Simple DocOpt Install for Python & Bash

docopt.py

"""Pythonic command-line interface parser that will make you smile.

 * http://docopt.org
 * Repository and issue-tracker: https://github.com/docopt/docopt
 * Licensed under terms of MIT license (see LICENSE-MIT)
 * Copyright (c) 2013 Vladimir Keleshev, [email protected]

Modified by Erik Stephens to provide curl-installable docopt for both
python & bash usage (python -m docopt ...)
"""
import sys
import re


__all__ = ['docopt']
__version__ = '0.6.1'


class DocoptLanguageError(Exception):

    """Error in construction of usage-message by developer."""


class DocoptExit(SystemExit):

    """Exit in case user invoked program with incorrect arguments."""

    usage = ''

    def __init__(self, message=''):
        SystemExit.__init__(self, (message + '\n' + self.usage).strip())


class Pattern(object):

    def __eq__(self, other):
        return repr(self) == repr(other)

    def __hash__(self):
        return hash(repr(self))

    def fix(self):
        self.fix_identities()
        self.fix_repeating_arguments()
        return self

    def fix_identities(self, uniq=None):
        """Make pattern-tree tips point to same object if they are equal."""
        if not hasattr(self, 'children'):
            return self
        uniq = list(set(self.flat())) if uniq is None else uniq
        for i, c in enumerate(self.children):
            if not hasattr(c, 'children'):
                assert c in uniq
                self.children[i] = uniq[uniq.index(c)]
            else:
                c.fix_identities(uniq)

    def fix_repeating_arguments(self):
        """Fix elements that should accumulate/increment values."""
        either = [list(c.children) for c in self.either.children]
        for case in either:
            for e in [c for c in case if case.count(c) > 1]:
                if type(e) is Argument or type(e) is Option and e.argcount:
                    if e.value is None:
                        e.value = []
                    elif type(e.value) is not list:
                        e.value = e.value.split()
                if type(e) is Command or type(e) is Option and e.argcount == 0:
                    e.value = 0
        return self

    @property
    def either(self):
        """Transform pattern into an equivalent, with only top-level Either."""
        # Currently the pattern will not be equivalent, but more "narrow",
        # although good enough to reason about list arguments.
        ret = []
        groups = [[self]]
        while groups:
            children = groups.pop(0)
            types = [type(c) for c in children]
            if Either in types:
                either = [c for c in children if type(c) is Either][0]
                children.pop(children.index(either))
                for c in either.children:
                    groups.append([c] + children)
            elif Required in types:
                required = [c for c in children if type(c) is Required][0]
                children.pop(children.index(required))
                groups.append(list(required.children) + children)
            elif Optional in types:
                optional = [c for c in children if type(c) is Optional][0]
                children.pop(children.index(optional))
                groups.append(list(optional.children) + children)
            elif AnyOptions in types:
                optional = [c for c in children if type(c) is AnyOptions][0]
                children.pop(children.index(optional))
                groups.append(list(optional.children) + children)
            elif OneOrMore in types:
                oneormore = [c for c in children if type(c) is OneOrMore][0]
                children.pop(children.index(oneormore))
                groups.append(list(oneormore.children) * 2 + children)
            else:
                ret.append(children)
        return Either(*[Required(*e) for e in ret])


class ChildPattern(Pattern):

    def __init__(self, name, value=None):
        self.name = name
        self.value = value

    def __repr__(self):
        return '%s(%r, %r)' % (self.__class__.__name__, self.name, self.value)

    def flat(self, *types):
        return [self] if not types or type(self) in types else []

    def match(self, left, collected=None):
        collected = [] if collected is None else collected
        pos, match = self.single_match(left)
        if match is None:
            return False, left, collected
        left_ = left[:pos] + left[pos + 1:]
        same_name = [a for a in collected if a.name == self.name]
        if type(self.value) in (int, list):
            if type(self.value) is int:
                increment = 1
            else:
                increment = ([match.value] if type(match.value) is str
                             else match.value)
            if not same_name:
                match.value = increment
                return True, left_, collected + [match]
            same_name[0].value += increment
            return True, left_, collected
        return True, left_, collected + [match]


class ParentPattern(Pattern):

    def __init__(self, *children):
        self.children = list(children)

    def __repr__(self):
        return '%s(%s)' % (self.__class__.__name__,
                           ', '.join(repr(a) for a in self.children))

    def flat(self, *types):
        if type(self) in types:
            return [self]
        return sum([c.flat(*types) for c in self.children], [])


class Argument(ChildPattern):

    def single_match(self, left):
        for n, p in enumerate(left):
            if type(p) is Argument:
                return n, Argument(self.name, p.value)
        return None, None

