1# -*- coding: utf-8 -*- 2""" 3 markupsafe 4 ~~~~~~~~~~ 5 6 Implements a Markup string. 7 8 :copyright: (c) 2010 by Armin Ronacher. 9 :license: BSD, see LICENSE for more details. 10""" 11import re 12from ._compat import text_type, string_types, int_types, \ 13 unichr, PY2 14 15 16__all__ = ['Markup', 'soft_unicode', 'escape', 'escape_silent'] 17 18 19_striptags_re = re.compile(r'(<!--.*?-->|<[^>]*>)') 20_entity_re = re.compile(r'&([^;]+);') 21 22 23class Markup(text_type): 24 r"""Marks a string as being safe for inclusion in HTML/XML output without 25 needing to be escaped. This implements the `__html__` interface a couple 26 of frameworks and web applications use. :class:`Markup` is a direct 27 subclass of `unicode` and provides all the methods of `unicode` just that 28 it escapes arguments passed and always returns `Markup`. 29 30 The `escape` function returns markup objects so that double escaping can't 31 happen. 32 33 The constructor of the :class:`Markup` class can be used for three 34 different things: When passed an unicode object it's assumed to be safe, 35 when passed an object with an HTML representation (has an `__html__` 36 method) that representation is used, otherwise the object passed is 37 converted into a unicode string and then assumed to be safe: 38 39 >>> Markup("Hello <em>World</em>!") 40 Markup(u'Hello <em>World</em>!') 41 >>> class Foo(object): 42 ... def __html__(self): 43 ... return '<a href="#">foo</a>' 44 ... 45 >>> Markup(Foo()) 46 Markup(u'<a href="#">foo</a>') 47 48 If you want object passed being always treated as unsafe you can use the 49 :meth:`escape` classmethod to create a :class:`Markup` object: 50 51 >>> Markup.escape("Hello <em>World</em>!") 52 Markup(u'Hello <em>World</em>!') 53 54 Operations on a markup string are markup aware which means that all 55 arguments are passed through the :func:`escape` function: 56 57 >>> em = Markup("<em>%s</em>") 58 >>> em % "foo & bar" 59 Markup(u'<em>foo & bar</em>') 60 >>> strong = Markup("<strong>%(text)s</strong>") 61 >>> strong % {'text': '<blink>hacker here</blink>'} 62 Markup(u'<strong><blink>hacker here</blink></strong>') 63 >>> Markup("<em>Hello</em> ") + "<foo>" 64 Markup(u'<em>Hello</em> <foo>') 65 """ 66 __slots__ = () 67 68 def __new__(cls, base=u'', encoding=None, errors='strict'): 69 if hasattr(base, '__html__'): 70 base = base.__html__() 71 if encoding is None: 72 return text_type.__new__(cls, base) 73 return text_type.__new__(cls, base, encoding, errors) 74 75 def __html__(self): 76 return self 77 78 def __add__(self, other): 79 if isinstance(other, string_types) or hasattr(other, '__html__'): 80 return self.__class__(super(Markup, self).__add__(self.escape(other))) 81 return NotImplemented 82 83 def __radd__(self, other): 84 if hasattr(other, '__html__') or isinstance(other, string_types): 85 return self.escape(other).__add__(self) 86 return NotImplemented 87 88 def __mul__(self, num): 89 if isinstance(num, int_types): 90 return self.__class__(text_type.__mul__(self, num)) 91 return NotImplemented 92 __rmul__ = __mul__ 93 94 def __mod__(self, arg): 95 if isinstance(arg, tuple): 96 arg = tuple(_MarkupEscapeHelper(x, self.escape) for x in arg) 97 else: 98 arg = _MarkupEscapeHelper(arg, self.escape) 99 return self.__class__(text_type.__mod__(self, arg)) 100 101 def __repr__(self): 102 return '%s(%s)' % ( 103 self.__class__.__name__, 104 text_type.__repr__(self) 105 ) 106 107 def join(self, seq): 108 return self.__class__(text_type.join(self, map(self.escape, seq))) 109 join.__doc__ = text_type.join.__doc__ 110 111 def split(self, *args, **kwargs): 112 return list(map(self.__class__, text_type.split(self, *args, **kwargs))) 113 split.__doc__ = text_type.split.__doc__ 114 115 def rsplit(self, *args, **kwargs): 116 return list(map(self.