1

I have a class called EasyUrl() that is derived from urlparse.Parseresult(). ParseResult() is instantiated when calling urlparse.urlparse(url), I have a static method inside EasyUrl() that changes the class type of the instantiated ParseResult() object into a EasyUrl() object. I wrap the urlparse.urlparse() function and the class type conversion into a function parse_url().

The reason behind such a function, is my attempt to hack around a separate problem I don't require an answer to but would like one, is I get a TypeError when __new__ is called during the instantiating process, which lets me know I have an invalid number of arguments.

Error received when instantiating EasyUrl() directly

# snippet 
url = 'stackoverflow.com'
url = EasyUrl(url)
# snippet end

Output:
TypeError: __new__() takes exactly 7 arguments (2 given)

The ParseResult() class inherits from a namedtuple().

Excerpt from urlparse library

class ParseResult(namedtuple('ParseResult', 'scheme netloc path params query fragment'), ResultMixin):

    __slots__ = ()

    def geturl(self):
        return urlunparse(self)

Now that I have described a little functionality of the code, here is the problem. I can't access the named tuple's (ParseResult) attributes. I'm trying to implement a default scheme for ParseResult() if it is missing.

But I can't access the attributes in the class definition.

import urlparse


def parse_url(url):
    """ Return a parsed EasyUrl() object"""
    parse_result = urlparse.urlparse(url)
    return EasyUrl.EvolveParseResult(parse_result)


class EasyUrl(urlparse.ParseResult):

    @staticmethod
    def EvolveParseResult(parse_result):
        """ Change the type of class into a EasyUrl() Object."""
        parse_result.__class__ = EasyUrl
        easy_url = parse_result # For readabilty
        easy_url.init()
        return easy_url

    def __init__(self, url):
        self = parse_url(url) # doesn't work

    def init(self):
        self.url = self.geturl()
        #self.set_scheme_if_non() # Uncomment when no error is raised

    def set_scheme_if_non(self, scheme='http'):
        if not self.scheme:
            self.scheme = scheme
            self.url = self.geturl() # Rebuild our url with the new scheme



# Passes the set_scheme_if_non trigger
#url = 'https://stackoverflow.com'
# Fails if statment, then attempts to set the variable,
# but error is raised: AttributeError: can't set attribute
url = 'stackoverflow.com'

# Will give the error: TypeError: __new__() takes exactly 7 arguments (2 given)
#url = EasyUrl(url)

# works fine, I don't know why. Except that I can't access
# the tuples attributes in the class definition
url = parse_url(url) 

print url.scheme # Works fine

url.set_scheme_if_non() # Raises an error

Output

File "/home/crispycret/easyurl.py", line 50, in <module>
  url.set_scheme_if_non() # Raises an error
File "/home/crispycret/easyurl.py", line 29, in set_scheme_if_non
  self.scheme = scheme

AttributeError: can't set attribute
4

1 回答 1

1

为什么不从头开始创建一个新类并从上面转移所有属性ParseResult呢？

import urlparse

class EasyURL(object):
    def __init__(self, parse_result):
        self.scheme = parse_result.scheme
        self.netloc = parse_result.netloc
        self.path = parse_result.path
        self.params = parse_result.params
        self.query = parse_result.query
        self.fragment = parse_result.fragment

    @classmethod
    def from_url(cls, url):
        return cls(urlparse.urlparse(url))

if __name__ == '__main__':
    url = 'http://foo.bar.com:8888/path/to/script.php?a=1&b=2'

    # Call from_url class method, which is very convenient
    easy_url = EasyURL.from_url(url)

    # Or, do it yourself
    easy_url = EasyURL(urlparse.urlparse(url))

您现在可以根据需要向此类添加任何其他方法。

更新

  • namedtuple是一个动态创建新类型的函数。顺便说一句，EasyURL除非我们添加一些代码，否则我们的对象不会充当元组__getitem__()
  • ParseResult不允许您更改您的属性，例如scheme，因此没有理由从它继承。
于 2015-06-06T02:41:02.237 回答