I am playing with onGenerate for my experimental Nim target. Some things are not completely clear for me:
- I can’t see any difference if I use --no-output or not. Does it make any change when I use it ?
- Can I give a platform_name for that target - or what platform_name is used when no platform is specified ?
- In order to build a standard library I want to create a _std directory in the platform directory to put the shadowed modules there. Is this possible from onGenerate/onAfterGenerate. Currently I have the modules in the root path, but this is a bit painful, because I need a lot of #if in order to distinguish between macro and target
- Is -D analyzer-optimize applied for onGenerate/onAfterGenerate. I see that the inliner is applied, but sometimes I get a chain of unnecessary assignments. For what I thought is, that it is applied before onAfterGenerate. But maybe I do something wrong to get the final types there.
- I am thinking of some optimizations. Can I analyze the code in onGenerate and for example mark a function to be inlined and then get it inlined in onAfterGenerate ?