    @classmethod
    def parse(class_, source):
        name = re.findall('(<\S*?>)', source)[0]
        value = re.findall('\[default: (.*)\]', source, flags=re.I)
        return class_(name, value[0] if value else None)


class Command(Argument):

    def __init__(self, name, value=False):
        self.name = name
        self.value = value

    def single_match(self, left):
        for n, p in enumerate(left):
            if type(p) is Argument:
                if p.value == self.name:
                    return n, Command(self.name, True)
                else:
                    break
        return None, None


class Option(ChildPattern):

    def __init__(self, short=None, long=None, argcount=0, value=False):
        assert argcount in (0, 1)
        self.short, self.long = short, long
        self.argcount, self.value = argcount, value
        self.value = None if value is False and argcount else value

    @classmethod
    def parse(class_, option_description):
        short, long, argcount, value = None, None, 0, False
        options, _, description = option_description.strip().partition('  ')
        options = options.replace(',', ' ').replace('=', ' ')
        for s in options.split():
            if s.startswith('--'):
                long = s
            elif s.startswith('-'):
                short = s
            else:
                argcount = 1
        if argcount:
            matched = re.findall('\[default: (.*)\]', description, flags=re.I)
            value = matched[0] if matched else None
        return class_(short, long, argcount, value)

    def single_match(self, left):
        for n, p in enumerate(left):
            if self.name == p.name:
                return n, p
        return None, None

    @property
    def name(self):
        return self.long or self.short

    def __repr__(self):
        return 'Option(%r, %r, %r, %r)' % (self.short, self.long,
                                           self.argcount, self.value)


class Required(ParentPattern):

    def match(self, left, collected=None):
        collected = [] if collected is None else collected
        l = left
        c = collected
        for p in self.children:
            matched, l, c = p.match(l, c)
            if not matched:
                return False, left, collected
        return True, l, c


class Optional(ParentPattern):

    def match(self, left, collected=None):
        collected = [] if collected is None else collected
        for p in self.children:
            m, left, collected = p.match(left, collected)
        return True, left, collected


class AnyOptions(Optional):

    """Marker/placeholder for [options] shortcut."""


class OneOrMore(ParentPattern):

    def match(self, left, collected=None):
        assert len(self.children) == 1
        collected = [] if collected is None else collected
        l = left
        c = collected
        l_ = None
        matched = True
        times = 0
        while matched:
            # could it be that something didn't match but changed l or c?
            matched, l, c = self.children[0].match(l, c)
            times += 1 if matched else 0
            if l_ == l:
                break
            l_ = l
        if times >= 1:
            return True, l, c
        return False, left, collected


class Either(ParentPattern):

    def match(self, left, collected=None):
        collected = [] if collected is None else collected
        outcomes = []
        for p in self.children:
            matched, _, _ = outcome = p.match(left, collected)
            if matched:
                outcomes.append(outcome)
        if outcomes:
            return min(outcomes, key=lambda outcome: len(outcome[1]))
        return False, left, collected


class TokenStream(list):

    def __init__(self, source, error):
        self += source.split() if hasattr(source, 'split') else source
        self.error = error

    def move(self):
        return self.pop(0) if len(self) else None

    def current(self):
        return self[0] if len(self) else None


def parse_long(tokens, options):
    """long ::= '--' chars [ ( ' ' | '=' ) chars ] ;"""
    long, eq, value = tokens.move().partition('=')
    assert long.startswith('--')
    value = None if eq == value == '' else value
    similar = [o for o in options if o.long == long]
    if tokens.error is DocoptExit and similar == []:  # if no exact match
        similar = [o for o in options if o.long and o.long.startswith(long)]
    if len(similar) > 1:  # might be simply specified ambiguously 2+ times?
        raise tokens.error('%s is not a unique prefix: %s?' %
                           (long, ', '.join(o.long for o in similar)))
    elif len(similar) < 1:
        argcount = 1 if eq == '=' else 0
        o = Option(None, long, argcount)
        options.append(o)
        if tokens.error is DocoptExit:
            o = Option(None, long, argcount, value if argcount else True)
    else:
        o = Option(similar[0].short, similar[0].long,
                   similar[0].argcount, similar[0].value)
        if o.argcount == 0:
            if value is not None:
                raise tokens.error('%s must not have an argument' % o.long)
        else:
            if value is None:
                if tokens.current() is None:
                    raise tokens.error('%s requires argument' % o.long)
                value = tokens.move()
        if tokens.error is DocoptExit:
            o.value = value if value is not None else True
    return [o]


