Re: How Does JavaScript Call Forth CSS??

dhtml wrote:

On Apr 30, 6:59 pm, Prisoner at War <prisoner_at_... (AT) yahoo (DOT) com> wrote:
On Apr 27, 2:20 pm, "david.karr" <davidmichaelk... (AT) gmail (DOT) com> wrote:
It's probably best to use reusable code for operations like this, and
fortunately all of the major JS frameworks include this. The
following URL provides a good summary of this:
http://www.openjs.com/scripts/dom/cl...nipulation.php

Wow, I guess libraries are still using capturing groups. I thought it
was pretty common knowledge to use non-capturing groups. They've got
that right in the tutorial and I can even see in Prototype JS:
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Therefore, chances are it is a really bad idea to do that. In fact, it
turns out that non-capturing parentheses are not supported before JavaScript
1.5, JScript 5.5, ECMAScript Ed. 3. Use them and your script will break in
IE before version 5.5, for example.

A capturing group stores the whitespace:
(\\s+|$)

A non-capturing group does not:
(?:\\s+|$)

True, after a fashion.

It is less efficient to use a capturing group.

Nonsense. The tiny amount of memory saved by not storing the backreference
is bought with the tiny amount of runtime that is required more to parse the
non-capturing expression.


F'up2 cljs

PointedEars
--
Use any version of Microsoft Frontpage to create your site.
(This won't prevent people from viewing your source, but no one
will want to steal it.)
-- from <http://www.vortex-webdesign.com/help/hidesource.htm>

On May 1, 5:03 pm, dhtml <dhtmlkitc... (AT) gmail (DOT) com> wrote:

On Apr 30, 6:59 pm, Prisoner at War <prisoner_at_... (AT) yahoo (DOT) com> wrote:

On Apr 27, 2:20 pm, "david.karr" <davidmichaelk... (AT) gmail (DOT) com> wrote:

It's probably best to use reusable code for operations like this, and
fortunately all of the major JS frameworks include this. The
following URL provides a good summary of this:

http://www.openjs.com/scripts/dom/cl...nipulation.php

Wow, I guess libraries are still using capturing groups. I thought it
was pretty common knowledge to use non-capturing groups. They've got
that right in the tutorial and I can even see in Prototype JS:

removeClassName: function(element, className) {
if (!(element = $(element))) return;
element.className = element.className.replace(
new RegExp("(^|\\s+)" + className + "(\\s+|$)"), ' ').strip();
return element;
}

A capturing group stores the whitespace:
(\\s+|$)

A non-capturing group does not:
(?:\\s+|$)

It is less efficient to use a capturing group.
A capturing group increases the count of the backreference number, so
for more complicated expressions, it would be harder to keep track of
which group is which number, e.g. \1 or $1.

It would actually be possible to make the Prototype JS function even
less efficient by adding the batch() functionality of YUI. But I'd
have to day they've done a pretty bad for a major library. The
assignment in the conditional, the extra call to $(), the capturing
regexp (^|\\s+), the misnaming of trim() as "strip()" (and the adding
of this misnamed trim() function to String.prototype).

At first guess, it seems that Prototype JS's "strip()" is a
normalize() function, but it's not. It's a trim() function, misnamed.
trim() (or strip()) doesn't wipe out the whitespace in the middle, so
you end up with "foo bar baz quux" -> remove "bar" -> "foo baz quux" -

remove "baz -> "foo quux". The add

It's quite surprising to see such code in such a popular library, at
least to me.

In takinga look at jQuery, i find another algorithm that is even
slower. So I have to take back what I said above about "the only thing
that could make the Prototype JS function slower". I didn't think
about the jQuery approach. Create a method "addClass" that delegates
to className.add, which splits the classname to an array, then uses an
each() function, then inside that function, check to make sure the
nodeType is 1.

Here it is:http://code.jquery.com/jquery-latest.js

className: {
// internal only, use addClass("class")
add: function( elem, classNames ) {
jQuery.each((classNames || "").split(/\s+/), function(i, className)
{
if ( elem.nodeType == 1 && !jQuery.className.has( elem.className,
className ) )
elem.className += (elem.className ? " " : "") + className;
});
},

The function is intended to be called on a Decorated collection, which
would mean an extra call to each(). In most cases, that could be
replaced by adding a fragment identifier to the css file, and then
adding a class to the ancestor.

.window .link {}

/* adding a class to .window object's changes the links */
.window-raised .link { color: red }

/* jQuery tends to encourage this instead */
.window .link-active { color: red }

Cool, thanks! I'd actually managed to google up JQuery but I wasn't
sure about JavaScript "libraries" and all that....

