Armin Ronacher | 07bc684 | 2008-03-31 14:18:49 +0200 | [diff] [blame] | 1 | # -*- coding: utf-8 -*- |
| 2 | """ |
| 3 | jinja2.utils |
| 4 | ~~~~~~~~~~~~ |
| 5 | |
| 6 | Utility functions. |
| 7 | |
| 8 | :copyright: 2008 by Armin Ronacher. |
| 9 | :license: BSD, see LICENSE for more details. |
| 10 | """ |
Christoph Hack | 8090986 | 2008-04-14 01:35:10 +0200 | [diff] [blame] | 11 | import re |
Benjamin Wiegand | 9682855 | 2008-05-03 22:27:29 +0200 | [diff] [blame] | 12 | import sys |
Christoph Hack | 8090986 | 2008-04-14 01:35:10 +0200 | [diff] [blame] | 13 | import string |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 14 | try: |
| 15 | from thread import allocate_lock |
| 16 | except ImportError: |
| 17 | from dummy_thread import allocate_lock |
Armin Ronacher | 76c280b | 2008-05-04 12:31:48 +0200 | [diff] [blame] | 18 | from htmlentitydefs import name2codepoint |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 19 | from collections import deque |
| 20 | from copy import deepcopy |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 21 | from itertools import imap |
Armin Ronacher | 8edbe49 | 2008-04-10 20:43:43 +0200 | [diff] [blame] | 22 | |
| 23 | |
Armin Ronacher | be4ae24 | 2008-04-18 09:49:08 +0200 | [diff] [blame] | 24 | _word_split_re = re.compile(r'(\s+)') |
| 25 | _punctuation_re = re.compile( |
| 26 | '^(?P<lead>(?:%s)*)(?P<middle>.*?)(?P<trail>(?:%s)*)$' % ( |
| 27 | '|'.join(imap(re.escape, ('(', '<', '<'))), |
| 28 | '|'.join(imap(re.escape, ('.', ',', ')', '>', '\n', '>'))) |
| 29 | ) |
| 30 | ) |
| 31 | _simple_email_re = re.compile(r'^\S+@[a-zA-Z0-9._-]+\.[a-zA-Z0-9._-]+$') |
Armin Ronacher | 76c280b | 2008-05-04 12:31:48 +0200 | [diff] [blame] | 32 | _striptags_re = re.compile(r'(<!--.*?-->|<[^>]*>)') |
| 33 | _entity_re = re.compile(r'&([^;]+);') |
| 34 | _entities = name2codepoint.copy() |
| 35 | _entities['apos'] = 39 |
Armin Ronacher | be4ae24 | 2008-04-18 09:49:08 +0200 | [diff] [blame] | 36 | |
Armin Ronacher | 7259c76 | 2008-04-30 13:03:59 +0200 | [diff] [blame] | 37 | # special singleton representing missing values for the runtime |
| 38 | missing = type('MissingType', (), {'__repr__': lambda x: 'missing'})() |
| 39 | |
| 40 | |
Armin Ronacher | 7ceced5 | 2008-05-03 10:15:31 +0200 | [diff] [blame] | 41 | # concatenate a list of strings and convert them to unicode. |
| 42 | # unfortunately there is a bug in python 2.4 and lower that causes |
| 43 | # unicode.join trash the traceback. |
Armin Ronacher | cda43df | 2008-05-03 17:10:05 +0200 | [diff] [blame] | 44 | _concat = u''.join |
Armin Ronacher | 7ceced5 | 2008-05-03 10:15:31 +0200 | [diff] [blame] | 45 | try: |
| 46 | def _test_gen_bug(): |
| 47 | raise TypeError(_test_gen_bug) |
| 48 | yield None |
Armin Ronacher | cda43df | 2008-05-03 17:10:05 +0200 | [diff] [blame] | 49 | _concat(_test_gen_bug()) |
Armin Ronacher | 7ceced5 | 2008-05-03 10:15:31 +0200 | [diff] [blame] | 50 | except TypeError, _error: |
Armin Ronacher | cda43df | 2008-05-03 17:10:05 +0200 | [diff] [blame] | 51 | if not _error.args or _error.args[0] is not _test_gen_bug: |
Armin Ronacher | 7ceced5 | 2008-05-03 10:15:31 +0200 | [diff] [blame] | 52 | def concat(gen): |
| 53 | try: |
Armin Ronacher | cda43df | 2008-05-03 17:10:05 +0200 | [diff] [blame] | 54 | return _concat(list(gen)) |
Armin Ronacher | 7ceced5 | 2008-05-03 10:15:31 +0200 | [diff] [blame] | 55 | except: |
| 56 | # this hack is needed so that the current frame |
| 57 | # does not show up in the traceback. |
| 58 | exc_type, exc_value, tb = sys.exc_info() |
| 59 | raise exc_type, exc_value, tb.tb_next |
