xpra help: present less options when subcommand is wrong

[stdedos]

I usually don't remember all the xpra sub-commands. Apart from opening ​http://xpra.org/manual, or xpra --help | head, I cannot "easily enough" see the options.

Plain xpra --help is 597 lines, too big to navigate.

xpra help gives:

$ xpra help
xpra for python 2.7 is not installed
 retrying with python3
xpra initialization error:
 invalid mode 'help'
$ xpra wrong
xpra for python 2.7 is not installed
 retrying with python3
xpra initialization error:
 invalid mode 'wrong'

---

Proposal: make xpra help or xpra wrong-command to present:

$ xpra update
Usage: 
	xpra start [DISPLAY]
	xpra start-desktop [DISPLAY]
	xpra upgrade [DISPLAY]
	xpra attach [DISPLAY]
	xpra detach [DISPLAY]
	xpra screenshot filename [DISPLAY]
	xpra info [DISPLAY]
	xpra control DISPLAY command [arg1] [arg2]..
	xpra print DISPLAY filename
	xpra version [DISPLAY]
	xpra showconfig
	xpra list
	xpra sessions
	xpra launcher
	xpra stop [DISPLAY]
	xpra exit [DISPLAY]
	xpra list-mdns
	xpra mdns-gui
	xpra shadow [DISPLAY]

See more options with `xpra --help`

For bonus points:

• Add a "did you mean" feature?
• Group similar-argument commands together, if that is not the case at this time.

[Antoine Martin]

Not sure how easy it is to hook into the command line parsing code. We'll see.

[stdedos]

I am only proposing it because I think it used to be like that (a year ago? a year and a half?). I don't know if you switched argparse systems, which then made it be like that and "cannot be done".

Feel free to ignore this if some of the above are not true.

[Antoine Martin]

Updates:

• r24806 re-order help
• r24807 adds "help" subcommand, which is also shown for "wrong-subcommand"

[stdedos]

Fixed in r24825:

Xpra-Python3-x86_64_4.0-r24825\xpra_cmd" sh
invalid subcommand 'sh'
Usage:
        xpra shadow [DISPLAY]
        xpra (This xpra installation does not support starting local servers.)
        xpra attach [DISPLAY]
        xpra detach [DISPLAY]
        xpra info [DISPLAY]
        xpra version [DISPLAY]
        xpra stop [DISPLAY]
        xpra exit [DISPLAY]
        xpra screenshot filename [DISPLAY]
        xpra control DISPLAY command [arg1] [arg2]..
        xpra print DISPLAY filename
        xpra showconfig
        xpra list
        xpra sessions
        xpra launcher
        xpra list-mdns
        xpra mdns-gui
        xpra

see 'man xpra' or 'xpra --help' for more details

Three nitpicks:

• First capital in invalid, see
• Empty xpra (except if you want to emphasize that plain xpra is a valid command i.e. opens the GUI)
• xpra (This xpra installation

I am not sure why would you write it like that:

trunk/src/xpra/scripts/parsing.py:209

def get_usage():
    command_options = []
    if supports_server:
        command_options = ["start [DISPLAY]",
                           "start-desktop [DISPLAY]",
                           "upgrade [DISPLAY]",
                           ] + command_options
    if supports_shadow:
        command_options.append("shadow [DISPLAY]")
    if not supports_server:
        command_options.append("(This xpra installation does not support starting local servers.)")

and not e.g.

def get_usage():
    command_options = []
    if not supports_server:
        command_options.append("(This xpra installation does not support starting local servers.)")
    else:  # supports_server
        command_options = ["start [DISPLAY]",
                           "start-desktop [DISPLAY]",
                           "upgrade [DISPLAY]",
                           ] + command_options
    if supports_shadow:
        command_options.append("shadow [DISPLAY]")

I also don't see what is the usage of ] + command_options.

[Antoine Martin]

First capital in invalid, see

r24840

Empty xpra (except if you want to emphasize that plain xpra is a valid command i.e. opens the GUI)

I do.

I am not sure why would you write it like that:

Because this code evolved from somewhere else and I was trying to minimize the changes, see r24807.
r24841 addresses that, and more.

[stdedos]

I guess this is not going in v3?

[Antoine Martin]

I guess this is not going in v3?

No, this is not a bug fix.

This would be suitable for milestone/3.1 however.