Python之字符串

2021-05-18 06:30

阅读:587

标签:elf   search   cme   diff   子序列   move   ESS   eth   rsa   

字符串:str

如"hzw","xidian university"

每个字符串拥有的功能:

class str(basestring):
    """
    str(object=‘‘) -> string
    
    Return a nice string representation of the object.
    If the argument is a string, the return value is the same object.
    """
    def capitalize(self):  
        """ 首字母变大写 """
        """
        S.capitalize() -> string
        
        Return a copy of the string S with only its first character
        capitalized.
        """
        return ""

    def center(self, width, fillchar=None):  
        """ 内容居中,width:总长度;fillchar:空白处填充内容,默认无 """
        """
        S.center(width[, fillchar]) -> string
        
        Return S centered in a string of length width. Padding is
        done using the specified fill character (default is a space)
        """
        return ""

    def count(self, sub, start=None, end=None):  
        """ 子序列个数 """
        """
        S.count(sub[, start[, end]]) -> int
        
        Return the number of non-overlapping occurrences of substring sub in
        string S[start:end].  Optional arguments start and end are interpreted
        as in slice notation.
        """
        return 0

    def decode(self, encoding=None, errors=None):  
        """ 解码 """
        """
        S.decode([encoding[,errors]]) -> object
        
        Decodes S using the codec registered for encoding. encoding defaults
        to the default encoding. errors may be given to set a different error
        handling scheme. Default is ‘strict‘ meaning that encoding errors raise
        a UnicodeDecodeError. Other possible values are ‘ignore‘ and ‘replace‘
        as well as any other name registered with codecs.register_error that is
        able to handle UnicodeDecodeErrors.
        """
        return object()

    def encode(self, encoding=None, errors=None):  
        """ 编码,针对unicode """
        """
        S.encode([encoding[,errors]]) -> object
        
        Encodes S using the codec registered for encoding. encoding defaults
        to the default encoding. errors may be given to set a different error
        handling scheme. Default is ‘strict‘ meaning that encoding errors raise
        a UnicodeEncodeError. Other possible values are ‘ignore‘, ‘replace‘ and
        ‘xmlcharrefreplace‘ as well as any other name registered with
        codecs.register_error that is able to handle UnicodeEncodeErrors.
        """
        return object()

    def endswith(self, suffix, start=None, end=None):  
        """ 是否以 xxx 结束 """
        """
        S.endswith(suffix[, start[, end]]) -> bool
        
        Return True if S ends with the specified suffix, False otherwise.
        With optional start, test S beginning at that position.
        With optional end, stop comparing S at that position.
        suffix can also be a tuple of strings to try.
        """
        return False

    def expandtabs(self, tabsize=None):  
        """ 将tab转换成空格,默认一个tab转换成8个空格 """
        """
        S.expandtabs([tabsize]) -> string
        
        Return a copy of S where all tab characters are expanded using spaces.
        If tabsize is not given, a tab size of 8 characters is assumed.
        """
        return ""

    def find(self, sub, start=None, end=None):  
        """ 寻找子序列位置,如果没找到,返回 -1 """
        """
        S.find(sub [,start [,end]]) -> int
        
        Return the lowest index in S where substring sub is found,
        such that sub is contained within S[start:end].  Optional
        arguments start and end are interpreted as in slice notation.
        
        Return -1 on failure.
        """
        return 0

    def format(*args, **kwargs): # known special case of str.format
        """ 字符串格式化,动态参数,将函数式编程时细说 """
        """
        S.format(*args, **kwargs) -> string
        
        Return a formatted version of S, using substitutions from args and kwargs.
        The substitutions are identified by braces (‘{‘ and ‘}‘).
        """
        pass

    def index(self, sub, start=None, end=None):  
        """ 子序列位置,如果没找到,报错 """
        S.index(sub [,start [,end]]) -> int
        
        Like S.find() but raise ValueError when the substring is not found.
        """
        return 0

    def isalnum(self):  
        """ 是否是字母和数字 """
        """
        S.isalnum() -> bool
        
        Return True if all characters in S are alphanumeric
        and there is at least one character in S, False otherwise.
        """
        return False

    def isalpha(self):  
        """ 是否是字母 """
        """
        S.isalpha() -> bool
        
        Return True if all characters in S are alphabetic
        and there is at least one character in S, False otherwise.
        """
        return False

    def isdigit(self):  
        """ 是否是数字 """
        """
        S.isdigit() -> bool
        
        Return True if all characters in S are digits
        and there is at least one character in S, False otherwise.
        """
        return False

    def islower(self):  
        """ 是否小写 """
        """
        S.islower() -> bool
        
