URL Parameters to meet W3C validation causes incorrect _GET parameterresults

[The KwikOne]

Since to have a page validate (HTML and XHTML) you should have your
URL parameters separated by the appropriate separator. However, if an
ampersand is used it needs to be expressed as & (example: ./
index.php?a=1&B=2). But, if that is done, then the _GET or
_REQUEST array element names are incorrect. In the example the _GET
array will be:
_GET['amp;a'] = 1 (instead of _GET['a'] = 1) and _GET['amp;B'] = 2
(instead of _GET['B'] = 2).

So, how do I get around this? Is there any other way other than
looping and stripping the 'amp;' part prior to actually looking for
parameters? What do others do in this situation?

[Olaf Schinkel]

The KwikOne schrieb:

Since to have a page validate (HTML and XHTML) you should have your
URL parameters separated by the appropriate separator. However, if an
ampersand is used it needs to be expressed as & (example: ./
index.php?a=1&B=2). But, if that is done, then the _GET or
_REQUEST array element names are incorrect. In the example the _GET
array will be:
_GET['amp;a'] = 1 (instead of _GET['a'] = 1) and _GET['amp;B'] = 2
(instead of _GET['B'] = 2).

So, how do I get around this? Is there any other way other than
looping and stripping the 'amp;' part prior to actually looking for
parameters? What do others do in this situation?

You shold read:
http://www.w3.org/TR/html4/appendix/notes.html#h-B.2.2
and:
urlencode();

[Michael Fesser]

..oO(The KwikOne)

>Since to have a page validate (HTML and XHTML) you should have your
URL parameters separated by the appropriate separator. However, if an
ampersand is used it needs to be expressed as & (example: ./
index.php?a=1&B=2).

You only do this if the URL is printed to an HTML page, for example in
an 'a' element:

<a href="./?a=1&amp;B=2">foo</a>

BTW: Don't write "index.html" or "index.php" in URLs, it's not necessary
and bad style.

>But, if that is done, then the _GET or
_REQUEST array element names are incorrect.

No. The &amp; is interpreted by the browser, the server will never see
it if you click that link.

>In the example the _GET
array will be:
_GET['amp;a'] = 1 (instead of _GET['a'] = 1) and _GET['amp;B'] = 2
(instead of _GET['B'] = 2).

So, how do I get around this?

There's a bug somewhere in your code. Can you post a short code example
or page snippet which causes this problem? How does the final HTML code
for such a link look? Do you see something like

<a href="./?a=1&amp;amp;B=2">foo</a>

Notice the "double-encoding" of the ampersand. This could be one reason
for your problem.

>Is there any other way other than
looping and stripping the 'amp;' part prior to actually looking for
parameters? What do others do in this situation?

Nothing, because it shouldn't happen.

Micha

[The KwikOne]

On Jul 18, 1:54*am, Michael Fesser <[e-mail]:

.oO(The KwikOne)

Since to have a page validate (HTML and XHTML) you should have your
URL parameters separated by the appropriate separator. However, if an
ampersand is used it needs to be expressed as &amp; (example: ./
index.php?a=1&amp;B=2).

You only do this if the URL is printed to an HTML page, for example in
an 'a' element:

<a href="./?a=1&amp;B=2">foo</a>

BTW: Don't write "index.html" or "index.php" in URLs, it's not necessary
and bad style.

But, if that is done, then the _GET or
_REQUEST array element names are incorrect.

No. The &amp; is interpreted by the browser, the server will never see
it if you click that link.

In the example the _GET
array will be:
_GET['amp;a'] = 1 (instead of _GET['a'] = 1) and _GET['amp;B'] = 2
(instead of _GET['B'] = 2).

So, how do I get around this?

There's a bug somewhere in your code. Can you post a short code example
or page snippet which causes this problem? How does the final HTML code
for such a link look? Do you see something like

<a href="./?a=1&amp;amp;B=2">foo</a>

Notice the "double-encoding" of the ampersand. This could be one reason
for your problem.

Is there any other way other than
looping and stripping the 'amp;' part prior to actually looking for
parameters? What do others do in this situation?

Nothing, because it shouldn't happen.

Micha

Micha,
Actually I found that this problem only occurred when doing a
window.open in javascript;
such as window.open('/x.php?a=1&amp;b=B', etc);
So, I have found my workaround (personally I think it is a bug within
javascript in Firefox v3
since it did not seem to see failures when IE v7 was used).
So, instead of window.open('/x.php?a=1&amp;b=B', etc);
I now have it as: window.open('/x.php?a=1&'+'b=B', etc);
and it works OK and validation does not fail.

[The KwikOne]

On Jul 17, 5:19*pm, Olaf Schinkel <[e-mail]:

The KwikOne schrieb:Since to have a page validate (HTML and XHTML) you should have your

URL parameters separated by the appropriate separator. However, if an
ampersand is used it needs to be expressed as &amp; (example: ./
index.php?a=1&amp;B=2). But, if that is done, then the _GET or
_REQUEST array element names are incorrect. In the example the _GET
array will be:
_GET['amp;a'] = 1 (instead of _GET['a'] = 1) and _GET['amp;B'] = 2
(instead of _GET['B'] = 2).

