We all know and, presumably, love the way “Formatting: XHTML” handles our weblog entries. Still, I’d think there was room for some improvement, because in a number of cases blindly replacing every linebreak with <br /> and every double line break with <p> ... </p> is not necessarily the most elegant solution.

I’d go as far as saying that in the cases of <table>, <ul>, <ol>, <dl> and probably a few others this behavior is definitely unwanted by most people, even though the rest of the entry could profit from normal formatting: this is why neither choosing “Formatting: None” nor “keeping your table on a single line” are really satisfying solutions, all things considered.

I know, of course, that all that XHTML and typography parsing is a difficult issue, so we won’t get it for 1.6, I’m sure. Still, I think this FR might prove somewhat popular, and I wanted to make it. Thanks for considering.

Most of these tags already have the “blind inserted” <p> and <br /> tags removed from them—you do not have to always keep the markup on a single line.  Tables being an exception, we currently have no special handling for table tags within Auto-XHTML fields.  But typically only when you begin more complex nesting of these other tags does the current parser not handle these situations.

That said, overall, yes, typography will likely be getting some major enhancement in 2.0.

Hi

I believe the best feature of Auto XTHML is its automatic handling of “fancy quotes” etc. That’s the only reason I use it. However, its dodgy behaviour with ul,li,table etc is irritating. I think the engine should just switch off when it sees any <div>, <ul> <ol> or <table> tag and switch on again after it.

Clealy the person using those tags knows HTML and wants to do manual coding within those tags.

So what about this, Richard?

Is it not reasonable to expect that the user will want a <br /> after “Bar Avenue.”, but not want one after “on Bar)”?  Parsing text unfortunately is not such a straightforward matter of “switching off” when encountering tags.

Clealy the person using those tags knows HTML and wants to do manual coding within those tags.

The only time it is honestly safe to make that assumption is with a formatting type of “None”.  However, if you find yourself only using XHTML typography for quote handling, can I recommend that instead of Auto XHTML typography that you use the XHTML Typography plugin, with the “light” setting?  I think you’ll end up with precisely what you are looking for.

Thanks for the points; the plugin will help!