Design an In-Memory Logger with Search and Deduplication

Implement an in-memory Logger class that stores log messages in insertion order. You do not need to consider persistence, concurrency, distributed logging, or production operational concerns; focus on OOP design and optimization trade-offs for different read/write workloads. Implement the following core operations: logger = Logger() logger.add(message: str) -> None logger.remove(text: str) -> None logger.truncate(n: int) -> None logger.capitalize() -> None logger.messages() -> list[str] Use the following semantics for this problem: add(message): append a message to the internal list; duplicate messages are allowed. remove(text): remove every occurrence of text from every stored message. If text is empty, make no change. truncate(n): truncate every message to at most n characters. n >= 0. capitalize(): uppercase the first character of each non-empty message and leave all remaining characters unchanged. messages(): return all current messages in insertion order without exposing a mutable reference to internal state. Then implement these query operations: logger.search(keywords: list[str]) -> list[str] logger.unique_messages() -> list[str] search(keywords): return messages containing every supplied keyword, preserving original log order. Matching is case-sensitive. An empty keyword list matches every message. unique_messages(): return messages with duplicates removed while preserving the first occurrence and its order. This method must not mutate the stored logs. Finally, discuss: How would you optimize if reads (search and unique_messages) greatly outnumber writes? How would you optimize if writes greatly outnumber reads? How would the design change if duplicates do not matter? CLI Input Format The first line contains an integer q, the number of operations. Each following line is one operation: ADD <message> REMOVE <text> TRUNCATE <n> CAPITALIZE SEARCH <keyword1>|<keyword2>|...|<keywordK> UNIQUE PRINT Keywords in SEARCH are separated by |. A SEARCH command with no text after it searches all messages. Every SEARCH, UNIQUE, and PRINT command outputs one JSON array of strings. Constraints 1 <= q <= 100,000 The total number of characters across all input messages, removal strings, and keywords is at most 1,000,000. 0 <= n <= 1,000,000. Example Input: 9 ADD error: disk full ADD warning: disk nearly full ADD error: disk full SEARCH disk|error REMOVE disk TRUNCATE 12 CAPITALIZE UNIQUE PRINT Output: ["error: disk full"] ["Error: full","Warning: nea"] ["Error: full","Warning: nea","Error: full"] Example Input 6 ADD hello world ADD hello logger SEARCH hello REMOVE hello PRINT UNIQUE Output ["hello world", "hello logger"] ["world", "logger"] ["world", "logger"]