Katkat
Back

Refactor a chat message processing function

OpenAISoftware Engineer · Technical Screen
HardSoftware EngineerTechnical ScreenCoding & Algorithms

Last updated

This problem tests your ability to refactor procedural, error-prone code into a robust, extensible pipeline for parsing and managing chat message state.

Problem Statement

You are tasked with refactoring a legacy message processing function. The current implementation uses a monolithic block of nested if-else statements to handle three types of messages: standard (text), command (starting with /), and system (metadata updates).

Implement a MessageProcessor class that processes a list of raw strings and updates an internal state object.

Requirements

  1. Commands: If a message starts with /, it must be parsed into a command name and arguments. Example: /set_user_status active becomes {'type': 'command', 'name': 'set_user_status', 'args': ['active']}.
  2. Standard: Normal text should be trimmed, and messages over 280 characters should be truncated to the first 280 characters with a '...' suffix.
  3. System: Messages starting with [SYSTEM] should be marked as type: 'system' and the system message content stripped of the prefix.
  4. State: Maintain a list of processed messages in a list self.history.

Function Signature

class MessageProcessor:
    def __init__(self):
        self.history = []

    def process_messages(self, raw_messages: list[str]) -> None:
        # Implementation here

Constraints

  • raw_messages length: 0 to 10,000.
  • Each string length: 0 to 1,000.
  • Handle empty strings gracefully (ignore them).

Example

Input: [' Hello world ', '/set_theme dark', '[SYSTEM] User joined'] Output self.history:

[
  {'type': 'standard', 'content': 'Hello world'},
  {'type': 'command', 'name': 'set_theme', 'args': ['dark']},
  {'type': 'system', 'content': 'User joined'}
]

Hints

Hint 1

Define a clear mapping or a strategy pattern to dispatch messages based on their prefix or structure rather than using a long series of if-else checks.

Hint 2

Use a helper method or private function to transform each raw string into a structured dictionary before appending it to the history.

Hint 3

Consider using regular expressions or Python's `.startswith()` and `.split()` methods to handle the parsing logic cleanly.

Follow-up questions

  • How would you modify this class to support undoing the last command if a specific `undo` command is received?
  • If this class needed to process millions of messages, how would you optimize the memory usage of `self.history`?
  • How would you implement a validation layer to ensure that commands have the required number of arguments before they are added to the history?