插入由特定用户保存到数据库中的多个URL的最佳方法是什么？

Question

i am making a bookmarking site where user saves his urls. so what is the best approach to insert all urls of a user into a database(mysql).

1)create individual table for individual user. in each table create 1 column and put 1 url per row(per cell).

2) create 1 table and insert all urls into a single cell according to the row of the user.

if 2nd option is best then can anyone explain me how to do it? i can insert one url in one cell but how to insert all urls into a single cell and again retrieve it to display in a proper manner. Thank you for taking time and reading my question. I tried hard to do this myself but failed. that's why asking here.

Answer 1

Create a bookmark table with the following columns: (id, fk_user_id, url) . Each entry would have its own id, and the foreign-key would be the id of the user.

You don't want to create an individual table per user. That won't scale! Imagine you have a thousand users - that's at least a thousand tables that you have!

The second approach could work, but has the following problems:

• To add a URL, you have to read in the existing value, append the new value, and then write it back. That's two database operations you are doing just to add a new value. If you take the foreign-key approach, you simply have to perform one write.
• While perhaps not as likely, you could have a case where the user is using two clients and you end up getting a request to add a bookmark from both of them at close to the same time. This can lead to consistency issues. For example, assume you already had http://google.com in the column, and Client #1 wants to add http://reddit.com and Client #2 wants to add http://stackoverflow.com . If Client #2 reads in the value before Client #1's data has been persisted, you will either end up with http://google.com, http://reddit.com or http://google.com, http://stackoverflow.com depending on who writes to the table last. This could even happen if you get multiple add requests from the same client at close to the same time.
• Add deleting bookmarks to the mix, and your consistency issues could get even worse. Deleting also becomes much more difficult, because instead of simply deleting by bookmark id, you now have to pull in the data, split it, and then perform string comparisons on the URLs to figure out what you need to delete, and then write the modified data back. What if you end up getting multiple delete requests close to the same time? You run into the same problem that you did with the case where you were adding new URLs.