Armin Ronacher | cda43df | 2008-05-03 17:10:05 +0200 | [diff] [blame] | 60 | else: |
| 61 | concat = _concat |
Armin Ronacher | 7ceced5 | 2008-05-03 10:15:31 +0200 | [diff] [blame] | 62 | del _test_gen_bug, _error |
| 63 | |
| 64 | |
Armin Ronacher | 4f7d2d5 | 2008-04-22 10:40:26 +0200 | [diff] [blame] | 65 | def contextfunction(f): |
Armin Ronacher | 9bb7e47 | 2008-05-28 11:26:59 +0200 | [diff] [blame] | 66 | """This decorator can be used to mark a function or method context callable. |
| 67 | A context callable is passed the active :class:`Context` as first argument when |
| 68 | called from the template. This is useful if a function wants to get access |
| 69 | to the context or functions provided on the context object. For example |
| 70 | a function that returns a sorted list of template variables the current |
| 71 | template exports could look like this:: |
| 72 | |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 73 | @contextfunction |
Armin Ronacher | 9bb7e47 | 2008-05-28 11:26:59 +0200 | [diff] [blame] | 74 | def get_exported_names(context): |
| 75 | return sorted(context.exported_vars) |
Armin Ronacher | 4f7d2d5 | 2008-04-22 10:40:26 +0200 | [diff] [blame] | 76 | """ |
| 77 | f.contextfunction = True |
| 78 | return f |
| 79 | |
| 80 | |
Armin Ronacher | 203bfcb | 2008-04-24 21:54:44 +0200 | [diff] [blame] | 81 | def environmentfunction(f): |
Armin Ronacher | 9bb7e47 | 2008-05-28 11:26:59 +0200 | [diff] [blame] | 82 | """This decorator can be used to mark a function or method as environment |
| 83 | callable. This decorator works exactly like the :func:`contextfunction` |
| 84 | decorator just that the first argument is the active :class:`Environment` |
| 85 | and not context. |
Armin Ronacher | 203bfcb | 2008-04-24 21:54:44 +0200 | [diff] [blame] | 86 | """ |
| 87 | f.environmentfunction = True |
| 88 | return f |
| 89 | |
| 90 | |
Armin Ronacher | 9bb7e47 | 2008-05-28 11:26:59 +0200 | [diff] [blame] | 91 | def is_undefined(obj): |
| 92 | """Check if the object passed is undefined. This does nothing more than |
| 93 | performing an instance check against :class:`Undefined` but looks nicer. |
| 94 | This can be used for custom filters or tests that want to react to |
| 95 | undefined variables. For example a custom default filter can look like |
| 96 | this:: |
| 97 | |
| 98 | def default(var, default=''): |
| 99 | if is_undefined(var): |
| 100 | return default |
| 101 | return var |
| 102 | """ |
| 103 | from jinja2.runtime import Undefined |
| 104 | return isinstance(obj, Undefined) |
| 105 | |
| 106 | |
Armin Ronacher | 187bde1 | 2008-05-01 18:19:16 +0200 | [diff] [blame] | 107 | def clear_caches(): |
| 108 | """Jinja2 keeps internal caches for environments and lexers. These are |
| 109 | used so that Jinja2 doesn't have to recreate environments and lexers all |
| 110 | the time. Normally you don't have to care about that but if you are |
| 111 | messuring memory consumption you may want to clean the caches. |
| 112 | """ |
| 113 | from jinja2.environment import _spontaneous_environments |
| 114 | from jinja2.lexer import _lexer_cache |
| 115 | _spontaneous_environments.clear() |
| 116 | _lexer_cache.clear() |
| 117 | |
| 118 | |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 119 | def import_string(import_name, silent=False): |
| 120 | """Imports an object based on a string. This use useful if you want to |
| 121 | use import paths as endpoints or something similar. An import path can |
| 122 | be specified either in dotted notation (``xml.sax.saxutils.escape``) |
