d3ro-voice/server/supabase/migrations/20260410000002_knowledge_documents.sql

-- ============================================================================
-- Phase V2-M: Knowledge Base — RAG 기초 (텍스트 청크 + 단순 검색)
-- V1 RAGService의 Supabase 버전. pgvector 임베딩은 V2-M+1에서 추가 예정.
-- ============================================================================

-- pgvector extension (미래 임베딩 검색용, 현재는 미사용)
-- CREATE EXTENSION IF NOT EXISTS vector;

-- 지식 문서 메타데이터
CREATE TABLE public.knowledge_documents (
  id             uuid PRIMARY KEY DEFAULT gen_random_uuid(),
  user_id        uuid NOT NULL REFERENCES auth.users(id) ON DELETE CASCADE,
  team_id        uuid REFERENCES public.teams(id) ON DELETE SET NULL,
  title          text NOT NULL,
  file_name      text,
  file_type      text CHECK (file_type IN ('txt', 'md', 'pdf', 'docx', 'html', 'url')),
  source_url     text,
  storage_key    text,  -- Supabase Storage의 경로 (업로드된 경우)
  chunk_count    integer NOT NULL DEFAULT 0,
  indexed        boolean NOT NULL DEFAULT false,
  indexed_at     timestamptz,
  created_at     timestamptz NOT NULL DEFAULT now(),
  updated_at     timestamptz NOT NULL DEFAULT now()
);

CREATE INDEX idx_knowledge_documents_user_id ON public.knowledge_documents(user_id);
CREATE INDEX idx_knowledge_documents_team_id ON public.knowledge_documents(team_id);
CREATE INDEX idx_knowledge_documents_created ON public.knowledge_documents(user_id, created_at DESC);

-- 청크 — 원문을 분할한 단위. 임베딩은 추후 vector 컬럼 추가.
CREATE TABLE public.knowledge_chunks (
  id             uuid PRIMARY KEY DEFAULT gen_random_uuid(),
  document_id    uuid NOT NULL REFERENCES public.knowledge_documents(id) ON DELETE CASCADE,
  chunk_index    integer NOT NULL,
  content        text NOT NULL,
  -- embedding vector(768),  -- V2-M+1에서 pgvector 활성화
  created_at     timestamptz NOT NULL DEFAULT now()
);

CREATE INDEX idx_knowledge_chunks_document ON public.knowledge_chunks(document_id);
CREATE UNIQUE INDEX idx_knowledge_chunks_uniq ON public.knowledge_chunks(document_id, chunk_index);

-- 임시: 전문 검색을 위한 tsvector (pgvector 적용 전 대체재)
ALTER TABLE public.knowledge_chunks
  ADD COLUMN tsv tsvector
  GENERATED ALWAYS AS (to_tsvector('simple', content)) STORED;

CREATE INDEX idx_knowledge_chunks_tsv ON public.knowledge_chunks USING gin(tsv);

-- RLS ------------------------------------------------------------------------
ALTER TABLE public.knowledge_documents ENABLE ROW LEVEL SECURITY;
ALTER TABLE public.knowledge_chunks    ENABLE ROW LEVEL SECURITY;

CREATE POLICY "knowledge_documents_read" ON public.knowledge_documents
  FOR SELECT USING (
    user_id = auth.uid()
    OR (team_id IS NOT NULL AND team_id IN (
      SELECT team_id FROM public.team_members WHERE user_id = auth.uid()
    ))
  );

CREATE POLICY "knowledge_documents_insert" ON public.knowledge_documents
  FOR INSERT WITH CHECK (user_id = auth.uid());

CREATE POLICY "knowledge_documents_update" ON public.knowledge_documents
  FOR UPDATE USING (user_id = auth.uid());

CREATE POLICY "knowledge_documents_delete" ON public.knowledge_documents
  FOR DELETE USING (user_id = auth.uid());

-- 청크는 부모 문서 권한 경유
CREATE POLICY "knowledge_chunks_read" ON public.knowledge_chunks
  FOR SELECT USING (
    document_id IN (SELECT id FROM public.knowledge_documents)
  );

CREATE POLICY "knowledge_chunks_insert" ON public.knowledge_chunks
  FOR INSERT WITH CHECK (
    document_id IN (SELECT id FROM public.knowledge_documents WHERE user_id = auth.uid())
  );

CREATE POLICY "knowledge_chunks_delete" ON public.knowledge_chunks
  FOR DELETE USING (
    document_id IN (SELECT id FROM public.knowledge_documents WHERE user_id = auth.uid())
  );

-- updated_at trigger
CREATE TRIGGER set_updated_at_knowledge_documents
  BEFORE UPDATE ON public.knowledge_documents
  FOR EACH ROW EXECUTE FUNCTION public.moddatetime();