TargetJS: The "State → Transition → Render" Framework

Code order defines the UI sequence.

targetjs.io

TargetJS is a JavaScript UI framework that transforms how applications handle state and execution flow. Instead of treating state merely as data for rendering, TargetJS integrates the active UI runtime including in-progress animation and execution flow directly into the state itself. This unified state approach allows developers to capture and restore the exact state of an interface at any millisecond.

TargetJS also lets code order directly define the UI sequence. UI, transitions, API calls, event handling, and state are unified into self-contained Targets that chain together through Code-Ordered Reactivity.

It can be used as a full-featured framework or as a lightweight library alongside other frameworks. It is also a highly performant web framework, as shown in the framework benchmark.

What problems TargetJS solves

UI frameworks model the final result, not the transition

Traditional frameworks model the UI as a function of data: change data, re-render the UI. When data changes from A → B, the UI immediately jumps to render B. The framework doesn’t naturally represent the journey from A → B. But modern, rich user experiences are more like: A → transition → B.

TargetJS redefines state to include both the transition and the active UI runtime. This allows transitions to adapt dynamically on the fly, for example, pausing when an element leaves the screen or changing speed on scroll. It also allows the entire UI to be captured mid-animation and restored later, resuming exactly where it left off.

Fragmentation across multiple mental models

In many applications, state, animation, events, loading, timing, and callbacks are all handled through separate concepts or APIs. This creates glue code and a mental split between them.

TargetJS unifies them under one concept and one model. Methods and fields are unified and both become reactive units with their own state, lifecycle, timing, execution conditions, looping, and callbacks. This shifts fields from passive values to active participants, reducing boilerplate and keeping application logic consolidated.

UI sequences are difficult to trace in code

UIs often follow sequences like this:

Click → animate button → fetch data → render results → animate items → highlight one item

In traditional code, that sequence is often scattered across different places such as event handlers, effects, promises, and callbacks.

TargetJS code order and target reactivity allow the implementation to more closely mirror the actual UI sequence.

🚀 Why TargetJS?

1. Unified State: UI values, transitions, and runtime execution are part of the same state.
2. Restorable UI and Runtime: UI and runtime state can be captured and later the complete state can be restored including transitions and execution flow from the saved point.
3. UI as Sequence: Code describes the UI story from top to bottom, just like the user experiences the interaction: "When this finishes, do that."
4. Adaptive Transitions: Transitions can pause, resume, change speed, and respond to input.
5. Ultra-Compact: Minimal code, with no coordination variables.
6. Zero Boilerplate Async: Targets handle waiting for nested asynchronous operations automatically.
7. Animation by Default: Turn value into high-performance animations simply by adding steps.

⚡ Quick Start (30 Seconds)

1. Install

npm install targetj

2. Example

This creates the following sequence:

bounce → move → turn red → log

Notice how the code reads in the same order as the UI sequence. The $$ suffix makes each target wait for all preceding targets to complete.

Click the square while it is animating to capture its current state. Click it again to restore that state and resume the animation from the exact point at which it was saved.

The checkpoint captures more than the square's visual state. It captures the runtime behind it including transition progress and pending targets. Restoring it resumes the whole sequence from that exact point, not just what was on screen. You can try the example at https://targetjs.io/examples/firstExample.html

import { App, state } from "targetj";

App({
  width: 100,
  height: 100,
  backgroundColor: "blue",
  cursor: "pointer",

  // Starts immediately.
  scale: { value: [0.5, 1.2, 1], steps: 100 },

  // $$ waits for the previous target to finish.
  x$$: { value: [0, 180], steps: 200 },

  backgroundColor$$: {
    value: "crimson",
    steps: 200
  },

  done$$() {
    console.log("Sequence complete");
  },

  // 1st click: save a checkpoint. 2nd click: restore it and resume.
  onClick() {
    state().toggle();
  }
}).mount("#app");

Targets

In TargetJS, targets are the fundamental unit of behavior instead of methods. Methods and properties both are internally transformed into targets that the framework schedules and executes.

