creativesar/face

1from fastapi import APIRouter, HTTPException, BackgroundTasks2from pydantic import BaseModel3from typing import List, Dict, Any4import logging5import os6from dotenv import load_dotenv7 8# Load environment variables9load_dotenv()10 11# Import services12from services.optimized_openrouter_rag_service import OptimizedOpenRouterRAGService13from services.qdrant_service import QdrantService14from services.openrouter_service import OpenRouterService15 16# Configure logging17logging.basicConfig(level=logging.INFO)18logger = logging.getLogger(__name__)19 20# Initialize services21openrouter_service = OpenRouterService()22qdrant_service = QdrantService()23rag_service = OptimizedOpenRouterRAGService(openrouter_service, qdrant_service)24 25router = APIRouter()26 27class ContentIndexRequest(BaseModel):28    content: str29    chapter_id: str30    section_title: str31    source_url: str32 33class BatchContentIndexRequest(BaseModel):34    contents: List[ContentIndexRequest]35 36class PersonalizeContentRequest(BaseModel):37    chapter_id: str38    user_id: str39    user_background: Dict[str, Any]40 41@router.post("/index")42async def index_content(content_data: ContentIndexRequest):43    """44    Index a single piece of content for RAG retrieval45    """46    try:47        logger.info(f"Indexing content for chapter: {content_data.chapter_id}")48 49        result = await rag_service.index_content(50            content=content_data.content,51            chapter_id=content_data.chapter_id,52            section_title=content_data.section_title,53            source_url=content_data.source_url54        )55 56        return {57            "success": True,58            "document_id": result["document_id"],59            "indexed_content_length": result["indexed_content_length"]60        }61    except Exception as e:62        logger.error(f"Error indexing content: {str(e)}")63        raise HTTPException(status_code=500, detail=str(e))64 65@router.post("/batch-index")66async def batch_index_content(content_data: BatchContentIndexRequest):67    """68    Index multiple pieces of content at once69    """70    try:71        logger.info(f"Batch indexing {len(content_data.contents)} content items")72 73        # Convert to the format expected by the service74        contents = [75            {76                "content": item.content,77                "chapter_id": item.chapter_id,78                "section_title": item.section_title,79                "source_url": item.source_url80            }81            for item in content_data.contents82        ]83 84        results = await rag_service.batch_index_content(contents)85 86        return {87            "success": True,88            "indexed_count": len(results),89            "results": results90        }91    except Exception as e:92        logger.error(f"Error batch indexing content: {str(e)}")93        raise HTTPException(status_code=500, detail=str(e))94 95@router.get("/status")96async def content_status():97    """98    Get status of the content indexing system99    """100    try:101        # Test Qdrant connection and get point count102        point_count = await qdrant_service.count_points()103 104        return {105            "status": "ready",106            "indexed_documents_count": point_count,107            "services": {108                "qdrant": await qdrant_service.test_connection(),109                "openrouter": await openrouter_service.test_connection()110            }111        }112    except Exception as e:113        logger.error(f"Error getting content status: {str(e)}")114        raise HTTPException(status_code=500, detail=str(e))115 116@router.post("/ingest-from-docs")117async def ingest_from_docs(background_tasks: BackgroundTasks):118    """119    Ingest content from the docs directory (for textbook content)120    This would read the Docusaurus docs and index them121    """122    try:123        # This would typically read from the docs directory124        # and index all the content for RAG125        docs_path = os.getenv("DOCS_PATH", "/app/docs")  # Default path in Docker126 127        if not os.path.exists(docs_path):128            raise HTTPException(status_code=404, detail=f"Docs path not found: {docs_path}")129 130        # This is a simplified version - in a real implementation, you'd want to131        # parse all the markdown files and extract content properly132        import os133        import glob134 135        markdown_files = glob.glob(f"{docs_path}/**/*.md", recursive=True)136        markdown_files += glob.glob(f"{docs_path}/**/*.mdx", recursive=True)137 138        content_items = []139        for file_path in markdown_files:140            try:141                with open(file_path, 'r', encoding='utf-8') as f:142                    content = f.read()143 144                # Extract chapter/section info from file path145                relative_path = os.path.relpath(file_path, docs_path)146                chapter_id = relative_path.replace('/', '_').replace('\\', '_').replace('.md', '').replace('.mdx', '')147                section_title = os.path.basename(file_path).replace('.md', '').replace('.mdx', '')148 149                content_items.append({150                    "content": content,151                    "chapter_id": chapter_id,152                    "section_title": section_title,153                    "source_url": f"/docs/{relative_path}"154                })155            except Exception as e:156                logger.warning(f"Could not process file {file_path}: {str(e)}")157 158        # Index all content items159        results = await rag_service.batch_index_content(content_items)160 161        return {162            "success": True,163            "processed_files": len(content_items),164            "indexed_documents": len(results),165            "results": results166        }167    except Exception as e:168        logger.error(f"Error ingesting from docs: {str(e)}")169        raise HTTPException(status_code=500, detail=str(e))170 171@router.post("/personalize")172async def personalize_content(personalize_data: PersonalizeContentRequest):173    """174    Get personalized content based on user background175    """176    try:177        # In a real implementation, this would adapt content based on user background178        # For now, we'll return the original content with some basic adaptation info179        logger.info(f"Personalizing content for chapter {personalize_data.chapter_id} for user {personalize_data.user_id}")180 181        # This would typically search for content related to the chapter_id182        # and adapt it based on the user's background183        # For now, we'll return a placeholder response184        return {185            "chapter_id": personalize_data.chapter_id,186            "user_id": personalize_data.user_id,187            "adaptation_info": "Content adaptation based on user background would happen here",188            "original_content": "Original textbook content would be returned here, adapted based on user background",189            "user_background": personalize_data.user_background190        }191    except Exception as e:192        logger.error(f"Error personalizing content: {str(e)}")193        raise HTTPException(status_code=500, detail=str(e))