Vietnamese editors have reported cases where content is copied out of Content Translation to be published as new articles and avoid the current system of limits. This ticket proposes to measure how frequently people copy content to the clipboard in Content Translation. This should apply only to the contents of the translation document (i.e., not counting when the user copies from the source text or other parts of the UI).

This would allow to:

• Identify potential problems. Which wikis may have an unusual activity.
• Evaluate potential solutions. Check how much the copy activity has decreased after limiting the access to less experienced users (T299636) or limiting the possibility of publishing such content (T301339).

Additional considerations:

• Copying events do not identify a problem directly. Content an be copied by several (good) reasons. However, we expect to see a higher than usual activity when copy is used to massively create new articles.
• As we count copy events, we may want to differentiate (or add metadata) about how much these translations copied have been modified. For example, we may wan to know how many times the initial unmodified machine translations are copied. Our main focus in this case is for copied content that won't be allowed to be publish in the tool rather than content that has been heavily edited and could be published with the tool anyways.