I am using argparse's action to add various data to a class. I would like to use that action on the default value if that arg is not provided at the command line. Is this possible?
Thanks!
argparse does not use the action when applying the default. It just uses setattr. It may use the type if the default is a string. But you can invoke the action directly.
Here I use a custom action class borrowed from the documentation. In the first parse_args nothing happens. Then I create a new namespace, and invoke the action on the default. Then I pass that namespace to parse_args. To understand this, you many need to import it into an interactive shell, and examine the attributes of the namespace and action.
# sample custom action from docs
class FooAction(argparse.Action):
def __call__(self, parser, namespace, values, option_string=None):
print('Setting: %r %r %r' % (namespace, values, option_string))
setattr(namespace, self.dest, 'action:'+values)
p = argparse.ArgumentParser()
a1 = p.add_argument('--foo', action=FooAction, default='default')
print 'action:',a1
print p.parse_args([])
ns = argparse.Namespace()
a1(p, ns, a1.default, 'no string') # call action
print p.parse_args([],ns)
print p.parse_args(['--foo','1'],ns)
which produces:
action: FooAction(option_strings=['--foo'], dest='foo', nargs=None, const=None, default='default', type=None, choices=None, help=None, metavar=None)
Namespace(foo='default')
Setting: Namespace() 'default' 'no string'
Namespace(foo='action:default')
Setting: Namespace(foo='action:default') '1' '--foo'
Namespace(foo='action:1')
I tailored the output to highlight when the action is being used.
Here's a way of performing a special action on an argument that isn't given on the command line (or given with a value == to the default). It's a simplification of the class given in https://stackoverflow.com/a/24638908/901925.
class Parser1:
def __init__(self, desc):
self.parser = argparse.ArgumentParser(description=desc)
self.actions = []
def milestone(self, help_='milestone for latest release.', default=None):
action = self.parser.add_argument('-m', '--milestone', help=help_, default=default)
self.actions.append(action)
return self
def parse(self):
args = self.parser.parse_args()
for a in self.actions:
if getattr(args, a.dest) == a.default:
print 'Please specify', a.dest
values = raw_input('>')
setattr(args, a.dest, values)
return args
print Parser1('desc').milestone(default='PROMPT').parse()
The prompting is done after parse_args. I don't see any reason to call parse_args again.
I needed to prompt the user if an option was not specified - that's how I did it:
class _PromptUserAction(argparse.Action):
def __call__(self, parser, namespace, values, option_string=None):
if values == self.default:
print 'Please specify', self.dest
values = raw_input('>')
setattr(namespace, self.dest, values)
class Parser:
def __init__(self, desc, add_h=True):
self.parser = argparse.ArgumentParser(description=desc, add_help=add_h,
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
#actions to be run on options if not specified (using default to check)
self.actions = []
#staticmethod
def new(description, add_help=True):
return Parser(description, add_help)
# ...
def milestone(self, help_='Specify the milestone for latest release.'):
action = self.parser.add_argument('-m', '--milestone',
dest='milestone',
action=_PromptUserAction,
default='PROMPT', # needed I think
type=str,
help=help_)
self.actions.append(action)
return self
def parse(self):
"""
Return an object which can be used to get the arguments as in:
parser_instance.parse().milestone
:return: ArgumentParser
"""
args = self.parser.parse_args()
# see: http://stackoverflow.com/a/21588198/281545
dic = vars(args)
ns = argparse.Namespace()
for a in self.actions:
if dic[a.dest] == a.default:
a(self.parser, ns, a.default) # call action
# duh - can I avoid it ?
import sys
return self.parser.parse_args(sys.argv[1:],ns)
I am interested if this can somehow be done without having to reparse the args (the import sys part). Maybe some constructor options for argparse.Action ?
Related
import argparse
class A:
def __init__(self,val=1):
self.val = val
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--val', default=1)
args = parser.parse_args()
a = A(val=args.val)
Quite often I find myself with a structure as shown above, where a class takes an optional argument with a default value. For the argparse I need to define this default value again. If it changes, I have to change it in multiple spots.