Mental Model

A target can:

• execute a method
• hold a value
• move toward a new value over time
• pause while moving toward a value
• wait for previous targets to complete
• react when previous targets update
• fetch data
• respond to events
• create children
• run callbacks
• control its own lifecycle

This lets UI code follow the same order as the user experience.

Target Controls

A target can also be defined as an object with optional controls that manage its lifecycle and execution.

| Property | Description | |------|------| | value | The data or function that determines the target's state. | | steps | Turns a value change into an animation. | | interval | Delay (ms) between steps or executions. | | cycles | Number of times the target repeats. | | loop | Controls repetition, either actively for continuous execution or passively when the value changes. | | active | Boolean property controlling when value is executed. | | enabledOn | Determines whether the target is enabled for execution. | | pauseOn | Pauses execution while the target is in progress. | | easing | Predefined easing function controlling how values update over steps. | | onComplete | Callback triggered when this target (and its children) finishes. | | onValueChange | Callback triggered when the target emits a new value. | | on<PropertyName>Step | Callback triggered on every step of a specific property. |

Compact Execution Syntax

Target names can include special suffixes that determine when they execute. This provides a compact alternative to coordinating the same behavior with callbacks.

| Syntax | Name | Behavior | | -------- | -------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | name | Standard | Runs immediately in code order unless active is false. | | name$ | Reactive | Runs whenever the preceding sibling Target updates. Similar to activating the next Target from onValueChange() or on<PropertyName>Step(). | | name$$ | Deferred | Runs after the entire preceding Target chain including children, animations, and API calls has completed. Similar to activating the next Target from onComplete(). |

To prevent a Target from running automatically, set active: false and activate it later with .activateTarget().

Examples: Like Button → Animated Like (in 3 Steps)

Let’s see how TargetJS handles a complex interaction that would usually require 50+ lines of React/CSS. The example demonstrates how to run four asynchronous operations in a strict sequence. In other words, each step has to wait for all the previous ones to complete.

1) Like button

One object defines a UI element without separate HTML/CSS. Static targets map directly to DOM styles/attributes. You can still use CSS if wanted.

<div id="likeButton"></div>

import { App } from "targetj";

App({
  width: 220,
  height: 60,
  lineHeight: 60,
  textAlign: "center",
  borderRadius: 10, 
  html: "♡ Like",
  // Runs immediately on mount
  scale: { value: [1.2, 1], steps: 12, interval: 12 },
  backgroundColor: { value: ["#ffe8ec", "#f5f5f5"], steps: 12, interval: 12 }
}).mount("#likeButton");

2) Adding the Interaction

We move the animation into an onClick and add a deferred heart animation.

<div id="likeButton"></div>

import { App } from "targetj";

App({
  width: 220, height: 60, lineHeight: 60, textAlign: "center",
  borderRadius: 10, backgroundColor: "#f5f5f5",
  cursor: "pointer", userSelect: "none",
  html: "♡ Like",
  onClick() {
    this.setTarget('scale', { value: [1.2, 1], steps: 8, interval: 12 });
    this.setTarget('backgroundColor', { value: [ '#ffe8ec', '#f5f5f5' ], steps: 12, interval: 12 });
  },
  heart$$: {  // Wait for the button animation to finish, THEN add and animate the heart.
    html: "♥", color: "crimson", fontSize: 20,
    fly() {
      const cx = (this.parent.getWidth() - this.getWidth()) / 2;
      this.setTarget('x', { value: [cx, cx + 22, cx - 16, cx + 10, cx ], steps: 50, cycles: 2 }); // Repeat it twice
      this.setTarget('y', { value: [0, -120], steps: 400 });
    }
  }  
}).mount("#likeButton");

3) The Full Async Workflow

We handle UI, two animations, a POST request, and a cleanup.

