dynamic-schema-sanitizer

A fully dynamic, deeply recursive sanitization + validation engine for Node.js. Designed to replace Mongoose validation, work with dynamic schemas, and sanitize payloads safely before hitting the database.

💡 Ideal for headless CMS, multi-tenant apps, form builders, no-SQL dynamic models, or any system where schemas are stored in DB instead of code.

✨ Features

• ⚡ Deep, recursive sanitization for any nested structure
• 🔍 Dynamic runtime schemas (no Mongoose models needed)
• 🔄 Type coercion: String, Number, Boolean, Date, ObjectId, Mixed
• 🛡 Validation support: required, enum, match, min/max
• 🧹 Clean null/undefined/empty objects/arrays
• 🚫 Unknown field handling: strip, keep, or error
• 🌐 Multilingual, customizable error messages
• 🧨 Fail-fast mode (stopOnFirstError)
• 🧰 Full bulk array processing with sanitizeBulk()
• 🔥 Stable enterprise-grade error codes
• ✔ Only dependency: mongoose (for ObjectId validation)

📦 Installation

npm install dynamic-schema-sanitizer

yarn add dynamic-schema-sanitizer

🚀 Quick Example

import { sanitizePayloadWithFields } from "dynamic-schema-sanitizer";

const schema = {
  name: { type: "String", required: true, trim: true },
  age: { type: "Number", min: 0, max: 120 },
  isActive: { type: "Boolean", default: true },
  tags: { type: ["String"], default: [] },
};

const payload = {
  name: "  John Doe  ",
  age: "25",
  isActive: "false",
  tags: ["nodejs", 123, true],
};

const result = sanitizePayloadWithFields(payload, schema, {
  validate: true,
  unknownFields: "strip",
});

console.log(result);

Output

{
  "value": {
    "name": "John Doe",
    "age": 25,
    "isActive": false,
    "tags": ["nodejs", "123", "true"]
  },
  "errors": []
}

🧩 Bulk Sanitization (sanitizeBulk)

import { sanitizeBulk } from "dynamic-schema-sanitizer";

const result = sanitizeBulk(
  [
    { name: "A", age: "10" },
    { name: "B", age: "invalid" },
  ],
  schema
);

console.log(result);

Bulk Output Format

{
  "records": [
    { "value": { "name": "A", "age": 10 }, "errors": [] },
    { "value": { "name": "B", "age": null }, "errors": [ ... ] }
  ],
  "valid": [ { ... } ],
  "invalid": [ { ... } ]
}

🧠 Schema Format

Primitive

name: {
  type: "String";
}

Required

email: { type: "String", required: true }

Enum

status: { type: "String", enum: ["active", "blocked"] }

Min/Max

score: { type: "Number", min: 1, max: 100 }

Regex

email: { type: "String", match: /.+@.+\..+/ }

ObjectId

userId: {
  type: "ObjectId";
}

Nested object

profile: {
  age: { type: "Number" },
  gender: { type: "String" }
}

Array of primitives

tags: {
  type: ["String"];
}

Array of objects

logs: [
  {
    message: { type: "String" },
    at: { type: "Date" },
  },
];

⚙️ Options

🔥 Stable Error Codes

• FIELD_REQUIRED
• INVALID_NUMBER
• INVALID_BOOLEAN
• INVALID_DATE
• INVALID_OBJECT_ID
• INVALID_TYPE
• ENUM_MISMATCH
• REGEX_MISMATCH
• MIN_VIOLATION
• MAX_VIOLATION
• EXPECTED_ARRAY
• EXPECTED_OBJECT
• UNKNOWN_FIELD

🌍 Multilingual Errors

sanitizePayloadWithFields(payload, schema, {
  language: "gu",
  messages: {
    gu: {
      FIELD_REQUIRED: "આ ફીલ્ડ જરૂરી છે",
      INVALID_NUMBER: "અમાન્ય નંબર",
    },
  },
});

📚 sanitizeBulk()

Bulk sanitization with full reporting:

import { sanitizeBulk } from "dynamic-schema-sanitizer";

const result = sanitizeBulk(records, schema);

Output

{
  records: SanitizeResult[],
  valid: SanitizeResult[],
  invalid: SanitizeResult[]
}

• records = ALL results
• valid = only valid rows
• invalid = only rows containing errors

Perfect for: ✔ CSV import ✔ XLSX bulk upload ✔ API bulk insert ✔ Queue processing ✔ prepare insertMany() safe payload

🧨 Using with MongoDB insertMany()

MongoDB behavior:

• ordered: true → stops on first error
• ordered: false → continues, skips invalid docs

You can do:

const { valid } = sanitizeBulk(records, schema);

await collection.insertMany(valid, { ordered: false });

🛠 Author

Sahil Trambadiya Creator & maintainer of dynamic-schema-sanitizer

⭐ Support

If this library saves you time, consider ⭐ starring the repo!