word counter? 1

[GUJUm0deL]

Is there a JS version of a word counter that ignores html tags like <strong> and <i> or html tags like <div id="something"> or <p>

_____________________________
Just Imagine.

http://www.imajinarts.com

 

[feherke]

Hi

Theoretically that could be reduced to :

document[teal].[/teal]body[teal].[/teal]textContent[teal].[/teal][COLOR=darkgoldenrod]match[/color][teal]([/teal][fuchsia]/\w+/g[/fuchsia][teal]).[/teal]length

But unfortunately that will not work correctly :
[ul]
[li]tag pair content like between [tt]<script>[/tt] and [tt]</script>[/tt] is kept[/li]
[li]words previously separated by tags get concatenated[/li]
[/ul]
( Beside those, Explorer does not know [tt]textContent[/tt], it uses the proprietary [tt]innerText[/tt]. )

So you will have to do it the hard way :

document[teal].[/teal]body[teal].[/teal]innerHTML
  [teal].[/teal][COLOR=darkgoldenrod]replace[/color][teal]([/teal][fuchsia]/<script[\s\S]*?>[\s\S]*?<\/script\s*>/gi[/fuchsia][teal],[/teal][green][i]' '[/i][/green][teal])[/teal] [gray]// remove scripts[/gray]
  [teal].[/teal][COLOR=darkgoldenrod]replace[/color][teal]([/teal][fuchsia]/<style[\s\S]*?>[\s\S]*?<\/style\s*>/gi[/fuchsia][teal],[/teal][green][i]' '[/i][/green][teal])[/teal]   [gray]// remove styles[/gray]
  [teal].[/teal][COLOR=darkgoldenrod]replace[/color][teal]([/teal][fuchsia]/<\/?\w+[\s\S]*?>/gi[/fuchsia][teal],[/teal][green][i]' '[/i][/green][teal])[/teal]                      [gray]// remove tags[/gray]
  [teal].[/teal][COLOR=darkgoldenrod]replace[/color][teal]([/teal][fuchsia]/&.+?;/g[/fuchsia][teal],[/teal][green][i]''[/i][/green][teal])[/teal]                                   [gray]// remove character entities[/gray]
  [teal].[/teal][COLOR=darkgoldenrod]match[/color][teal]([/teal][fuchsia]/\w+/g[/fuchsia][teal])[/teal]                                          [gray]// get the words[/gray]
  [teal].[/teal]length                                                 [gray]// get their count[/gray]

But even will also fail as the character entities may evaluate to word characters or not :
[ul]
[li]El Ni&ntilde;o - word character, should be kept[/li]
[li]Hello&nbsp;World - non-word character, should be removed[/li]
[/ul]
Personally I know about no good ( are at least satisfactory ) way to handle the character entities.

Maybe you can get some hints from there anyway.


Feherke.

http://free.rootshell.be/~feherke/

 

[whosrdaddy]

with jQuery you can get the text?

var renderedText = $('#element').text();

googling "jquery word counter" may help you out?

Cheers,
Daddy

-----------------------------------------------------
What You See Is What You Get
Never underestimate tha powah of tha google!

 

[GUJUm0deL]

Thanks. I looked at some examples for jQuery word counter and saw that it did still counted the <a href> tag as part of the wording.

For example:
hello my name is <a href="a.html">bob</a>

Came back with 6 words, and not 5.

_____________________________
Just Imagine.

http://www.imajinarts.com

 

[whosrdaddy]

<div id="test">
 hello my name is <a href="a.html">bob</a>
</div>

alert($('#test').text());

gives me "hello my name is bob".
5 words?

/Daddy



-----------------------------------------------------
What You See Is What You Get
Never underestimate tha powah of tha google!

 

[feherke]

Hi

gives me "hello my name is bob".

And so it should for "hello my name is[red]<br>[/red]<a href="a.html">bob</a>" too, but it will concatenate "is" and "bob" into 4 words.


Feherke.

http://free.rootshell.be/~feherke/

 

[interfacetricks]

One approach you can take is to loop through each node, and count the words if it's a text node, or loop through the child nodes if it's not.

This requires a few extra details, for example some elements you want to skip (<script> tags, for example), some you want the "value" instead, etc. But with a few minor ammendments this algorithm seems to work pretty well.

I've made a testing version of it here:

http://webinterfacetricks.com/word_count/

It uses the regular expression (/\w+/g) which means find a sequence of word characters (\w), of any length (+) and don't stop after the first match (g for global).

There's more explanation in the source-code at the link. Let me know if anything is unclear

 

[feherke]

Hi

Good idea. That way you not need to care about the character entities. I like it.


Feherke.

http://free.rootshell.be/~feherke/