def parse_shorts(tokens, options):
    """shorts ::= '-' ( chars )* [ [ ' ' ] chars ] ;"""
    token = tokens.move()
    assert token.startswith('-') and not token.startswith('--')
    left = token.lstrip('-')
    parsed = []
    while left != '':
        short, left = '-' + left[0], left[1:]
        similar = [o for o in options if o.short == short]
        if len(similar) > 1:
            raise tokens.error('%s is specified ambiguously %d times' %
                               (short, len(similar)))
        elif len(similar) < 1:
            o = Option(short, None, 0)
            options.append(o)
            if tokens.error is DocoptExit:
                o = Option(short, None, 0, True)
        else:  # why copying is necessary here?
            o = Option(short, similar[0].long,
                       similar[0].argcount, similar[0].value)
            value = None
            if o.argcount != 0:
                if left == '':
                    if tokens.current() is None:
                        raise tokens.error('%s requires argument' % short)
                    value = tokens.move()
                else:
                    value = left
                    left = ''
            if tokens.error is DocoptExit:
                o.value = value if value is not None else True
        parsed.append(o)
    return parsed


def parse_pattern(source, options):
    tokens = TokenStream(re.sub(r'([\[\]\(\)\|]|\.\.\.)', r' \1 ', source),
                         DocoptLanguageError)
    result = parse_expr(tokens, options)
    if tokens.current() is not None:
        raise tokens.error('unexpected ending: %r' % ' '.join(tokens))
    return Required(*result)


def parse_expr(tokens, options):
    """expr ::= seq ( '|' seq )* ;"""
    seq = parse_seq(tokens, options)
    if tokens.current() != '|':
        return seq
    result = [Required(*seq)] if len(seq) > 1 else seq
    while tokens.current() == '|':
        tokens.move()
        seq = parse_seq(tokens, options)
        result += [Required(*seq)] if len(seq) > 1 else seq
    return [Either(*result)] if len(result) > 1 else result


def parse_seq(tokens, options):
    """seq ::= ( atom [ '...' ] )* ;"""
    result = []
    while tokens.current() not in [None, ']', ')', '|']:
        atom = parse_atom(tokens, options)
        if tokens.current() == '...':
            atom = [OneOrMore(*atom)]
            tokens.move()
        result += atom
    return result


def parse_atom(tokens, options):
    """atom ::= '(' expr ')' | '[' expr ']' | 'options'
             | long | shorts | argument | command ;
    """
    token = tokens.current()
    result = []
    if token in '([':
        tokens.move()
        matching, pattern = {'(': [')', Required], '[': [']', Optional]}[token]
        result = pattern(*parse_expr(tokens, options))
        if tokens.move() != matching:
            raise tokens.error("unmatched '%s'" % token)
        return [result]
    elif token == 'options':
        tokens.move()
        return [AnyOptions()]
    elif token.startswith('--') and token != '--':
        return parse_long(tokens, options)
    elif token.startswith('-') and token not in ('-', '--'):
        return parse_shorts(tokens, options)
    elif token.startswith('<') and token.endswith('>') or token.isupper():
        return [Argument(tokens.move())]
    else:
        return [Command(tokens.move())]


def parse_argv(tokens, options, options_first=False):
    """Parse command-line argument vector.

    If options_first:
        argv ::= [ long | shorts ]* [ argument ]* [ '--' [ argument ]* ] ;
    else:
        argv ::= [ long | shorts | argument ]* [ '--' [ argument ]* ] ;

    """
    parsed = []
    while tokens.current() is not None:
        if tokens.current() == '--':
            return parsed + [Argument(None, v) for v in tokens]
        elif tokens.current().startswith('--'):
            parsed += parse_long(tokens, options)
        elif tokens.current().startswith('-') and tokens.current() != '-':
            parsed += parse_shorts(tokens, options)
        elif options_first:
            return parsed + [Argument(None, v) for v in tokens]
        else:
            parsed.append(Argument(None, tokens.move()))
    return parsed


def parse_defaults(doc):
    # in python < 2.7 you can't pass flags=re.MULTILINE
    split = re.split('\n *(<\S+?>|-\S+?)', doc)[1:]
    split = [s1 + s2 for s1, s2 in zip(split[::2], split[1::2])]
    options = [Option.parse(s) for s in split if s.startswith('-')]
    #arguments = [Argument.parse(s) for s in split if s.startswith('<')]
    #return options, arguments
    return options


