add backend for content search

Author: yangjianfeng1208

education-ai-suite/backend/.gitignore

@@ -0,0 +1,2 @@
+__pycache__
+ext_components/storage_minio/content_search_minio

education-ai-suite/backend/README.md

@@ -0,0 +1,198 @@
+# Backend
+
+### System Workflow
+The system employs a Producer-Consumer pattern to decouple heavy AI inference from the API response cycle.
+
+```mermaid
+graph TD
+    %% Component Styles
+    classDef storage fill:#f9f,stroke:#333,stroke-width:2px;
+    classDef queue fill:#ff9,stroke:#333,stroke-width:2px;
+    classDef app fill:#bbf,stroke:#333,stroke-width:2px;
+
+    User((User / Frontend)) 
+
+    subgraph FastAPI_App [FastAPI Service Layer]
+        API_Upload[Video Upload API]:::app
+        API_Summary[Task Summary API]:::app
+        Processor[Sync Processor Module]:::app
+    end
+
+    subgraph Data_Layer [Infrastructure Layer]
+        DB[(PostgreSQL)]:::storage
+        S3[(MinIO Object Storage)]:::storage
+        Redis{Redis Stream}:::queue
+    end
+
+    subgraph AI_Cluster [AI Execution Layer]
+        Worker[Async Worker Process]
+        AI_Service[AI Mock Service]
+    end
+
+    %% Flows
+    User -->|1. Upload Video| API_Upload
+    API_Upload -->|2. Store File| S3
+    API_Upload -->|3. Save Metadata| DB
+    API_Upload -->|4. Push Task| Redis
+
+    User -->|1. Submit JSON| API_Summary
+    API_Summary -->|2. Log Task| DB
+    
+    %% Sync vs Async Logic
+    API_Summary -->|Mode: Sync| Processor
+    Processor -->|HTTP Req| AI_Service
+    Processor -->|Update Result| DB
+    Processor -->|Direct Resp| User
+
+    API_Summary -->|Mode: Async| Redis
+    Redis -->|Immediate Ack| User
+
+    %% Worker Logic
+    Redis -.->|Subscribe/Pull| Worker
+    Worker -->|Fetch Stream| S3
+    Worker -->|Inference| AI_Service
+    Worker -->|Writeback Result| DB
+```
+### Project Structure
+```bash
+backend/
+├── api/              # API Routes and Business Logic
+├── core/             # Database Models (SQLAlchemy)
+├── mock_services/    # Independent AI Mock Provider
+├── tests/            # Pytest Suite (Unit & Integration)
+├── processor.py      # Synchronous AI Task Handler
+├── database.py       # Database Engine & Session Config
+├── worker_run.py     # Redis Stream Consumer (Worker)
+├── main.py           # Application Entry Point
+└── pytest.ini        # Testing Configuration
+```
+### Prerequisites
+#### Hardware & OS
+OS: Windows 11 / Linux
+
+Python: 3.12.x
+
+#### Infrastructure
+Redis: Task queuing (v5.0+)
+
+PostgreSQL: Metadata storage (v16+)
+
+MinIO: Large file object storage
+
+### Environment Setup
+#### Conda Environment (Miniforge)
+install conda (miniforge) and setup env
+```bash
+conda create -n edu-ai python=3.12
+conda activate edu-ai
+```
+#### .condarc
+```bash
+channels:
+  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge/
+  - https://mirrors.ustc.edu.cn/anaconda/cloud/conda-forge/
+  - conda-forge
+mirrored_channels:
+  conda-forge:
+    - https://conda.anaconda.org/conda-forge
+    - https://prefix.dev/conda-forge
+ssl_verify: false
+proxy_servers:
+  http: http://proxy-dmz.intel.com:911
+  https: http://proxy-dmz.intel.com:912
+```
+```
+conda install -c conda-forge psycopg2 redis-py
+pip install fastapi uvicorn httpx pydantic-settings
+```
+python (3.12.x) C:\Users\user\miniforge3\python.exe
+
+```powershell
+conda activate edu-ai
+pip install -r requirements.txt
+```
+#### backup env
+conda env export > environment.yml
+#### restore env
+conda env create -f environment.yml
+
+### Running the System
+#### Prepare 3rdpart modules
+copy `education-ai-suite/content-search/content_search_minio` under `backend/ext_components/storage_minio` or just create a softlink,
+like
+```bash
+tree -L 3 ext_components/
+ext_components/
+├── readme.md
+└── storage_minio
+    └── content_search_minio -> /xx/x/edge-ai-suites/education-ai-suite/content-search/content_search_minio
+```
+#### Launch the serives
+```powershell
+# Terminal A
+& "C:\Users\user\miniforge3\envs\edu-ai\python.exe" .\main.py
+
+# Terminal B
+& "C:\Users\user\miniforge3\envs\edu-ai\python.exe" .\worker_run.py
+
+# Terminal C
+& "C:\Users\user\miniforge3\envs\edu-ai\python.exe" .\mock_services\dummy_ai_provider.py
+```
+
+### API Usage & Testing
+#### Synchronous Summary (Immediate Result)
+Method: POST
+
+Endpoint: http://127.0.0.1:8000/api/tasks/video-summary
+
+Body (JSON):
+```json
+{
+    "video_url": "C:/videos/classroom_test.mp4",
+    "sync": true
+}
+```
+
+#### Asynchronous Summary (Webhook Notification)
+Method: POST
+
+Endpoint: http://127.0.0.1:8000/api/tasks/video-summary
+
+Body (JSON):
+```json
+{
+    "video_url": "C:/videos/test.mp4",
+    "sync": false,
+    "callback_url": "[https://webhook.site/your-unique-id](https://webhook.site/your-unique-id)"
+}
+```
+webhook.site
+https://webhook.site/ unique URL: e.g. https://webhook.site/28865adb-376c-4a0a-ac59-5204a60f9fe3
+
+#### Video Upload (MinIO Integration)
+Method: POST
+
+Endpoint: http://127.0.0.1:8000/api/tasks/video-upload
+
+Body: form-data | key: video_file | type: File
+
+### Automated Tests
+```powershell
+pip install -r .\tests\requirements.txt
+pytest .\tests\pytest -v
+```
+
+### Debug tools
+pgadmin 4
+tiny RDM
+
+
+### Others
+install Redis
+https://github.com/tporadowski/redis/releases
+
+install PostgreSQL
+16.11.3 https://www.postgresql.org/download/windows/ passwd: edu-ai port: 5432
+
+install postman
+xxx

