./www/ruby-capybara, Integration testing tool for rack based web applications

[ CVSweb ] [ Homepage ] [ RSS ] [ Required by ] [ Add to tracker ]


Branch: CURRENT, Version: 3.12.0, Package name: ruby24-capybara-3.12.0, Maintainer: pkgsrc-users

Capybara is an integration testing tool for rack based web applications. It
simulates how a user would interact with a website.


Required to run:
[textproc/ruby-nokogiri] [net/ruby-addressable] [www/ruby-rack-test] [textproc/ruby-xpath] [www/ruby-rack16] [lang/ruby24-base] [mail/ruby-mini_mime]

Required to build:
[pkgtools/cwrappers]

Master sites:

SHA1: a656a3f223f6baf959108fd26aa07aa35106a554
RMD160: fb354c7471401e92fcd392b2831cc5346edde511
Filesize: 306 KB

Version history: (Expand)


CVS history: (Expand)


   2019-04-15 08:11:33 by Thomas Klausner | Files touched by this commit (10)
Log message:
*: remove references to ruby 2.3
   2019-02-03 14:45:07 by Takahiro Kambe | Files touched by this commit (11)
Log message:
Allow Ruby 2.6 to build.
   2018-12-17 16:12:50 by Takahiro Kambe | Files touched by this commit (3) | Package updated
Log message:
www/ruby-capybara: update to 3.12.0

# Version 3.12.0
Release date: 2018-11-28

### Added

* Support Ruby 2.6 endless range in Result#[] and query `:between` option
* Pre-registered headless firefox driver :selenium_headless [Andrew Havens]
* Selenium driver now defaults to clearing `sessionStorage` and `localStorage`. \ 
To disable pass `clear_local_storage: false` and/or `clear_session_storage: \ 
false` when creating Capybara::Selenium::Driver instance in your driver \ 
registration

### Fixed

* Raise error if only :x or :y are passed as an offset to click methods

### Removed

* Support for RSpec < 3.5

# Version 3.11.1
Release date: 2018-11-16

###Fixed

* Fixed :link_or_button XPath generation when it has had an expression filter added

# Version 3.11.0
Release date: 2018-11-14

### Added

* Ability for node filters to set detailed error messages
* `Capybara::HTML` Will use `nokogumbo` for HTML parsing if installed
* `Selector#locator_filter` added to allow for dynamic locator in selectors

### Fixed

* Node filters are evaluated in the context of the Selector they are used in to \ 
ensure the correct options are used

# Version 3.10.1
Release date: 2018-11-03

### Fixed

* Fix `aria-label` and `test_id` matching for `link_or_button` selector type - \ 
Issue #2125
* Fixed crash in element path creation for matcher failure messages - Issue #2120

# Version 3.10.0
Release date: 2018-10-23

### Added

* :class filter can now check for class names starting with !
* Selector `xpath`/`css` expression definitions will get filter names from block \ 
parameters if not explicitly provided
* `any_of_selectors` assertions and matchers to complement `all_of_selectors` \ 
and `none_of_selectors`

### Fixed

* Selector `css` expression definiton declared filters now work again
* Cleaned up warnings [Yuji Yaginuma]
* Workaround installation of rspec matcher proxies under jruby by reverting to \ 
the old solution not using prepend, so jruby bugs are not hit - Issue #2115

# Version 3.9.0
Release date: 2018-10-03

### Added

* Selenium with Chrome removes all cookies at session reset instead of just \ 
cookies from current domain if possible
* Support for Regexp for system :id and :class filters where possible
* `using_session` now accepts a session object as well as the name of the \ 
session for users who  manually manage sessions
* The `:field` selector will now find `type = "hidden"` fields if the \ 
`type: "hidden"` filter option is provided
   2018-09-24 18:03:42 by Takahiro Kambe | Files touched by this commit (2) | Package updated
Log message:
www/ruby-capybara: update to 3.8.1

# Version 3.8.1
Release date: 2018-09-22

### Fixed

* Filling in of date fields with a string when using selenium chrome
  regression [Micah Geisel]
   2018-09-23 18:50:34 by Takahiro Kambe | Files touched by this commit (3) | Package updated
Log message:
www/ruby-capybara: update to 3.8.0

