_id
+            if 'search_id' in search_data:
+                search_id = search_data['search_id']
+            elif 'interactions' in search_data and len(search_data['interactions']) > 0:
+                first_interaction = search_data['interactions'][0]
+                search_id = first_interaction.get('search_id')
+
+            if not search_id:
+                return await self._fallback_direct_query(query, page, size)
+
+            # 第二步：使用 search_id 获取论文结果
+            papers_url = f"https://consensus.app/api/paper_search/{search_id}/?page={page-1}&size={size}"
+            papers_headers = {
+                'Referer': 'https://consensus.app',
+                'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36',
+                'Accept': 'application/json, text/plain, */*'
+            }
+
+            papers_response = requests.get(
+                papers_url, headers=papers_headers, timeout=15)
+
+            if papers_response.status_code != 200:
+                return await self._fallback_direct_query(query, page, size)
+
+            papers_data = papers_response.json()
+            papers = papers_data.get("papers", [])
+
+            return papers[:size]
+
+        except Exception as e:
+            # 降级到直接查询方式
+            return await self._fallback_direct_query(query, page, size)
+
+    async def _fallback_direct_query(self, query: str, page: int = 1, size: int = 10) -> list:
+        """降级方案：直接查询论文（原来的方法）"""
+        try:
+            url = "https://consensus.app/api/paper_search/?query=" + \
+                query + "&page=" + str(page) + "&size=" + str(size)
+            payload = {}
+            headers = {
+                'Referer': 'https://consensus.app',
+                'User-Agent': 'Apifox/1.0.0 (https://apifox.com)'
+            }
+            response = requests.request(
+                "GET", url, headers=headers, data=payload).json()
+
+            # return top 10 result in claims
+            top10Claims = response["papers"][:10]
+            return top10Claims
+        except Exception as e:
+            return []
+
+    async def _get_openai_summary(self, query: str, paper_list: list) -> str:
+        """内部方法：获取 OpenAI 总结"""
+        paperContext = ""
+        for paper in paper_list:
+            # 组织成，这是一篇名为<title>的论文，发表在<journal>上，相关的内容为<display_text>。
+            paperContext += "This is a paper named: " + \
+                paper["title"] + " published in: " + paper["journal"] + \
+                ". The related content is: " + paper["display_text"] + "."
+
+        completion = self.client.chat.completions.create(
+            model="gpt-4.1-nano",
+            messages=[
+                {"role": "system", "content": "You are a helpful AI researcher assistant. User is now conducting scientific geoscience research. The modeling history is: " + self.modelingContext +
+                    ". He/she is now asking a question about " + query + ".  And here are some information related to this question: " + paperContext + ". And your task is to summarize the information and answer the question. Please answer the question in English."},
+                {
+                    "role": "user",
+                    "content": "Besides your summary, you should not reply any other information. Below is your answer to the question based on the information provided: "
+                }
+            ]
+        )
+        return completion.choices[0].message.content