"Quick Fix" isn't new at all.
Various editors have had this for a long time for a variety of languages (XML and otherwise). The oldest reference I found in 5 minutes of Googling was some documentation about how to implement Quick Fix for your programming language in Eclipse IDE (2009; https://wiki.eclipse.org/FAQ_How_do_I_implement_Quick_Fixes_for_my_own_language%3F).
Eclipse has been doing all of the above for ages:
- Detect an error: check
- Display the location and the error: check
- Compute possible remedies: check
- Display information about the error: check
- Displaying possible remedies: check
- Having the user select a remedy: check
- Actually fix the error: check (Eclipse offers a lot more options than just replacing the error with a fix, though)
The information about the error must contain the following:
- A link to the location of the error: check
- A link to the definition on the corresponding schema: not necessarily available in Eclipse. Depends on the implementation of the QuickFixer.
- Links to information in the W3C specification: For Java, this is usually a link to the relevant JavaDoc or whatever the equivalent of that is in the language being edited.
The difference between the functionality that has been patented and the long existing prior art seems to be providing links to XML Schema and W3C standards. Since existing Quick Fix functionality already provides for links to documentation, I cannot understand how this could in any way be described as "new".
Even though this isn't patenting "Auto Correct" in general, it is narrowing a long existing functionality (instead of for various languages, it's just XML now) and calling that new. My common sense tells me this patent is trivial. The functionality itself is not, but this patent was filed long after many IDEs had already been offering Quick Fix functionality for years.
[edit]
After searching around a little more, I found the patent 7657832 - Correcting validation errors in structured documents as a pretty clear example of prior art from 2003.