扩展argparse以在可选参数选择的帮助文本中写入集名称,并在结尾处定义这些集合

Question

问题的例子

如果我有一个在多个参数之间共享的有效选项字符串列表,则该列表将写入帮助字符串中的多个位置.让它更难阅读:

def main():
    elements = ['a', 'b', 'c', 'd', 'e', 'f']

    parser = argparse.ArgumentParser()
    parser.add_argument(
        '-i',
        nargs='*',
        choices=elements,
        default=elements,
        help='Space separated list of case sensitive element names.')
    parser.add_argument(
        '-e',
        nargs='*',
        choices=elements,
        default=[],
        help='Space separated list of case sensitive element names to '
        'exclude from processing')

    parser.parse_args()

使用命令行参数运行上述函数时,--help它显示:

usage: arguments.py [-h] [-i [{a,b,c,d,e,f} [{a,b,c,d,e,f} ...]]]
                    [-e [{a,b,c,d,e,f} [{a,b,c,d,e,f} ...]]]

optional arguments:
  -h, --help            show this help message and exit
  -i [{a,b,c,d,e,f} [{a,b,c,d,e,f} ...]]
                        Space separated list of case sensitive element names.
  -e [{a,b,c,d,e,f} [{a,b,c,d,e,f} ...]]
                        Space separated list of case sensitive element names
                        to exclude from processing

---

什么会好的

如果可以定义选项列表名称并且在帮助输出中在多个位置写入选项列表名称并最后定义它将会很好.理论上它会像这样工作:

def main_optionlist():
    elements = ['a', 'b', 'c', 'd', 'e', 'f']

    # Two instances of OptionList are equal if and only if they
    # have the same name (ALFA in this case)

    ol = OptionList('ALFA', elements)

    parser = argparse.ArgumentParser()
    parser.add_argument(
        '-i',
        nargs='*',
        choices=ol,
        default=ol,
        help='Space separated list of case sensitive element names.')
    parser.add_argument(
        '-e',
        nargs='*',
        choices=ol,
        default=[],
        help='Space separated list of case sensitive element names to '
        'exclude from processing')

    parser.parse_args()

当使用命令行参数运行上述函数时--help,它将显示类似于:

usage: arguments.py [-h] [-i [ALFA [ALFA ...]]]
                    [-e [ALFA [ALFA ...]]]

optional arguments:
  -h, --help            show this help message and exit
  -i [ALFA [ALFA ...]]
                        Space separated list of case sensitive element names.
  -e [ALFA [ALFA ...]]
                        Space separated list of case sensitive element names
                        to exclude from processing
sets in optional arguments:
  ALFA                  {a,b,c,d,e,f}

---

题

我需要:

• 用可选参数替换选项名称所示的{'l','i','s','t','s'}.
• 在帮助文本的末尾显示了一个部分,说明了每个选项名称包含哪些元素.

所以我问:

1. 使用argparse可以吗？
2. 我必须继承哪些类以及需要覆盖哪些方法？

我曾尝试查看argparse的源代码,但由于这种修改感觉非常先进,我不知道该如何开始.

Answer 1

我的回答并不是试图扩展argparse,而是使用argparse的可用选项......这样可以解决你的情况吗？

import argparse
import textwrap

def main():
    elements = ['a', 'b', 'c', 'd', 'e', 'f']

    parser = argparse.ArgumentParser(
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog = textwrap.dedent('''\
            sets in optional arguments: 
                ALFA\t\t{a,b,c,d,e,f}"
            '''))

    parser.add_argument(
        '-i',
        nargs='*',
        choices=elements,
        default=elements,
        metavar="ALFA",
        help='Space separated list of case sensitive element names.')
    parser.add_argument(
        '-e',
        nargs='*',
        choices=elements,
        default=[],
        metavar="ALFA",
        help='Space separated list of case sensitive element names to '
        'exclude from processing')

    parser.parse_args()

产量

usage: test.py [-h] [-i [ALFA [ALFA ...]]] [-e [ALFA [ALFA ...]]]

optional arguments:
  -h, --help            show this help message and exit
  -i [ALFA [ALFA ...]]  Space separated list of case sensitive element names.
  -e [ALFA [ALFA ...]]  Space separated list of case sensitive element names
                        to exclude from processing

sets in optional arguments: 
    ALFA                {a,b,c,d,e,f}"

这种方法的好处在于,您可以随意为每个标志命名metavar,并且您可以手动设置epilog的格式以反映任何格式描述.不需要子类化.

Answer 2

根据现在删除的赏金捐赠者的请求,与其他答案形成对比的完全通用的解决方案:

import argparse
from operator import itemgetter

class OptionListGroup(object):
  class GroupAction(object):
    def __init__(self, left, right):
      self.help = right
      self.option_strings = [left]
      self.nargs = 0

  def __init__(self, lists):
    self.description = None
    self.title = "referenced sets"
    self._group_actions = [self.GroupAction(name, self.format_list(lst))
                           for name, lst in sorted(lists)]

  def format_list(self, lst):
    return '{%s}' % ', '.join(map(str, lst))

class MyArgParser(argparse.ArgumentParser):
  def __init__(self, *args, **kwargs):
    self._option_lists = {}
    super(MyArgParser, self).__init__(*args, **kwargs)

  def parse_args(self, *args, **kw):
    self._action_groups.append(OptionListGroup(self._option_lists.values()))
    return super(MyArgParser, self).parse_args(*args, **kw)

  def add_option_list(self, name, lst):
    if name in map(itemgetter(0), self._option_lists.values()):
      raise ValueError, "Name already existing"
    self._option_lists[id(lst)] = (name, lst)

  def add_argument(self, *args, **kw):
    name_list = self._option_lists.get(id(kw.get('choices')))
    if name_list:
      kw['metavar'] = name_list[0]
    return super(MyArgParser, self).add_argument(*args, **kw)

用法示例:

alfa = ['a', 'b', 'c', 'd', 'e', 'f']
num = [1, 2, 3]

parser = MyArgParser()

parser.add_option_list('ALFA', alfa)
parser.add_option_list('NUM', num)

parser.add_argument(
  '-a',
  nargs='*',
  choices=alfa,
  default=alfa,
  help='Characters (defaults to include all)')

parser.add_argument(
  '-e',
  nargs='*',
  choices=num,
  default=[],
  help='Digits (defaults to exclude all)')

parser.parse_args()

帮助输出:

usage: argparse-optionlist.py [-h] [-a [ALFA [ALFA ...]]]
                              [-e [NUM [NUM ...]]]

optional arguments:
  -h, --help            show this help message and exit
  -a [ALFA [ALFA ...]]  Characters (defaults to include all)
  -e [NUM [NUM ...]]    Digits (defaults to exclude all)

referenced sets:
  ALFA                  {a, b, c, d, e, f}
  NUM                   {1, 2, 3}