Tidy can find unbalanced quotes but won't fix them beyond containing the problem. Right now, I'm not post-Tidy processing to fix them. Should I? We already own most of the code to do that. Frankly, I don't know why Tidy throws in the towel on that one.
Mistyped tags are found by Tidy. It will report
t bogus='100'
out as an unknown tag. No one can automatically fix that problem. There is a message I could recast about the condition, and I'll do so.
Mistyped tags are found by Tidy.
No need, unless we're revisiting the decision to strip all but a subset of tags.
Is there any way that the poster could be notified before the post ...posts? Like, if the HTML is too screwy for Tidy, instead of posting, take the poster to the Edit page with a note at the top saying "Your HTML can't be Tidied, you'll have to fix it yourself."
I think I like this idea. I might even extend it to posts that
can
be tidied. Whenever a post is sumitted, it would get run through Tidy and Gus' TidyPlus©. If they do anything to the post, the user gets sent to a special editpost page with a comment on top that says, "Your post contained some unfoamy html. We tried to fix it for you. Here's what it will look like. Either accept the changes we made, or fix it yourself in the text box below. Loser."
I think I like this idea. I might even extend it to posts that can be tidied. Whenever a post is sumitted, it would get run through Tidy and Gus' TidyPlus©. If they do anything to the post, the user gets sent to a special editpost page with a comment on top that says, "Your post contained some unfoamy html. We tried to fix it for you. Here's what it will look like. Either accept the changes we made, or fix it yourself in the text box below. Loser."
Doable. We could also add the Tidy error output as a comment in their text.
Any opposed?
What does the Tidy error output look like? Is it friendly for non-techies?
It looks like this:
line 1 column 1 - Error:
t bogus
is not recognized!
line 1 column 1 - Warning: missing
t !DOCTYPE
declaration
line 1 column 1 - Warning: discarding unexpected
t bogus
line 1 column 1 - Warning: inserting missing 'title' element
Info: Document content looks like HTML 3.2
5 warnings, 1 error were found!
Gus - The !DOCTYPE, title, and HTML 3.2 error messages are irrelevent to our posts and would befuddle many users. Unless we can limit the error message to only things that "matter", I don't think we should include it.
t off to research Tidy config options re messages