def printable_usage(doc):
    # in python < 2.7 you can't pass flags=re.IGNORECASE
    usage_split = re.split(r'([Uu][Ss][Aa][Gg][Ee]:)', doc)
    if len(usage_split) < 3:
        raise DocoptLanguageError('"usage:" (case-insensitive) not found.')
    if len(usage_split) > 3:
        raise DocoptLanguageError('More than one "usage:" (case-insensitive).')
    return re.split(r'\n\s*\n', ''.join(usage_split[1:]))[0].strip()


def formal_usage(printable_usage):
    pu = printable_usage.split()[1:]  # split and drop "usage:"
    return '( ' + ' '.join(') | (' if s == pu[0] else s for s in pu[1:]) + ' )'


def extras(help, version, options, doc):
    if help and any((o.name in ('-h', '--help')) and o.value for o in options):
        print(doc.strip("\n"))
        sys.exit()
    if version and any(o.name == '--version' and o.value for o in options):
        print(version)
        sys.exit()


class Dict(dict):
    def __repr__(self):
        return '{%s}' % ',\n '.join('%r: %r' % i for i in sorted(self.items()))


def docopt(doc, argv=None, help=True, version=None, options_first=False):
    """Parse `argv` based on command-line interface described in `doc`.

    `docopt` creates your command-line interface based on its
    description that you pass as `doc`. Such description can contain
    --options, <positional-argument>, commands, which could be
    [optional], (required), (mutually | exclusive) or repeated...

    Parameters
    ----------
    doc : str
        Description of your command-line interface.
    argv : list of str, optional
        Argument vector to be parsed. sys.argv[1:] is used if not
        provided.
    help : bool (default: True)
        Set to False to disable automatic help on -h or --help
        options.
    version : any object
        If passed, the object will be printed if --version is in
        `argv`.
    options_first : bool (default: False)
        Set to True to require options preceed positional arguments,
        i.e. to forbid options and positional arguments intermix.

    Returns
    -------
    args : dict
        A dictionary, where keys are names of command-line elements
        such as e.g. "--verbose" and "<path>", and values are the
        parsed values of those elements.

    Example
    -------
    >>> from docopt import docopt
    >>> doc = '''
    Usage:
        my_program tcp <host> <port> [--timeout=<seconds>]
        my_program serial <port> [--baud=<n>] [--timeout=<seconds>]
        my_program (-h | --help | --version)

    Options:
        -h, --help  Show this screen and exit.
        --baud=<n>  Baudrate [default: 9600]
    '''
    >>> argv = ['tcp', '127.0.0.1', '80', '--timeout', '30']
    >>> docopt(doc, argv)
    {'--baud': '9600',
     '--help': False,
     '--timeout': '30',
     '--version': False,
     '<host>': '127.0.0.1',
     '<port>': '80',
     'serial': False,
     'tcp': True}

    See also
    --------
    * For video introduction see http://docopt.org
    * Full documentation is available in README.rst as well as online
      at https://github.com/docopt/docopt#readme

    """
    if argv is None:
        argv = sys.argv[1:]
    DocoptExit.usage = printable_usage(doc)
    options = parse_defaults(doc)
    pattern = parse_pattern(formal_usage(DocoptExit.usage), options)
    # [default] syntax for argument is disabled
    #for a in pattern.flat(Argument):
    #    same_name = [d for d in arguments if d.name == a.name]
    #    if same_name:
    #        a.value = same_name[0].value
    argv = parse_argv(TokenStream(argv, DocoptExit), list(options),
                      options_first)
    pattern_options = set(pattern.flat(Option))
    for ao in pattern.flat(AnyOptions):
        doc_options = parse_defaults(doc)
        ao.children = list(set(doc_options) - pattern_options)
        #if any_options:
        #    ao.children += [Option(o.short, o.long, o.argcount)
        #                    for o in argv if type(o) is Option]
    extras(help, version, argv, doc)
    matched, left, collected = pattern.fix().match(argv)
    if matched and left == []:  # better error message if left?
        return Dict((a.name, a.value) for a in (pattern.flat() + collected))
    raise DocoptExit()


if __name__ == '__main__':
    doc = """Bash interface for docopt, the CLI description language.

Usage: docopt.py [options] -h <msg> : [<argv>...]

Options:

  -h, --help MSG  : The help message in docopt format. If - is given,
   read the help message from standard input. If no argument is given,
   print docopts's own help message and quit.

  -V, --version MSG  : A version message. If - is given, read the
   version message from standard input.  If the help message is also
   read from standard input, it is read first. If no argument is
   given, print docopts's own version message and quit.

  -O, --options-first  : Disallow interspersing options and positional
   arguments: all arguments starting from the first one that does not
   begin with a dash will be treated as positional arguments.

