Merge branch 'master' of github.com:johnbintz/persistent_selenium

This commit is contained in:
John Bintz 2013-01-27 14:42:27 -05:00
commit 8552f6a336
5 changed files with 44 additions and 9 deletions

View File

@ -19,11 +19,19 @@ require 'persistent_selenium/driver'
Capybara.default_driver = :persistent_selenium
```
If you're using Cucumber, you can also install that hook:
``` bash
persistent_selenium install
```
Should work just the same as if you used the standard Capybara Selenium driver, except for
these two differences:
* The browser starts up first thing and sticks around.
* The last page you were on before your tests passed/failed stays there, so you can inspect it.
* The browser starts up first thing and sticks around, so you don't pay the startup/shutdown
penalty with each test run.
* The last page you were on before your tests passed/failed stays there, so you can inspect it
and adjust your tests.
The browser's cookies and such are reset before the next test runs, so you still get the state
cleared out before your next set of tests.
@ -32,3 +40,10 @@ cleared out before your next set of tests.
It's DRb, which Just Works (tm), and a little reshuffling of the default Capybara Selenium driver's code.
#### When DRb doesn't Just Work (tm)
You may see recycled object errorrs. Capybara is complex enough that, eventually, something
will get garbage collected on the server and an object on the client will notice.
I turned off garbage collection on the server but it still happens.
Just rerun the test. It doesn't happen too often. It's the price you pay for speed.

View File

@ -4,4 +4,5 @@ $: << File.expand_path('../../lib', __FILE__)
require 'persistent_selenium/cli'
GC.disable
PersistentSelenium::CLI.start

View File

@ -6,6 +6,7 @@ module PersistentSelenium
class Browser < Capybara::Selenium::Driver
def initialize(browser_type)
@browser_type = browser_type
@__found_elements__ = []
end
def browser
@ -56,10 +57,6 @@ module PersistentSelenium
# instead.
end
end
GC.enable
GC.start
GC.disable
end
def starting_page

View File

@ -3,6 +3,12 @@ require 'persistent_selenium'
module PersistentSelenium
class CLI < Thor
include Thor::Actions
def self.source_root
File.expand_path('../../../skel', __FILE__)
end
desc "start", "Start the server"
method_options :port => PersistentSelenium.port, :browser => PersistentSelenium.browser
def start
@ -16,14 +22,19 @@ module PersistentSelenium
puts "Starting persistent_selenium on #{PersistentSelenium.port} with #{PersistentSelenium.browser}"
browser = Browser.new(PersistentSelenium.browser)
browser.browser
@browser = Browser.new(PersistentSelenium.browser)
@browser.browser
DRb.start_service PersistentSelenium.url, browser
DRb.start_service PersistentSelenium.url, @browser
DRb.thread.join
end
desc "install", "Install Cucumber hook for using persistent selenium"
def install
directory '.', '.'
end
default_task :start
end
end

View File

@ -0,0 +1,11 @@
# get the driver ready for use
require 'persistent_selenium/driver'
# set it for all tests
# Capybara.default_driver = Capybara.javascript_driver = :persistent_selenium
#
# or, set it via the ENV, with it being the default
# Before do
# Capybara.default_driver = Capybara.javascript_driver = (ENV['DRIVER'] || 'persistent_selenium').to_sym
# end