Fixing users incorrectly formatted text

[chadmg]

Users will never always enter data the way you desire all of the time. I've seen someone select Florida as their state and then in the notes section say "I'm actually in Puerto Rico." People type ALL CAPS titles, put their full name in the first name field, fill in bogus email addresses. It's a problem when you are trying to keep your databases pristine and accurate. So I've started thinking about how to prevent incorrect or incorrectly formatted data.

I've implemented a tool that automatically adds acronym tags to known acronyms, so newbies understand them. I've thought about adding a tool that automatically capitalizes the first word of every sentence. But do you always capitalize words at the beginning of the sentence? Does iPod become IPod?

A problem has come up when users enter in their titles in all caps. The best solution is to train your users to enter the data correctly, and then fix errors by hand. But I am trying to figure out a more automated solution. However, you can't simply just convert titles to lowercase and then capitalize the first letter of every word. Some words don't get capitalized, like and, or, and the. Then there are abbreviations and roman numerals and words with capitals in the middle like iPod. So there's really no way to automatically correctly format a mized case title 100% of the time.

The best thing I have come up with is to throw up a flag if the title is in all caps and show a warning to the user like "You have entered a title that is all caps, which is frowned upon. Please reformat your title. If your title needs to be in all caps, please hit submit." After all, the title could just be an abbreviation that should be in all caps.

Is there a better solution? Is there anyone else that looks at these things as a problem? Am I just too anal? How do you get your users to enter information in correctly?

[rogerd]

I like the idea of an all caps warning for titles, although I can imagine some users saying, "Yeah, dude, I need all caps so other people see it and answer me." :)

Automatically assigning proper case would probably work most of the time, though you'd get things like "an IBM computer or an iPod" being converted to "An Ibm Computer Or An Ipod" Maybe assigning proper case only to posts that have been typed all in caps would be a good compromise - relatively few posts would be affected, and most of the time the correction would be fine.

Beyond titles, there's the issue of poorly formatted posts. A mod can fix a bad title in seconds, but who wants to edit a lengthy post that is all in lower or upper case, or has some other problem?