        Return True if all cased characters in S are lowercase and there is
        at least one cased character in S, False otherwise.
        """
        return False

    def isspace(self):  
        """
        S.isspace() -> bool
        
        Return True if all characters in S are whitespace
        and there is at least one character in S, False otherwise.
        """
        return False

    def istitle(self):  
        """
        S.istitle() -> bool
        
        Return True if S is a titlecased string and there is at least one
        character in S, i.e. uppercase characters may only follow uncased
        characters and lowercase characters only cased ones. Return False
        otherwise.
        """
        return False

    def isupper(self):  
        """
        S.isupper() -> bool
        
        Return True if all cased characters in S are uppercase and there is
        at least one cased character in S, False otherwise.
        """
        return False

    def join(self, iterable):  
        """ 连接 """
        """
        S.join(iterable) -> string
        
        Return a string which is the concatenation of the strings in the
        iterable.  The separator between elements is S.
        """
        return ""

    def ljust(self, width, fillchar=None):  
        """ 内容左对齐,右侧填充 """
        """
        S.ljust(width[, fillchar]) -> string
        
        Return S left-justified in a string of length width. Padding is
        done using the specified fill character (default is a space).
        """
        return ""

    def lower(self):  
        """ 变小写 """
        """
        S.lower() -> string
        
        Return a copy of the string S converted to lowercase.
        """
        return ""

    def lstrip(self, chars=None):  
        """ 移除左侧空白 """
        """
        S.lstrip([chars]) -> string or unicode
        
        Return a copy of the string S with leading whitespace removed.
        If chars is given and not None, remove characters in chars instead.
        If chars is unicode, S will be converted to unicode before stripping
        """
        return ""

    def partition(self, sep):  
        """ 分割,前,中,后三部分 """
        """
        S.partition(sep) -> (head, sep, tail)
        
        Search for the separator sep in S, and return the part before it,
        the separator itself, and the part after it.  If the separator is not
        found, return S and two empty strings.
        """
        pass

    def replace(self, old, new, count=None):  
        """ 替换 """
        """
        S.replace(old, new[, count]) -> string
        
        Return a copy of string S with all occurrences of substring
        old replaced by new.  If the optional argument count is
        given, only the first count occurrences are replaced.
        """
        return ""

    def rfind(self, sub, start=None, end=None):  
        """
        S.rfind(sub [,start [,end]]) -> int
        
        Return the highest index in S where substring sub is found,
        such that sub is contained within S[start:end].  Optional
        arguments start and end are interpreted as in slice notation.
        
        Return -1 on failure.
        """
        return 0

    def rindex(self, sub, start=None, end=None):  
        """
        S.rindex(sub [,start [,end]]) -> int
        
        Like S.rfind() but raise ValueError when the substring is not found.
        """
        return 0

    def rjust(self, width, fillchar=None):  
        """
        S.rjust(width[, fillchar]) -> string
        
        Return S right-justified in a string of length width. Padding is
        done using the specified fill character (default is a space)
        """
        return ""

    def rpartition(self, sep):  
        """
        S.rpartition(sep) -> (head, sep, tail)
        
        Search for the separator sep in S, starting at the end of S, and return
        the part before it, the separator itself, and the part after it.  If the
        separator is not found, return two empty strings and S.
        """
        pass

    def rsplit(self, sep=None, maxsplit=None):  
        """
        S.rsplit([sep [,maxsplit]]) -> list of strings
        
        Return a list of the words in the string S, using sep as the
        delimiter string, starting at the end of the string and working
        to the front.  If maxsplit is given, at most maxsplit splits are
        done. If sep is not specified or is None, any whitespace string
        is a separator.
        """
        return []

    def rstrip(self, chars=None):  
        """
        S.rstrip([chars]) -> string or unicode
        
        Return a copy of the string S with trailing whitespace removed.
        If chars is given and not None, remove characters in chars instead.
        If chars is unicode, S will be converted to unicode before stripping
        """
        return ""

    def split(self, sep=None, maxsplit=None):  
        """ 分割, maxsplit最多分割几次 """
        """
        S.split([sep [,maxsplit]]) -> list of strings
        
        Return a list of the words in the string S, using sep as the
        delimiter string.  If maxsplit is given, at most maxsplit
        splits are done. If sep is not specified or is None, any
        whitespace string is a separator and empty strings are removed
        from the result.
        """
        return []

    def splitlines(self, keepends=False):  
        """ 根据换行分割 """
        """
        S.splitlines(keepends=False) -> list of strings
        
        Return a list of the lines in S, breaking at line boundaries.
        Line breaks are not included in the resulting list unless keepends
        is given and true.
        """
        return []

    def startswith(self, prefix, start=None, end=None):  
        """ 是否起始 """
        """
        S.startswith(prefix[, start[, end]]) -> bool
        