<div id="likeButton"></div>

import { App } from "targetj";

App({
  width: 220, height: 60, lineHeight: 60, textAlign: "center",
  borderRadius: 10, backgroundColor: "#f5f5f5", cursor: "pointer", userSelect: "none",
  role: "button", tabIndex: 0,
  html: "♡ Like",
  onClick() {
    this.setTarget('scale', { value: [1.2, 1], steps: 8, interval: 12 });
    this.setTarget('backgroundColor', { value: [ '#ffe8ec', '#f5f5f5' ], steps: 12, interval: 12 });
  },
  heart$$: {
    html: "♥", color: "crimson", fontSize: 20,
    fly() {
      const cx = (this.parent.getWidth() - this.getWidth()) / 2;
      this.setTarget('x', { value: [cx, cx + 22, cx - 16, cx + 10, cx ], steps: 50, cycles: 2 }); // Repeat it twice
      this.setTarget('y', { value: [0, -120], steps: 400 });
    }
  },
  // Wait for the nested animation in heart$$ to finish , then fetch
  fetch$$: { method: "POST", id: 123, url: "/api/like" },
  // Wait for fetch to finish, then cleanup
  removeHearts$$() { this.removeChildren(); },
  onKey(e) { if (e.key === "Enter") this.activateTarget("onClick"); } 
}).mount("#likeButton");

Summary

Each target has its own state and lifecycle. Targets execute automatically in the order they are written. $$ defers execution until all prior sibling targets (including their children) are fully complete. Animations, API calls, event handling, and child creation are all treated uniformly as targets. Complex asynchronous flows can be structured by organizing work into parent and child targets. In addition, targets provide built-in capabilities such as onComplete callbacks, pauseOn, enabledOn, looping with delays, and more. This also makes the code more compact, as it avoids using extra variables to track progress and reduces the need for loops and conditional statements.

Table of Contents

1. 📦 Alternative Installation Via CDN
2. Using TargetJS as a Library
3. Deeper Examples:
   • Search → Fetch → Replace → Highlight Example
   • Infinite Loading and Scrolling Example
4. Special Target Names
5. How to Debug in TargetJS
6. Documentation
7. License
8. Contact
9. 💖 Support TargetJS

📦 Alternative Installation Via CDN

Add the following <script> tag to your HTML to load TargetJS from a CDN:

<script src="https://unpkg.com/targetj@latest/dist/targetjs.js"></script>

This exposes TargetJS on window, so you can initialize your app with TargetJS.App(...).

Ensure your code runs after the DOM is ready (use defer, place your script at the bottom of the <body>, or wrap it in a DOMContentLoaded listener).

<div id='redbox'></div>

<script>
    TargetJS.App({
        backgroundColor: 'red',
        width: { value: [100, 250, 100], steps: 20 },
        height: { value: [100, 250, 100], steps: 20 }
    }).mount('#redbox');
</script>

Zero-JS Declarative HTML

TargetJS can also be used as a "no-code" library. Elements with tg- attributes are discovered and activated automatically.

<div
   tg-background="red"
   tg-width="{ value: [100, 250, 100], steps: 20 }"
   tg-height="{ value: [100, 250, 100], steps: 20 }">
</div>

Using TargetJS as a Library

TargetJS can run inside an existing app mounted into a DOM element managed by another framework.

React (mount + cleanup)

import React, { useLayoutEffect, useRef } from "react";
import { App as TApp } from "targetj";

export default function TargetIsland() {
  const hostRef = useRef(null);

  useLayoutEffect(() => {
    const el = hostRef.current;
    if (!el) return;

    TApp({
      width: { value: [100, 500], steps: 100 },
      height: 200,
      backgroundColor: "purple",
      onClick() { console.log("click"); }
    }).mount(el);

    return () => {
      TApp.unmount();
    };
  }, []);

  return <div ref={hostRef} style={{ width: 100, height: 200, overflow: "hidden" }} />;
}

Deeper Examples

Search → Fetch → Replace → Highlight

This example shows how TargetJS models a UI workflow directly in code order: Click → animate button → fetch users → remove old results → add new results → pause → highlight one result

The fetch target is initially set to active: false, which means it waits for an explicit trigger. When the user clicks, the fetch target is activated. TargetJS understands that fetching data is an asynchronous operation.

