Find step impl when there are no markers in gherkin about behave.vim HOT 5 CLOSED

I have been thinking about this for some time and one of the possible solutions is to use a custom formatter, so you can get all the available steps, and where they are. Whenever I have time, I will test this idea, if you haven't found a solution already.

from behave.vim.

Very Interesting. I was thinking about something in the lines of behave --steps-catalog or behave -d --show-source to gather this information but I hadn't tried that yet.

Take a look at these options, they already show the steps and its locations.

from behave.vim.

This is the way I search steps

behave --steps-catalog <feature file/folder> | egrep "(Given|When|Then)" | awk '{$1=$1};1' | cut -d " " -f2- | uniq | grep fill

from behave.vim.

Hej folks,
do you think the steps.usage formatter could be of any help? Example in the current repo:

$ behave --dry-run -f steps.usage --no-summary tests/fixtures
@given('I am using "{prg}"')              # tests/fixtures/steps/steps.py:4
  Given I am using "emacs"                # tests/fixtures/test.feature:3

@step('I am '{perc}' '{mood}'')           # tests/fixtures/steps/steps.py:9
  And I am '100%' 'not happy'             # tests/fixtures/test.feature:4

@when('I switch to "{prg}"')              # tests/fixtures/steps/steps.py:14
  When I switch to "vim"                  # tests/fixtures/test.feature:5

@then('I will be happy')                  # tests/fixtures/steps/steps.py:19
  Then I will be happy                    # tests/fixtures/test.feature:6

which gives the following possibilites:

1. Going from step to implementation
   When invoked in a feature file, remember the current line, and find the path:line in the steps.usage output. Then jump to the corresponding implementation as given by the second column of the first line of the block.
2. Listing steps acc. to implementation
   When invoked from inside a step implementation function, find the decorator line in the output of steps.usage, and populate the qf list with the entries in the second column of the block.

Sooo my vimscript foo is not the best, but unless you say you'd have this implemented super-quickly, I'd attempt an implementation :)
What do you think about the approach? I use behave 1.2.6

from behave.vim.

Hi again, I have been looking for a solution. I implemented something for FZF.

@pylipp thanks to your comment I found the steps formatter.

behave features/ -d -f steps 2> /dev/null | \
awk -F " *# " '/\s*(Given|When|Then|\*)/ {print $1"\t"$2}' | \
fzf -d "\t" --with-nth=1 \
   --bind 'enter:execute(echo {} | cut -f2 | pbcopy )' \
   --bind 'tab:execute(echo {} | cut -f1 | awk "{\$1=\$1};1" | pbcopy )'

Tab copy the step name.
Enter copy the step location

Note: It is a bit slow... 😃
Note2: Grep could be used instead.
Note3: The output from awk could be saved to a file.
Note4: Added in fzf wiki

PD: I am working on a step formatter for Behave with the output of vim-grep, so I can remove the awk from the command.

from behave.vim.