        Return True if S starts with the specified prefix, False otherwise.
        With optional start, test S beginning at that position.
        With optional end, stop comparing S at that position.
        prefix can also be a tuple of strings to try.
        """
        return False

    def strip(self, chars=None):  
        """ 移除两段空白 """
        """
        S.strip([chars]) -> string or unicode
        
        Return a copy of the string S with leading and trailing
        whitespace removed.
        If chars is given and not None, remove characters in chars instead.
        If chars is unicode, S will be converted to unicode before stripping
        """
        return ""

    def swapcase(self):  
        """ 大写变小写,小写变大写 """
        """
        S.swapcase() -> string
        
        Return a copy of the string S with uppercase characters
        converted to lowercase and vice versa.
        """
        return ""

    def title(self):  
        """
        S.title() -> string
        
        Return a titlecased version of S, i.e. words start with uppercase
        characters, all remaining cased characters have lowercase.
        """
        return ""

    def translate(self, table, deletechars=None):  
        """
        转换,需要先做一个对应表,最后一个表示删除字符集合
        intab = "aeiou"
        outtab = "12345"
        trantab = maketrans(intab, outtab)
        str = "this is string example....wow!!!"
        print str.translate(trantab, xm)
        """

        """
        S.translate(table [,deletechars]) -> string
        
        Return a copy of the string S, where all characters occurring
        in the optional argument deletechars are removed, and the
        remaining characters have been mapped through the given
        translation table, which must be a string of length 256 or None.
        If the table argument is None, no translation is applied and
        the operation simply removes the characters in deletechars.
        """
        return ""

    def upper(self):  
        """
        S.upper() -> string
        
        Return a copy of the string S converted to uppercase.
        """
        return ""

    def zfill(self, width):  
        """方法返回指定长度的字符串,原字符串右对齐,前面填充0。"""
        """
        S.zfill(width) -> string
        
        Pad a numeric string S with zeros on the left, to fill a field
        of the specified width.  The string S is never truncated.
        """
        return ""

    def _formatter_field_name_split(self, *args, **kwargs): # real signature unknown
        pass

    def _formatter_parser(self, *args, **kwargs): # real signature unknown
        pass

    def __add__(self, y):  
        """ x.__add__(y)  x+y """
        pass

    def __contains__(self, y):  
        """ x.__contains__(y)  y in x """
        pass

    def __eq__(self, y):  
        """ x.__eq__(y)  x==y """
        pass

    def __format__(self, format_spec):  
        """
        S.__format__(format_spec) -> string
        
        Return a formatted version of S as described by format_spec.
        """
        return ""

    def __getattribute__(self, name):  
        """ x.__getattribute__(name)  x.name """
        pass

    def __getitem__(self, y):  
        """ x.__getitem__(y)  x[y] """
        pass

    def __getnewargs__(self, *args, **kwargs): # real signature unknown
        pass

    def __getslice__(self, i, j):  
        """
        x.__getslice__(i, j)  x[i:j]
                   
                   Use of negative indices is not supported.
        """
        pass

    def __ge__(self, y):  
        """ x.__ge__(y)  x>=y """
        pass

    def __gt__(self, y):  
        """ x.__gt__(y)  x>y """
        pass

    def __hash__(self):  
        """ x.__hash__()  hash(x) """
        pass

    def __init__(self, string=‘‘): # known special case of str.__init__
        """
        str(object=‘‘) -> string
        
        Return a nice string representation of the object.
        If the argument is a string, the return value is the same object.
        # (copied from class doc)
        """
        pass

    def __len__(self):  
        """ x.__len__()  len(x) """
        pass

    def __le__(self, y):  
        """ x.__le__(y)  x"""
        pass

    def __lt__(self, y):  
        """ x.__lt__(y)  x"""
        pass

    def __mod__(self, y):  
        """ x.__mod__(y)  x%y """
        pass

    def __mul__(self, n):  
        """ x.__mul__(n)  x*n """
        pass

    @staticmethod # known case of __new__
    def __new__(S, *more):  
        """ T.__new__(S, ...) -> a new object with type S, a subtype of T """
        pass

    def __ne__(self, y):  
        """ x.__ne__(y)  x!=y """
        pass

    def __repr__(self):  
        """ x.__repr__()  repr(x) """
        pass

    def __rmod__(self, y):  
        """ x.__rmod__(y)  y%x """
        pass

    def __rmul__(self, n):  
        """ x.__rmul__(n)  n*x """
        pass

    def __sizeof__(self):  
        """ S.__sizeof__() -> size of S in memory, in bytes """
        pass

    def __str__(self):  
        """ x.__str__()  str(x) """
        pass

str

str

 

Python之字符串

标签:elf   search   cme   diff   子序列   move   ESS   eth   rsa   

原文地址:https://www.cnblogs.com/pythonlearing/p/9744964.html


评论


亲,登录后才可以留言!