education-ai-suite/backend/api-map.md

@@ -0,0 +1,15 @@
+# Edge AI Libraries Sample Applications API Map
+
+本表用于前端对接与兼容性对照，来源于 `edge-ai-libraries/sample-applications`。
+
+| 子项目 | 目录 | 用途 | Base Path / 服务 | API 列表 |
+|---|---|---|---|---|
+| chat-question-and-answer | `sample-applications/chat-question-and-answer/` | RAG 问答微服务（FastAPI） | `/v1/chatqna` | `GET /`（docs 入口），`GET /health`，`GET /model`，`POST /chat`（SSE 流式） |
+| chat-question-and-answer-core | `sample-applications/chat-question-and-answer-core/` | 一体化 RAG 服务（FastAPI） | `/v1/chatqna` | `GET /health`，`GET /model`，`GET/POST/DELETE /documents`，`POST /chat`（SSE/非流式），OpenVINO: `GET /devices`、`GET /devices/{device}`，Ollama: `GET /ollama-models`、`GET /ollama-model` |
+| document-summarization | `sample-applications/document-summarization/` | 文档摘要服务（FastAPI） | `http://localhost:8090/v1/docsum` | `GET /version`，`POST /summarize/`（multipart + SSE） |
+| video-search-and-summarization (pipeline-manager) | `sample-applications/video-search-and-summarization/pipeline-manager/` | 视频检索与摘要编排（NestJS） | 服务根路径 | `GET /app/config`，`GET /app/features`，`GET /pipeline/frames`，`GET /pipeline/evam`，`GET /audio/models`，`GET/POST /videos`，`GET /videos/{videoId}`，`POST /videos/search-embeddings/{videoId}`，`GET /tags`，`DELETE /tags/{tagId}`，`GET /search`，`GET /search/watched`，`GET /search/{queryId}`，`POST /search`，`POST /search/{queryId}/refetch`，`POST /search/query`，`PATCH /search/{queryId}/watch`，`DELETE /search/{queryId}`，`GET /summary`，`GET /summary/ui`，`GET /summary/{stateId}`，`GET /summary/{stateId}/raw`，`POST /summary`，`DELETE /summary/{stateId}`，`GET /health`；内部隐藏：`POST /llm`、`POST /vlm`、`GET /states/{stateId}`、`GET /states/raw/{stateId}` |
+| video-search-and-summarization (search-ms) | `sample-applications/video-search-and-summarization/search-ms/` | 向量检索服务（FastAPI） | 服务根路径 | `POST /query`，`GET /watcher-last-updated`，`GET /health`，`GET /initial-upload-status` |
+| plcopen-benchmark | `sample-applications/plcopen-benchmark/` | PLCopen 基准测试（C++） | N/A | 未发现 HTTP API |
+| plcopen-databus | `sample-applications/plcopen-databus/` | PLCopen 数据总线（C++） | N/A | 未发现 HTTP API |
+
+> 备注：如需导出为 OpenAPI 或进一步拆分字段级请求/响应说明，请告知前端对接格式需求。

