InteractiveConsole does not support -q option

[abalkin]

BPO	21966
Nosy	@abalkin, @benjaminp, @bitdancer, @swarmer
Files	code.patch code_flags_argparse.patch code_flags_argparse_v2.patch

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = 'https://github.com/abalkin'
closed_at = <Date 2014-07-12.20:38:01.865>
created_at = <Date 2014-07-12.16:05:23.710>
labels = ['easy', 'type-feature', 'library']
title = 'InteractiveConsole does not support -q option'
updated_at = <Date 2014-07-12.20:38:01.864>
user = 'https://github.com/abalkin'

bugs.python.org fields:

activity = <Date 2014-07-12.20:38:01.864>
actor = 'belopolsky'
assignee = 'belopolsky'
closed = True
closed_date = <Date 2014-07-12.20:38:01.865>
closer = 'belopolsky'
components = ['Library (Lib)']
creation = <Date 2014-07-12.16:05:23.710>
creator = 'belopolsky'
dependencies = []
files = ['35932', '35933', '35934']
hgrepos = []
issue_num = 21966
keywords = ['patch', 'easy']
message_count = 13.0
messages = ['222850', '222862', '222863', '222865', '222866', '222867', '222868', '222869', '222870', '222871', '222872', '222880', '222882']
nosy_count = 5.0
nosy_names = ['belopolsky', 'benjamin.peterson', 'r.david.murray', 'python-dev', 'anton.barkovsky']
pr_nums = []
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'enhancement'
url = 'https://bugs.python.org/issue21966'
versions = ['Python 3.5']

[abalkin]

When invoked with -q option, python3 prints no banner:
$ python3 -q

>>

However, code.InteractiveConsole does not implement this feature:
$ python3 -mcode -q
Python 3.4.1 (default, May 19 2014, 13:10:29)
[GCC 4.2.1 Compatible Apple LLVM 5.1 (clang-503.0.40)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
(InteractiveConsole)

>>

[swarmer]

Here's a patch.

[abalkin]

That was quick! (Well - I knew it would not be hard.)

I have two questions:

1. How should python3 -q -mcode behave?
2. If we add this patch, should we also attempt to emulate other command line options (-V, -h, etc.)?

[bitdancer]

Whether or not other options are emulated, unimplemented ones should probably be rejected.

[swarmer]

1. How should python3 -q -mcode behave?

I've only now found out about sys.flags. I think we should check for -q both before -m and after, because why not?

2. If we add this patch, should we also attempt to emulate other command line options (-V, -h, etc.)?

As I see it, the module is only concerned with REPL functionality, making these options a bit out of scope.

[abalkin]

In order to implement reasonable rejection behavior, we probably need to add some support for -h.

$ python3 -z
Unknown option: -z
usage: python3 [option] ... [-c cmd | -m mod | file | -] [arg] ...
Try `python -h' for more information.

I don't think we should condition acceptance of this patch on extra features.

AFAICT, the main use of the code module is in embedded situations and the if __name__ == "__main__" behavior is mostly there for demonstration purposes.

On the other hand, something like

$ python3 -mcode -z
Unknown option: -z
usage: python3 -mcode [-q]

is not hard to implement.

[swarmer]

Here's a patch that checks both sys.flags and sys.argv and uses argparse.

[abalkin]

I think we should check for -q both before -m and after, because why not?

If we check for sys.flags.quiet, wouldn't it be surprising to have

$ python3 -mcode -q
>>> import sys; sys.flags.quiet
0

[swarmer]

That's not a very likely scenario and I think the distinction between arguments that are passed to the script and interpreter flags is fairly obvious.

[abalkin]

A nitpick: banner=banner in

+ interact(banner=banner)

is redundant.

+ interact(banner)

would work and is just as clear.

[swarmer]

Yeah, my love for keyword arguments is a bit too big sometimes.