__class__, text_type.rsplit(self, *args, **kwargs))) 117 rsplit.__doc__ = text_type.rsplit.__doc__ 118 119 def splitlines(self, *args, **kwargs): 120 return list(map(self.__class__, text_type.splitlines(self, *args, **kwargs))) 121 splitlines.__doc__ = text_type.splitlines.__doc__ 122 123 def unescape(self): 124 r"""Unescape markup again into an text_type string. This also resolves 125 known HTML4 and XHTML entities: 126 127 >>> Markup("Main » <em>About</em>").unescape() 128 u'Main \xbb <em>About</em>' 129 """ 130 from _constants import HTML_ENTITIES 131 def handle_match(m): 132 name = m.group(1) 133 if name in HTML_ENTITIES: 134 return unichr(HTML_ENTITIES[name]) 135 try: 136 if name[:2] in ('#x', '#X'): 137 return unichr(int(name[2:], 16)) 138 elif name.startswith('#'): 139 return unichr(int(name[1:])) 140 except ValueError: 141 pass 142 return u'' 143 return _entity_re.sub(handle_match, text_type(self)) 144 145 def striptags(self): 146 r"""Unescape markup into an text_type string and strip all tags. This 147 also resolves known HTML4 and XHTML entities. Whitespace is 148 normalized to one: 149 150 >>> Markup("Main » <em>About</em>").striptags() 151 u'Main \xbb About' 152 """ 153 stripped = u' '.join(_striptags_re.sub('', self).split()) 154 return Markup(stripped).unescape() 155 156 @classmethod 157 def escape(cls, s): 158 """Escape the string. Works like :func:`escape` with the difference 159 that for subclasses of :class:`Markup` this function would return the 160 correct subclass. 161 """ 162 rv = escape(s) 163 if rv.__class__ is not cls: 164 return cls(rv) 165 return rv 166 167 def make_wrapper(name): 168 orig = getattr(text_type, name) 169 def func(self, *args, **kwargs): 170 args = _escape_argspec(list(args), enumerate(args), self.escape) 171 #_escape_argspec(kwargs, kwargs.iteritems(), None) 172 return self.__class__(orig(self, *args, **kwargs)) 173 func.__name__ = orig.__name__ 174 func.__doc__ = orig.__doc__ 175 return func 176 177 for method in '__getitem__', 'capitalize', \ 178 'title', 'lower', 'upper', 'replace', 'ljust', \ 179 'rjust', 'lstrip', 'rstrip', 'center', 'strip', \ 180 'translate', 'expandtabs', 'swapcase', 'zfill': 181 locals()[method] = make_wrapper(method) 182 183 # new in python 2.5 184 if hasattr(text_type, 'partition'): 185 def partition(self, sep): 186 return tuple(map(self.__class__, 187 text_type.partition(self, self.escape(sep)))) 188 def rpartition(self, sep): 189 return tuple(map(self.__class__, 190 text_type.rpartition(self, self.escape(sep)))) 191 192 # new in python 2.6 193 if hasattr(text_type, 'format'): 194 format = make_wrapper('format') 195 196 # not in python 3 197 if hasattr(text_type, '__getslice__'): 198 __getslice__ = make_wrapper('__getslice__') 199 200 del method, make_wrapper 201 202 203def _escape_argspec(obj, iterable, escape): 204 """Helper for various string-wrapped functions.""" 205 for key, value in iterable: 206 if hasattr(value, '__html__') or isinstance(value, string_types): 207 obj[key] = escape(value) 208 return obj 209 210 211class _MarkupEscapeHelper(object): 212 """Helper for Markup.__mod__""" 213 214 def __init__(self, obj, escape): 215 self.obj = obj 216 self.escape = escape 217 218 __getitem__ = lambda s, x: _MarkupEscapeHelper(s.obj[x], s.escape) 219 __unicode__ = __str__ = lambda s: text_type(s.escape(s.obj)) 220 __repr__ = lambda s: str(s.escape(repr(s.obj))) 221 __int__ = lambda s: int(s.obj) 222 __float__ = lambda s: float(s.obj) 223 224 225# we have to import it down here as the speedups and native 226# modules imports the markup type which is define above. 227try: 228 from _speedups import escape, escape_silent, soft_unicode 229except ImportError: 230 from ._native import escape, escape_silent, soft_unicode 231 232if not PY2: 233 soft_str = soft_unicode 234 __all__.append('soft_str') 235