Fix(javascript): Handle namespace imports correctly

[Stansom]

This change addresses a bug in how JavaScript namespace imports (e.g., import * as name from "module") are transformed.

The previous implementation incorrectly assumed the structure of the AST node for a namespace import, leading to errors when trying to get the alias. This has been corrected by using tsc.find-child-by-type to reliably find the identifier node.

This ensures that namespace imports are correctly converted to require statements.

[Stansom]

I also want to add a commit that improves the error handling in the JavaScript and TypeScript transformers.

Previously, when a transformation failed, a generic error was thrown, making it difficult to debug issues. This PR introduces the following changes to provide more informative error messages:

• The handle-transform-error function now includes the line and column number of the problematic code in the error message.
• A new function, body-contains-forbidden-keyword?, has been added to prevent the transformation of arrow functions that contain this, super, arguments, or new.target.

These changes will make it easier for users to identify and fix errors in their code.

[Olical]

Tried out the before and after locally, works fine for me and the error is printed nicely.

[Olical]

Instead of using a var and mutating it I'm pretty sure this can be implemented with an if statement and returning the result of (traverse node) - inside the recursive function we can return at the point you (set found n) . We don't actually use a return keyword though which breaks out of the control flow, instead we should have two steps.

In the let block we can do the lookup, if we find a value, we end the recursion and return the found value in the tail position. I think we also have to check the result of each sub-call to traverse in that each so we return if something is returned from the child call.

I think this current code actually iterates more than it needs to and doesn't exit early when it finds a match? Just a small optimisation but I think this can be written in a true recursive (although maybe not tail recursive because of the tree nature, that would require more state to be passed along in the recursion which I don't care about so much) function that would be easier to follow and perform less work when it its an error.

I could be wrong about this though because I've been pretty sick all week and I'm still kind of out of it today but I wanted to get a response to you 😅

If you disagree or would rather not try to implement this in a recursive way, please let me know, I'm open to being swayed.

[Stansom]

Thank you for sharing your thoughts! Unfortunately, I don't have time to look into it right now, but I will as soon as I can.

I hope you feel better soon!