calling functions from cgi

[Tony Rice]

If I've loaded a set of functions in a webdocument like this:
<script language="JavaS cript" type="text/javascript"
src="/js/foo.js"</script>

Shouldn't I be able to call a function from inside foo.js later in that
same document by doing something like this:

<script language="JavaS cript" type="text/javascript">myf unction(myargs)
</script>

Even though it's generated from a CGI script? The all loaded javascript
files have the same scope do they not? It's not like the functions in
foo.js available only to other functions within that file? Right?

[RobG]

Tony Rice wrote:

RobG <rg***@iinet.ne t.au> wrote in
news:qG******** *********@news. optus.net.au:

Tony Rice wrote:

RobG <rg***@iinet.ne t.au> wrote in news:zvBDf.251$ Nv4.46029
@news.optus. net.au:

That sounds like you are trying to access elements before the page
has created them, either onload or after the AJAX stuff.

Code snippet? URL?
I'm using jah (http://microformats.org/wiki/rest/ahah) to load URLs
inline dynamically. The perl script looks like this:

my $ajaxcall="jah( 'myotherscript. cgi?myarg=arg', 'contentdiv')";
if (param('myparam ')){
print "<script type=\"text/javascript\">$a jaxcall</script>\n";

}else{
print a({href=>"javas cript:$ajaxcall "}, 'click to load');
}
print "<div id='contentdiv' ></div>\n";
Don't post server code, post what your client gets - this is a
JavaScript news group, not Perl :-)

What content is written to the script element? What is written as the
value of the A element's href attribute? What is the surrounding
HTML?

I have no idea what is ending up in your A element attributes, so it's
hard to help. Use view source on the client to get the actual HTML
that is delivered to your page (a minimal test case will do).

Sorry, thought it was obvious from that code snippet and it's important
to show how the javascript is generated differently depending on the
context.

It's never safe to assume what the client gets from the server based on
server code. Only by posting what the client gets can the code be
debugged with any certainty.

Use view source, or one of the 'view innerHTML' bookmarklets, e.g. copy
and paste the following as a singe line in the address:

javascript:'<co de>'+(document. documentElement ||
document.body). innerHTML.repla ce(/&/g,%22&amp;%22). replace(/</g,
%22&lt;%22).rep lace(/%20%20/g, %22&nbsp;%20%22 ).replace(/\n/g,
%22<br>%22)+'<\/code>';

The problem has been fixed by putting

<A HREF="javascrip t:jah('somethin g.cgi', 'contentdiv')>c lick to load</A>

inside <div id='contentdiv' > rather than outside.

I have no idea why this fixed it. This is the only change made but now
I'm getting no further errors from getElementById when calling the ahah
function directly.
There is no ahah function, there is jah() and jahDone() (which is called
by jah()).

Until you post the page source that the client gets, is anyone's guess
as to why it now 'works'. jahDone() should replace the content of
'contentdiv' with whatever is sent back from the server, so your link
will be replaced.
[...]

I'm getting the following errors:

document.get ElementById(tar get) has no properties

The value of 'target' should be the ID of an element on your page. If
the page hasn't finished loading (unlikely), or the ID doesn't exist
(much more likely), then you will get the error you get.

I checked that angle out as well. didn't matter where the target div was
placed in relation ship to the ahah call.

The ahah code runs from a click (as far as I can tell) therefore the
page /should/ be loaded and contentdiv should be able to be anywhere in
the page.

Not sure why it works not, but at least it works.

Unless you post the page source actually received by the client...
--
Rob