Add alt longname and multiple char support #151
Conversation
| @@ -636,7 +708,10 @@ def parse(_paramlist, _neg_given) | |||
| def type_format ; "" ; end | |||
|
|
|||
| def educate | |||
| (short? ? "-#{short}, " : " ") + "--#{long}" + type_format + (flag? && default ? ", --no-#{long}" : "") | |||
| optionlist = [] | |||
There was a problem hiding this comment.
Noting that this is an undo of the short-argument padding/alignment.
If there can be multiple short args, then the padding doesnt make sense anymore or it would be a bit more complicated to align.
| if sopt =~ INVALID_ARG_REGEX | ||
| raise ArgumentError, "short option name '#{sopt}' can't be a number or a dash" | ||
| end |
There was a problem hiding this comment.
numeric options are interesting though. I'm reminded of git log -1 - would be cool to support a singular pure number as some sort of new type. 🤔
For this piece of code though I agree that the short opt can't be a number
There was a problem hiding this comment.
This is a reformat of this
raise ArgumentError, "a short option name can't be a number or a dash" if sopt =~ ::Optimist::Parser::INVALID_SHORT_ARG_REGEX
but yeah I agree. Implementing commands like ls (e.g. ls -1 or ls -d1) with optimist is impractical as-is due to the numerical short-opt restriction. I found this out trying to implement a version of colorls some time back. OTOH, there are unintended consequences to supporting this, bc it's harder to tell what's a negative integer vs. an option, especially since optmist allows ambiguous integer lists.
| @auto = false | ||
| return | ||
| end | ||
| raise ArgumentError, "Cannot use :none with any other values in short option: #{values.inspect}" |
There was a problem hiding this comment.
Updated to prevent :none from being used with any other value (including itself)
|
Gah I'm sorry - I keep merging the other ones and this keeps conflicting 😅 |
Added: - Align the short and long forms into their own columns in the help output (#145 - thanks akhoury6) - Add support for DidYouMean when long options are spelled incorrectly (#150 - thanks nanobowers) - Using `permitted:` restricts the allowed values that a end-user inputs to a pre-defined list (#147 - thanks akhoury6) - Add exact_match to settings, defaulting to inexact matching (#154 - thanks nanobowers) - Add setting to disable implicit short options (#155 - thanks nanobowers) - Add alt longname and multiple char support (#151 - thanks nanobowers) - Permitted regexp/range support (#158, #159- thanks nanobowers) - Add some examples (#161 - thanks nanobowers) Changed: - Enable frozen_string_literal for future-ruby support (#149, #153 - thanks nanobowers) - Refactor constraints (#156 - thanks nanobowers) - Fix assert_raises to assert_raises_errmatch (#160 - thanks nanobowers)
It can be useful to have more than one option name for the same option. E.g. you may have a
--concatoption, but intuitively to some users they would rather--append. Or sometimes you want to change names of the option but keep backward compatibility.This enables short options and long options to have more than one name.
For short options, the
chars:can now take a list of strings/symbols (previously was just a symbol or char)For long options, the
alt:takes additional alternate long options. The value provided inlong:is still respected, any name(s) given inalt:are used in addition to the default long-argument or what was provided inlong:There was some refactoring of how short and long options work, which pushed them down into a
ShortNamesandLongNamesclass. This caused an update a couple of existing tests. I think this is fine unless someone out there was accessing internals of theOptionclass, probing theshortandlonginstance variables and expecting them to be strings.@miq-bot add-label enhancement
@miq-bot add-reviewer @Fryguy