So, how do I get around this? Is there any other way other than
looping and stripping the 'amp;' part prior to actually looking for
parameters? What do others do in this situation?

You shold read:http://www.w3.org/TR/html4/appendix/notes.html#h-B.2.2
and:
urlencode();

It states QUOTE...
For example, to use the URI "http://host/?x=1&y=2" as a linking URI,
it must be written <A href="http://host/?x=1&y=2"or <A
href="http://host/?x=1&amp;y=2">
END QUOTE
That is exactly what I stated above. However, it is now resolved, (see
my response to micha)

[John Dunlop]

The KwikOne:

Actually I found that this problem only occurred when doing a
window.open in javascript;
such as window.open('/x.php?a=1&amp;b=B', etc);

Context is everything. Character references are not recognised in the
script element, so the "amp;" is part of the URL. Compare attributes
such as href, in which character references are recognised. See HTML
4.01 sec. 6.2:

http://www.w3.org/TR/html401/types.html#type-cdata

So, I have found my workaround (personally I think it is a bug within
javascript in Firefox v3
since it did not seem to see failures when IE v7 was used).

Firefox gets it right; Internet Explorer gets it wrong.

--
Jock

[Michael Fesser]

..oO(The KwikOne)

>Actually I found that this problem only occurred when doing a
window.open in javascript;
such as window.open('/x.php?a=1&amp;b=B', etc);
So, I have found my workaround (personally I think it is a bug within
javascript in Firefox v3

Depends.

In an external JS file character references like &amp; would not be
resolved, but they will be resolved within a 'script' element if the
document is written and delivered as XHTML.

In normal HTML and XHTML delivered as text/html this won't happen, so
IMHO the behaviour of Firefox is correct.

>since it did not seem to see failures when IE v7 was used).
So, instead of window.open('/x.php?a=1&amp;b=B', etc);
I now have it as: window.open('/x.php?a=1&'+'b=B', etc);
and it works OK and validation does not fail.

OK.

Micha

[The KwikOne]

On Jul 23, 9:11*am, [e-mail] wrote:

On Tue, 22 Jul 2008 14:06:38 -0700 (PDT), The KwikOne <[e-mail]>
wrote:

On Jul 17, 5:19*pm, Olaf Schinkel <[e-mail]:

The KwikOne schrieb:Since to have a page validate (HTML and XHTML) you should have your
URL parameters separated by the appropriate separator. However, if an
ampersand is used it needs to be expressed as &amp; (example: ./
index.php?a=1&amp;B=2). But, if that is done, then the _GET or
_REQUEST array element names are incorrect. In the example the _GET
array will be:
_GET['amp;a'] = 1 (instead of _GET['a'] = 1) and _GET['amp;B'] = 2
(instead of _GET['B'] = 2).

So, how do I get around this? Is there any other way other than
looping and stripping the 'amp;' part prior to actually looking for
parameters? What do others do in this situation?

You shold read:http://www.w3.org/TR/html4/appendix/notes.html#h-B.2.2
and:
urlencode();

It states QUOTE...
For example, to use the URI "http://host/?x=1&y=2" as a linking URI,
it must be written <A href="http://host/?x=1&y=2"or <A
href="http://host/?x=1&y=2">
END QUOTE
That is exactly what I stated above. However, it is now resolved, (see
my response to micha)

I dont think the W3C people even know the diference between a URI and URL

A matter of symantics only since a URL is a linked URI (or link to the
URI)

[The KwikOne]

On Jul 23, 5:09*am, Michael Fesser <[e-mail]:

.oO(The KwikOne)

Actually I found that this problem only occurred when doing a
window.open in javascript;
such as window.open('/x.php?a=1&amp;b=B', etc);
So, I have found my workaround (personally I think it is a bug within
javascript in Firefox v3

Depends.

In an external JS file character references like &amp; would not be
resolved, but they will be resolved within a 'script' element if the
document is written and delivered as XHTML.

In normal HTML and XHTML delivered as text/html this won't happen, so
IMHO the behaviour of Firefox is correct.

since it did not seem to see failures when IE v7 was used).
So, instead of window.open('/x.php?a=1&amp;b=B', etc);
I now have it as: window.open('/x.php?a=1&'+'b=B', etc);
and it works OK and validation does not fail.

OK.

Micha

Micha,
Well, one of them (IE or FF) gets it wrong (with FF amp; was prefixed
to the argument name, and IE worked OK) and when I split up the url as
above they both work OK. (PS: the page was XHTML 1.0 Transitional)
And, the original problem was noted within an onclick (<img ...
onclick="javascript:window.open( ..."), and
also noticed within an inline script element function. Whether or not
it works in a linked script is
unknown (I have not had the time to set up a test case).