Re: Tidy bug

Went a bit overboard yesterday (did not get my cold undead fingers
pried from the machine, that is) and the latest local does it better,
i.e. adheres more to the original preferences collected in this
thread: some stuff now remains /after/ the table instead of being
thrown up front.
Okay, those preferences were a majority vote of 1- or 2-and-a-half
(counting myself as .5 since I was dithering) against ...? , but it's
nice to see a bit more 'sensible' output anyhow.

Since I have the problem that my copy contains several changes that,
ah, blend a bit, extracting a clean diff is sort of a rather-not,
until I really trust what I got is final, that is.

Would you opt for the alternative way to test pre-alpha code -- that
is if you or anyone wishes to test this -- where you get a distro
(tar.gz or some such) emailed or URLed, which can then be built with
./configure && make , ye gods willing?

(OT FYI: extra edits in there: preliminary <meta charset> decoding
support through some wicked stream rewinding + svg/math(MathML)/canvas
tag support + a few odds an' ends)


./test/inputs/
has been enhanced with several test files to see what's going on (
in_2900790-[1-7].html )


Side note (I do those wholesale, it seems ;-) ): tidy acts (acted)
quite a bit different when encountering 'really' empty nodes versus
nodes which contain comment(s). Compare tidied output for

<table><tr></tr></table>

versus

<table><tr><!-- ook! --></tr></table>

latter one doesn't get inferred <td>s in CVS HEAD, while former does.
This has also been taken care of. node->content=3D=3DNULL does not equal
'empty node' any more.



Sigh. Now if I could find a reliable way to introduce a wee bit of
mandatory-yet-not-&nbsp; whitespace in place of the <tr> and </tr>
here:

word0<table><tr>word1</tr>word2

versus

word0
<table>
<tr>word1</tr>
word2

I would be overjoyed. Bloody CleanWhitespace() is squeezing my potatoes. :-(




On Mon, Nov 23, 2009 at 6:59 AM, Ben McCann <[email protected]> wrote:
> Thanks Ger! =A0I really appreciate you taking the time to look at this. =
=A0I
> agree I would not expect the=A0extra 'gobbledigook' text to land before t=
he
> table. =A0And if the <tr>content</tr> is going to be turned into a full t=
able
> then I imagine I would expect the content to remain within the table. =A0I
> think what you've changed thus far is an improvement over the previous
> functioning though since we are not losing any data as before (the table
> id).
> Thanks!
> Ben



-- =

Met vriendelijke groeten / Best regards,

Ger Hobbelt

--------------------------------------------------
web:    http://www.hobbelt.com/
        http://www.hebbut.net/
mail:   [email protected]
mobile: +31-6-11 120 978
--------------------------------------------------

---------------------------------------------------------------------------=
---
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day =

trial. Simplify your report design, integration and deployment - and focus =
on =

what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july