Fixed parser bug assuming value parsed is part of command#6636
Fixed parser bug assuming value parsed is part of command#6636
Conversation
Codecov Report
@@ Coverage Diff @@
## dev #6636 +/- ##
===================================
Coverage 0% 0%
===================================
Files 11 11
Lines 145 145
Branches 11 11
===================================
Misses 145 145Continue to review full report at Codecov.
|
| value=value) | ||
| if not self.command_source: | ||
| # parser has no `command_source`, value is part of command itself | ||
| error_msg = "{prog}: '{value}' is not in the '{prog}' command-group. See '{prog} --help'.".format( |
There was a problem hiding this comment.
No need to hyphenate command group.
|
Can we add a test for this? |
| telemetry.set_user_fault(error_msg) | ||
| logger.error(error_msg) | ||
| candidates = difflib.get_close_matches(value, action.choices, cutoff=0.8) | ||
| candidates = difflib.get_close_matches(value, action.choices, cutoff=0.7) |
There was a problem hiding this comment.
What's the rationale behind this change?
There was a problem hiding this comment.
I felt it a little too strict; 0.8 allows zero edits for words under 5 letters; and at most 1 edit until words of length 10 or more.
0.7 allows a edit for four-character words and 2 edits for words of length 7, 8, and 9.
This isn't the main purpose of this PR, however, and I can change this back if you guys prefer
There was a problem hiding this comment.
I'm fine with it--just wanted to know the reason.
|
@troydai I'll look into adding a test |
4 participants
Added testing for spellchecker and msg choice
Fixes: #6562