You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've considered how this support should be and decided to introduce a new DSL to describe a query, apart from the current one, to search over/transform codes. Note that this decision still needs to be self-reviewed carefully.
Reason
Our existing DSL is good at describing a query in the almost same form as codes to search over/transform. However, for some languages that are used for describing a data structure instead of a control flow, such as JSON or YAML, this kinda DSL could be expressive too much! What we really need to handle them is a simpler one with minimum expressivity.
Initial Design of Our New DSL
Now I'm thinking about syntax & semantics for our new DSL. It can be something like jq or yq; their DSLs are expressive enough to describe what kind of data should be matched and how the data should be like.
Another choice could be some kinda language with a notation to describe the commutativity of components. I may have read an academic paper on this topic.
tree-sitter grammar for YAML is here. It would be helpful to validate YAML quickly.
The text was updated successfully, but these errors were encountered: