SQL: Unique Words Between Messages | Dataford Interview Questions

Your question is SQL: Unique Words Between Messages. Start with the requirements and the two tables on the right.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem

Discord wants to compare two messages from a channel and identify words that appear in only one of them. Write a PostgreSQL query that tokenizes the selected messages, removes words shared by both messages, and returns each remaining word once.

Requirements

Read the two target messages from discord_messages, joining discord_channels to verify they belong to the requested Discord channel.
Split message content into words using PostgreSQL string functions and preserve each word's message order.
Treat exact spellings as separate tokens when determining whether a word is shared. After identifying words unique to one message, deduplicate case-insensitively while preserving the first spelling encountered.
Return the words in their first-occurrence order across the two messages.

Schema

discord_channels

Column	Type	Description
channel_idPK	INTEGER	Discord channel identifier
channel_name	VARCHAR(100)	Human-readable Discord channel name

discord_messages

Column	Type	Description
message_idPK	INTEGER	Discord message identifier
channel_id	INTEGER	Channel containing the message
content	TEXT	Message text