JS works in Safari, but not FF - DOM error

[sdustinh]

I've got a problem with a script that is working perfectly in Safari, but won't work in FF. The error it spits out is

Exception ``[Exception... "Node was not found" code: "8" nsresult: "0x80530008 (NS_ERROR_DOM_NOT_FOUND_ERR)" location: "https://***.net/~dhaggard/portal/js/checkPartition.js Line: 355"]'' thrown from function deletePartition(aPartID=string:"hdpart-1/1243", partHDrive=string:"hdpart-1") in <https://***.net/~dhaggard/portal/js/checkPartition.js> line 355.

Exception ``[Exception... "Node was not found" code: "8" nsresult: "0x80530008 (NS_ERROR_DOM_NOT_FOUND_ERR)" location: "https://***.net/~dhaggard/portal/js/checkPartition.js Line: 355"]'' thrown from function (null)() in <javascript:deletePartition('hdpart-1/1243','hdpart-1')> line 1.

Here is the function I'm trying to call...

function deletePartition (aPartID, partHDrive) {
 
    var oldPartHD = document.getElementById(partHDrive);
    var oldPartition = document.getElementById(aPartID);
 
    oldPartHD.tBodies[0].removeChild(oldPartition);
 
}

I'm trying to convert a script written in Safari over to work in FF/IE. I've been using http://www.quirksmode.org/dom/w3c_core.html that page to make sure that all the functions I used are even supported, but I'm still getting a couple of errors. Does anyone know what may be a fix for this? Thanks!

[acoder]

I'm assuming the removeChild line is causing the error. Can you show the table with id "hdpart-1" and tr with id "hdpart-1/1243"?

[sdustinh]

Sure...

<table id="hdpart-1">
    <tbody>
        <tr id="hdpart-1/1234">
            <td>Cell 1</td>
            <td>Cell 2</td>
            <td>Cell 3</td>
            <td>Cell 4</td>
            <td>Cell 5</td>
            <td>Cell 6</td>
            <td>Cell 7</td>
            <td>Cell 8</td>
            <td>Cell 9</td>
        </tr>
    </tbody>
</table>

It's the removeChild() that's generating the error, though I don't know if that's what's causing it. All of the errors I've generated, while tinkering with it, have been the result of FF not being able to find the node to remove. Safari, however, can find it. So I'm fairly certain it's an easy fix, I just can't seem to figure it out.

Thanks for the help!

[gits]

hi ...

is oldPartition ok? i mean is the node found before you try to remove it?

kind regards

[mrhoo]

strictly speaking, oldPartHD.tBodies[0].firstChild is a newline text node, not a tr element, and firefox is pretty strict with nodes.

[gits]

strictly speaking, oldPartHD.tBodies[0].firstChild is a newline text node, not a tr element, and firefox is pretty strict with nodes.

hmmm ... that was my first thought too ... but firstChild isn't used here?

[mrhoo]

Your right gits- I think the trauma of working with a function named removePartition was too much for me!
Could it be that firefox isn't honoring an id with a slash in it?
"hdpart-1/1234">

[gits]

Your right gits- I think the trauma of working with a function named removePartition was too much for me!
Could it be that firefox isn't honoring an id with a slash in it?
"hdpart-1/1234">

hmmm ... i tested it ... and it seems to be working with FF 2.0.0.8 ... strange problem indeed?

[mrhoo]

I could remove all but the last row in a table in firefox .
When the tBody.getElementsByTagName('tr') length is 1 I had to remove the tbody itself from the table.

Edit- nope, this works-

var T=document.getElementsByTagName('table')[0];
var TB=T.tBodies[0];
while(TB.lastChild)TB.removeChild(TB.lastChild)

removes all the trs and text nodes and leaves an empty tbody.

It is a puzzle.

[drhowarddrfine]

Could it be that firefox isn't honoring an id with a slash in it?
"hdpart-1/1234">

You are right. Slashes are not allowed in id names but I don't know what FF does when it encounters it.

[gits]

You are right. Slashes are not allowed in id names but I don't know what FF does when it encounters it.

hi ...

without a doctype (in quirksmode) FF found the element by id ... may be with a doctype it could be a problem ... FF could be more strict with the code? Why is Safari doing it? as i said ... seems to be strange ...

@sdustinh: do you use a doctype? ...

kind regards

[drhowarddrfine]

With a doctype, Firefox is not being more strict, it's just following the rules of the 'document type definition'. Quirks mode, after all, is just an IE bug and not a real 'mode'. So whenever you code to a bug, you never know what will happen in any browser, such as Safari in this case.

Not that I'm saying that is the problem here, but a doctype is required of all modern web pages