TextMate-style Grammar for Terraform Plan Diff Output
This repository contains a TextMate-style language grammar for tokenizing the
human-oriented output from the terraform show PLANFILE command in
HashiCorp Terraform.
This is not a typical programming language but instead more of a domain-specific diff representation intended for human consumption. The goal here is therefore just to help call out some interesting tokens that might be interesting to highlight when including a Terraform plan diff in a place which uses inferred highlighting information, such as Markdown literal blocks in systems which use TextMate-style grammars to format their contents.
The goal is to provide at least enough tokens to roughly recreate the formatting Terraform would typically produce itself in a virtual terminal using escape sequences, though it also includes some additional detail in case it's useful to achieve a more detailed highlighting result than Terraform would typically be able to achieve within the constraints of a terminal.
Because this is not a typical use-case for TextMate-style grammars, the scopes selected are rather idiosyncratic and intended to stand a reasonable chance of matching rules in existing themes designed for more "normal" languages, because it's unlikely that most rendering environments would have rules specifically for this very-specialized grammar.
Testing
This grammar has test cases under the tests directory which are intended for
use with
vscode-tmgrammar-test.
To run those tests, install that tool and then run a command like the following:
vscode-tmgrammar-test -s 'text.tfplandiff' -g 'terraformPlanDiff.tmLanguage.json' -t './tests/*'The test only catches incorrect and missing tokens, and will not catch new tokens that were not previously mentioned in the test. When adding new tokens to the grammar it's important to visit the existing test cases and introduce new assertions for locations where those new tokens ought to apply.