Suggest most simillar command #1272
Conversation
federicoemartinez
requested review from
tleonhardt,
kotfu and
kmvanbrunt
as code owners
cmd2/cmd2.py
Outdated
| @@ -529,6 +531,9 @@ def __init__( | |||
| # Add functions decorated to be subcommands | |||
| self._register_subcommands(self) | |||
|
|
|||
| self.suggest_similar_command = suggest_similar_command | |||
| self.default_suggestion_message = "The most similar command is {}" | |||
There was a problem hiding this comment.
When I doubt I tend to fall back on what bash does. How about:
', did you mean: {}'
There was a problem hiding this comment.
It was my first choice. I changed it because of git:
git comit
git: 'comit' is not a git command. See 'git --help'.
The most similar command is
commit
But sure, I can change it.
cmd2/cmd2.py
Outdated
| if self.suggest_similar_command: | ||
| suggested_command = self._suggest_similar_command(statement.command) | ||
| if suggested_command: | ||
| err_msg = err_msg + os.linesep + self.default_suggestion_message.format(suggested_command) |
There was a problem hiding this comment.
With my alternate suggestion message, no need for line separator.
There was a problem hiding this comment.
I can use a space
Codecov ReportPatch coverage:
Additional details and impacted files@@ Coverage Diff @@
## master #1272 +/- ##
=======================================
Coverage 98.54% 98.55%
=======================================
Files 22 22
Lines 5705 5728 +23
=======================================
+ Hits 5622 5645 +23
Misses 83 83
☔ View full report in Codecov by Sentry. |
|
I see doc jobs are failing but I don't think that's related to this branch |
|
It would be great to have this merged |
|
My PR just merged which also addresses many of the issues you probably ran into. You may want to rebase over latest to see. |
|
I'm not sure what's going on with your branch but my changes are now showing up as changes in your branch. I highly suggest squashing related commits and then rebasing it all on top of latest master. |
federicoemartinez
force-pushed
the
feature/1271
branch
from
b38757c
to
f0b0580
Compare
I think it's done now. I made a mistake rebasing, I'm much more used to mercurial to be honest. |
|
@anselor I applied the changes from the formatter jobs and the doc job. Don't know about the linter. It looks like it is not from my PR |
There pyinvoke commands you can run locally to check |
|
I'm trying
They passed now (Though with python 3.8 in my local setup). |
|
@federicoemartinez The linter problem looks spurious and not applicable to the situation. Would you mind just adding |
|
done! |
|
As far as I'm aware it passed before the last PR. What I've sometimes see happen is the validation library updates and new errors are found that did not previously fail. |
closes #1271
The app can be configured to suggest the most similar command. In that case, the most similar command of the visible commands (more similar than a threshold) will be suggested and added to the default error message.
By default SequenceMatcher.ratio() from the standard library is used but a subclass can define _suggest_similar_command with a new algorithm.