my code works in mozilla browser,but when i give it in chrome it is not opening chrome browser. I downlaoded chromedriver and put it in %HOMEPATH%\Local Settings\Application Data\Google\Chrome\Application\chrome.exe folder,also i did the same in ruby/bin folder.

But still i am getting the error "undefined method downcase for nil nilclass nomethoderror"

@browser=Watir::Browser.new :chrome

it works in mozilla

@browser=Watir::Browser.new :ff

Gem::InstallError: babosa requires Ruby version >= 2.0.0.
An error occurred while installing babosa (1.0.2), and Bundler cannot continue.
Make sure that gem install babosa -v '1.0.2' succeeds before bundling.

On Ruby 2.7, running the tests yield the following failure

$ ruby --version
ruby 2.7.0p0 (2019-12-25 revision 647ee6f091) [x86_64-linux]

$ bundle exec rspec
................................F...............................................

Failures:

  1) Babosa::Identifier#to_ruby_method should raise an error when it would generate an impossible method name
     Failure/Error: expect {"1".to_identifier.to_ruby_method}.to raise_error(Babosa::Identifier::Error)
     
       expected Babosa::Identifier::Error, got #<FrozenError: can't modify frozen String: ""> with backtrace:
         # ./lib/babosa/identifier.rb:172:in `downcase!'
         # ./lib/babosa/identifier.rb:172:in `to_ruby_method!'
         # ./lib/babosa/identifier.rb:289:in `send_to_new_instance'
         # ./lib/babosa/identifier.rb:263:in `to_ruby_method'
         # ./spec/babosa_spec.rb:151:in `block (4 levels) in <top (required)>'
         # ./spec/babosa_spec.rb:151:in `block (3 levels) in <top (required)>'
     # ./spec/babosa_spec.rb:151:in `block (3 levels) in <top (required)>'

Finished in 0.06649 seconds (files took 0.20925 seconds to load)
80 examples, 1 failure

Failed examples:

rspec ./spec/babosa_spec.rb:149 # Babosa::Identifier#to_ruby_method should raise an error when it would generate an impossible method name

In Ruby 2.7, casting a nil object to String yields a frozen object, thanks to https://bugs.ruby-lang.org/issues/16150?next_issue_id=16184.

So, if leader and trailer in to_ruby_method method are nils, as is the case in the failing test, to_s yields frozen strings, on which we are attempting downcase!, which fails (obviously).

It seems that Babosa is not properly loading in a Rails 5 app, resulting in 'undefined method' errors.
Any ideas?

First off - Thanks for this gem!

For data consistency reasons would be nice to validate format of the slug attribute in some cases.

Any suggestions for a regexp or other approach for such a validation?

I recently introduced a regression that could have been caught if Babosa were testing against Active Support 3.2 in CI. We should set this up using appraisals or something similar.

According to the docs for word_chars! it is not supposed to strip numbers from the string, but it is doing so as of 11e0f33

In babosa 1.0.4 emojis were kept when using normalize, now they seem to be removed.

Example:
In 1.0.4 'Emojis on slug 😋'.to_slug.normalize.to_s # => 'emojis-on-slug-😋'
In 2.0.0 'Emojis on slug 😋'.to_slug.normalize.to_s # => 'emojis-on-slug'

The change seems related to the function word_chars!, before was matching specific chars to to strip out, now is a lot more broad by using [^\p{letter}]

Is this intended? Would it be possible to keep support for the emojis?

While Chinese works well enough with 1:1 substitutions to produce something meaningful, this is impossible for Japanese. Consider the following three sentences:

• 生まれた子供たち。
• 生きる意味を見失う。
• 生意気な小僧。

Notice that they all begin with 生. But (one possible transliteration) is:

• umareta kodomo─tachi.
• ikiru imi o miushinau.
• namaiki na kozoo.

So depending on the context 生? can be eitheru,i, ornama`. And it gets worse, there are more possibilities (see file below). Admittedly, there are less possibilities for other characters, but the principle stays the same.

