I try to truncate this word "testtesttesttesttesttest", but it can't. This is my code:
<%= truncate_html("testtesttesttesttesttest", :length => 11, :omission => '..') %>

If I try to truncate a string using an omission, the string may get truncated even when it is shorter than the length. This happens when the combined length of the string and the omission is greater than the :length option. For instance, if you add the following test to html_truncator_spec.rb ...

it 'does not truncate a string shorter than length' do
  truncate('some string', length: 12, omission: '...').should == 'some string'
end

... it will fail with the following error:

1) TruncateHtml::HtmlTruncator does not truncate a string shorter than length
   Failure/Error: truncate('some string', length: 12, omission: '...').should == 'some string'
     expected: "some string"
          got: "some..." (using ==)
   # ./spec/truncate_html/html_truncator_spec.rb:62:in `block (2 levels) in <top (required)>'

This happens even though "some string" is shorter than 12 characters. Is this expected behavior? It differs from the behavior of truncate from Active Support, where "some string".truncate(12, omission: '...') == "some string".

Hello guys,
I am trying to use truncate_html on such strings,
<p>T h a n k s</p>
issue with this is , it removes the space added purposely.
so when i do truncate_html("<p>T h a n k s</p>", length: 15000)
it gives me output as =>
<p>Thanks</p>
Is there any configuration, so escape such kind of characters ?
Any help will be appreciated.
Thanks

I have tried installing this gem with both bundle install and gem install and my rails app is unable to find the truncate_html method. I keep getting a nomethod error.

It looks like the omission string is applied before the word_boundry regex is applied, so if you have the default omission '...' or any other omission with a '.' character in it, the result is word boundry truncation instead of the regex param passed in.

As I'm looking at the code, :word_boundary seems to have changed from a true/false to a regex expression. At least, that's the default configuration. Now we can no longer explicitly pass in true to :word_boundary, since it errors out with the statement: "NoMethodError: undefined method `source' for true:TrueClass". I'm using Refinery, which does that in its code.

Could you add a check to see if :word_boundary is true and then switch it to the default regex if it is?

Thanks,
Justin

Unlike the Rails truncate method, length passed to truncate_html does not include the omission text. So for example:

truncate("a b c", 4, "...")
#= > "a..."

truncate_html("a b c", :length => 4, :omission => "...")
#=> "a b ..."

At the very least, this should be documented somewhere.

Is length calculated based on the text outside of the html tags? Or does it include the HTML tags?

If it does include the HTML tags, then it would be very useful to be able to set the length of the text outside of HTML tags.

Thanks for an awesome gem!

Hi @hgmnz,

A client is running some content with Unicode characters (namely, an up arrow) through truncate_html and noticing that those characters are disappearing.

I've narrowed it down to the scan in TruncateHtml::HtmlString. However, that's a hell of a regex to read, so I was wondering if you wouldn't mind walking me through it.

You can paste this code into an .rb file and run it to see what I mean:

# encoding: utf-8
unicode_string = "Up Arrow (↑) points up."

# From TruncateHtml::HtmlString
# 
def regex
  /(?:<script.*>.*<\/script>)+|<\/?[^>]+>|[[[:alpha:]]\w\|`~!@#\$%^&*\(\)\-_\+=\[\]{}:;'",\.\/?]+|\s+|[[:punct:]]/
end

# scan normally respects unicode.
puts unicode_string.scan(/.*/).join

# but this regex does not.
puts unicode_string.scan(regex).join

The result at the command line is

Up Arrow (↑) points up.
Up Arrow () points up.

Thanks!

by adding a backslash. See https://github.com/hgimenez/truncate_html/commit/36c2edfa3acd40782600ac11e52a18ce30f19a1f#commitcomment-1246688

In the README an example regex for splitting on strings is config.word_boundary = /\S[\.\?\!]/.

This regex fails, however, when an HTML tag is at the end of the string. Testing on the string 'Here is a string.', I get:

Minitest::Assertion: --- expected
+++ actual
@@ -1 +1 @@
-"<strong>Here is a string.</strong>"
+"<strong>Here is a string."

The regex should read /\S([\.\?\!]|\z)/. This will also match against the end of the string.

Sometimes there is a requirement to toggle the rest of given html text using Javascript(show more/show less functionality). In order to implement this without duplicating truncated html I need the rest of it.
I tried to get the rest of html as following:

truncated_content = truncate_html(item.content, length: 300, omission: '')
trunc_length = truncated_content.length
rest = item.content[trunc_length, item.content.length]

But when there are extra consecutive whitespace in the given html, above code won't produce correct output. This is because in TruncateHtml::HtmlString#html_tokens method consecutive whitespace are being reduced to single whitespace.

Therefore I propose to create new helper "slice_html" that returns both truncated html and
the rest as following:

truncated_content, rest = slice_html(item.content, length: 300, omission: '')

All the Cyrillic characters from my truncated html are missing.

I'm realizing as I write this that this could perhaps be fixed with the regex provided in the documentation for sentence boundaries, but as it stands the regex can cut off the resulting string mid href. As such, you end up with an open quotation that is not closed and many resulting formatting problems on the page. I got something like this as a result.

<p>Sample text with link to <a href=\"http://www.example.</a></p>

Invalid ticket, sorry.

In this series, we’re interviewing NYCDA graduates to talk about their program...
gets truncated to
In this series, we’re interviewingNYCDA graduates to talk about their program...

