In-Memory File System: Strict mkdir, touch, ls, rm, rmdir and a Command Runner

Implement an in-memory file system with directories and files under a single root, /. The task is delivered as a multi-file project with provided test cases, so reading the tests to pin down expected behavior, running them, and debugging failures are part of the work. It comes in parts; the first two are below.

Clarifying Questions

• Are all paths absolute? How should a relative path, an empty path, a trailing slash, repeated slashes, or components such as . and .. be treated?
• Which characters may a file or directory name contain?
• How should an operation report failure: by raising an exception, returning a boolean, or returning an error string?
• Do files have contents, or only names?

Part 1 — mkdir, touch and ls

Implement:

• mkdir(path): create a directory.
• touch(path): create an empty file.
• ls(path): list a directory.

Every operation must check that the path is valid. Neither mkdir nor touch creates missing directories along the path: every component except the last must already exist as a directory, and only the final component is created. For example, mkdir("/a/b") fails if /a does not exist.

Clarifying Questions for this Part

• What should mkdir do when the final component already exists as a directory, or as a file?
• What should touch do on an existing file (a no-op, as in Unix, or an error), and on an existing directory?
• What does ls return for a directory: names only, and in what order? What does it return when the path names a file?
• Should ls("/") work, and can mkdir("/") ever succeed?

What This Part Should Cover

• A tree representation that distinguishes files from directories.
• Path parsing and validation, including failure when an intermediate component is missing or is a file.
• Deterministic ls output and clear error behavior.

Part 2 — rm, rmdir and the execution function

Implement rm(path) to remove a file and rmdir(path) to remove a directory. Then write the execution function: it takes the commands the tests supply, dispatches each one to the matching operation, and returns the results.

Clarifying Questions for this Part

• Does rmdir remove only empty directories, or everything beneath it? Can it remove the root?
• Does rm refuse directories, or also remove them recursively?
• What is the command format (one string per command, such as mkdir /a, or a structured list), and what does the execution function return: one output per command, or only the outputs of ls?
• When one command fails, does execution continue with the next one?

What This Part Should Cover

• Removal semantics that keep the tree consistent, with explicit rules for non-empty directories and the root.
• A dispatcher that handles unknown commands and wrong argument counts.
• Consistent reporting of each command's result or error.