Thus transliterating Japanese is almost as hard as transliterating it - it can be done in software only approximately (like automated translations), and it isn't going to be perfect.

Fortunately, ruby is quite popular in Japan and there are some ruby ports of some morphological analyzers, which pretty much do all the hard work. I made a simple script (mainly for private use) that illustrates how Japanese transliterations might be achieved.

Run this ruby script (more notes/comments and links to the gems at the beginning)
http://pastebin.com/hgJjqYcb

with

$ ruby japanese_to_ascii.rb < test_text > output

on this file
http://pastebin.com/CGTHiE58

to get this output
http://pastebin.com/aknZyH0Q

As you can see it isn't perfect, but for the most part correct and readable. So this raises the question, should Japanese be supported in the first place, seeing how hard and possibly open-ended it is...?

As there are some open pull requests for ukrainian from last summer that fail the Travis CI build, it seems as if there's a problem with CI. @norman could you take a look?

Master fails on rbx in CI, this needs to be fixed.

Hello, while working with the library some words (like Yáñez) make the program to give an error, and happen when 2 "special" characters are togeder.

Thanks in advance

I am using babosa to create correct Chinese in slug. babosa works perfectly for new records, but I can't update existing records using babosa.

I ran Product.find_each(&:save) just like the usual way to update slug column in Friendly_id for existing records, but it doesn't update the slug column.

When I added custom transliterators, for example:

module Babosa
module Transliterator
class German < Latin
APPROXIMATIONS = {
"ä" => "ae",
"ö" => "oe",
"ü" => "ue",
"Ä" => "Ae",
"Ö" => "Oe",
"Ü" => "Ue"
}
end
end
end

I received the following warning:

...config/initializers/babosa.rb:5: warning: already initialized constant Babosa::Transliterator::German::APPROXIMATIONS
...gems/babosa-0.3.11/lib/babosa/transliterator/german.rb:5: warning: previous definition of APPROXIMATIONS was here

How can I add custom transliterators?

I bisected and commit 8892b93 broke support for ruby 1.8.7/rbx/jruby

... need to be handled.

I have a button on the bottom of the page. I need to scrolldown before clicking the link.I need to automate this.
Tried different things but no success.

These are the below snippets i tried..

@ff.goto("javascript:window.moveTo(10,20)") --getting timeout error
@ff.div(:id,"right-pane-footer").document.scrollIntoView -
@ff.button(:id, "save").scrollIntoView

button = @ff.button(:id,"save")
button.locate
button.ole_object.invoke('scrollIntoView')
@ff.button(:id,"save").ole_object.invoke('scrollIntoView')

@ff.element(:id,"save").scrollIntoView

Default latin transliterator is enough for Turkish support.
Maybe it should be mentioned in readme as supported with latin or a Turkish class could be added directly inheriting from latin.
I tested with Turkish special characters (Ç,ç,Ü,ü,Ö,ö,İ,ı,Ş,ş,Ğ,ğ) and characters with circumflex (â,î,û,ê,ô):

"Ç,ç,Ü,ü,Ö,ö,İ,ı,Ş,ş,Ğ,ğ,â,î,û,ê,ô,Â,Î,Û,Ê,Ô".to_slug.transliterate.to_s
=> "C,c,U,u,O,o,I,i,S,s,G,g,a,i,u,e,o,A,I,U,E,O"

In the Danish language the three letters æ, ø, å is usually serialized as ae, oe, aa. One might consider adding this as a language option...

Hi there,

I saw you have already fixed this 21 days ago. I'm wondering if there is a plan to release a new version ?

Last version was like 2 years ago.

Thanks.

Hello,

Firstly, thank you for Babosa! :)

Now, to business: I've seen this issue in production when Babosa is used within a threaded Sidekiq worker:

NameError: uninitialized constant Babosa::Transliterator::Latin

It's a very sporadic error and cannot be reproduced easily.

I'm fairly sure that it's happening because Babosa uses autoload and autoload has thread-safety problems.

I would suggest switching to just using require_relative.