Examples

Ten runnable programs live in the examples directory of the repository. Each is a complete Zig program with a main function, and each is compiled by zig build check.

zig build run-basic_compression

Run every one:

zig build run-all-examples

The programs

Each example covers one thing. None repeats another's demonstration.

Example	What it covers
basic_compression	One-shot round trip with verification and header introspection
compression_levels	All nine levels side by side with ratios and timings
reusable_contexts	One Compressor and one Decoder, many streams
caller_buffers	Caller-provided output buffers, sizing and retry on overflow
chunked_input	Irregular write sizes producing a byte-identical stream
streaming_pipelines	compressStream and decompressStream as bounded pipelines
file_compression	File to file with staging, plus the damaged-stream failure path
large_file	A 16 MiB round trip in constant memory, hash-verified
decoder_semantics	Statistics, CRC verification, corruption taxonomy, concatenation
custom_allocator	A counting custom allocator proving the byte balance

Run one by prefixing its name with run-:

zig build run-compression_levels
zig build run-large_file

The examples that write files put them in a directory under the system temporary area they create themselves and remove again at the end, so a run leaves the working directory as it found it.

Output

The examples write to standard error through std.debug.print, which keeps them usable in a pipeline where standard output might be redirected.

Source layout

Tests live at the bottom of the module they cover, so a behaviour described in the documentation can be read next to the code that implements it and the tests that pin it down. The examples are the executable half of the same idea: each one is short enough to read in full and produces output you can check against the claim in its name.