Intelligent and more correct Html `addCssClass`

[kartik-v]

I will be submitting a PR for improving and making the Html::addCssClass more intelligent and more practical to use for all use cases - similar or better than jquery addClass.

Enhancements:

• Parse the original $options['class'] correctly irrespective of whitespaces (tabs, newlines or spaces)
• Parse the $class param correctly irrespective of whitespaces (tabs, newlines or spaces)
• If the $class property is separated by spaces... it will validate each class separately already added and
• Intelligently do not append css class if it already exists irrespective of the whitespace

Current Implementation

Example

$options = ['class' => '    kartik']; // note the whitespace
Html::addCssClass($options, ' hide    kartik  robot');
Html::addCssClass($options, "\tkartik   hide fancy    box     holy");

Output

// generated output
$options['class'] =  '    kartik  hide    kartik  robot         kartik   hide fancy    box     holy';

Proposed Implementation (After PR Change)

Example

$options = ['class' => '    kartik']; // note the whitespace
Html::addCssClass($options, ' hide    kartik  robot');
Html::addCssClass($options, "\tkartik   hide fancy    box     holy");

Output

// generated output
$options['class'] =  'kartik hide robot fancy box holy';

[cebe]

how realistic is it that this method will receive unclean input? In general improvement is good but it should not add too much overhead.

[kartik-v]

The purpose is not just for unclean input... but it currently does not correctly support adding CSS class separated by white spaces. Let's say I want to add 5 different CSS classes to an options... the current approach is to call this method 5 times.

[cebe]

sure, that should be possible.

[klimov-paul]

how realistic is it that this method will receive unclean input? In general improvement is good but it should not add too much overhead.

Agreed.

Let's say I want to add 5 different CSS classes to an options... the current approach is to call this method 5 times.

Will be resolved by #8458

Closed in favor of #6043