Bot Proposal: HTML Encoding Bot

[teusday]

In my time editing Open Library records, I've run into a handful of records (Authors, Editions, and Works) that have HTML encoding as part of their names. Often that causes duplicate records - one with the correct name and one with the HTML-encoded name.

I'd like to write a bot that finds and fixes HTML encodings in record names. Here's a sample of the kind of change the bot would be making.

Technical details:

Loosely, my plan is:

1. Find records with &...; in the name
2. Use a library like FTFY to get the correct characters
3. Verify that the unescaped name actually changed, then write that change back to Open Library

Potential Issues

• This may result in text that uses non-normalized letters (see this sample). My plan above doesn't account for that. I'm not sure whether fixing that belongs in this bot or a separate bot.
• While unlikely, it's possible there does exist a record that legitimately has an HTML encoding in its name. That would be hard to determine before-the-fact, but this bot could avoid editing any records it had already edited, allowing users to manually correct the name without getting into an edit war with the bot.