Thoughts on LLM based tech support bot?

[jonchurch]

Thoughts on LLM based tech support bot?

[jonchurch]

im closing this as it's not something I expect us to seriously prioritize

[jonchurch]

for reference it was this issue, and the dev had a good outcome from my human written but LLM assisted debugging: pillarjs/routington#29

the output the LLM gave me with just their issue (which was a pretty dang good issue, it included enough code to know exactly what they were doing):

The issue you're encountering with matching a query string (/name?age=50) using routington is because routington is designed to match URL paths, not query strings. The query string is not considered part of the URL path; it's a separate component used to pass additional data to the server. Therefore, when you try to match /name?age=50, it won't work as expected because routington is looking for a path that matches exactly, without considering the query string.

To fix this issue, you need to separate the path from the query string before matching. You can use the URL or another method to parse the incoming request URL, extract the path without the query string, and then pass it to your match function. Here's how you could adjust the dispatcher function to handle this:

 function dispatcher(request: { path: string; method: string }): { param: string; handler: Function } | string | never {
  // Extract path without query string
  const url = new URL(request.path, `http://${request.headers?.host || 'localhost'}`);
  const pathWithoutQueryString = url.pathname;

  const match = router.match(pathWithoutQueryString);
  if (!match) { console.log('404 - Not Found'); return "" };

  const node = match.node;
  const cb = node[request.method];

  if (!cb) { console.log('405 - Method Not Allowed'); return ""; }

  return {
    param: match.param,
    handler: cb // You might need to check if cb is not undefined before calling it
  }
}

This adjustment uses the URL object to parse the incoming request.path, extracting the pathname (which excludes the query string) before it's passed to router.match. Note that this example assumes request.path includes the full URL; you might need to adjust this based on how your actual request object is structured. If request.path already contains only the path, you won't need to prepend a base URL.

Also, ensure you handle cases where the method does not exist on the matched route to avoid runtime errors.