Visounday Server

https://visoundayserver.azurewebsites.net/

Topics

• Tech Stack
• Schema DB
• API
• How to Run the Application

Tech Stack

1. Use Database Azure Cosmos DB for MongoDB
2. Node.js + Express
3. Mongoose ORM
4. Open AI
5. Firebase Authentication for Microsoft Login
6. Json Web Token
7. Canvas
8. ffmpeg
9. Multer
10. Cloudinary for Video Indexer
11. Axios

For more detail check on package.json

Schema DB

User Schema

• name: Stores the user's name.
• email: Stores the user's email address, validated to ensure it matches a standard email format.
• provider: Indicates how the user is authenticated or registered (e.g., Google, Facebook).
• videos: An array of video IDs (ObjectId), referencing videos uploaded by the user.

Both schemas include:

• timestamps: Automatically adds createdAt and updatedAt fields to track document creation and updates.
• versionKey: Prevents adding a __v field to track document versions.

Videos Schema

• user: References a user by their unique ID (ObjectId), linked to the 'User' model.
• cloudinaryId: Stores the ID associated with the video file on Cloudinary (a cloud-based media management platform).
• url: Stores the URL where the video can be accessed.
• description: Optional field for a text description of the video.
• gpt: Stores a string related to GPT (likely output from a GPT model).
• tags: An array of strings representing tags associated with the video.
• gptChats: An array of objects containing role, content, and type, likely representing chat data from a GPT model.
• cover: Stores a URL for the cover image of the video.
• tagImages: An array of objects containing keyword and URLs, likely representing images associated with specific tags.
• frames: An array of objects containing img_url, id, and text, likely representing frames or segments within the video.
• bingStatus: Boolean field indicating the status of something related to Bing (Microsoft's search engine), defaulting to false.

Additional Notes:

• These schemas are defined using Mongoose, a library for MongoDB and Node.js, facilitating schema creation and model definition.
• They establish the structure for documents stored in MongoDB collections, with relationships (ref fields) defined between User and Video models for querying and population purposes.

API

Short explanation of visounday server API.

Method	Routes	Auth	Authz	Header	Body	Response	Description
POST	/verify				Token from the client obtained through Microsoft Social Login.	Acces Token	To get access token from Visounday Server.
GET	/videos	✅		Access Token		List of Videos	Get list videos for logged user.
POST	/videos	✅		Access Token	Url of video and video id (from cloudinary unique name)	Data Video	Generate collage cover photo, video indexer tag, framing video and computer vision. Response is schema Video.
GET	/videos/:id	✅	✅	Access Token		Data Video	Get data video by id.
POST	/videos/:id/gpt	✅	✅	Access Token	Content : message as input to gpt from user	gptChats	Interact with GPT-4.

How to Run the Application

You need serviceAccountKey.json for Microsoft Social Login from Firebase

Use Docker & Create .envbased.env.template`

OR

Run code manually with node version 22.2.0

• Create .env based .env.template
• Navigate terminal to this folder
• Install package npm install
• Run application npm run dev
• Open localhost