When running a test file using shoulda-context, the names found are not the full names of the test, and therefore cannot be run via t.

Say I have a rails model post.rb and a spec post_spec.rb.

If I input <leader>T while the post_spec.rb buffer is open, the command is as follows:

rspec spec/models/post_spec.rb  -l 16

This works just fine.

However, let's say I input <leader>t while my cursor is on line 16 of the spec. This is the following command:

rspec spec/models/post_spec.rb  -l 16

However, the actual command should be:

rspec spec/models/post_spec.rb:16

I have the current version of rspec installed, 3.4.1. Let me know if you need any other pieces of information.

After setting let g:rubytest_in_quickfix = 1 everything looked to function as expected:

• Run test
• Errors added to quickfix

However when the output finishes and the first error is opened the following happens:

It looks like the filename is getting prepended with [ which is leading to the inability to open the actual file.

I am a quickfix noob and tried to use :set modifiable in hopes that I could just remove the [ char and the file would open as desired. Hopefully someone has seen this issue before and there is a quick fix. (in the meantime I'll dig into the code and see if anything stands out)

Investigation
How quickfix list is created in the ExecTest(cmd) function

cex system(cmd)

The code I'm editing is at "/home/joe/Rails Projects/myproject/test/unit/auth_test.rb" When I run t I get a "ruby: No such file or directory /home/joe/Rails" error.

It looks like rubytest can't handle spaces in the directory

I'm sure this is my local problem, but when I run a single unit test (leader-t, which I have remapped to leader-\ ) the tests run in a new split window at the bottom of macvim, but as soon as they complete, they disappear - so you can't see the results.

This happens regardless of whether the test passes or fails.

This is not in the quickfix window, this is in the default mode.

I'm just trying to get to grips with vim, and I really can't figure this out. Any pointers?

I have the janus plugin bungle installed, and my .vimrc.before looks like this:

map \ RubyTestRun " change from t to
map ] RubyFileRun " change from T to ]
map / RubyTestRunLast " change from l to /

and the .vimrc.after looks like this:

" launch peepopen via command T
if has("gui_macvim")
map PeepOpen
end

" execute the current file as ruby with command R
map :!ruby %

" don't run the rubytest actions in the quickfix window
let g:rubytest_in_quickfix = 0

color pastelsondark
set gfn=Monaco:h12

" disable the macvim toolbar
if has("gui_macvim")
set guioptions=egmrt
endif

" show / hide NERDTree with \d
nnoremap d :NERDTreeToggle

" remove trailing spaces with \s
nnoremap s :%s/\s+$//e

" remove trailing spaces when saving any ruby file
autocmd BufWritePre *rb :%s/\s+$//e

I use this plugin with MiniTest::Spec.
I added new variables to my vimrc as follows:

let g:rubytest_cmd_spec = "ruby %p"
let g:rubytest_cmd_example = "ruby %p -n /%c/"

When I run the testcase(spec) by pressing t,
It seems it doesn't reflect proper spec.
Seems doesn't detect the proper line properly.

I'm getting this segfault when hitting -t in a cucumber feature:

cucumber features/sign_up.feature -n Create an account from the homepage                                                                                         
/Users/adarsh/.rvm/gems/ruby-1.9.3-p125/gems/json-1.6.5/ext/json/ext/json/ext/parser.bundle: [BUG] Segmentation fault
ruby 1.8.7 (2010-01-10 patchlevel 249) [universal-darwin11.0]

Command terminated

Press ENTER or type command to continue

I'm using ruby 1.9.3p125 so this error confuses me.

Upon executing the command, # gets evaluated by Vim and becomes the working directory.

Here's an example:

test "#something" do

end

will execute:

ruby test.rb -n /test_/Users/cjoudreysomething/

I use Vim 7.4 and writing cucumber features for a rails app. I use the default mapping, ,t nad ,T works both fine, but ,l not. it only shows :call <SNR>52_RunLast() under the status line, but not running the last feature.

I suppose this maybe a bug, please take a look into it, thanks.

I am currently running the git master version of ruby-test and using a
rvm gemset with 1.9.3 ruby. From the cli everything works but when hitting \t or \T

I get an error in the form of:
~/.rvm/rubies/ruby-1.9.3-p392/lib/ruby/site_ruby/1.9.1/rubygems/custom_require.rb|36| in `require': cannot load such file -- spork (LoadError)

Is this a known issue or am I doing something wrong? (which I guess seems very plausible)

On line 88, I see the test command defined by the user (or the script default) is replaced with one that includes -Itest -rtest_helper.

This is bad for me for two reasons:

1. My work app already requires test_helper fully qualified in-test so this results in warnings about reassigning constants
2. Some of my apps don't have a test_helper so I can't use this plugin without touching an empty one

If -Iwhatever and -rwhatever are needed, the user can add them using your configuration options.

This looked intentional so I thought an Issue vs a Pull req would be better since there might be an explanation for it.

Thanks!

When attempting to output results to Quickfix, the plugin simply echos call <72>Run(1). Nothing else happens and the quickfix buffer remains empty.

I've noticed that many Vim ruby test runners use something like this to run individual Test::Unit cases:

let g:rubytest_cmd_testcase = "ruby %p -n '/%c/'"

I'm not sure if there's a reason for the // in this command, but it causes multiple tests to run when the name of one test is a substring of the name of another test. For example, if you have the following:

def test_foo; end
def test_foo_bar; end
def test_foo_baz; end

...then there is no way to run test_foo alone. It will run all three tests.

Simply omitting the // here will make it run the desired single test.

Is there something I'm missing or is this just a convention that's been carried over from earlier test runners?

The problem is that the "test" directory is not added to my LOAD_PATH when the rubytest_cmd_* does not start with "ruby".

In my vimrc I use bundle exec when running my ruby tests:

let g:rubytest_cmd_test = "bundle exec ruby %p"
let g:rubytest_cmd_testcase = "bundle exec ruby %p -n '/%c/'"

When I run a single test case I get the following:

bundle exec ruby test/controllers/training_controller_test.rb -n /test_get_content_inserts_base_tag_if_none_exists/
test/controllers/training_controller_test.rb:1:in `require': cannot load such file -- test_helper (LoadError)
        from test/controllers/training_controller_test.rb:1:in `<main>'

shell returned 1

Changing this line to: let cmd = substitute(cmd, 'ruby ', 'ruby -Itest ', '') fixes my issue. I can submit a PR but wanted to ask for thoughts on a better way to handle this.

When I run my tests the tests run fine, but where it should wait for me to press a key to close the terminal window it does not wait, so I can never see the results of my tests.

Steps to reproduce:

1. Go to rails project root directory.
2. Launch vim
3. type :e spec/<path_to_spec_file>
4. Launch all tests.

OS Windows XP

Error:
d:\s\ruby\lib\ruby\gems\1.8\gems\rspec-1.2.7\lib\spec\runner\options.rb|255| in `files_to_load': File or directory not found: speccontrollershome_controller_spec.rb (RuntimeError)