I'm not so sure about those JavaScript libraries, either.

Assignment in conditional is done to extend ('normalize') an element
(so that you can simply pass string into Element#removeClassName) and
"return" if it's not a valid element. Such "normalization" is
consistent throughout a library. Capturing groups, afaik, are not
supported by some of the browsers that Prototype supports. #strip is
named after Ruby's method - as you might know, prototype follows its
conventions in certain cases.

Best,
kangax

On May 1, 6:34 pm, kangax <kan... (AT) gmail (DOT) com> wrote:

On May 1, 5:03 pm, dhtml <dhtmlkitc... (AT) gmail (DOT) com> wrote:



On Apr 30, 6:59 pm, Prisoner at War <prisoner_at_... (AT) yahoo (DOT) com> wrote:

On Apr 27, 2:20 pm, "david.karr" <davidmichaelk... (AT) gmail (DOT) com> wrote:

It's probably best to use reusable code for operations like this, and
fortunately all of the major JS frameworks include this. The
following URL provides a good summary of this:

http://www.openjs.com/scripts/dom/cl...nipulation.php

Wow, I guess libraries are still using capturing groups. I thought it
was pretty common knowledge to use non-capturing groups. They've got
that right in the tutorial and I can even see in Prototype JS:

removeClassName: function(element, className) {
if (!(element = $(element))) return;
element.className = element.className.replace(
new RegExp("(^|\\s+)" + className + "(\\s+|$)"), ' ').strip();
return element;
}

A capturing group stores the whitespace:
(\\s+|$)

A non-capturing group does not:
(?:\\s+|$)

It is less efficient to use a capturing group.
A capturing group increases the count of the backreference number, so
for more complicated expressions, it would be harder to keep track of
which group is which number, e.g. \1 or $1.

It would actually be possible to make the Prototype JS function even
less efficient by adding the batch() functionality of YUI. But I'd
have to day they've done a pretty bad for a major library. The
assignment in the conditional, the extra call to $(), the capturing
regexp (^|\\s+), the misnaming of trim() as "strip()" (and the adding
of this misnamed trim() function to String.prototype).

At first guess, it seems that Prototype JS's "strip()" is a
normalize() function, but it's not. It's a trim() function, misnamed.
trim() (or strip()) doesn't wipe out the whitespace in the middle, so
you end up with "foo bar baz quux" -> remove "bar" -> "foo baz quux" -

remove "baz -> "foo quux". The add

It's quite surprising to see such code in such a popular library, at
least to me.

In takinga look at jQuery, i find another algorithm that is even
slower. So I have to take back what I said above about "the only thing
that could make the Prototype JS function slower". I didn't think
about the jQuery approach. Create a method "addClass" that delegates
to className.add, which splits the classname to an array, then uses an
each() function, then inside that function, check to make sure the
nodeType is 1.

Here it is:http://code.jquery.com/jquery-latest.js

className: {
// internal only, use addClass("class")
add: function( elem, classNames ) {
jQuery.each((classNames || "").split(/\s+/), function(i, className)
{
if ( elem.nodeType == 1 && !jQuery.className.has( elem.className,
className ) )
elem.className += (elem.className ? " " : "") + className;
});
},

The function is intended to be called on a Decorated collection, which
would mean an extra call to each(). In most cases, that could be
replaced by adding a fragment identifier to the css file, and then
adding a class to the ancestor.

.window .link {}

/* adding a class to .window object's changes the links */
.window-raised .link { color: red }

/* jQuery tends to encourage this instead */
.window .link-active { color: red }

Cool, thanks! I'd actually managed to google up JQuery but I wasn't
sure about JavaScript "libraries" and all that....

I'm not so sure about those JavaScript libraries, either.

Assignment in conditional is done to extend ('normalize') an element
(so that you can simply pass string into Element#removeClassName) and
"return" if it's not a valid element. Such "normalization" is
consistent throughout a library. Capturing groups, afaik, are not
supported by some of the browsers that Prototype supports. #strip is

Non-capturing groups that is.

@dhtml

remove "baz -> "foo quux".

I believe your assertion is false, it still only leaves 1 space.
The pattern matches the beginning of the string OR one or more
<spaces> then the className then the end of the string OR one of more
<spaces> after.
This would remove <space>baz<space> and replace with <space> turning
the className value into "foo<space>quux".

Info not directed at anyone:
Prototype doesn't officially support IE 5.5
http://www.w3schools.com/browsers/browsers_stats.asp
At some point ya have to let that version go

- JDD