Hi Harold,

I found you thanks to github search feature, as your gem on rubygems.org is pointing the source code to your old address at: http://github.com/hgimenez/truncate_html. I don't really know how they handle it but I think if you loose your access to rubygems, you could contact them and try to reclaim the ownership of that specific gem.

Here is the address to help you with it:
http://rubygems.org/gems/truncate_html

Thanks for your time

s = "This is bad <script type=text/javascript>document.write('lum dee dum');</script>"
truncate_html(s, :length => 20, :omission => "... <a href='foo'>read more</a>")

=>

"This is bad <script type=text/javascript>document.write('lum... <a href='foo'>read more</a></script>"

I figured out that texts with multibyte characters truncates wrongly.

>> truncate_html("blah", :length => 2, :omission => "longer")
NoMethodError: You have a nil object when you didn't expect it!
The error occurred while evaluating nil.rstrip
        from C:/Ruby/lib/ruby/gems/1.8/gems/truncate_html-0.2.1/lib/truncate_html/html_truncator.rb:30:in `truncate'
        from C:/Ruby/lib/ruby/gems/1.8/gems/truncate_html-0.2.1/lib/truncate_html/html_truncator.rb:17:in `each'
        from C:/Ruby/lib/ruby/gems/1.8/gems/truncate_html-0.2.1/lib/truncate_html/html_truncator.rb:17:in `truncate'
        from C:/Ruby/lib/ruby/gems/1.8/gems/truncate_html-0.2.1/lib/app/helpers/truncate_html_helper.rb:4:in `truncate_html'

it will truncate in the middle of a href="" string and create invalid html...

You have a nil object when you didn't expect it!
The error occurred while evaluating nil.word_boundary

truncate_html(name, :length => 40, :omission => "...")

/usr/local/lib/ruby/gems/1.8/gems/truncate_html-0.3.1/lib/truncate_html/html_truncator.rb:11:in truncate' /usr/local/lib/ruby/gems/1.8/gems/truncate_html-0.3.1/lib/app/helpers/truncate_html_helper.rb:6:intruncate_html'

regression from truncate_html 0.2.2

Running on Rails 2.2.2

Could you include in the README how to use this independently of rails?

The following test fails with truncate_html 0.5.4:

truncate_html("One two three", :length => 12)

should return

One two t...

but instead it returns

One two t

Changing the value of :word_boundary makes no difference.

This plugin is truncating chinese characters

The behavior I expect is that omission is used any time the text is truncated, regardless of how the end point of the truncation is determined.
#40 had a fix for the omission missing when break_token is used, but was closed with no discussion for some reason.

haven't had the free time to dig into why this is happening yet, but wanted to note this here.

1.9.3-p194 :008 > helper.truncate_html('hello and goodbye', length: 15)
=> "hello and ..."

When I upgraded the gem to 0.9.1 or 0.9.2 it threw an exception I had to go back to 0.5.5.

Right now this fails. I've added support:

http://github.com/bcardarella/truncate_html/commit/77119b6de3eb6d096f23ec7dac14159b0ff49fe0

It's common to see HTML code like this:

<p>Hello
World</p>

Which, for example, when scrapped, results in the following Ruby string:

"<p>Hello\nWorld</p>"

If we render this code in a view, the browser will display it as Hello World.
However, truncate_html strips the \n, so the string will be rendered incorrectly as HelloWorld.

I'm using helpers inside mailer views and this it works fine:

class UserMailer < Devise::Mailer

  helper :home
  helper :post
  .
  .
  .
end

However, I need to use truncate_html gem inside mailer views

How can I add this helper to mailer views?

Thanks

Hi,

We are having some ajax call to controller, and controller return json to front-end.

Is there anyway that we call truncate_html inside controller

Thanks

I know there was this issue: #43

But should it not just interpret the new line?

Take my website for example: http://www.dchapman.io/ (here the code examples look wonky because the new line chars are getting replaced with spaces).

Here's what the post should look like: http://www.dchapman.io/posts/changing-column-type-in-postgresql-rails

Here's what the index looks like where truncate_html is being used: https://github.com/dchapman1988/dchapman.io/blob/master/app/views/posts/index.html.slim#L10

It will be awesome to be able to exclude special tags ( html or not ) from the character count. Example

truncate_html("Please take care of truncating this <script type='text/javascript'>function call_me_function() { console.debug('whatsup'); call_me_function();} </script> big and long portion of text", :length=>40)

instead of outputting:

Please take care of truncating this <script type='text/javascript'>function call_me_function() { console.debug('whatsup'); call_me_function();} </script>...

Will do

Please take care of trunkating this <script type='text/javascript'>function call_me_function() { console.debug('whatsup'); call_me_function();} </script> big and long portion of text

because I set in some truncate_html settings that tags like <script> should not be counted.

This could apply to other tags that are already invisible to the users:

It is common in blogs to place a page break in the content. Maybe it is after the first paragraph, maybe it is after the first line, but the point is it could be anywhere.

I suggest a feature where you could look for a configurable tag, maybe <break /> that if found, treat the character count as 0 and truncate the html, cleaning up just as you do if the character limit was reached.

It would probably be an alteration to this if block looking something like

    if @chars_remaining <= 0 || token == TruncateHtml.configuration.break_token
       close_open_tags
       break
    else
      process_token(token)
    end

This allows the gem to be data-driven for the people who want it as well as truncated at a pre-set value for existing users of the gem

e.g. helper.truncate_html('non breaking spaces') will return 'nonbreakingspaces'

This might be a very simple thing, or it might be possible already, but I would like to have an easy way to know if a given string was actually truncated by truncate_html.

Thanks!