# Version 3.8.0
Release date: 2018-09-20

### Added

* Workaround gecodriver 0.22 issue with undefined pause durations
* :element selector ignores XML namespaces

### Fixed

* Added Errno::ECONNRESET to the errors which will allows https server detection

# Version 3.7.2
Release date: 2018-09-12

### Fixed

* Fix MatchQuery based matchers when used on a root element found using any type \ 
of parent/ancestor query - Issue #2097

* Fix Chrome/FF HTML5 drag simulation for elements (a, img) which default to \ 
draggable - Issue #2098

# Version 3.7.1
Release date: 2018-09-05

### Fixed

* Restored ability to pass symbol as the CSS selector when calling \ 
`has_css?`/`have_css`/etc - Issue #2093

# Version 3.7.0
Release date: 2018-09-02

### Added

* `Capybara.disable_animation` can be set to a CSS selector to identify which \ 
elements will have animation disabled [Michael Glass]
* `Capybara.default_normalize_ws` option which sets whether or not text \ 
predicates and matchers (`has_text?`, `has_content?`, `assert_text`, etc) use \ 
`normalize_ws` option by default. Defaults to false. [Stegalin Ivan]
* Selector based predicates, matchers, and finders now support the \ 
`:normalize_ws` option for the `:text`/`:exact_text` filters. Defaults to the \ 
`Capybara.default_normalize_ws`setting above.
* Element `choose`/`check`/`uncheck`/`attach_file`/`fill_in` can now operate on \ 
the element they're called on or a descendant if no locator is passed.

### Fixed

* All CSS styles applied by the `Element#attach_file` `:make_visible` option \ 
will now have `!important` priority set to ensure they override any other \ 
specified style.
* Firefox file inputs are only manually cleared when necessary.

# Version 3.6.0
Release date: 2018-08-14

### Added

* Workaround geckodriver/firefox send_keys issues as much as possible using the \ 
Selenium actions API
* Workaround lack of HTML5 native drag and drop events when using Selenium \ 
driver with Chrome and FF >= 62
* `Capybara.predicates_wait` option which sets whether or not Capybaras matcher \ 
predicate methods (`has_css?`, `has_selector?`, `has_text?`, etc.) default to \ 
using waiting/retrying behavior (defaults to true)

# Version 3.5.1
Release date: 2018-08-03

### Fixed

* Fixed misspelled method name `refute_matches_elector` => \ 
`refute_matches_selector`

# Version 3.5.0
Release date: 2018-08-01

### Added

* text predicates and matchers (`has_text?`, `has_content?`, `assert_text`, etc) \ 
now support a `normalize_ws` option

### Fixed

* `attach_file` with Selenium and local Firefox 62+ now correctly generates only \ 
one change event when attaching multiple files

# Version 3.4.2
Release date: 2018-07-24

### Fixed

* `match_xxx` selectors and `matches_xxx?` predicates work correctly with \ 
elements found using a sibling selector - Issue #2073

# Version 3.4.1
Release date: 2018-07-20

### Fixed

* `Session#evaluate_script` now strips the script in `Session` rather than only \ 
in the Selenium driver

# Version 3.4.0
Release date: 2018-07-19

### Fixed

* Make selenium driver :backspace clear stategy work even if caret location is \ 
in middle of field content [Champier Cyril]
* Selenium issue with fieldset nested in disabled fieldset not being considered \ 
disabled
* `Session#evaluate_script` and `Element#evaluate_script` now strip \ 
leading/trailing whitespace from scripts [Ian Lesperance]

### Added

* Work around Selenium lack of support for `file_detector` with remote geckodriver
* `#within_frame` locator is optional when only one frame exists
* `Capybara.test_id` option that allows for matching the Capybara provided \ 
selector types on an arbitrary attribute
  (defaults to nil), set to your test id attribute ('data-test-id, etc) if using \ 
test id attributes in your project

# Version 3.3.1
Release date: 2018-06-27

### Fixed

* `selenium-webdriver` version check [ahorek]
* Selenium driver correctly responds to `disabled?` for fieldset elements - \ 
Issue #2059 [Thomas Walpole]

