[Imps] reconstruct the active formatting elements

Sat Dec 23 02:06:33 PST 2006

On Sat, 23 Dec 2006 08:01:00 +0100, Ian Hickson <ian at hixie.ch> wrote:
>> Is there any testcase that could uncover this bug?
>
>    <p><b><i><u></p>
>    <p>X

So we pass "X" but not X" it seems.
Given that the latter actually does generate:

<html>
    <head>
    <body>
      <p>
        <b>
          <i>
            <u>
      " "
      <p>
        <b>
          <i>
            <u>
              "X"

as you seem to suggest.

> ...should get the same parse tree as:
>
>    <p><b><i><u></u></i></b></p>
>    <p><b><i><u>X
>
>
>> Or is it a bug in the specification? Note that if we change it to "entry
>> not in self.openElements" several testcases will no longer pass and
>> exceptions will be thrown for trying to access an element not in the
>> list...
>
> The error in the spec is that it walks off the start of the list, I  
> guess. I should add a clause that covers that case. Fixed (by adding  
> some steps).

Ok, I'll rewrite our algorithm to match that and see if it helps, thanks!

-- 
Anne van Kesteren
<http://annevankesteren.nl/>
<http://www.opera.com/>