The $$ postfix means that a target waits for the preceding sibling targets to complete before running. In this example, removeChildren$$ waits for fetch to complete before it begins. addChildren$$ begins after both fetch and removeChildren$$ are completed.

Notice how fetch, removeChildren$$, and addChildren$$ appear in the same order as the UI sequence. The code is organized around the experience itself.

Lastly, pause$$ adds a short pause before highlighting the first user with an animation. setTarget is an imperative way to implement targets within methods.

import { App } from "targetj";

App({
      containerOverflowMode: 'always',
      searchButton: {
        element: 'button',
        type: 'button',
        width: 220,
        height: 60,
        lineHeight: 60,
        border: 0,
        borderRadius: 10,
        cursor: 'pointer',
        textAlign: 'center',
        backgroundColor: '#f5f5f5',
        html: 'Search',
        onClick() {
          this.setTarget('html', 'Searching...');
          const users = this.parent.getChild('users');
          this.setTarget('scale', { value: [1, 1.15, 1], steps: 12 });
          this.setTarget('backgroundColor', { value: ['#ffe8ec', '#f5f5f5'], steps: 12 });
          users.activateTarget('search', { reset: true });
        },
      },
      users: {
        gap: 10,
        marginTop: 20,
        containerOverflowMode: 'always',
        search: {
          active: false,
          value() {
            this.removeChildren();
          },
        },
        fetch$$: 'https://targetjs.io/api/randomUsers',
        addChildren$$: {
          cycles() {
            return this.val('fetch').length;
          },
          value(i) {
            const user = this.val('fetch')[i];
            return {
              width: 340,
              borderRadius: 10,
              backgroundColor: '#fafafa',
              boxShadow: '0 6px 16px rgba(0, 0, 0, 0.08)',
              padding: 14,
              containerOverflowMode: 'always',
              scale: { value: [0.96, 1], steps: 10 },
              userName: {
                marginTop: 5,
                html: user.name,
                fontWeight: 700,
              },
              userEmail: {
                marginTop: 5,
                html: user.email,
                opacity: 0.7,
              },
            };
          },
        },
        relabelButton$$() {
          const button = this.parent.getChild('searchButton');
          button.setTarget('html', 'Search');
        },
        pause$$: { interval: 150 },
        highlightFirst$$() {
          const firstUser = this.getChild(0);
          if (firstUser) {
            firstUser.setTarget('scale', { value: [1, 1.04, 1], steps: 10 });
            firstUser.setTarget('backgroundColor', { value: '#fff1a8', steps: 10 });
          }
        },
      },
    }).mount('#app');

Infinite Loading and Scrolling Example

In this advanced example, we implement an infinite-scrolling application.

• addChildren is a special target that adds multiple items to the container’s children each time it executes. The onVisibleChildrenChange event detects changes in the visible children and activates addChildren to insert new items and fill any gaps.

• photo and userName each add a div element inside every item, serving as placeholders for the photo and user name.

• pause$$ delays the execution of all targets that follow it by 300 ms. It also has pauseOn: 'hidden', which means the delay is paused if the item becomes invisible, such as when the user scrolls down quickly. This also puts all following $$ targets on hold, so fetch$$ will not execute until the item becomes visible again.

• fetch$$ retrieves the user’s details.

• reveal$$ executes after fetch$$, revealing the user name and populating the photo with a random color.

• wave executes when all visible children have completed their targets or when scrolling ends, giving each user item a coordinated animation.

TargetJS employs a tree-like structure to track visible branches, optimizing scroller performance.