Leaving the default value in the class is not an option, as the class might be initialised from somewhere else as well.
I considered to set the argparse default value to None and only pass it when it's not None, but with multiple variables that will lead to convoluted code.
The only other thing coming to my mind is a section on the top of the page to set constants like "DEFAULT_val" so the defaults are defined only once, but I'm wondering if there's a better solution.
import argparse
import inspect
def get_default_args(func):
signature = inspect.signature(func)
return {
k: v.default
for k, v in signature.parameters.items()
if v.default is not inspect.Parameter.empty
}
class A:
def __init__(self,val=1):
self.val = val
if __name__ == '__main__':
default_values = get_default_args(A.__init__)
parser = argparse.ArgumentParser()
parser.add_argument('--val', default=default_values['val'])
args = parser.parse_args()
a = A(val=args.val)
Found the solution in the answers to another question
I want to create a command line flag that can be used as
./prog.py --myarg=abcd,e,fg
and inside the parser have this be turned into ['abcd', 'e', 'fg'] (a tuple would be fine too).
I have done this successfully using action and type, but I feel like one is likely an abuse of the system or missing corner cases, while the other is right. However, I don't know which is which.
With action:
import argparse
class SplitArgs(argparse.Action):
def __call__(self, parser, namespace, values, option_string=None):
setattr(namespace, self.dest, values.split(','))
parser = argparse.ArgumentParser()
parser.add_argument('--myarg', action=SplitArgs)
args = parser.parse_args()
print(args.myarg)
Instead with type:
import argparse
def list_str(values):
return values.split(',')
parser = argparse.ArgumentParser()
parser.add_argument('--myarg', type=list_str)
args = parser.parse_args()
print(args.myarg)
The simplest solution is to consider your argument as a string and split.
#!/usr/bin/env python3
import argparse
parser = argparse.ArgumentParser()
parser.add_argument("--myarg", type=str)
d = vars(parser.parse_args())
if "myarg" in d.keys():
d["myarg"] = [s.strip() for s in d["myarg"].split(",")]
print(d)
Result:
$ ./toto.py --myarg=abcd,e,fg
{'myarg': ['abcd', 'e', 'fg']}
$ ./toto.py --myarg="abcd, e, fg"
{'myarg': ['abcd', 'e', 'fg']}
I find your first solution to be the right one. The reason is that it allows you to better handle defaults:
names: List[str] = ['Jane', 'Dave', 'John']
parser = argparse.ArumentParser()
parser.add_argument('--names', default=names, action=SplitArgs)
args = parser.parse_args()
names = args.names
This doesn't work with list_str because the default would have to be a string.
Your custom action is the closest way to how it is done internally for other argument types. IMHO there should be a _StoreCommaSeperatedAction added to argparse in the stdlib since it is a somewhat common and useful argument type,
It can be used with an added default as well.
Here is an example without using an action (no SplitArgs class):
class Test:
def __init__(self):
self._names: List[str] = ["Jane", "Dave", "John"]
#property
def names(self):
return self._names
#names.setter
def names(self, value):
self._names = [name.strip() for name in value.split(",")]
test_object = Test()
parser = ArgumentParser()
parser.add_argument(
"-n",
"--names",
dest="names",
default=",".join(test_object.names), # Joining the default here is important.
help="a comma separated list of names as an argument",
)
print(test_object.names)
parser.parse_args(namespace=test_object)
print(test_object.names)
Here is another example using SplitArgs class inside a class completely
"""MyClass
Demonstrates how to split and use a comma separated argument in a class with defaults
"""
import sys
from typing import List
from argparse import ArgumentParser, Action
class SplitArgs(Action):
def __call__(self, parser, namespace, values, option_string=None):
# Be sure to strip, maybe they have spaces where they don't belong and wrapped the arg value in quotes
setattr(namespace, self.dest, [value.strip() for value in values.split(",")])
class MyClass:
def __init__(self):
self.names: List[str] = ["Jane", "Dave", "John"]
self.parser = ArgumentParser(description=__doc__)
self.parser.add_argument(
"-n",
"--names",
dest="names",
default=",".join(self.names), # Joining the default here is important.
