toscator

Tosca 1.3 compiler

Getting Started

Please read this entire file before you start writing code.

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

Clone the repository
Install the dependencies: npm install
Check that the tests run: npm run test

IDE

Please use Webstorm, as its understanding of types is well better than in VSCode, unless you really know what you are doing.

Webstorm is free for students : https://www.jetbrains.com/community/education/

ESLint

To keep the code clean, we use ESLint. You should enable the option to automatically run eslint when you save a file.

Testing

We use Jest for unit tests. Open the *.spec.js files for examples.

You must run the test with npm run test after each git pull and before each git commit.

If the tests do not pass after a git pull, either your environment is broken or someone made a mistake.

If your tests passed after the previous git pull but do not pass anymore after your changes, and you want to git commit, please fix the tests before doing so.

Parsing

If you simply want to parse a TOSCA file, then follow this instructions:

Write a javascript script to parse a file like in the example below which we will name parse_file.js

import {Parser} from "../parser/parse.js";

// Provide the relative path to the tosca file to parse.
const parser = new Parser(new NodeJsFileManager());
const result = Parser.parse('./tests/data_manual/ToscaNextcloud.yml');
console.log(result);

Compilation

TOSCA is not just parsed: There are extra steps such as type resolution or substitution.

See src/compilation.js:compile

Update the grammar

To update the grammar, there are 2 methods:

modify tosca_1_3.yaml
create a new tosca_x.yaml file

In both cases, you should preprocess the file to generate a new tosca_x.js file that will be used to parse. Otherwise it will preprocess it at every run.

To preprocess, you have to follow these steps:

Run this

import { preprocess } from 'lidy-js/parser/node_parse.js';
preprocess('./schemas/___your_schema_here___.yaml');

In the output file, modify the path used to import in the new generated file by "lidy-js"

import { parse as parse_input } from "lidy-js"

Replace imports if necessary in Toscator's code.

Prerequisites

Tested with:

node v18.13.0
npm 9.2.0

May work with other versions.

Built With

Tosca 1.0 - OASIS Tosca 1.0 documentation
Tosca 1.3 - OASIS Tosca 1.3 documentation

To-Do

Check "TODO" in each file, in comments.
Write tests for what is not tested yet.
Parsing
- (unknown state) ??
Types
- Work in progress, see below.
Substitution
- (unknown state) ??
Orchestration
- nothing done yet
Refactor
- ToscaImport and file loading is unclear (path management)
Implement the specification:
- Open tests/unit/spec_examples/spec_examples.spec.js
- Uncomment tests (check the file: some examples may not be valid)
- Implement until they pass.
- Write for each example a unit test to verify that the result is exact (and not just that it does not crash)
- Repeat until everything is implemented.
Types inheritance
- What about TOSCA types defined by the specification? (ie. tosca.nodes.*)

Type	Feature	Unit test	Integration test (YAML)	Comment
Data
Artifact
Capabilities
Requirement
Relationship
Interface
Group
Policy
Node	wip	x	wip	see table below

Node Type Inheritance

Field	Feature	Unit test	Integration test (YAML)	Comment
properties	x	x	x
capabilities	x	x	x
attributes	x	x	x
interfaces	x	x	x
requirements	?	x		what is the expected behavior ?
artifacts	?	x		what is the expected behavior ?

ditrit/toscator