Unless otherwise noted, these tasks are all recursive: they will run on the project you are in and all subprojects contained within.
buildr gettext:extractrunsxgettextbuildr gettext:mergerunsmsgmergebuildr msgfmtrunsmsgfmt
The msgfmt task is run during compilation and it can take awhile to run on
every different locale we support. To alleviate the slowness, the task looks
at the environment variable nopo. If the variable is set to a locale or
comma separated list of locales, msgfmt will only run against those locales.
Setting nopo to anything else will prevent msgfmt from running at all.
If you keep forgetting to set nopo you can have Buildr do it for you
automatically by placing something like the following in ~/.buildr/buildr.rb:
#! /usr/bin/env ruby
ENV['nopo'] ||= 'de'Buildr will automatically evaluate that file and set nopo to "de" unless
the variable is already set.
buildr checkstyle
Buildr provides a Checkstyle task, but we have our own that reads from the
Eclipse Checkstyle Plugin configuration. The Eclipse configuration defines
several variables that are then passed in to the project_conf/checks.xml
(which is the actual Checkstyle configuration). This practice allows us to
have slightly different style requirements for tests versus production code.
The Eclipse Checkstyle Plugin defaults to reading from a file named
.checkstyle in the root of the Eclipse project and that file points to the
location of checks.xml. Unfortunately, checks.xml isn't in the Eclipse
project root and the plugin doesn't know how to look outside of the Eclipse
project directory except by using an absolute path.
To solve this problem, we generate the .checkstyle file programmatically when
running the buildr eclipse task. The template is located at
project_conf/.checkstyle and uses an XML entity to represent the location of
checks.xml. When you run buildr eclipse, we set the value of the
conf_dir entity in project_conf/eclipse-checkstyle.xml to the absolute
path to checks.xml and drop the result into .checkstyle in your Eclipse
project directory.
buildr rspecruns RSpec tests seriallybuildr rspec:parallelruns RSpec tests in parallel when possiblebuildr rspec:failuresruns the tests that failed on the last runbuildr rspec:my_spec_name:my_test_namerunsmy_test_namein themy_spec_namefile
The spec tests are our integration tests. You can run them serially with
buildr rspec. If you want to speed things up use buildr rspec:parallel.
That task will run most of the tests in parallel. A few must still be run
serially to prevent errors (generally import tests are run serially).
You can run specific tests by appending items to the rspec task name. For
example, buildr rspec:vcpu,consumer will run any spec file that begins with
"vcpu" or "consumer". You can exclude tests with a minus sign in front of the
identifier. E.g. buildr rspec:-vcpu will run all spec files that do not
begin with "vcpu".
Additionally, you can provide either line numbers or test names to the task.
For example, buildr rspec:vcpu:62,41 will run the tests on line 62 and 41 of
the vcpu spec file. Likewise, buildr rspec:vcpu,consumer:consumer will run
all tests in the vcpu and consumer specs that have the word "consumer" in the
test name.
The general syntax is
rspec:test_name[,test_name ...][:signifier[,signifier ...]]
where the signifier is either a string or an integer.
Please note that if you need to use a phrase to single out a test, you will
need to quote the task name: buildr "rspec:vcpu:should be valid" to prevent
the shell from interfering. Also note that any phrase or line number you
specify will be applied to all tests. So buildr rspec:vcpu,consumer:62
will only run tests that begin on line 62 in either the vcpu or consumer specs.
This is a limitation of RSpec itself.
When you run RSpec, failed tests are recorded in target/rspec.failures. You
can then use the rspec:failures task to just run failed tests which will then
update the list of failures again. Thus, you can keep running rspec:failures
until the list is empty.
buildr "changeset:my changeset name"Much like therspectask, thechangesettask is followed by a colon and an argument. In this case the argument is a brief description of the nature of the changeset. Be sure to quote the task name to prevent the shell from interpreting the spaces.
buildr erbrenders any templates found under theerbdirectory
This plugin is discussed in detail at http://www.candlepinproject.org/docs/candlepin/auto_conf.html
buildr syntasticcreates.syntastic_class_pathfor the Vim Syntastic pluginbuildr pomcreates apom.xmlfile with the project dependencies in itbuildr rpmlintrunsrpmlinton all*.specfiles