A feature tour followed by complete programs from hello-world to a recursive FFT.
Two parts to this section.
The Tour is every language feature shown in small, focused snippets — read it if you want to see what the language is. The Programs are complete real programs combining features — read them if you want to see what programs look like.
All code is valid against the language reference. Built-in functions (sqrt, sum, map, …) and constants (pi, e, i, …) come from the implicit prelude.
Pages
Tour— Every language feature in small, focused snippets — literals, bindings, blocks, arithmetic, arrays, structs, lambdas, functions, modes, closures, control flow, prelude, testing, modules.
Hello, world— The first program — print to stdout — followed by a function and a calculation.
Arrays and Structs— Element-wise array operations that fuse to one loop, and a small struct with a function over it.
Multi-File Modules— Descriptive statistics across three files in one module folder, with inline tests that see private helpers.
Numerical Methods— Newton's method for square roots, then one Runge-Kutta 4 step for a vector ODE.
Three Ways to Express the Same Computation— The same `normalize` algorithm written three ways — functional (fuses to one loop), explicit-loop with a local `var`, and slice assignment for partial in-place updates.
Mandelbrot Set— Rank-2 matrix plus complex iteration — escape-time rendered as ASCII art. Row-vector slice assignment writes a whole row of the image with one statement.
Sum of Sinusoids— A three-term wave function and a partial Fourier sum, sampled over [0, 1].
Power Iteration— Dominant eigenvalue and eigenvector of a square matrix via the `@` operator and `dot`.
Dedicated @test Modules— A sibling test-only module with support code and table-driven fixtures, stripped from non-test builds.
FFT— Cooley-Tukey radix-2 recursive FFT for any power-of-2 N — complex numbers, strided slicing for the even/odd split, element-wise array arithmetic, and slice assignment to stitch the two output halves.
Sum Types and Match— Modeling a Newton-style solver's outcomes with an enum, then dispatching with match.
Writing a PPM image with `[byte]`— Build a small gradient image as a `[byte]` buffer in P6 (binary PPM) format and dump it to disk with `write_bytes`.