action=SplitArgs,
help="a comma separated list of names as an argument",
)
self.parser.parse_args(namespace=self)
if __name__ == "__main__":
print(sys.argv)
my_class = MyClass()
print(my_class.names)
sys.argv = [sys.argv[0], "--names", "miigotu, sickchill,github"]
my_class = MyClass()
print(my_class.names)
And here is how to do it in a function based situation, with a default included
class SplitArgs(Action):
def __call__(self, parser, namespace, values, option_string=None):
# Be sure to strip, maybe they have spaces where they don't belong and wrapped the arg value in quotes
setattr(namespace, self.dest, [value.strip() for value in values.split(",")])
names: List[str] = ["Jane", "Dave", "John"]
parser = ArgumentParser(description=__doc__)
parser.add_argument(
"-n",
"--names",
dest="names",
default=",".join(names), # Joining the default here is important.
action=SplitArgs,
help="a comma separated list of names as an argument",
)
parser.parse_args()
I know this post is old but I recently found myself solving this exact problem. I used functools.partial for a lightweight solution:
import argparse
from functools import partial
csv_ = partial(str.split, sep=',')
p = argparse.ArgumentParser()
p.add_argument('--stuff', type=csv_)
p.parse_args(['--stuff', 'a,b,c'])
# Namespace(stuff=['a', 'b', 'c'])
If you're not familiar with functools.partial, it allows you to create a partially "frozen" function/method. In the above example, I created a new function (csv_) that is essentially a copy of str.split() except that the sep argument has been "frozen" to the comma character.
I have to either store the command line argument in a variable or assign a default value to it.
What i am trying is the below
import sys
Var=sys.argv[1] or "somevalue"
I am getting the error out of index if i don't specify any argument. How to solve this?
Var=sys.argv[1] if len(sys.argv) > 1 else "somevalue"
The builtin argparse module is intended for exactly these sorts of tasks:
import argparse
# Set up argument parser
ap = argparse.ArgumentParser()
# Single positional argument, nargs makes it optional
ap.add_argument("thingy", nargs='?', default="blah")
# Do parsing
a = ap.parse_args()
# Use argument
print a.thingy
Or, if you are stuck with Python 2.6 or earlier, and don't wish to add a requirement on the backported argparse module, you can do similar things manually like so:
import optparse
opter = optparse.OptionParser()
# opter.add_option("-v", "--verbose") etc
opts, args = opter.parse_args()
if len(args) == 0:
var = "somevalue"
elif len(args) == 1:
var = args[0]
else:
opter.error("Only one argument expected, got %d" % len(args))
print var
Good question.
I think the best solution would be to do
try:
var = sys.argv[1]
except IndexError:
var = "somevalue"
Try the following with a command-line-processing template:
def process_command_line(argv):
...
# add your option here
parser.add_option('--var',
default="somevalue",
help="your help text")
def main(argv=None):
settings, args = process_command_line(argv)
...
print settings, args # <- print your settings and args
Running ./your_script.py with the template below and your modifications above prints {'var': 'somevalue'} []
For an example of a command-line-processing template see an example in Code Like a Pythonista: Idiomatic Python (http://python.net/~goodger/projects/pycon/2007/idiomatic/handout.html#command-line-processing):
#!/usr/bin/env python
"""
Module docstring.
"""
import sys
import optparse
def process_command_line(argv):
"""
Return a 2-tuple: (settings object, args list).
`argv` is a list of arguments, or `None` for ``sys.argv[1:]``.
