Requirements:

• Call-context query using Create search api queries from old query api queries #1314
• Resolver (flowr search resolvers and mappers #1312) for params/args of a function (Info.Parameters) and function call (Info.Arguments)
• Some type of collection of read and write functions similar to the dependencies query
  • -> we could also edit the dependencies query to not be specific to entire files to allow using that one directly in our rule through fromQuery? Then we'd need a Info.LastCall resolver to link these dependencies to the last call of an associated write/read in case the script creates files it later expects itself!
  • -> load these infos from a static file (like a csv/json) in the flowr sources rather than having them hard-coded in code?

We then check the file path's validity as part of the linting rule itself, rather than as part of the query API. ie we only use the query api to gather all of the paths we may need to inspect.

// using a modified dependencies query:
Q.fromQuery({ type: 'dependencies', filter: 'read' })
 .with(Info.LastCall, { callName: 'setwd', cascadeIf: CascadeAction.Continue })
 .with(Info.Arguments)

// then run the query and select the nth argument based on our list of functions

Things to consider:

• Configurability? file read/write queries can be quite slow so maybe we allow users to make the linting rule "less specific