add FailSubcommand, WriteUsageForSubcommand, WriteHelpForSubcommand

[alexflint]

Closes #134 #135 #139

This pr makes it possible to print usage and help text for explicit subcommands. It adds three new functions:

Parser.WriteUsageForSubcommand
Parser.WriteHelpForSubcommand
Parser.FailSubcommand

These behave like their non-subcommand cousins, except that they also take a subcommand name (or a sequence of subcommand names in the case of nested subcommands)

[codecov-commenter]

Codecov Report

Merging #156 (f2f8764) into master (66cb696) will not change coverage.
The diff coverage is 100.00%.

@@            Coverage Diff            @@
##            master      #156   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            5         5           
  Lines          650       679   +29     
=========================================
+ Hits           650       679   +29     

Impacted Files	Coverage Δ
parse.go	100.00% <100.00%> (ø)
usage.go	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 66cb696...f2f8764. Read the comment docs.

[Keithcat1]

Thanks for this. Is there a way to enable printing help messages by default for nested subcommands? As I understand it, even with this pull request merged I would still have to call Parser.WriteHelpForSubcommand in each problematic subcommands code.

[Keithcat1]

Why isn't this merged yet? It doesn't seem to be blocking on anything.

[chrisgoffinet]

@alexflint could we merge this?

[xvello]

Hello @alexflint, thanks a lot for your work on this!

It does look like #139 is not completely fixed yet though: we are still printing p.description, which is obtained by calling Description() on the root command, not the subcommand selected by the user.

I am experimenting on this branch, but I'd love your input. Do you have any opinion on how that should be implemented?

[alexflint]

@xvello you are exactly right. I've reopened #139.