Question

I am currently creating a database, it allows a user to upload a publication, which is stored in a table called paper, it stores the paper_id, title, abstract filename and topic_id .I have a table called topic which has topic_id and topic_name which i use for the user to select a topic for their publication. However i want the user to be able to select at least 3 topics, is this possible using this system? I have run out of ideas of how to do it and help would be greatly appreciated

Answer 1

Don’t store topic_id in the paper table. Instead, create another normalized (many-to-many) table which links topic_id to paper_id.

/* Each paper can exist in this table as many times as necessary for all its topics */
CREATE TABLE paper_topics (
  paper_id INT NOT NULL,
  topic_id INT NOT NULL,
  FOREIGN KEY (paper_id) REFERENCES paper (paper_id),
  FOREIGN KEY (topic_id) REFERENCES topic (topic_id),
  PRIMARY KEY (paper_id, topic_id)
);

This will allow you to store as many topics per paper as necessary.

To retrieve the topics for a paper, use:

SELECT 
  paper.*,
  topic_name
FROM 
  paper
  LEFT JOIN paper_topics ON paper.paper_id = topic.paper_id
  LEFT JOIN topic ON topic.topic_id = paper_topic.topic_id
WHERE paper.paper_id = <some paper id value>

It is just about never a good idea to attempt to store multiple values in one column (such as a comma-separated list of topic_id in the paper table). The reason is that in order to query against it, you must use FIND_IN_SET() which drives up the complexity of performing joins and makes it impossible to utilize a column index when querying.