Changes

Use acorn-loose as fallback parser for logic rules with imports and top-level returns.

This change attempts to parse logic rules with acorn-loose if the standard acorn parser fails. The acorn-loose parser is designed to handle syntactically invalid JavaScript by constructing a valid AST even when there are syntax errors. Since we only use the AST to locate and inline import statements, the AST remains usable for our purposes even when the syntax is technically invalid. acorn-loose is meant for our use case of parsing code during transformation steps. We still try to parse using acorn first as recommended by the acorn team.

This is necessary because logic rules can contain both import statements (for shared code) and top-level return statements, which is technically invalid JavaScript syntax. Standard acorn throws a SyntaxError when encountering this combination. However, this syntax is acceptable in our context because we transform the code before evaluation by inlining the imports and wrapping everything in a function, which makes it valid JavaScript.

Checklist before requesting a review

• Cleaned up debug prints, comments, and unused code
• Tested end to end
• Included screenshots or walkthrough video of change if impacts UX

Related pull requests

Link other PRs here that are related to this change