utils.py

LINUX下

"""
General Utilities
(part of web.py)
"""

__all__ = [
  "Storage", "storage", "storify", 
  "iters", 
  "rstrips", "lstrips", "strips", "utf8",
  "TimeoutError", "timelimit",
  "Memoize", "memoize",
  "re_compile", "re_subm",
  "group",
  "IterBetter", "iterbetter",
  "dictreverse", "dictfind", "dictfindall", "dictincr", "dictadd",
  "listget", "intget", "datestr",
  "numify", "denumify", "dateify",
  "CaptureStdout", "capturestdout", "Profile", "profile",
  "tryall",
  "ThreadedDict",
  "autoassign",
  "to36",
  "safemarkdown"
]

import re, sys, time, threading
try: import datetime
except ImportError: pass

class Storage(dict):
    """
    A Storage object is like a dictionary except `obj.foo` can be used
    in addition to `obj['foo']`.
    
        >>> o = storage(a=1)
        >>> o.a
        1
        >>> o['a']
        1
        >>> o.a = 2
        >>> o['a']
        2
        >>> del o.a
        >>> o.a
        Traceback (most recent call last):
            ...
        AttributeError: 'a'
    
    """
    def __getattr__(self, key): 
        try:
            return self[key]
        except KeyError, k:
            raise AttributeError, k

    def __setattr__(self, key, value): 
        self[key] = value

    def __delattr__(self, key):
        try:
            del self[key]
        except KeyError, k:
            raise AttributeError, k

    def __repr__(self):     
        return '<Storage ' + dict.__repr__(self) + '>'

storage = Storage

def storify(mapping, *requireds, **defaults):
    """
    Creates a `storage` object from dictionary `mapping`, raising `KeyError` if
    d doesn't have all of the keys in `requireds` and using the default 
    values for keys found in `defaults`.

    For example, `storify({'a':1, 'c':3}, b=2, c=0)` will return the equivalent of
    `storage({'a':1, 'b':2, 'c':3})`.
    
    If a `storify` value is a list (e.g. multiple values in a form submission), 
    `storify` returns the last element of the list, unless the key appears in 
    `defaults` as a list. Thus:
    
        >>> storify({'a':[1, 2]}).a
        2
        >>> storify({'a':[1, 2]}, a=[]).a
        [1, 2]
        >>> storify({'a':1}, a=[]).a
        [1]
        >>> storify({}, a=[]).a
        []
    
    Similarly, if the value has a `value` attribute, `storify will return _its_
    value, unless the key appears in `defaults` as a dictionary.
    
        >>> storify({'a':storage(value=1)}).a
        1
        >>> storify({'a':storage(value=1)}, a={}).a
        <Storage {'value': 1}>
        >>> storify({}, a={}).a
        {}
    
    """
    def getvalue(x):
        if hasattr(x, 'value'):
            return x.value
        else:
            return x
    
    stor = Storage()
    for key in requireds + tuple(mapping.keys()):
        value = mapping[key]
        if isinstance(value, list):
            if isinstance(defaults.get(key), list):
                value = [getvalue(x) for x in value]
            else:
                value = value[-1]
        if not isinstance(defaults.get(key), dict):
            value = getvalue(value)
        if isinstance(defaults.get(key), list) and not isinstance(value, list):
            value = [value]
        setattr(stor, key, value)

    for (key, value) in defaults.iteritems():
        result = value
        if hasattr(stor, key): 
            result = stor[key]
        if value == () and not isinstance(result, tuple): 
            result = (result,)
        setattr(stor, key, result)
    
    return stor

iters = [list, tuple]
import __builtin__
if hasattr(__builtin__, 'set'):
    iters.append(set)
try: 
    from sets import Set
    iters.append(Set)
except ImportError: 
    pass

class _hack(tuple): pass
iters = _hack(iters)
iters.__doc__ = """
A list of iterable items (like lists, but not strings). Includes whichever
of lists, tuples, sets, and Sets are available in this version of Python.
"""

def _strips(direction, text, remove):
    if direction == 'l': 
        if text.startswith(remove): 
            return text[len(remove):]
    elif direction == 'r':
        if text.endswith(remove):   
            return text[:-len(remove)]
    else: 
        raise ValueError, "Direction needs to be r or l."
    return text

def rstrips(text, remove):
    """
    removes the string `remove` from the right of `text`

        >>> rstrips("foobar", "bar")
        'foo'
    
    """
    return _strips('r', text, remove)

def lstrips(text, remove):
    """
    removes the string `remove` from the left of `text`
    
        >>> lstrips("foobar", "foo")
        'bar'
    
    """
    return _strips('l', text, remove)

def strips(text, remove):
    """removes the string `remove` from the both sides of `text`

        >>> strips("foobarfoo", "foo")
        'bar'
    
    """
    return rstrips(lstrips(text, remove), remove)

def utf8(text):
    """Encodes text in utf-8.
        
