- relax rspec dependecy upper limit
- add ruby 3.0 to test matrix
- add ruby 3.1 to test matrix
- remove 1.8 ruby from test matrix, this is the first step to remove it's support completely
- html/body matching from now is forbidden (#75)
- make ruby 2.7 possible to fail on CI
- fixed have_empty_tag matcher on jruby
- added ruby 2.7 to build matrix (no official support until stable release though)
- added ruby 2.6 to build matrix (#70)
infrastructure/misc:
- migrated to travis-ci.com instead of travis-ci.org (thanks Hamada Takayuki for inspiration: #70)
- refactor travis-ci build setup
- updated status badges to SVG (thanks to Olle Jonsson: #71)
- re-added ruby support back to 1.8.7 (supporting same versions as rspec 3)
- split lib in few files
- fixed README (thanks to Rodrigo Castro)
- fixed deep nesting (thanks to Misha Gorodnitzky)
- Rspec differ fix (thanks to Misha Gorodnitzky)
- "seen" text matching (thanks to Alex Shi)
- move license information into separate
LICENSE.md
file
- add support for
frozen_string_literals
- fix for rdoc links and not only
- become compatible with RSpec pre-release versions
- new, explicit configuration, refer to README
- added ruby 2.2.0 to CI
- rspec 3 version update
- added ruby 2.1.2 to CI
- introduced rspec 3.0.0 compatibility (thanks to Tiago)
- new "expect" syntax support
- options for have_tag now support Regexp (thanks to Ian C. Anderson)
- added license to gemspec
- added :without to have_tag? like have_tag('div', :without => { :class => 'test' }) NOTE: currently only classes supported
- ruby 2.0.0 support
- added with_text matcher
- some code refactoring, but a lot of refactoring left for future
- rewritten README, added more usage examples
- removed dealing with whitespaces (#11), too much magic for usage (#16)
- some attempt to improve documentation
- Fix for content matching regexps with single and double quotes (#14 thanks to watsonbox)
- capybara support
- added simple #description method for "it { should have_tag }" cases
- fix for unicode text matching (issue #13)
- leading and trailing whitespaces are ignored in tags where they should be ignored(#11, and again thanks to Simon Schoeters)
- whitespaces ignoring as browser does in :text matching
- have_tag backwards compability(thanks to Felix Tjandrawibawa, #12)
- make possible use non-string as :text option(#10, thanks for idea to Simon Schoeters)
- a little bit refactoring
- added some html5 inputs
- added message for should_not
- raise exception when wrong parametres specified(:count and :minimum (or :maximum) simultaneously)
- support all versions of nokogiri since 1.4.4
- updated nokogiri to 1.5.0 version, for nokokiri less than 1.5.0 use 0.0.6 release of this gem
- allow for single quotes in content matchers (thanks to Kelly Felkins).
- added some experimental matchers:
- have_form
- with_hidden_field
- with_text_field
- with_password_field
- with_file_field?
- with_text_area
- with_check_box
- with_radio_button
- with_select
- with_option
- with_submit
- have_form
- additional parameters(:count,:text,:with) rely on each other to match what exactly tester want to match(link to comment)
-
now following will work:
rendered.should have_tag('div') do with_tag('p') end
-
tags can be specified via symbol
-
classes can be specified via array or string(class-names separated by spaces), so following will work:
'<div class="one two">'.should have_tag('div', :with => { :class => ['two', 'one'] }) '<div class="one two">'.should have_tag('div', :with => { :class => 'two one' })
- documented source code
- added changelog
- all needed options and error messages