Settings for changing optimist behavior

[nanobowers]

Per longstanding PR #83, I have some ideas on ways to enhance behavior of optimist in ways that may not be backward compatible (see below). In order to make these new behaviors opt-in, it would be nice to have a way to tell optimist how to enable them.

Some implementation ideas include:

1. inside Optimist::options block, e.g.:

opts = Optimist.options do
  enable_some_setting
  opt :elk, "elk"
end

2. as arguments to Optimist.options

opts = Optimist.options(some_setting: true) do
  opt :yak, "yak"
end

3. Via class methods to set a global

Optimist.some_setting = true
opts = Optimist.options do
  opt :moose, "moose"
end

Downsides are:

• Alternative-1 causes order dependence (of the setting enablement and the opt command) which feels like it's not in the spirit of most DSLs.
• Alternative-2 causes problems with the cloaker function that is in Optimist.new
• Alternative-3 is ugly since it applies globally.

Would like maintainer/interested parties from previous PR's opinions ( @kbrock @Fryguy @jrafanie @chrisarcand ) on the best implementation before trying to resurrect #83.

For reference, suggested behavior enhancements would include the following and will probably be in separate PRs going forward:

• :enable_inexact_match : Allow minimum unambigous number of characters to match a long option (similar to Perl Getopt::Long)
• :disable_auto_short_opts : Prevent automatic creation of short options
• :enable_suggestions : Enables suggestions when unknown arguments are given and DidYouMean is installed.

[nanobowers]

Went with option 2 above, PR is #108 which adds inexact_match and suggestions feature. Would be reasonably straightforward to do option 1 above.
This question and associated requests for features has been sitting around over two years now in some form (#83 , etc) if feature additions are not of interest to the maintainers (@kbrock , @Fryguy ), please let me know.

[nanobowers]

Feature forking. https://github.com/nanobowers/optimist_xl