I heard, but cannot find any document to double check, that this technique helps reduce the computer processes with specificity of CSS, when the browser has to calculate which class to apply when a tag has IDs or Classes in the document.
This is an extract from Google Material Design Lite, which uses the technique such:
<!-- Colored FAB button -->
<button class="mdl-button mdl-js-button mdl-button--fab mdl-button--colored">
<i class="material-icons">add</i>
</button>
You will notice the class
mdl-button
This one is used as an ID basically. While
mdl-button--colored
is the ID mentioned above, with the class
colored
So, practically this is the same to:
#mdl-button.colored
With the difference that the computer wont have to process the specificity with the ID and Class, because now we only have a unique class.
So again.. the question is "how is this technique called" or if there is no name, maybe can someone post a link to documents that explain deeply the concept?
Cheers guys!
You might want to check out the Block Element Modifier methodology for CSS. The idea behind it helps you achieve reusable components and code sharing in the front-end.
Check out this link for more information:
http://getbem.com/introduction/
descendant, child or adjacent, and overly qualified selectors
Which are what Google says not to use.
Related
I have the following HTML code, I tend to have a 3 code project identifier at the start
<span class="abc-user-overview__header__title">
<span class="abc-user-overview__header__title__name">
{{ name }}
<span class="abc-user-overview__header__active">true</span>
</span>
Although the 'name' element is a child of title, how strict does the naming convention have to be? As I think abc-user-overview__header__title__name is too long and would prefer to drop the __title, giving me:
<span class="abc-user-overview__header__title">
<span class="abc-user-overview__header__name">
{{ name }}
<span class="abc-user-overview__header__active"true</span>
</span>
Is this valid and acceptable BEM?
Sadly your code is not valid according to BEM convention. This is the official name structure block-name__elem-name_mod-name_mod-val.
Here is your code with valid BEM naming:
<span class="abc-user-overview__title">
<span class="abc-user-overview__name">{{ name }}</span>
<span class="abc-user-overview__status abc-user-overview__status_active">true</span>
</span>
Few tips:
Avoid naming your blocks according to their content. Try to be generic for blocks that can be reused. For example, let us have a block that represents a list content. On one of the pages we may display News .news-list, but on other we may display Products, so reusing block with name .news-list with Products inside isnt very nice. In this case a simple class like .list will be enough.
If something can be reused, make it block, not an element.
For boolean modifiers, the value is not included in the name.
Full documentation and great examples can be found in the official website: https://en.bem.info/methodology/naming-convention/#naming-rules
Those names aren't too long. Sometimes when names get too long it's because you've got too much going on in that page. When I see that in angular (for example) I consider extracting components, then you can chop off the block and half the element name from the existing css classes and replace with a shorter block name. it makes the parent component so much easier to read when you move a lot of unrelated complexity to a different component. Good candidates for this could include
mobile menus
containers (anything with responsive widths, padding or icons to collapse or expand some region)
footers
complicated widgets that show some stats
any repeated card with a title, photo, description and hyperlink.
Some of these components might be useful for the whole application, or maybe just for that page. Etc.
I have to create a primary heading component, below is my markup along with CSS classes. I'm following BEM naming convention for class name.
I have h1 element consists of two spans. One span for main heading text, and second span for sub heading text. The main and sub are variations of my heading.
I have not specified the Element class (Which could be heading-primary__text ) and i have directly attached modifier classes to span elements.
<h1 class="heading-primary">
<span class="heading-primary--main">Video Background Option</span>
<span class="heading-primary--sub">One Page Parallax</span>
</h1>
Is that a right way to follow BEM methodology without specifying Elements classes & attaching Block's modifiers classes to Elements(span)? Because i don't need elements classes.
Is there any alternate?
While this is subjective, and as per the convinience of the project . i'd recommend doing something like this- as you already have a header-primary_text element class
<h1 class="heading-primary">
<span class="heading-primary_text">Video Background Option</span>
<!--create a modifier -->
<span class="heading-primary_text--sub">One Page Parallax</span>
</h1>
In this way you can make the sub a modifier class for the subtext.
More info can be seen here https://en.bem.info/methodology/quick-start/#modifier
Hope this helps :)
I think there is a much simpler way to do this just using basic HTML. You only want to have one h1 per page and since you said that your second span of your h1 is a "subheading" I feel like you would be way better off marking that one as an h2 instead of two spans of different context within one h1 heading! Always good to use the built in benefits of HTML first if you can.
No, it is not the right way. You cannot use block (or element) modifier alone on the HTML tag without specifying the block (or element) class itself.
Please refer to BEM documentation here: https://en.bem.info/methodology/quick-start/#modifier
A modifier can't be used alone From the BEM perspective, a modifier
can't be used in isolation from the modified block or element. A
modifier should change the appearance, behavior, or state of the
entity, not replace it.
