@mummareddy_mohanreddy/jsone-core

Universal JSON table inference and generation engine.

Convert ANY JSON (objects, arrays, primitives) into structured tables automatically. Zero dependencies, fully typed TypeScript.

Features

• 🎯 Universal JSON Support — Objects, arrays, primitives, nested structures — all become valid tables
• ⚡ Zero Dependencies — Ultra-lightweight (9.1 kB), no external packages
• 📊 Automatic Inference — Detects table structures and generates columns without configuration
• 🔍 Smart Type Detection — Automatically identifies dates, numbers, booleans, and strings
• 🪶 Fully Typed — Complete TypeScript support with full type definitions included
• 📈 Nested Object Flattening — Converts nested objects to dot-notation columns
• 🏆 Quality-Based Ranking — Multiple tables? Ranks by row count and column quality

Installation

npm install @mummareddy_mohanreddy/jsone-core

Quick Start

import { tableFromJsone } from "@mummareddy_mohanreddy/jsone-core";

// Single object
const json1 = { name: "Alice", age: 30, joined: "2024-01-15" };
const table1 = tableFromJsone(json1);
// Result: 1 row, 3 columns (name, age, joined)

// Array of objects
const json2 = [
  { id: 1, name: "Alice", role: "Admin" },
  { id: 2, name: "Bob", role: "User" }
];
const table2 = tableFromJsone(json2);
// Result: 2 rows, 3 columns

// Array of primitives
const json3 = ["apple", "banana", "cherry"];
const table3 = tableFromJsone(json3);
// Result: 3 rows, 2 columns (#index, value)

// Nested objects
const json4 = {
  users: [
    { name: "Alice", profile: { city: "NYC", zip: "10001" } },
    { name: "Bob", profile: { city: "LA", zip: "90001" } }
  ]
};
const table4 = tableFromJsone(json4);
// Result: 2 rows with flattened columns (name, profile.city, profile.zip)

API Reference

tableFromJsone(data: any, viewId?: string): TableResult

Converts any JSON structure into a table representation.

Parameters:

• data — Any JSON value (object, array, primitive, nested structure)
• viewId — Optional string for table identification

Returns: TableResult object containing:

• rows: any[][] — 2D array of cell values
• columns: Column[] — Column definitions with names and inferred types
• messages: string[] — Any warnings or metadata messages

Example:

const result = tableFromJsone(jsonData);
console.log(result.rows);      // [[cell1, cell2, ...], ...]
console.log(result.columns);   // [{ name: "col1", type: "string" }, ...]

inferColumns(rows: any[][], maxSamples?: number): Column[]

Detects column types from row data using statistical analysis.

Returns: Array of Column objects with inferred types.

flattenRow(row: any, prefix?: string): Record<string, any>

Recursively flattens nested objects into dot-notation keys.

Example:

flattenRow({ user: { name: "Alice", profile: { city: "NYC" } } });
// Returns: { "user.name": "Alice", "user.profile.city": "NYC" }

Type Detection

The library automatically detects:

• Dates — ISO 8601 format (80% heuristic threshold)
• Numbers — Integers and floats
• Booleans — true/false values
• Strings — Default fallback
• Null/Undefined — Preserved as-is

Optional Metadata

Wrap JSON in $meta to provide explicit table hints:

const json = {
  $meta: {
    title: "Sales Report",
    description: "Q1 2024 sales data"
  },
  data: [
    { product: "Widget", revenue: 10000 },
    { product: "Gadget", revenue: 15000 }
  ]
};

const table = tableFromJsone(json);
// Uses data array as table source, applies metadata hints

TypeScript Support

Full TypeScript support with exported types:

import type { 
  TableResult, 
  Column, 
  ColumnType 
} from "@mummareddy_mohanreddy/jsone-core";

const result: TableResult = tableFromJsone(data);

Performance

• Parsing: O(n) time complexity where n = JSON depth
• Inference: O(r × c) where r = rows, c = columns
• Memory: No external dependencies, minimal overhead

Use Cases

• 📄 Convert JSON API responses to viewable tables
• 📊 Transform database exports into structured data
• 🔍 Explore complex nested JSON structures
• 📈 Generate reports from JSON data
• 🎨 Display semi-structured data in tables

License

MIT

Links

• 📦 npm Package
• 🐙 GitHub Repository
• 📖 Full Format Specification
• 🚀 Viewer Web App