This fixes issue #980 and other similar issues where given that multiple imports are being simultaneously parsed, the insertion order of the modules in NodeMap is not correctly preserved when using a Vec. This comes from the fact that the parser generates the corresponding ModuleId. To respect the order of insertion, we use a map between the generated ModuleId and the parsed Modules.
This fixes issue #980 and other similar issues where given that multiple imports are being simultaneously parsed, the insertion order of the modules in
NodeMapis not correctly preserved when using aVec. This comes from the fact that the parser generates the correspondingModuleId. To respect the order of insertion, we use a map between the generatedModuleIdand the parsedModules.fixes #980 fixes #982