Geometry

To run any of these examples, you’ll want to save a bit of code in a python file, with any name, e.g. geometry.py, and then run that file by navigating to it on the command line and constructing a call like, coldtype geometry.py

Dividing Rect(angles)

One of the core concepts of Coldtype is the use of the coldtype.geometry.Rect class to encapsulate rectangles and methods for slicing & dicing them.

The most basic rectangle is the one passed to a renderable, i.e. the r variable you get when you define a renderable function, like def rect(r) below. So to fill the entire canvas with a single random color, you can do something like this:

from coldtype import *

@renderable((700, 300))
def rect(r):
    return DATPen().rect(r).f(hsl(random()))

All @renderables have a rectangle associated with them (the full rectangle of the artifact canvas), and all rendering functions are passed rectangles, either via the first and only argument, or as a property of the first argument, as is the case with @animation renderables, which pass a Frame argument that makes the rectangle accessible via f.a.r (where f is the Frame).

But we’re getting ahead of ourselves.

A Rect has lots of methods, though the most useful ones are inset, offset, take, divide, and subdivide.

Here’s a simple example that insets, offsets, and takes.

@renderable((700, 300))
def iot(r):
    return (DATPen()
        .rect(r
            .take(0.5, "mnx") # "mnx" refers to the edge of the rectangle "minimum x"
            .inset(20, 20)
            .offset(0, 10))
        .f(hsl(0.5)))

More complex slicing & dicing

N.B. You may have noticed that the rect functions that a mix of float and int arguments. That’s because a value less than 1.0 will be treated, by the dividing-series of rect functions, as percentages of the dimension implied by the edge argument. So in that take(0.5, "mnx") above, the 0.5 specifies 50% of the width of the rectangle (width because of the mnx edge argument).

Here’s an example that divides a rectangle into left and right rectangles, and shows another useful method, square (which takes the largest square possible from the center of the given rectangle).

@renderable((700, 300))
def lr(r):
    ri = r.inset(50, 50)
    left, right = ri.divide(0.5, "mnx")
    return DATPenSet([
        (DATPen()
            .rect(ri)
            .f(None)
            .s(0.75)
            .sw(2)),
        (DATPen()
            .oval(left
                .square()
                .offset(100, 0))
            .f(hsl(0.6, a=0.5))),
        (DATPen()
            .oval(right
                .square()
                .inset(-50))
            .f(hsl(0, a=0.5)))])

Here’s an example using subdivide to subdivide a larger rectangle into smaller pieces, essentially columns.

@renderable((700, 300))
def columns(r):
    dps = DATPenSet()
    for c in r.inset(10).subdivide(5, "mnx"):
        dps += DATPen().rect(c.inset(10)).f(hsl(random()))
    return dps

Of course, columns like that aren’t very typographic. Here’s an example using subdivide_with_leading, a useful method for quickly getting standard rows or columns with classic spacing.

@renderable((700, 500))
def columns_leading(r):
    dps = DATPenSet()
    for c in r.subdivide_with_leading(5, 20, "mxy"):
        dps += DATPen().rect(c).f(hsl(random()))
    return dps