Here is a code example from the docs:
<!-- Correct. The `search-form` block has the `theme` modifier with the value `islands` -->
<form class="search-form search-form_theme_islands">
<input class="search-form__input">
<button class="search-form__button">Search</button>
</form>
<!-- Incorrect. The modified class `search-form` is missing -->
<form class="search-form_theme_islands">
<input class="search-form__input">
<button class="search-form__button">Search</button>
</form>
You mentioned that you don't need an element class, this topic is also covered in BEM docs
https://en.bem.info/methodology/faq/#why-include-the-block-name-in-modifier-and-element-names
semuzaboi's suggestion sounds as a good alternative to me.
First of all, elements are specified after two __ like block__element_modifier.
Secondly, yes. Blocks may not have any elemenets inside, but rather have modifiers (most common case a block with lang modifiers for Internationalization (block_lang_ru))
PS as well as element may not have any modifiers inside. But block can not be nested inside another one. They should be placed inside one directory on the same level.
Is it possible to transform a phrase to plural using just CSS, based on the number within the element?
I'm guessing this breaks a presentation/data separation boundary philosophy, but just wondering if it's possible. It's easy with JavaScript, but wondering if CSS can take care of it.
<span class="plural">0 book</span>
outputs: 0 books
<span class="plural">1 book</span>
outputs: 1 book
<span class="plural">2 book</span>
outputs: 2 books
No. CSS cannot read the content of the element to determine what number it is, and CSS doesn't have if conditions.
If you would only apply .plural to items that should be regularly pluralized, you could do:
.plural:after {content: "s"}
But then we run into the problem of how English is irregular, and not all plurals are guaranteed to end in "s".
Going to try answer this myself, as it seems a pure CSS solution is not possible. I've found a small workaround here that works for me. It's not ideal, but may work for others too..
I just render the number behind the plural class name and wrap this around the s. Then I use a class for plural1 that hides the s.
This works for me because I use templates to render my html and it's easy to slot the numbers in.
.plural1 {
display: none;
}
<div>-1 book<span class="plural-1">s</span>
</div>
<div>0 book<span class="plural0">s</span>
</div>
<div>1 book<span class="plural1">s</span>
</div>
<div>2 book<span class="plural2">s</span>
</div>
Pros: This can work for other endings and other languages too. Negative numbers are supported.
Cons: Adds unused classes to your elements (plural2, plural3, etc.) that the browser will need to read and ignore. Not really an issue for a small number of object though (eg. 100).
I found many different views on many articles on presentation tags, with some people thinking all tags are presentational, but some others do not think so.
For example: in the HTML 5 specification, they do not think <small> is presentational.
In this list of tags - which are all HTML 5 supported - which tag is presentational and which is not?
<abbr>
<address>
<area>
<b>
<bdo>
<blockquote>
<br>
<button>
<cite>
<dd>
<del>
<dfn>
<dl>
<dt>
<em>
<hr>
<i>
<ins>
<kbd>
<map>
<menu>
<pre>
<q>
<samp>
<small>
<span>
<strong>
<sub>
<sup>
<var>
Who decides which HTML tag is presentational and Which is not - and how do they make that decision? Is it a particularly large group such as the W3C or is it based on groups of web developers, i.e. the web community? Also, between the two, which advice we should follow for deciding which tags are presentational?
If a tag is valid as according to the W3C in accepted doctypes, then what are the pros to not using any xhtml tag from any point of view?
in user/usability/accessibility point of view
if we use more HTML tags then pages without CSS will better.
in developer point of view
if we make use of more available tags in HTML, than we do not need to use <span class=className">
it takes more time to write and it uses more charter space than tags in HTML and CSS both.
For example:
instead of using:
<span class="boldtext">Some text<span>
.boldtext {font-weight:700}
We can use:
<b>Some text<b>
b {font-weight:700}
it looks cleaner, it is easier to use , it uses less characters - which will reduce the page size - and it is more readable in source. It also does not break the rule of content and presentation separation.
We can also do this:
<b class="important">Some text<b>
b.important {font-weight:700}
and whenever we want to change font-weight then we can change css only in both examples.
If a tag is considered valid by w3c in their recognized doctypes, then what are the pros to not using any X/HTML presentational tags which are not directly recognized by either the W3C, or by the HTML specifications?
Can we change any design parameters without changing anything in HTML? Does this fit within the meme of content and presentation separation?
If any HTML tag breaks the rule of separation, then does not the css property Content break as well?
see this article.
Why are the HEIGHT and WIDTH attributes for the IMG element permitted?. does it not break the rule of separation? A good debate on this matter can be found here.
W3C decides the semantics of tags. The specification documents of HTML5 gives conditions on the use of the various tags.
