Python walkthrough#

This complete lesson is the checked-in source used by the native package example. It is paired with the Rust lesson, rather than being a separate demonstration of an imagined API.

The complete scene#

"""Paired with noon::example_scenes::first_lesson; no external assets required."""

from noon import *


def step(scene, shapes):
    scene.play(shapes.animate.shift(0.75 * RIGHT), run_time=0.5, rate_func=linear)


class FirstLesson(Scene):
    def construct(self):
        title = Text("One scene, ordinary functions", font_size=36).shift(2.7 * UP)
        equation = MathTypst("x(t) = x_0 + v t", font_size=48).shift(1.6 * UP)
        self.add(title, equation)

        square = Square(side_length=0.8).set_fill(BLUE, opacity=1)
        circle = Circle(radius=0.4).set_fill(ORANGE, opacity=1)
        square.shift(0.8 * LEFT + 0.4 * DOWN)
        circle.shift(0.8 * RIGHT + 0.4 * DOWN)
        shapes = VGroup(square, circle)
        self.add(shapes)
        self.wait(0.5)
        for _ in range(2):
            step(self, shapes)

        caption = Text("Two steps, one shared scene", font_size=24).shift(2 * DOWN)
        self.add(caption)
        self.wait(0.5)

Objects, families, and membership#

Text, MathTypst, Square and Circle construct visual objects. shift places them in scene coordinates, not CSS pixels. VGroup makes the two shapes a family, so a single animation can move both. self.add makes those objects visible without introducing a Create animation.

The small step helper receives the scene and the family. It calls the same play method you would use inside construct; a helper function and an ordinary Python loop need no separate scheduling mechanism.

Authored time#

The first wait contributes half a second. Each of the two calls to step contributes half a second, and the final wait contributes another half second. The lesson therefore authors two seconds of content. Creating the last caption happens after the earlier plays; it is not necessary to create every object at time zero.

The Python host executes authoring code. Rust owns the shared scene semantics and runtime. Replaying a qualified finite scene should not be confused with rerunning its source, especially when source has side effects.

Run the same file natively#

PYTHONPATH=build/python python3 -m noon_native web/python/examples/first_lesson.py

For an installed wheel, python -m noon_native first_lesson.py or noon first_lesson.py uses the installed facade without PYTHONPATH. Follow native Python details for the wheel-build procedure; only install a candidate that matches your interpreter and platform.

Extend the lesson#

Change the shift amount or the loop count and run again. Changing a Python numeric parameter is ordinary source authoring. For interactive changes driven by input during playback, use the appropriate input/reactive APIs and consult playback and interaction.

Use the generated Python API index to find the actual signatures and declared methods of Scene, Mobject, VGroup and Text in this revision.