[Htmlparser-developer] Need Advice - peculiar bug

[Somik R. <so...@ya...>]

Hi Folks,
    As you know, Cedric Rosa has been giving some nice bug reports - and =
I was working on those today. A major problem has been solved-  we can =
now parse tags which are incorrectly ended, and those that have tags in =
inverted commas in them.   =20
    However, one problem remains. Although parsing doesent crash, when =
the tag was incorrectly ended, I am removing all the inverted commas =
from it, simply bcos I dont know which inverted comma was the wrong one.
e.g.
<Meta name=3D"sdsd" value=3D"sdsds"">
<Meta name=3D"sdsd" value=3D"sdsd"sds">
<Meta name=3D"sadd" value=3D"sdsd " sdsd  sds ">

This leads to complications. In the third case, if all inverted commas =
are removed, parseParameters cannot pick up the entire string in value =
(because of the spaces). The parser needs to be intelligent enough to =
know which inverted comma was the erroneous one - the same way as we can =
tell.

Can anyone suggest some logic that we could formulate to do this ?

Regards,
Somik