Hi Vincent, On Wed, Nov 26, 2008 at 1:49 PM, Vincent Massol <[email protected]> wrote:

On Nov 26, 2008, at 4:13 AM, Asiri Rathnayake wrote:

...

Hi Vincent,

On Tue, Nov 25, 2008 at 9:07 PM, Vincent Massol <[email protected]> wrote:

...

Guillaume,

I can see some other problems: * title is not correct

Did you mean the title content (spellings?) or the font size of the title? If it's the latter, it's because openoffice server in this case has used <font size=""></font> for sizing, and this is filtered in the strict filtering mode.

A title should be converted to a title (i.e. using the == syntax in wiki).

Wasn't this a real title in the original word doc?

It seems not. If the user editing the document was using headings instead of font sizes, this would result in a title.

...

I'm pretty sure you didn't mean the latter one, if so please explain what's exactly wrong :)

* lots of empty lines

Good point.

If any of you have some free time (which excludes vincent ;)) please try this :

Open a new text document in openoffice and:

1. Type one line of text.

2. Press enter twice and type another line of text.

So you'd have something like :

" Hi, This is the first line

And this is the second line"

3. Save the document as *html*.

4. Open the document in firefox and see the result.

The same thing happens when converting documents via openoffice server. The problem is, for each empty line in a word document, open office uses :

<p style="margin-bottom: 0in"><br/></p>

What is the effect of "margin-bottom: 0in"? If it removes the new line then we should check for this and remove the paragraph altogether (for all filterings IMO).

Well it is supposed to, but it doesn't work in either FireFox or IE. But anyway, since it is supposed to mean there is no newline in between, we could remove such paragraphs. But consider the following case : <p style="margin-bottom: 0in; background-color: red;" >Paragraph-1</p> <p>Paragraph-2</p> In this case if we blindly remove the first paragraph, it will loose other styles along with it. So we should only filter such paragraphs in strict/moderate filtering scenarios. In non-restricted mode, we can't do the same. Again, I'm thinking whether this filtering belongs to office-importer or xhtml parser ? WDYT ?

...

For spacing. And this is obviously not correct. May be I should take this to openoffice guys ?

Yes you should but in the meantime we can filter this (only when margin-bottom is 0).

...

Vincent, if you remember, I tried to replace these empty paragraphs with <br/> elements. But I didn't mention this to you at that time.

Yes you didn't mention that there was a CSS with margin-bottom: 0in which changes everything...

Well I saved the best for the end ;) Thanks. - Asiri

-Vincent

...

I simply didn't like these empty paragraphs ;)

Thanks.

- Asiri

...

Thanks -Vincent

On Nov 25, 2008, at 4:30 PM, Guillaume Lerouge wrote:

...

Hi Asiri,

I tested strict filtering with my usual test document and apparently something went wrong with links (see the bottom of the page): http://91.121.237.216/xwiki/bin/view/Main/GuillaumeTestProgrammez

Apart from this, it looks pretty good.

Guillaume

On Tue, Nov 25, 2008 at 4:14 PM, Asiri Rathnayake < [email protected]> wrote:

...

Hi Devs, Users,

I have uploaded yet another new version of office-importer @ http://91.121.237.216/xwiki/bin/view/Import/.

This version utilizes the latest rendering module which has improved a lot since the last upload.

Also a moderate style filtering option is introduced.

Please try it out and let us know what you think.

Thanks.

- Asiri

---

users mailing list [email protected] http://lists.xwiki.org/mailman/listinfo/users