HTML5
To continue with your example, there is nothing wrong with using <b> to bold some text unless:
The text being bolded is a single entity already represented by a tag:
Incorrect:
<label for="name"><b>Name:</b></label>
Correct: (Use CSS to style the element)
label { font-weight: bold; }
<label for="name">Name:</label>
The text is being bolded to put added emphasis and weight on a section or words of a block of text.
Incorrect:
<p>HTML has been created to <b>semantically</b> represent documents.</p>
Correct: (Use <strong>)
<p>HTML has been created to <strong>semantically</strong> represent documents.</p>
The following is an example of proper use of the <b> tag:
Correct:
<p>You may <b>logout</b> at any time.</p>
I realize that there doesn't seem to be a lot of difference between the above example and the one using <strong> as the proper example. To simply explain it, the word semantically plays an important role in the sentence and its emphasis is being strengthened by bold font, while logout is simply bolded for presentation purposes.
The following would be an improper usage.
Incorrect:
<p><b>Warning:</b> Following the procedure described below may irreparably damage your equipment.</p>
Correct: (This is used to add strong emphasis, therefore use <strong>)
<p><strong>Warning:</strong> Following the procedure described below may irreparably damage your equipment.</p>
Using <span class="bold"> is markup-smell and simply shouldn't be allowed. The <span> element is used to apply style on inline elements when a generic presentation tag (ie.: <b> doesn't apply) For example to make some text green:
Incorrect:
<p>You will also be happy to know <span class="bold">ACME Corp</span> is a <span class="eco-green">certified green</span> company.</p>
Correct: (Explanation below)
<p>You will also be happy to know <b>ACME Corp</b> is a <em class="eco-green">certified green</em> company.</p>
The reason here why you would want to use <em> as opposed to <span> for the word green is because the color green here is used to add emphasis on the fact that ACME Corp is a certified green company.
The following would be a good example of the use of a <span> tag:
Correct:
<p>You may press <kbd>CTRL+G</hbd> at any time to change your pen color to <span class="pen-green">green</span>.</p>
In this example, the word green is styled in green simply to reflect the color, not to add any emphasis (<em>) or strong emphasis (<strong>).
The whole distinction between "presentation" elements versus "structure" element is, in my opinion, a matter of common sense, not something defined by W3C or anyone else. :-P
An element that describes what its content is (as opposed to how it should look) is a structure element. Everything else is, by definition, not structural, and therefore a presentation element.
Now, I'll answer the second part of your post. I understand this is a contentious topic, but I'll speak my mind anyway.
Well-made HTML should not concern itself with how it should look. That's the job of the stylesheet. The reason it should leave it to the stylesheet, is so you can deliver one stylesheet for desktop computers, another one for netbooks, smartphones, "dumbphones" (for lack of a better term), Kindles, and (if you care about accessibility, and you should) screen readers.
By using presentation markup in your HTML, you force a certain "look" across all these different types of media, removing the ability of the designer to choose a look that works best for such devices. This is micromanagement of the worst sort, and designers will hate you for it. :-)
To use your example, instead of using <b>, you should ask yourself what the boldness is supposed to express. If you're trying to express a section title, use one of the header tags (<h1> through <h6>). If you're trying to express strong emphasis, use <strong>. You get the idea. Express the what, not the how; leave the how to the stylesheet designers.
</soapbox>
It's not that presentational elements should be avoided, it's that markup should be as semantic as possible. When designing a document structure, default styling should be considered a secondary affect. If an element is used solely for presentation, it's not semantic, no matter what element is used.
The example usage of <b> isn't semantic, because <b> imparts no meaning. <span class="boldtext"> also isn't semantic. As such, their usage is mixing presentation into the structure.
I've not seen a useful, authoritative web resource which says how to reliably enable rollover effects on a4j:commandButton elements in RichFaces JSF content.
I have found plenty of css resources re buttons, but then when they start talking about html I start wondering what the equivalent is in more recent web technologies.
Basically I want my buttons to look different on hover and then on click. I'd like to use images but client side js is out of scope for this q.
m
Use the following attributes as in this example below to your a4j:commandButton (have a look at the full list (v3.3.1.GA)):
<a4j:commandButton value="Submit" styleClass="ctlBtn"
id="myCtlBtn" type="submit" style="margin-bottom: 5px;"
reRender="allMyOtherplaces"
onmouseover="this.className='ctlBtn btnHover'"
onmousedown="this.className='ctlBtn btnDown'"
onmouseup="this.className='ctlBtn'"
onmouseout="this.className='ctlBtn'"
onclick="this.className='ctlBtn btnDown'"
oncomplete="this.className='ctlBtn'" />
And create classes ctlBtn, btnHover and btnDown in your included .css file.