alfonsoml wrote:The support for THead has been added in http://dev.fckeditor.net/ticket/2430#comment:10
Now there's a new patch to support TH in the cells dialog in http://dev.fckeditor.net/ticket/822 that needs some testing.

<table width="200" cellspacing="1" cellpadding="1" border="1">
    <tbody>
        <tr>
            <th style="">col 1</th>
            <th style="">Col 3</th>
            <th style="">Col 4</th>
        </tr>
        <tr>
            <td>Col 5</td>
            <td>Col 6</td>
            <td>Col 7</td>
        </tr>
        <tr>
            <td>Col 8</td>
            <td>Col 9</td>
            <td>Col 10</td>
        </tr>
    </tbody>
</table>

No.
2430 has been already applied to the trunk, so you should update your local copy and it will be applied.

Why should it?
you have selected a cell and mark it as a header or data. The rest of the cells will be whatever you want.

<table cellspacing="1" cellpadding="1" width="200" border="1">
    <tbody>
        <tr>
            <th>&nbsp;c</th>
            <th>&nbsp;c</th>
            <th>&nbsp;c</th>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
    </tbody>
</table>

<table cellspacing="1" cellpadding="1" width="200" border="1">
    <tbody>
        <tr>
            <th colspan="2">&nbsp;c&nbsp;c</th>
            <th>&nbsp;c</th>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
    </tbody>
</table>

<table cellspacing="1" cellpadding="1" width="200" border="1">
    <tbody>
        <tr>
            <th>&nbsp;c&nbsp;c</th>
            <td>&nbsp;</td>
            <th>&nbsp;c</th>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
    </tbody>
</table>

Sounds good. I agree about IE selection behavior. It is one of those "nice to have" lists but may not be possible to include everything.

I ran some more tests and didn't see any issues. The patch looks good.

alfonsoml wrote:Did you filed the ticket about splitting THs?

https://dev.fckeditor.net/ticket/2472

alfonsoml wrote:I wanted to fix next #1865, but it's a little complex, so for the moment I've attached a patch to fix http://dev.fckeditor.net/ticket/2472 and it does include some other little simplifications in the code.

Issue #1:



Result:

Issue #2:

https://dev.fckeditor.net/ticket/2486
https://dev.fckeditor.net/ticket/2487

Any chance of including drag-adjustable column and row borders?

alfonsoml, yes, just like that. I saw it in one version I downloaded, but can no longer find it. Users really like that capability.

alfonsoml wrote:Thanks for the testing.
I guess that I might fix them as part of #1865 because I think that it might need a great review of all the code involved.

https://dev.fckeditor.net/ticket/2486

https://dev.fckeditor.net/ticket/2486

The main problem is that in order to fix other bugs (that are triggered more easily like the context-menu at last row) I think that the _CreateTableMap and _InstallTableMap might need several revisions, that are more important than this issue at hand.

I can see why this is an issue now that I have been in the code for a while. Bear with me but I have a question about the create/install tablemap functions. What is the main reason driving the need for these functions in the code? Given that we already have access to the table DOM object, it seems to me that creating a map out of it is redundant.

I have been working on the fcktablehandler.js for a few days now and am trying to figure out if the create/install tablemap can be eliminated and re-write the code to work directly at the DOM level. So far I have been able to make the FCKTableHandler.HorizontalSplitCell function independent of the create/install tablemap function and it seems to be working pretty well. The advantage with this approach is that working with theads, tbodies and so on becomes easier (just my opinion). Again, pardon my ignorance if there is something missing, but is this option even worth looking into?

I would really appreciate some feedback before I go deeper into.

Thanks,
Shri

alfonsoml wrote:I guess that when Martinkou coded the new table code it seemed the best option at the moment. If you think about tables as holding only one tbody, then they are a matrix, so using a function to get the details about that matrix and handling that way all the rowspan and colspan seems indeed like a very good idea.

But when there are several of those matrix in the table, then we might need to review the code and check if it can be improved to handle properly the new situations, so if you think that you can improve the code go ahead, try to work little by little and that way we might find the code that works.

https://dev.fckeditor.net/ticket/1865

I'm sorry about not reviewing still your previous patch, but these days I'm busy and in order to test it I want to have some time to focus on it.

http://www.fckeditor.net/forums/viewtopic.php?f=6&t=11430

<p>This is some <strong>sample text</strong>. You are using <a href="http://www.fckeditor.net/">FCKeditor</a>.</p>

<p>This is some <strong>sample text</strong>.</p>
<table width="200" cellspacing="1" cellpadding="1" border="1">
    <tbody>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
    </tbody>
</table>
<p>You are using <a href="http://www.fckeditor.net/">FCKeditor</a>.</p>

<div>This is some <strong>sample text</strong>. You are using <a href="http://www.fckeditor.net/">FCKeditor</a>.</div>

<div>This is some <strong>sample text</strong>. </div>
<table width="200" cellspacing="1" cellpadding="1" border="1">
    <tbody>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
        <tr>
            <td>&nbsp;</td>
            <td>&nbsp;</td>
        </tr>
    </tbody>
</table>
<div>You are using <a href="http://www.fckeditor.net/">FCKeditor</a>.</div>

http://dev.fckeditor.net/ticket/1865

Well, considering the captions ... i've to make some adjustments to my code, since if there is one it's the firstchild of the table and not thead.

http://dev.fckeditor.net/ticket/1865

There are still some questions left that need to be answered before the handling of tables can be further improved.

1. Should it be possible to merge down cells from thead to tbody?
2. If so, should the merged cell be in thead or in tbody?
3. Take a default 3 * 2 table, split one cell horizontaly and merge it back to one cell. All cells in that column have a colspan="2". Should that stay that way or should the colspan be lowered (and in this case: removed)?
4. Should the merging of a th-cel with a td-cell result in a th-cell (it does at the moment) or a td-cell?

kwillems wrote:
1. Should it be possible to merge down cells from thead to tbody?
2. If so, should the merged cell be in thead or in tbody?
4. Should the merging of a th-cel with a td-cell result in a th-cell (it does at the moment) or a td-cell?

I believe those represent different groups of cell and the merging operations should be limited to a single group.

kwillems wrote:
3. Take a default 3 * 2 table, split one cell horizontaly and merge it back to one cell. All cells in that column have a colspan="2". Should that stay that way or should the colspan be lowered (and in this case: removed)?

Having clean HTML is good. There is no sense having a column with all cells at the same colspan, and the same for rows with rowspan. So, I think it would be nice if the splitting operations would lowerdown *span attributes, finally removing them at zero.