Merge Nu Scripts

does exactly what the name says.

since both nu -c ast and nu --ide-ast do not retain all important information this is for now regex-based and therefore limited.

Limitations:

• use and export use have to be the first command in a line or stand directly after a ;.
  • if either condition is met within a (multiline-) string it will get screwed up.
• $NU_LIB_DIRS are not supported. use relative imports instead.
• source is not supported.
• The conversion-script should be able to run on windows, but it has never been tested.

Usage:

python3 merge_nu_files.py main_file.nu

How does it work?

A few steps to understand why it does what:

1. FACT: use works the same on both files and in-file defined modules
2. SOLUTION: therefore this script just wraps files with export module NAME {}
3. PROBLEM: / can't be included in the name of a in-file module and duplicate names can exist with files.
4. SOLUTION: therefore this script generates randomised names for modules of the converted files
5. PROBLEM: the names of the modules changed
6. SOLUTION: since the name has changed we have to replace it wherever it has been importing (using regex)
7. PROBLEM: use foo.nu now no longer offers foo test, but instead RANDOM_TEXT test due to the randomisation
8. SOLUTION: use foo and use abc foo both result in foo, therefore just add a submodule with the original name below the generated name