Aqua

Aqua is a new-gen language for distributed systems.

Aqua programs are executed on many peers, sequentially or in parallel, forming a single-use coordination network.

Aqua's runtime is heterogeneous: it includes browsers, servers, devices, all involved in solving a single task. Therefore, Aqua scripts are compiled into several targets at once, with AIR and Typescript as a default.

Using Aqua

Please refer to Aqua Book to learn how to use Aqua.

Compiler CLI

To build the Aqua compiler, clone the repo & run sbt cli/assembly, or simply download the latest JAR file from the releases page.

It requires java to run Aqua compiler from the command line:

java -jar aqua-cli-%version_number%.jar -i path/to/input/dir -o path/to/output/dir

Input directory should contain files with aqua scripts.

Repository structure

• types – data types, arrows, stream types definitions and variance
• parser - parser, takes source text and produces a source AST
• model - middle-end, internal representation of the code, optimizations and transfromations
• model/transform - optimizations and transfromations, converting model to the result, ready to be rendered
• model/test-kit - tests and test helpers for the model and transformations
• semantics - rules to convert source AST into the model
• linker - checks dependencies between modules, builds and combines an abstract dependencies tree
• backend - compilation backend interface
• compiler - compiler as a pure function made from linker, semantics and backend
• backend/air – generates AIR code from the middle-end model
• backend/ts - generates AIR code and Typescript wrappers for use with Fluence JS SDK
• cli - CLI interface