"""
if argv is None:
argv = sys.argv[1:]
# initialize the parser object:
parser = optparse.OptionParser(
formatter=optparse.TitledHelpFormatter(width=78),
add_help_option=None)
# define options here:
parser.add_option( # customized description; put --help last
'-h', '--help', action='help',
help='Show this help message and exit.')
settings, args = parser.parse_args(argv)
# check number of arguments, verify values, etc.:
if args:
parser.error('program takes no command-line arguments; '
'"%s" ignored.' % (args,))
# further process settings & args if necessary
return settings, args
def main(argv=None):
settings, args = process_command_line(argv)
# application code here, like:
# run(settings, args)
return 0 # success
if __name__ == '__main__':
status = main()
sys.exit(status)
This is more like a code design question. what are good default values for optional options that are of type string/directory/fullname of files?
Let us say I have code like this:
import optparse
parser = optparse.OptionParser()
parser.add_option('-i', '--in_dir', action = "store", default = 'n', help = 'this is an optional arg')
(options, args) = parser.parse_args()
Then I do:
if options.in_dir == 'n':
print 'the user did not pass any value for the in_dir option'
else:
print 'the user in_dir=%s' %(options.in_dir)
Basically I want to have default values that mean the user did not input such option versus the actual value. Using 'n' was arbitrary, is there a better recommendation?
You could use an empty string, "", which Python interprets as being False; you can simply test:
if options.in_dir:
# argument supplied
else:
# still empty, no arg
Alternatively, use None:
if options.in_dir is None:
# no arg
else:
# arg supplied
Note that the latter is, per the documentation the default for un-supplied arguments.
How about just None?
Nothing mandates that the default values must be of the same type as the option itself.
import optparse
parser = optparse.OptionParser()
parser.add_option('-i', '--in_dir', default=None, help='this is an optional arg')
(options, args) = parser.parse_args()
print vars(options)
(ps. action="store" isn't required; store is the default action.)
I'm trying to get to know optparse a bit better, but I'm struggling to understand why the following code behaves the way it does. Am I doing something stupid?
import optparse
def store_test(option, opt_str, value, parser, args=None, kwargs=None):
print 'opt_str:', opt_str
print 'value:', value
op = optparse.OptionParser()
op.add_option('-t', '--test', action='callback', callback=store_test, default='test',
dest='test', help='test!')
(opts, args) = op.parse_args(['test.py', '-t', 'foo'])
print
print 'opts:'
print opts
print 'args:'
print args
Output:
opt_str: -t
value: None
opts:
{'test': 'test'}
args:
['foo']
Why is 'foo' not being passed to store_test() and instead being interpreted as an extra argument? Is there something wrong with op.parse_args(['-t', 'foo'])?
↓
http://codepad.org/vq3cvE13
Edit:
Here's the example from the docs:
def store_value(option, opt_str, value, parser):
setattr(parser.values, option.dest, value)
[...]
parser.add_option("--foo",
action="callback", callback=store_value,
type="int", nargs=3, dest="foo")
You're missing a "type" or "nargs" option attribute:
op.add_option('-t', '--test', action='callback', callback=store_test, default='test',
dest='test', help='test!', type='str')
This option will cause it to consume the next argument.
Reference:
http://docs.python.org/library/optparse.html#optparse-option-callbacks
type
has its usual meaning: as with the "store" or "append" actions, it instructs optparse
to consume one argument and convert it to type. Rather than storing the converted
value(s) anywhere, though, optparse passes it to your callback function.
nargs
also has its usual meaning: if it is supplied and > 1, optparse will consume nargs
arguments, each of which must be convertible to type. It then passes a tuple of converted
values to your callback.
This seems to be the relevant code from optparse.py:
def takes_value(self):
return self.type is not None
def _process_short_opts(self, rargs, values):
[...]
if option.takes_value():
# Any characters left in arg? Pretend they're the
# next arg, and stop consuming characters of arg.
if i < len(arg):
rargs.insert(0, arg[i:])
stop = True
nargs = option.nargs
if len(rargs) < nargs:
if nargs == 1:
self.error(_("%s option requires an argument") % opt)
else:
self.error(_("%s option requires %d arguments")
% (opt, nargs))
elif nargs == 1:
value = rargs.pop(0)
else:
value = tuple(rargs[0:nargs])
del rargs[0:nargs]
else: # option doesn't take a value
value = None
option.process(opt, value, values, self)