<div id="userList"></div>

import { App, getEvents, getScreenWidth, getScreenHeight } from "targetj";

App({
  preventDefault: true,
  width: 300,
  height() { return getScreenHeight(); },
  x() { return (getScreenWidth() - this.getWidth()) / 2; },
  containerOverflowMode: "always",
  canDeleteDom: false,
  overflow: 'scroll',
  onWindowScroll: true,
  addChildren: {
      value() {
          return Array.from({length: 10}, (_, i) => ({
              height: 56,
              width() { return this.parent.getWidth() - 30; },
              bottomMargin: 8,
              borderRadius: 12,
              backgroundColor: 'white',
              boxShadow: "0 8px 20px rgba(0,0,0,0.08)",
              pause: { 
                  interval() { return getEvents().isWindowScrolling() ? 600 : 0; },
                  pauseOn: 'hidden'
              },
              photo$$: {
                  x: 10,
                  y: 10,
                  width:34,
                  height:34,
                  borderRadius: '50%',
                  backgroundColor: '#ddd'
              },
              userName$$: {
                  x: 60,
                  y: 10,
                  width: 180,
                  height: 30,
                  overflow: 'hidden',
                  borderRadius: 5,
                  backgroundColor: '#ddd'
              },
              pause$$: { 
                  interval: 300,
                  pauseOn: 'hidden'
              },
              fetch$$: "https://targetjs.io/api/randomUser",
              reveal$$() {
                  const userName = this.getChild('userName$$');
                  userName.setTarget('html', this.val('fetch$$').name);
                  userName.setTarget('backgroundColor', { value: 'white', steps: 20 });
                  this.getChild('photo$$').setTarget('backgroundColor', { value: '#' + Math.random().toString(16).slice(-6), steps: 20 });
              }
          }));
      },
      onVisibleComplete() {
          if (!getEvents().isWindowScrolling()) {
              this.activateTarget('wave');
          }
      }
  },
  onWindowScrollTopEnd() {
      if (this.isTargetVisibleTreeComplete('addChildren')) {
          this.activateTarget('wave');
      }
  },
  wave: {
      active: false,
      interval: 30,
      cycles() { return this.visibleChildren.length; },
      value(i) {
          const child = this.visibleChildren[i];
          if (child) {
              child.setTarget("scale", { value: [1, 1.06, 1], steps: 18 });
              child.setTarget("opacity", { value: [1, 0.7, 1], steps: 18 });
          }
      }
  },
  onVisibleChildrenChange() {
      if (!this.visibleChildren.length) {
          return this.activateTarget("addChildren");
      }
      const scrollTop = this.$dom?.getScrollTop() ?? 0;
      const lastY = this.getLastChild().getY() - scrollTop - 500;
      if (lastY <= this.getHeight()) {
          this.activateTarget("addChildren");
      }
  }
}).mount("#userList");

Special Target Names

Some target names have built-in meaning and interact directly with the DOM, layout system, or browser events.
Because these behaviors are expressed as targets, they still participate in the same execution system and dependency flows as any other target.

Styles

These targets update CSS properties and transforms:

• width, height
• opacity
• x, y, z
• rotate, rotateX, rotateY, rotateZ
• scale
• backgroundColor, color

These can be animated simply by adding steps.

For custom CSS, use the css Target.

Structure

These targets define the structure of the interface:

• children or addChildren – adds new children each time the target executes
• html – inner HTML content, often simple text
• element – specify the DOM element type (e.g., div, canvas)

Events

These targets respond to browser events:

• onClick
• onScroll
• onKey
• onResize
• onEnter / onLeave
• onVisibleChildrenChange

How to Debug in TargetJS

TargetJS provides built-in debugging tools:

TargetJS.tApp.stop(); // Stop the application.
TargetJS.tApp.start(); // Restart the application
TargetJS.tApp.throttle = 0; // Slow down execution (milliseconds between cycles)
TargetJS.tApp.debugLevel = 1; // Log cycle execution

• Use t(id) in the browser console to find an object by its element id.
• Use t(id).debug() to inspect all the vital properties.
• Use t(id).logTree() to inspect the UI structure.

Documentation

Explore the potential of TargetJS and dive into our interactive documentation at www.targetjs.io.

License

Distributed under the MIT License. See LICENSE for more information.

Contact

Ahmad Wasfi - [email protected]

💖 Support TargetJS

If you would like to show some appreciation:

• ⭐ Star this repo on GitHub to show your support!
• 🐛 Report issues & suggest features.
• 📢 Share TargetJS with your network.