  -F, --for-func  : Set when using for functions.  Issues a `return`
   instead of an `exit`.

  -H, --no-help  : Don't handle --help and --version specially.

  -A, --array NAME  : Export the arguments as a Bash 4.x associative array
   called NAME.

  -s, --separator STR  : The string to use to separate the help message
   from the version message when both are given via standard
   input. [default: ----]
    """
    try:
        from io import StringIO
    except ImportError:
        from StringIO import StringIO

    # helper functions
    def shellquote(s):
        return "'" + s.replace("'", r"'\''") + "'"

    def isbashidentifier(s):
        return re.match(r'^([A-Za-z]|[A-Za-z_][0-9A-Za-z_]+)$', s)

    def to_bash(obj):
        return {
            type(None): lambda x: '',
            bool:       lambda x: 'true' if x else 'false',
            int:        lambda x: str(x),
            str:        lambda x: shellquote(x),
            list:       lambda x: '(' + ' '.join(map(shellquote, x)) + ')',
        }[type(obj)](obj)

    def name_mangle(elem):
        if elem == '-' or elem == '--':
            return None
        elif re.match(r'^<.*>$', elem):
            var = elem[1:-1]
        elif re.match(r'^-[^-]$', elem):
            var = elem[1]
        elif re.match(r'^--.+$', elem):
            var = elem[2:]
        else:
            var = elem
        var = var.replace('-', '_')
        if not isbashidentifier(var):
            raise ValueError(elem)
        else:
            return var

    try:
        args = docopt(doc, help=False, options_first=True)
    except DocoptExit as e:
        message = e.args[0]
        if message.startswith('-h') or message.startswith('--help'):
            print(doc.strip())
            sys.exit()
        if message.startswith('-V') or message.startswith('--version'):
            print(__version__.strip())
            sys.exit()
        else:
            sys.exit(message)

    argv = args['<argv>']
    doc = args['--help']
    version = args['--version']
    options_first = args['--options-first']
    help = not args['--no-help']
    name = args['--array']
    separator = args['--separator']

    if doc == '-' and version == '-':
        doc, version = (page.strip() for page in
                        sys.stdin.read().split(separator, 1))
    elif doc == '-':
        doc = sys.stdin.read().strip()
    elif version == '-':
        version = sys.stdin.read().strip()

    # parse options or abort if there is an error in docopt
    if args['--for-func']:
        exit = 'return'
    else:
        exit = 'exit'
    try:
        # temporarily redirect stdout to a StringIO so we can catch docopt()
        # output on --help and --version
        stdout = sys.stdout
        sys.stdout = StringIO()
        exit_message = None
        args = docopt(doc, argv, help, version, options_first)
    except DocoptLanguageError as e:
        # invalid docstring by user
        sys.exit("%s: invalid doc argument: %s" % (sys.argv[0], e))
    except DocoptExit as e:
        # invoked with invalid arguments
        exit_message = "echo %s >&2\n%s 64" % (shellquote(str(e)), exit,)
    except SystemExit as e:
        # --help or --version found and --no-help was not given
        exit_message = 'echo %s >&2\n%s 1' % (shellquote(sys.stdout.getvalue()), exit,)
    finally:
        # restore stdout to normal and quit if a docopt parse error happened
        sys.stdout.close()
        sys.stdout = stdout
        if exit_message:
            print(exit_message)
            sys.exit()

    if name is not None:
        if not isbashidentifier(name):
            sys.exit("%s: not a valid Bash identifier: %s" % (sys.argv[0], name))
        # fake nested Bash arrays for repeatable arguments with values
        arrays = dict((elem, value) for elem, value in args.items() if
                      isinstance(value, list))
        for elem, value in arrays.items():
            del args[elem]
            args[elem+',#'] = len(value)
            args.update(('%s,%d' % (elem, i), v) for i,v in enumerate(value))
        print('declare -A %s' % (name,))
        for elem, value in args.items():
            print('%s[%s]=%s' % (name, shellquote(elem), to_bash(value)))
    else:
        try:
            variables = dict(zip(map(name_mangle, args.keys()),
                                 map(to_bash, args.values())))
        except ValueError as e:
            sys.exit("%s: name could not be mangled into a valid Bash "
                     "identifier: %s" % (sys.argv[0], e))
        else:
            variables.pop(None, None)
            args.pop('-', None)
            args.pop('--', None)
        if len(variables) < len(args):
            sys.exit("%s: two or more elements have identically mangled names" %
                     (sys.argv[0],))
        for var, value in variables.items():
            print("%s=%s" % (var, value))