* forgot to remove 'no_arg' from Pry::Command::CodeCollector, this method no longer exists and so 'gist' (when not given any params) would break. Replaced no_arg with empty string "" so that 'gist' gists current object/method by default.
Final adjustments made to CHANGELOG and CONTRIBUTORS
This was a horrible mistake, it's stupidly annoying * in the case of 'rake pry' it warngs every time we do 'cd' (since 'cd' is a top-levle method defined by rake) That's all i can think of right now, but that's still annoying enough.
And rewrite edit-method / show-command to use this API.
This enables a simplified UI, i.e: gist my_file.rb instead of gist -f my_file.rb and `gist my_method` instead of `gist -m my_method`. Unfortunately, we had to kill our tests in the process of doing this (gist_spec.rb is just commented out) - however since play and save-file have tests and they use the same Pry::CodeCollector object 'gist' is indirectly tested. Nonetheless, we should re-add tests at some point
There was a bug where if 'edit --ex' was invoked in a patched method context it would instead to the equivalent of 'edit --method' (patching the current method rather than the exception)
This allows callers to know that CC is working on a file rather than ruby code.
It was creating a new Slop instance for its own needs, thus and so remove the previous abstraction for Slop commands. The `ClassCommand::Options` class should be removed because we don't need it anymore.
Make them use the new Slop syntax.
* use a plain old Slop instance instead of `Options`; * remove `ClassCommand::Options` and its tests; * add clearer subcommand example (for `Command#subcommand` method); * refactor `ClassCommand#slop`; * refactor `ClassCommand#complete`. Slop v3.4.0 has introduced full-featured subcommands. There is no need in Pry specific code anymore.
Slop v3.4.0 includes new syntax for subcommands. And we're going to utilise it.
Many plugins only define a puts method on the output object. This change removse the ability to customize the formatter used when outputting variables in `ls`. We should reconsider the best way to do this.
Basically, the condition was a bit wrong. Let's hope this commit fixes it.
Let's try this, if it annoys people, we'll just revert it in a patch level
This is because pry_tester doesn't work when the eval_string is mutated by a command
These implementations have slightly different APIs for getting your temp directory path.
The "edit" command already intelligently escapes arguments, if need. This commit also fixes four failing tests on Windows.
Some tests are still failing, though. This commit fixes wrong usage of temporary paths on Windows. Please, note: the `edit` command is working well; the issue is in poor tests. Prettify some ugly code in `spec/editor_spec.rb`.
Also, there was a failing test on Windows ("should shell-escape files"). You will see it if you're running tests on Windows. Make it conditional.
Unfortunately nested assignments to @old_editor meant that the original @old_editor was permanently lost. This caused sporadic failures of spec/editor_spec.rb which was relying on Pry.config.editor containing a string.
Also, move path escaping to `Editor` class. The test is currently failing (looks like a bug in `Editor`).
The completion proc returning nil causes errors in input systems that only expect strings (e.g. Coolline).
This reverts commit 1c33a73. Looks unpleasant for most of Pry users.