Too strict test on #wh6

[66Gramms]

createInterface has a member called completer which is exactly made for autocompletion, yet the tests are written in such a way that using the completer function is not accapted as it’s matches aren’t separated by 2 spaces and I haven’t found a way to change that.

So if you use the built-in way that is also recommended in the code that is given to you, you can’t complete this challange, or at least I haven’t found a way. Here is my relevant code:

const rl = createInterface({
  input: process.stdin,
  output: process.stdout,
  prompt: "$ ",
  completer: handleCompletion,
});

let previousLine = "";
function handleCompletion(line: string) {
  const matches = Array.from(Commands.keys())
    .filter((command) => command.startsWith(line))
    .map((input) => `${input} `);

  const executables: string[] = [];
  for (const dir of (process.env.PATH || "").split(PATH_SEPARATOR)) {
    try {
      const files = fs.readdirSync(dir);
      executables.push(
        ...files
          .filter((file) => file.startsWith(line))
          .map((input) => `${input} `)
      );
    } catch (err) {
      // Ignore errors (e.g., if a path does not exist)
    }
  }

  if (!matches.length && !executables.length) {
    process.stdout.write("\u0007"); // Ring bell if no matches
    return [[], line];
  }

  if (line !== previousLine && [...matches, ...executables].length > 1) {
    process.stdout.write("\u0007"); // Ring bell on first match
    previousLine = line;
    return [[], line];
  }

  return [[...matches, ...executables], line];
}

Results:

[andy1li]

Hey @66Gramms, you can manually write to stdout and return [] in the completer function to bypass the default behavior of displaying suggestions on separate lines.

Here’s an example to build on:

[andy1li]

Closing this thread due to inactivity. If you still need assistance, feel free to reopen or start a new discussion!