# Version 3.3.0
Release date: 2018-06-25

### Added

* RackTest driver now handles 307/308 redirects
* `execute_async_script` can now be called on elements to run the JS in the \ 
context of the element
* `:download` filter option on `:link' selector
* `Window#fullscreen`
* `Element#style` and associated matchers

### Changed

* Minimum "supported" `selenium-webdriver` is raised to 3.5.0 (but you \ 
really should be using newer than that)

### Fixes

* Selenium driver with Firefox workaround for clicking on table row - \ 
https://github.com/mozilla/geckodriver/issues/1228
* :class and :id filters applied to CSS based selectors now correctly handle the \ 
CSS comma
* Selenium driver handles namespaces when generating an elements `#path` - Issue \ 
#2048

# Version 3.2.1
Release date: 2018-06-04

### Fixes

* Only split CSS selectors when :class or :id options are given. Restores 3.1.1 \ 
functionality for now but the underlying issue
  will require a larger fix, hopefully coming soon. - Issue #2044 [Thomas Walpole]

# Version 3.2.0
Release date: 2018-06-01

### Changed

* Ruby 2.3.0+ is now required
* `ElementNotFound` errors raised in selector filters are interpreted as non-matches

### Added

* New global configuration `default_set_options` used in \ 
`Capybara::Node::Element#set` as default `options` hash [Champier Cyril]
* `execute_script` and `evaluate_script` can now be called on elements to run \ 
the JS in the context of the element [Thomas Walpole]
* Filters in custom selectors now support a `matcher` Regexp to handle multiple \ 
filter options [Thomas Walpole]
* `:element` selector type which will match on any attribute (other than the \ 
reserved names) passed as a filter option [Thomas Walpole]
* `:class` filter option now supports preceding class names with `!` to indicate \ 
not having that class [Thomas Walpole]
* `:class` and `:id` filter options now accept `XPath::Expression` objects to \ 
allow for more flexibility in matching [Thomas Walpole]
* `Capybara.disable_animation` setting which triggers loading of a middleware \ 
that attempts to disable animations in pages.
  This is very much a beta feature and may change/disappear in the future. \ 
[Thomas Walpole]

# Version 3.1.1
Release date: 2018-05-25

### Fixes

* Ensure keystrokes are sent when setting time/date fields to a string with the \ 
Selenium driver [Thomas Walpole]

# Version 3.1.0
Release date: 2018-05-10

### Added

* Support for using `select` with text inputs associated with a datalist element
* `type` filter on `:button` selector
* Support for server operating in https mode
* Selenium driver now uses JS to fill_in/set date and time fields when passed \ 
date or time objects [Aleksei Gusev, Thomas Walpole]
   2018-05-06 04:24:54 by Takahiro Kambe | Files touched by this commit (3) | Package updated
Log message:
www/ruby-capybara: update to 3.0.3

# Version 3.0.3
Release date: 2018-04-30

### Fixes

* Issue in `check` where the locator string could not be omitted
* Selenium browser type detection when using remote [Ian Ker-Seymer]
* Potential hang when waiting for requests to complete [Chris Zetter]

# Version 3.0.2
Release date: 2018-04-13

### Fixes

* Fix expression filter descriptions in some selector failure messages
* Fix compounding of negated matechers - Issue #2010

# Version 3.0.1
Release date: 2018-04-06

### Changed

* Restored ability for `Capybara.server=` to accept a proc which was \ 
accidentally removed in 3.0.0

# Version 3.0.0
Release date: 2018-04-05

### Changed

* Selenium driver only closes extra windows for browsers where that is known to \ 
work (Firefox, Chrome)
* "threadsafe" mode is no longer considered beta

### Fixes

* Multiple file attach_file with Firefox
* Use Puma::Server directly rather than Rack::Handler::Puma so signal handlers \ 
don't prevent test quitting

# Version 3.0.0.rc2
Release date: 2018-03-23

### Changed

* Visibile text whitespace is no longer fully normalized in favor of being more \ 
in line with the WebDriver spec for visible text
* Drivers are expected to close extra windows when resetting the session
* Selenium driver supports Date/Time when filling in date/time/datetime-local inputs
* `current_url` returns the url for the top level browsing context
* `title` returns the title for the top level browsing context

