Argparse，处理可重复的项目集 [英] Argparse, handle repeatable set of items

问题描述

是否可以创建一组相关且可重复的可选参数?

Is it possible to create a set of optional arguments that are related and repeatable?

比方说，我有三个参数-a，-b，-c组成一个集合， -a是必需的，但是-b和-c是可选的. (已更新)

Let's say I have three arguments -a, -b, -c that form a set, -a is required, but -b and -c are optional. (Updated)

我希望能够指定其中的多个集合.

I would like an ability to specify multiple set of these.

Script.py -a 1 -b 2 -c 3    -a 4 -c 6   -a 7 -b 8    -a 10

这将被解析为字典列表，如下所示:

This will be parsed as a list of dict as follows

[
    {"a":1, "b":2, "c":3},
    {"a":4, "c":6},
    {"a":7, "b":8},
    {"a":10}
]

推荐答案

使参数可重复的一种方法是使用'append'操作类型:

A way to make arguments repeatable is to use an 'append' action type:

import argparse
parser = argparse.ArgumentParser()
parser.add_argument('-a', action='append')
parser.add_argument('-b', action='append')
parser.add_argument('-c', action='append')
argv = '-a 1 -b 2 -c 3    -a 4 -c 6   -a 7 -b 8    -a 10'
args = parser.parse_args(argv.split())
print args

产生:

Namespace(a=['1', '4', '7', '10'], b=['2', '8'], c=['3', '6'])

不幸的是，它确实丢失了一些信息.无法将"4"与"6"而不是"8"相关联.

Unfortunately it does lose some information. There's no way to associate the '4' with the '6' instead of the '8'.

如果使用'-'分隔参数块，则此迭代解析器可能会完成此任务:

If you use '--' to separate blocks of arguments, then this iterative parser might do the job:

parser = argparse.ArgumentParser()
# SUPPRESS keeps the default None out of the namespace
parser.add_argument('-a', type=int, default=argparse.SUPPRESS, required=True)
parser.add_argument('-b', type=int, default=argparse.SUPPRESS)
parser.add_argument('-c', type=int, default=argparse.SUPPRESS)
argv = '-a 1 -b 2 -c 3  --  -a 4 -c 6 --  -a 7 -b 8  --  -a 10'

arglist = []
rest = argv.split()
while rest:
    args,rest = parser.parse_known_args(rest)
    rest = rest[1:]  # remove the 1st '--'
    print args
    arglist.append(vars(args))
print arglist

生产:

Namespace(a=1, b=2, c=3)
Namespace(a=4, c=6)
Namespace(a=7, b=8)
Namespace(a=10)

[{'a': 1, 'c': 3, 'b': 2}, 
 {'a': 4, 'c': 6}, 
 {'a': 7, 'b': 8}, 
 {'a': 10}]

我不确定它是否足够健壮.我将-a设置为必需，因此从其中一个组中省略它会引发错误.

I'm not sure if it is robust enough. I made -a required, so omitting it from one of the groups will raise an error.

或改编farzad的迭代器:

Or adapting farzad's iterator:

def by_sets(iterator, start):
    set = []
    for val in iterator:
        if set and val == start:
            yield set
            set = [val]
        else:
            set.append(val)
    yield set

argv = '-a 1 -b 2 -c 3  -a 4 -c 6 -a 7 -b 8  -a 10'
# print list(by_sets(argv.split(), '-a'))
# [['-a', '1', '-b', '2', '-c', '3'], ['-a', '4', '-c', '6'],... ['-a', '10']]

arglist = []
for aset in by_sets(argv.split(), '-a'):
    arglist.append(vars(parser.parse_args(aset)))
print arglist

产生:

[{'a': 1, 'c': 3, 'b': 2}, {'a': 4, 'c': 6}, {'a': 7, 'b': 8}, {'a': 10}]

循环也可以理解为:

[vars(parser.parse_args(aset)) for aset in by_sets(argv.split(), '-a')]

这篇关于Argparse，处理可重复的项目集的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！