Project Builder API #1780
Project Builder API #1780
Conversation
|
This crude prototype can be executed via: |
| return this; | ||
| } | ||
|
|
||
| public setEngine(engine : 'tree-sitter' | 'r-shell') { |
There was a problem hiding this comment.
This should be the KnownParser type which... Should be renamed to known engine :p
There was a problem hiding this comment.
So we want to have an extra field for the engine that has precedence over the options?
It gets a bit confusing when doing something like this:
await new FlowrAnalyzerBuilder(require("x <- 1"))
.setEngine(new RShell())
.amendConfig(c => {
c.defaultEngine = 'tree-sitter'
return c;
}).build();
There was a problem hiding this comment.
Correction: Take name property from KnownParser
There was a problem hiding this comment.
amendConfig should accept a function that can return FlowrConfigOptions | void.
Disabling the eslint warning is fine
src/project/flowr-analyzer.ts
Outdated
| this.parser = parser; | ||
| } | ||
|
|
||
| public async dataflow() : Promise<PipelineOutput<typeof DEFAULT_DATAFLOW_PIPELINE>> { |
There was a problem hiding this comment.
Using default data flow pipeline here is... Wrong in general we should select the type based on the engine :D in general we maybe want to embed a cache?
There was a problem hiding this comment.
in general we maybe want to embed a cache?
Yes, I think that would make sense. We need to have the dataflow analysis results for the query execution anyways. In theory, the slicing pipeline could re-use the dataflow step results, but I'm not sure whether we should enable this right now
No description provided.