🏷️ Handle Schemas and Dataset Types

Schemas are like content labels that describe what's inside your protected data.

They define the structure and types of your data automatically when you protect it, making it easy for iApps to know what they're working with.

Think of schemas as data fingerprints - they tell iApps "this protected data contains an email address and a phone number" without revealing the actual values.

How Schemas Work

When you protect data with DataProtector, the SDK automatically analyzes your JSON object and generates a schema. No manual schema definition needed - it's all handled for you.

const protectedData = await dataProtectorCore.protectData({
  name: 'User Contact',
  data: {
    email: 'alice@example.com',
    phoneNumber: '+1234567890',
    preferences: {
      newsletter: true,
      notifications: false,
    },
  },
});

console.log('✅ Protected data created!');
console.log('📍 Address:', protectedData.address);

🏷️ Generated Schema:

{
  "email": "string",
  "phoneNumber": "string",
  "preferences": {
    "newsletter": "bool",
    "notifications": "bool"
  }
}

Schema Structure The schema automatically maps your data structure to

types that iApps can understand and validate. :::

Supported Data Types

The schema automatically detects these types:

Type	Description	Example
string	Text data	"alice@example.com"
bool	Boolean values	true, false
f64	Numbers	42, 3.14
i128	Big integers	BigInt(123456789)
application/octet-stream	Binary data	File contents
image/jpeg, image/png, etc.	Media files	Images, videos

Auto-Detection The SDK automatically detects file types based on

content. No need to specify MIME types manually. :::

Why Schemas Matter

🎯 For iApp Development

Schemas let your iApps validate and process data safely:

// Inside your iApp
const email = await deserializer.getValue('email', 'string');
const preferences = await deserializer.getValue(
  'preferences.newsletter',
  'bool'
);

🛡️ For Type Safety

Prevents your iApps from processing incompatible data types.

🔍 For Data Discovery

Users can find relevant protected data without seeing the actual content:

const listProtectedData = await dataProtectorCore.getProtectedData({
  requiredSchema: {
    email: 'string',
  },
});

Real Examples

Simple User Profile

const userData = await dataProtectorCore.protectData({
  data: {
    email: 'user@example.com',
    age: 25,
    isSubscribed: true,
  },
});

🏷️ Generated Schema:

{
  "email": "string",
  "age": "f64",
  "isSubscribed": "bool"
}

Nested Contact Information

const contactData = await dataProtectorCore.protectData({
  data: {
    personal: {
      firstName: 'Alice',
      lastName: 'Smith',
    },
    contact: {
      email: 'alice@example.com',
      phone: '+1234567890',
    },
    preferences: {
      marketing: false,
      notifications: true,
    },
  },
});

🏷️ Generated Schema:

{
  "personal": {
    "firstName": "string",
    "lastName": "string"
  },
  "contact": {
    "email": "string",
    "phone": "string"
  },
  "preferences": {
    "marketing": "bool",
    "notifications": "bool"
  }
}

File Data

const fileBuffer = await createArrayBufferFromFile(file);

const fileData = await dataProtectorCore.protectData({
  data: {
    fileName: file.name,
    fileContent: fileBuffer,
    uploadDate: Date.now(),
  },
});

🏷️ Schema for file upload:

{
  "fileName": "string",
  "fileContent": "image/jpeg",
  "uploadDate": "f64"
}

Using Schemas in iApps

Once you have protected data with a schema, you'll want to process it inside an iApp.

Type Matching **Your iApp and frontend must use the same field names

and types.** If they don't match, you'll get runtime errors when processing the data. :::

→ Ready to build an iApp? Check out our detailed Inputs and Outputs guide to learn how to access schema fields inside your iApp using the deserializer.

Next Steps

You now understand how schemas work with protected data. Here's what to explore next:

• Build an iApp: Check out the iApp Generator guide to create your first data processor
• Process data: Learn about processProtectedData for running computations
• See it in action: Try our Hello World tutorial for a complete example