From 5503acc7a0ea1a983c9528cfe4f3355e0536f905 Mon Sep 17 00:00:00 2001
From: lonpatovaadelina <lonpatovaadelina@mail.ru>
Date: Tue, 26 May 2026 22:36:36 +0000
Subject: [PATCH] =?UTF-8?q?=D0=90=D0=B3=D0=B5=D0=BD=D1=82=20=D1=81=20RAG-?=
 =?UTF-8?q?=D0=BF=D0=B0=D0=BC=D1=8F=D1=82=D1=8C=D1=8E:=20solution.py?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .../solution.py                               | 155 ++++++++++++++++++
 1 file changed, 155 insertions(+)
 create mode 100644 solutions/6a02e23da6fe2e4ac16acf65_Агент_с_RAG-памятью/solution.py

diff --git a/solutions/6a02e23da6fe2e4ac16acf65_Агент_с_RAG-памятью/solution.py b/solutions/6a02e23da6fe2e4ac16acf65_Агент_с_RAG-памятью/solution.py
new file mode 100644
index 0000000..50c80bf
--- /dev/null
+++ b/solutions/6a02e23da6fe2e4ac16acf65_Агент_с_RAG-памятью/solution.py
@@ -0,0 +1,155 @@
+import os
+import logging
+from typing import List, Optional
+
+from fastapi import FastAPI, HTTPException
+from fastapi.responses import JSONResponse
+from pydantic import BaseModel
+from langchain.embeddings import OpenAIEmbeddings
+from langchain.vectorstores import Qdrant
+from langchain.llms import OpenAI
+from langchain.chains import RetrievalQA
+from langchain.prompts import PromptTemplate
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from rich import print as rprint
+from qdrant_client import QdrantClient
+from qdrant_client.http import models as qdrant_models
+
+# =======================
+# Конфигурация и логирование
+# =======================
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s: %(message)s",
+)
+logger = logging.getLogger("rag_agent")
+
+# =======================
+# Параметры
+# =======================
+QDRANT_HOST = os.getenv("QDRANT_HOST", "localhost")
+QDRANT_PORT = int(os.getenv("QDRANT_PORT", "6333"))
+QDRANT_COLLECTION = os.getenv("QDRANT_COLLECTION", "rag_collection")
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+if not OPENAI_API_KEY:
+    raise RuntimeError("OPENAI_API_KEY не задана в переменных окружения")
+
+# =======================
+# Инициализация Qdrant
+# =======================
+client = QdrantClient(host=QDRANT_HOST, port=QDRANT_PORT)
+try:
+    client.retreive_collection(name=QDRANT_COLLECTION)
+except Exception:
+    client.retreive_collection(name=QDRANT_COLLECTION, create=True)
+
+# =======================
+# Векторное хранилище
+# =======================
+embeddings = OpenAIEmbeddings(openai_api_key=OPENAI_API_KEY)
+vectorstore = Qdrant(
+    client=client,
+    collection_name=QDRANT_COLLECTION,
+    embeddings=embeddings,
+)
+
+# =======================
+# Модель и цепочка
+# =======================
+llm = OpenAI(openai_api_key=OPENAI_API_KEY, temperature=0.2)
+prompt_template = PromptTemplate(
+    input_variables=["context", "question"],
+    template=(
+        "Ты эксперт по теме. Используй следующий контекст для ответа на вопрос.\n"
+        "Контекст:\n{context}\n\nВопрос: {question}\nОтвет:"
+    ),
+)
+qa_chain = RetrievalQA.from_chain_type(
+    llm=llm,
+    chain_type="stuff",
+    retriever=vectorstore.as_retriever(search_kwargs={"k": 4}),
+    return_source_documents=True,
+    chain_type_kwargs={"prompt": prompt_template},
+)
+
+# =======================
+# FastAPI приложение
+# =======================
+app = FastAPI(title="RAG Agent")
+
+class DocumentIn(BaseModel):
+    text: str
+    chunk_size: Optional[int] = 1000
+    chunk_overlap: Optional[int] = 200
+
+class QueryIn(BaseModel):
+    question: str
+
+@app.post("/ingest")
+async def ingest(doc: DocumentIn):
+    """
+    Разбивает документ на чанки и сохраняет в Qdrant.
+    """
+    try:
+        splitter = RecursiveCharacterTextSplitter(
+            chunk_size=doc.chunk_size,
+            chunk_overlap=doc.chunk_overlap,
+            separators=["\n\n", "\n", " ", ""],
+        )
+        chunks = splitter.split_text(doc.text)
+        if not chunks:
+            raise ValueError("Документ не содержит текста для разбивки")
+        ids = [f"{hash(chunk)}" for chunk in chunks]
+        vectorstore.add_texts(chunks, ids=ids)
+        logger.info(f"Добавлено {len(chunks)} чанков")
+        return JSONResponse({"status": "ok", "chunks": len(chunks)})
+    except Exception as e:
+        logger.exception("Ошибка при загрузке документа")
+        raise HTTPException(status_code=400, detail=str(e))
+
+@app.post("/query")
+async def query(q: QueryIn):
+    """
+    Выполняет запрос к RAG-агенту.
+    """
+    try:
+        result = qa_chain({"question": q.question})
+        answer = result["result"]
+        sources = [doc.metadata.get("source", "unknown") for doc in result["source_documents"]]
+        logger.info(f"Ответ на вопрос: {q.question}")
+        return {"answer": answer, "sources": sources}
+    except Exception as e:
+        logger.exception("Ошибка при обработке запроса")
+        raise HTTPException(status_code=500, detail=str(e))
+
+# =======================
+# Тесты (используются при запуске как скрипт)
+# =======================
+if __name__ == "__main__":
+    import uvicorn
+    import pytest
+    import sys
+
+    @pytest.fixture(scope="module")
+    def client_app():
+        from fastapi.testclient import TestClient
+        return TestClient(app)
+
+    def test_ingest_and_query(client_app):
+        doc = {
+            "text": "Python – это язык программирования высокого уровня, созданный Гвидо ван Россумом. Он известен своей простотой и читаемостью.",
+            "chunk_size": 50,
+            "chunk_overlap": 10,
+        }
+        r = client_app.post("/ingest", json=doc)
+        assert r.status_code == 200
+        r = client_app.post("/query", json={"question": "Кто создал Python?"})
+        assert r.status_code == 200
+        data = r.json()
+        assert "Python" in data["answer"]
+
+    # Запуск тестов
+    if "test" in sys.argv:
+        sys.exit(pytest.main([__file__]))
+    else:
+        uvicorn.run(app, host="0.0.0.0", port=8000)
\ No newline at end of file