piped

A pipeline orchestrator for everyone to build softwares easier and faster.

Warning: This project is on early stage of development now. Any breaking changes can be made.

What is it?

Basically piped is a build system such as the cmake, MSBuild or bazel. But you can define custom pipelines to build or process assets in any formats with any tools.

Usage

piped <input>

The <input> is a valid pipeline file path.

Examples

See the examples directory.

Docs

Predefined variables

Name	Type	Description	Possible values
hostArch	String	Indicates the CPU's architecture of the host system.	x86, x86_64, arm
hostOS	String	Indicates the CPU's architecture of the host system.	linux, macos, windows

Syntax

Comments

Comments start by double slashes. They include all characters until the end of line.

// One-line comment

Variables & Set statements

Variables can be one of types of below.

Type	Description
Bool	An boolean type that can be true or false at a time.
Integer	A 64-bit signed integer.
String	A variable length string.
Array	An array that contains variable length of other values.
Dictionary	An collection of key-value pairs. Keys are always strings.

String literals can be escaped by \(back slash). Here're all escape sequences supported.

Escape sequence	Description
\n	Line feed character.
\r	Carriage return character.
\t	Tap character.
\\	Back space character.
\0	Null character.
\'	Single quote character.
\"	Double quote character.
\`	Grave character.

Variables can be set by set statements.

@set bar=0;
@set bar="Some string value";
@set foo=[bar, baz];

Print statements

You can write anything to the stdout or stderr as you need.

@print "A value of 'foo' is " foo;
@printErr "Error! the given path '" path "' is not valid!";

Pipeline invocations

Pipeline is a main concept of the piped - it is some logics that can be executed in parallel. You can do many things with them such as downloading, unzipping, clonning and compiling. To invoke them, just type its name and feed some arguments.

copy src="my_file.txt" dst="./some/path";

Background invocations

If it is ok to run a pipeline in background and execute next lines immediately, mark them as nonblock. The piped will execute them on background worker thread in parallel.

// It will not block execution!
@nonblock copy src="my_file.txt" dst="./some/path";

Also you can give it a name to refer it later.

@nonblock "my-copy" copy src="my_file.txt" dst="./some/path";

Awaiting invocations

Use await to wait for background invocations. There are 3 different types to do so.

@await;				// This will wait for all previous "unnamed" background invocations.
@await "my-copy";	// This will wait for all previous "named" background invocations.
@await all;			// This will wait for all previous background invocations regardless of named or unnamed.

Importing pipelines

Any pipelines can be imported by import statements. It is an error if a pipeline is imported more than once. You have to specify its name when importing.

@import "./sub-pipeline.piped" as sub_pipeline;

When you invoke a imported pipeline, any arguments you fed will be injected into that pipeline as variables. You can check if pipeline invocations have needed arguments when writing pipelines with functions.

// pipeline.piped
sub_pipeline some_parameter="some_argument";

// sub-pipeline.piped
@if is_exists("some_parameter") {
	// Do some logics
}

Documentations for all built-in pipelines are here.

If statements

if statements are quite typical.

@if equals(hostOS, "windows") {
	@print "Hello, windows!";
} @else if equals(hostOS, "macos") {
	@print "Hello, macos!";
} @else {
	@print "Hello, unknown!";
}

Since the piped does not support any operator in syntax, you should use functions for real world applications. Please refer functions for more details.

Functions

Functions are callable logics that always return a value.

some_function(1, 2, 3);

Unlike pipeline invocations, functions have positional parameters. It means position of each arguments are important.

Documentations for all functions are here.