education-ai-suite/backend/api/.gitignore

@@ -0,0 +1 @@
+__pycache__

education-ai-suite/backend/api/route.py

@@ -0,0 +1,145 @@
+# api/route.py
+from fastapi import APIRouter, Depends, HTTPException, File, UploadFile
+from sqlalchemy.orm import Session
+from database import get_db, SessionLocal
+from core.models import AITask
+
+from processor import run_dummy_ai_logic
+from services.storage_service import storage_service
+
+import redis
+import json
+
+router = APIRouter()
+
+# Initialize sync Redis client (tporadowski/redis default port 6379)
+redis_client = redis.Redis(host='localhost', port=6379, db=0, decode_responses=True)
+
+@router.post("/tasks/video-summary")
+async def submit_video_task(payload: dict, db: Session = Depends(get_db)):
+    # Whether to enable synchronous wait mode (default False, async)
+    is_sync = payload.get("sync", False)
+    callback_url = payload.get("callback_url")
+
+    # 1. Persist to PostgreSQL regardless of sync/async
+    new_task = AITask(
+        task_type="video_summary",
+        payload=payload,
+        status="PENDING",
+        user_id="admin"
+    )
+    db.add(new_task)
+    db.commit()
+    db.refresh(new_task)
+
+    if is_sync:
+        # --- Case A: synchronous mode ---
+        try:
+            # Update status to PROCESSING
+            new_task.status = "PROCESSING"
+            db.commit()
+
+            # Simulate worker processing (or call the real AI function)
+            # Note: dummy_ai_work should be an async function
+            result = await run_dummy_ai_logic(payload.get("video_url"))
+
+            # Write back result
+            new_task.status = "COMPLETED"
+            new_task.result = result
+            db.commit()
+
+            return {
+                "task_id": new_task.id, 
+                "status": "COMPLETED", 
+                "result": result,
+                "mode": "synchronous"
+            }
+        except Exception as e:
+            new_task.status = "FAILED"
+            db.commit()
+            raise HTTPException(status_code=500, detail=f"Sync processing failed: {str(e)}")
+
+    else:
+        # --- Case B: async callback mode (original logic) ---
+        try:
+            new_task.status = "QUEUED"
+            db.commit()
+            
+            redis_client.xadd(
+                "stream:video_processing",
+                {"task_id": str(new_task.id)}
+            )
+            return {
+                "task_id": new_task.id, 
+                "status": "QUEUED",
+                "mode": "asynchronous",
+                "callback_notified": "pending" if callback_url else "none"
+            }
+        except Exception as e:
+            raise HTTPException(status_code=500, detail=f"Redis error: {str(e)}")
+
+def get_db():
+    db = SessionLocal()
+    try:
+        yield db
+    finally:
+        db.close()
+
+@router.get("/tasks/{task_id}")
+async def get_task_status(task_id: str, db: Session = Depends(get_db)):
+    # Look up task in the database
+    task = db.query(AITask).filter(AITask.id == task_id).first()
+    
+    if not task:
+        raise HTTPException(status_code=404, detail="Task not found")
+    
+    return {
+        "task_id": task.id,
+        "status": task.status,
+        "payload": task.payload,
+        "result": task.result, # Null if not completed; otherwise the mock summary
+        "created_at": task.created_at
+    }
+
+@router.post("/tasks/video-upload")
+async def submit_upload_task(
+    video_file: UploadFile = File(...),
+    db: Session = Depends(get_db)
+):
+    """
+    File-upload task endpoint:
+    1. Video stream -> MinIO (via symlink module)
+    2. Path -> PostgreSQL
+    3. Task ID -> Redis Stream
+    """
+    try:
+        # 1. Call service to handle storage
+        minio_payload = await storage_service.upload_and_prepare_payload(video_file)
+        
+        # 2. Persist to PostgreSQL
+        new_task = AITask(
+            task_type="video_summary",
+            payload=minio_payload, # Stores MinIO path info
+            status="QUEUED",
+            user_id="admin"
+        )
+        db.add(new_task)
+        db.commit()
+        db.refresh(new_task)
+
+        # 3. Push to Redis queue
+        redis_client.xadd(
+            "stream:video_processing",
+            {"task_id": str(new_task.id)}
+        )
+
+        return {
+            "task_id": new_task.id,
+            "status": "QUEUED",
+            "storage": "minio",
+            "object_key": minio_payload["video_key"]
+        }
+        
+    except Exception as e:
+        db.rollback()
+        raise HTTPException(status_code=500, detail=f"Upload task failed: {str(e)}")

education-ai-suite/backend/config.py

@@ -0,0 +1,12 @@
+# config.py
+from pydantic_settings import BaseSettings
+
+class Settings(BaseSettings):
+    # Default points to localhost (default for native Windows install)
+    DATABASE_URL: str = "postgresql+psycopg2://postgres:password@localhost:5432/edu_ai"
+    REDIS_URL: str = "redis://localhost:6379/0"
+    
+    class Config:
+        env_file = ".env"
+
+settings = Settings()

education-ai-suite/backend/core/.gitignore

@@ -0,0 +1 @@
+__pycache__