### Added

* `Driver#frame_url` returns the url for the current frame
* `Driver#frame_title` returns the title for the current frame

# Version 3.0.0.rc1
Release date: 2018-03-02

### Added
* Support for libraries wrapping Capybara elements and providing a \ 
`#to_capybara_node` method

### Changed

* `first` now raises ElementNotFound, by default, instead of returning nil when \ 
no matches are found  - Issue #1507
* 'all' now waits for at least one matching element by default.  Pass `wait: \ 
false` if you want the previous
  behavior where an empty result would be returned immediately if no matching \ 
elements exist yet.
* ArgumentError raised if extra parameters passed to selector queries

### Removed

* Ruby < 2.2.2 support
* `Capybara.exact_options` no longer exists. Just use `exact: true` on relevant \ 
actions/finders if necessary.
* All previously deprecated methods removed
* RSpec 2.x support
* selenium-webdriver 2.x support
* Nokogiri < 1.8 support
* `field_labeled` alias for `find_field`
   2018-03-17 15:26:23 by Takahiro Kambe | Files touched by this commit (3) | Package updated
Log message:
www/ruby-capybara: update to 2.18.0

#Version 2.18.0
Release date: 2018-02-12

### Fixed

* Firefox/geckodriver setting of contenteditable childs contents
* Ignore Selenium::WebDriver::Error::SessionNotCreatedError when quitting driver \ 
[Tim Connor]

### Removed

* Headless chrome modal JS injection that is no longer needed for Chrome \ 
64+/chromedriver 2.35+

# Version 2.17.0
Release date: 2018-01-02

### Added

* `have_all_of_selectors`, `have_none_of_selectors` RSpec matchers for parity \ 
with minitest assertions [Thomas Walpole]

### Fixed

* Allow xpath 3.x gem [Thomas Walpole]
* Issue when drivers returned nil for `current_path` and a matcher was used with \ 
a Regexp [Thomas Walpole]
* Error message when visible element not found, but non-visible was [Andy Klimczak]

# Version 2.16.1
Release date: 2017-11-20

### Fixed

* Fix rack_test driver for rack_test 0.7.1/0.8.0 [Thomas Walpole]
* `accept_prompt` response text can contain quotes when using selenium with \ 
headless chrome [Thomas Walpole]

# Version 2.16.0
Release date: 2017-11-13

### Added

* Attempt to move element into view when selenium doesn't correctly do it - See \ 
PR #1917 [Thomas Walpole]
* `current_path` matchers will now autodetect path vs url based on string to be \ 
matched. Deprecates
  `:only_path` in favor of `:ignore_query` option [Thomas Walpole]
* Session#evaluate_async_script [Thomas Walpole]

### Fixed

* Default prompt value when using headless Chrome works correctly [Thomas Walpole]
* Support new modal error returned by selenium-webdriver 3.7 for W3C drivers \ 
[Thomas Walpole]
* Calling `respond_to?` on the object passed to `Capybara.configure` block - \ 
Issue #1935

# Version 2.15.4
Release date: 2017-10-07

### Fixed
*  Visiting an absolute URL shouldn't overwrite the port when no server or \ 
always_include_port=false - Issue #1921

# Version 2.15.3
Release date: 2017-10-03

### Fixed
*  Visiting '/' when Capybara.app_host has a trailing '/' - Issue #1918 [Thomas \ 
Walpole]

# Version 2.15.2
Release date: 2017-10-02

### Fixed

*  Include within scope description in element not found/ambiguous errors \ 
[Thomas Walpole]
*  Raise error when no activation block is passed to modal methods if using \ 
headless chrome [Thomas Walpole]
*  Don't retry element access when inspecting [Ivan Neverov]
*  Don't override a specified port (even if it is default port) in visited url \ 
[Thomas Walpole]
   2018-01-01 23:30:04 by Roland Illig | Files touched by this commit (537)
Log message:
Sort PLIST files.

Unsorted entries in PLIST files have generated a pkglint warning for at
least 12 years. Somewhat more recently, pkglint has learned to sort
PLIST files automatically. Since pkglint 5.4.23, the sorting is only
done in obvious, simple cases. These have been applied by running:

  pkglint -Cnone,PLIST -Wnone,plist-sort -r -F