        >> utf8(u'\u1234') # doctest doesn't seem to like utf-8
        '\xe1\x88\xb4'

        >>> utf8('hello')
        'hello'
        >>> utf8(42)
        '42'
    """
    if isinstance(text, unicode):
        return text.encode('utf-8')
    elif isinstance(text, str):
        return text
    else:
        return str(text)

class TimeoutError(Exception): pass
def timelimit(timeout):
    """
    A decorator to limit a function to `timeout` seconds, raising `TimeoutError`
    if it takes longer.
    
        >>> import time
        >>> def meaningoflife():
        ...     time.sleep(.2)
        ...     return 42
        >>> 
        >>> timelimit(.1)(meaningoflife)()
        Traceback (most recent call last):
            ...
        TimeoutError: took too long
        >>> timelimit(1)(meaningoflife)()
        42

    _Caveat:_ The function isn't stopped after `timeout` seconds but continues 
    executing in a separate thread. (There seems to be no way to kill a thread.)

    inspired by <http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/473878>
    """
    def _1(function):
        def _2(*args, **kw):
            class Dispatch(threading.Thread):
                def __init__(self):
                    threading.Thread.__init__(self)
                    self.result = None
                    self.error = None

                    self.setDaemon(True)
                    self.start()

                def run(self):
                    try:
                        self.result = function(*args, **kw)
                    except:
                        self.error = sys.exc_info()

            c = Dispatch()
            c.join(timeout)
            if c.isAlive():
                raise TimeoutError, 'took too long'
            if c.error:
                raise c.error[0], c.error[1]
            return c.result
        return _2
    return _1

class Memoize:
    """
    'Memoizes' a function, caching its return values for each input.
    
        >>> import time
        >>> def meaningoflife():
        ...     time.sleep(.2)
        ...     return 42
        >>> fastlife = memoize(meaningoflife)
        >>> meaningoflife()
        42
        >>> timelimit(.1)(meaningoflife)()
        Traceback (most recent call last):
            ...
        TimeoutError: took too long
        >>> fastlife()
        42
        >>> timelimit(.1)(fastlife)()
        42
    
    """
    def __init__(self, func): 
        self.func = func
        self.cache = {}
    def __call__(self, *args, **keywords):
        key = (args, tuple(keywords.items()))
        if key not in self.cache: 
            self.cache[key] = self.func(*args, **keywords)
        return self.cache[key]

memoize = Memoize

re_compile = memoize(re.compile) #@@ threadsafe?
re_compile.__doc__ = """
A memoized version of re.compile.
"""

class _re_subm_proxy:
    def __init__(self): 
        self.match = None
    def __call__(self, match): 
        self.match = match
        return ''

def re_subm(pat, repl, string):
    """
    Like re.sub, but returns the replacement _and_ the match object.
    
        >>> t, m = re_subm('g(oo+)fball', r'f\\1lish', 'goooooofball')
        >>> t
        'foooooolish'
        >>> m.groups()
        ('oooooo',)
    """
    compiled_pat = re_compile(pat)
    proxy = _re_subm_proxy()
    compiled_pat.sub(proxy.__call__, string)
    return compiled_pat.sub(repl, string), proxy.match

def group(seq, size): 
    """
    Returns an iterator over a series of lists of length size from iterable.

        >>> list(group([1,2,3,4], 2))
        [[1, 2], [3, 4]]
    """
    if not hasattr(seq, 'next'):  
        seq = iter(seq)
    while True: 
        yield [seq.next() for i in xrange(size)]

class IterBetter:
    """
    Returns an object that can be used as an iterator 
    but can also be used via __getitem__ (although it 
    cannot go backwards -- that is, you cannot request 
    `iterbetter[0]` after requesting `iterbetter[1]`).
    
        >>> import itertools
        >>> c = iterbetter(itertools.count())
        >>> c[1]
        1
        >>> c[5]
        5
        >>> c[3]
        Traceback (most recent call last):
            ...
        IndexError: already passed 3
    """
    def __init__(self, iterator): 
        self.i, self.c = iterator, 0
    def __iter__(self): 
        while 1:    
            yield self.i.next()
            self.c += 1
    def __getitem__(self, i):
        #todo: slices
        if i < self.c: 
            raise IndexError, "already passed "+str(i)
        try:
            while i > self.c: 
                self.i.next()
                self.c += 1
            # now self.c == i
            self.c += 1
            return self.i.next()
        except StopIteration: 
            raise IndexError, str(i)
iterbetter = IterBetter

def dictreverse(mapping):
    """
        >>> dictreverse({1: 2, 3: 4})
        {2: 1, 4: 3}
    """
    return dict([(value, key) for (key, value) in mapping.iteritems()])

def dictfind(dictionary, element):
    """
    Returns a key whose value in `dictionary` is `element` 
    or, if none exists, None.
    
        >>> d = {1:2, 3:4}
        >>> dictfind(d, 4)
        3
        >>> dictfind(d, 5)
    """
    for (key, value) in dictionary.iteritems():
        if element is value: 
            return key

def dictfindall(dictionary, element):
    """
    Returns the keys whose values in `dictionary` are `element`
    or, if none exists, [].
    
        >>> d = {1:4, 3:4}
        >>> dictfindall(d, 4)
        [1, 3]
        >>> dictfindall(d, 5)
        []
    """