Inmanta Language server

Provides the bridge between the Inmanta compiler and Visual Studio Code IDE, by implementing a Language Server.

For recent versions of inmanta-core (>=6), the server takes ownership of the Python environment it runs in and installs any modules and module dependencies in that environment. For older versions of inmanta-core, a separate compiler venv is used, which can be configured with the compilerVenv option.

This is alpha software, it may or may not work

Installation

1. mkvirtualenv -p python3 lstesting

2. pip install inmantals

3. install the inmanta extension via marketplace

4. change config for the extension (ctrl+,)

   1. set inmanta.pythonPath to the virtual env you just created ~/.virtualenvs/lstesting/bin/python3
   2. set inmanta.ls.enabled to true

Features

1. navigate-to-definition on types
2. docstring display on hover
3. find references to a symbol (e.g. right-click > Find All References)
4. supports working on an Inmanta project or an Inmanta module

Not supported yet

1. work with incorrect models (needs to compile, all or nothing)

Troubleshooting

No module named x

If compilation fails with the message "no module named x" where x is a Python module, you might need to clean up the virtual environments used by the compiler. This issue can be caused by running the compiler from multiple different environments. To clean up the virtual environment, remove the .env directory in the Inmanta project directory as well as the compiler venv specified by the client, if it exists.

References

https://microsoft.github.io/language-server-protocol/specification