// Slide 18: Retrievers + Vector Stores -- FAISS / Chroma / PGVector
// Code slide: index documents, embed, retrieve top-k, plug into a chain.

const ds = require('./design-system');

function createSlide(pres, theme) {
  const slide = pres.addSlide();
  ds.helpers.slideBase(slide, pres, theme);

  ds.helpers.addHeader(slide, pres, theme, {
    eyebrow: 'STAGE 1: CHAINS',
    section: 'Retrievers',
    title: 'Vector store -> retriever -> RAG-цепочка',
    sectionNumber: 1,
  });

  ds.helpers.addCodeBlock(slide, pres, theme, {
    x: 0.5, y: 1.5, w: 9.0, h: 3.25,
    language: 'python',
    filePath: 'examples/retriever_rag.py',
    code: [
      'from langchain_openai import OpenAIEmbeddings',
      'from langchain_community.vectorstores import FAISS',
      'from langchain_core.runnables import RunnablePassthrough',
      '',
      '# 1. Эмбеддинги и индекс',
      'embeddings = OpenAIEmbeddings(model="text-embedding-3-small")',
      'docs = ["Cats are mammals.", "Python is a programming language.",',
      '        "LangChain helps build LLM apps."]',
      'vectorstore = FAISS.from_texts(docs, embedding=embeddings)',
      '',
      '# 2. .as_retriever() превращает store в Runnable',
      'retriever = vectorstore.as_retriever(search_kwargs={"k": 2})',
      '',
      '# 3. RAG-цепочка через LCEL: context + question -> answer',
      'from langchain_core.prompts import ChatPromptTemplate',
      'from langchain.chat_models import init_chat_model',
      '',
      'prompt = ChatPromptTemplate.from_template(',
      '    "Answer based on context.\\nContext: {context}\\nQ: {question}"',
      ')',
      'model = init_chat_model("openai:gpt-4.1-mini")',
      'rag = (',
      '    {"context": retriever, "question": RunnablePassthrough()}',
      '    | prompt | model | StrOutputParser()',
      ')',
    ].join('\n'),
  });

  ds.helpers.addCallout(slide, pres, theme, {
    x: 0.5, y: 4.85, w: 9.0, h: 0.3,
    kind: 'info',
    text: 'Альтернативные сторы: Chroma (легковесный), PGVector (production Postgres), Pinecone, Weaviate, Qdrant.',
  });

  ds.helpers.addSourceLine(slide, pres, theme, {
    source: 'python.langchain.com/docs/concepts/retrievers/',
  });
  ds.helpers.addPageNumber(slide, pres, theme, 18);
}

module.exports = { createSlide };