| 123 | or with a colon as object delimiter (``xml.sax.saxutils:escape``). |
| 124 | |
| 125 | If the `silent` is True the return value will be `None` if the import |
| 126 | fails. |
| 127 | |
| 128 | :return: imported object |
Armin Ronacher | 9a027f4 | 2008-04-17 11:13:40 +0200 | [diff] [blame] | 129 | """ |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 130 | try: |
| 131 | if ':' in import_name: |
| 132 | module, obj = import_name.split(':', 1) |
| 133 | elif '.' in import_name: |
| 134 | items = import_name.split('.') |
| 135 | module = '.'.join(items[:-1]) |
| 136 | obj = items[-1] |
| 137 | else: |
| 138 | return __import__(import_name) |
| 139 | return getattr(__import__(module, None, None, [obj]), obj) |
| 140 | except (ImportError, AttributeError): |
| 141 | if not silent: |
| 142 | raise |
Armin Ronacher | 9a027f4 | 2008-04-17 11:13:40 +0200 | [diff] [blame] | 143 | |
| 144 | |
Christoph Hack | e9e43bb | 2008-04-13 23:35:48 +0200 | [diff] [blame] | 145 | def pformat(obj, verbose=False): |
Armin Ronacher | be4ae24 | 2008-04-18 09:49:08 +0200 | [diff] [blame] | 146 | """Prettyprint an object. Either use the `pretty` library or the |
Christoph Hack | e9e43bb | 2008-04-13 23:35:48 +0200 | [diff] [blame] | 147 | builtin `pprint`. |
| 148 | """ |
| 149 | try: |
| 150 | from pretty import pretty |
| 151 | return pretty(obj, verbose=verbose) |
| 152 | except ImportError: |
| 153 | from pprint import pformat |
| 154 | return pformat(obj) |
Christoph Hack | 8090986 | 2008-04-14 01:35:10 +0200 | [diff] [blame] | 155 | |
| 156 | |
Christoph Hack | 8090986 | 2008-04-14 01:35:10 +0200 | [diff] [blame] | 157 | def urlize(text, trim_url_limit=None, nofollow=False): |
Armin Ronacher | be4ae24 | 2008-04-18 09:49:08 +0200 | [diff] [blame] | 158 | """Converts any URLs in text into clickable links. Works on http://, |
Christoph Hack | 8090986 | 2008-04-14 01:35:10 +0200 | [diff] [blame] | 159 | https:// and www. links. Links can have trailing punctuation (periods, |
| 160 | commas, close-parens) and leading punctuation (opening parens) and |
| 161 | it'll still do the right thing. |
| 162 | |
| 163 | If trim_url_limit is not None, the URLs in link text will be limited |
| 164 | to trim_url_limit characters. |
| 165 | |
| 166 | If nofollow is True, the URLs in link text will get a rel="nofollow" |
| 167 | attribute. |
| 168 | """ |
| 169 | trim_url = lambda x, limit=trim_url_limit: limit is not None \ |
| 170 | and (x[:limit] + (len(x) >=limit and '...' |
| 171 | or '')) or x |
| 172 | words = _word_split_re.split(text) |
| 173 | nofollow_attr = nofollow and ' rel="nofollow"' or '' |
| 174 | for i, word in enumerate(words): |
| 175 | match = _punctuation_re.match(word) |
| 176 | if match: |
| 177 | lead, middle, trail = match.groups() |
| 178 | if middle.startswith('www.') or ( |
| 179 | '@' not in middle and |
| 180 | not middle.startswith('http://') and |
| 181 | len(middle) > 0 and |
| 182 | middle[0] in string.letters + string.digits and ( |
| 183 | middle.endswith('.org') or |
| 184 | middle.endswith('.net') or |
| 185 | middle.endswith('.com') |
| 186 | )): |
| 187 | middle = '<a href="http://%s"%s>%s</a>' % (middle, |
| 188 | nofollow_attr, trim_url(middle)) |
| 189 | if middle.startswith('http://') or \ |
| 190 | middle.startswith('https://'): |
| 191 | middle = '<a href="%s"%s>%s</a>' % (middle, |
| 192 | nofollow_attr, trim_url(middle)) |
| 193 | if '@' in middle and not middle.startswith('www.') and \ |
| 194 | not ':' in middle and _simple_email_re.match(middle): |
| 195 | middle = '<a href="mailto:%s">%s</a>' % (middle, middle) |
| 196 | if lead + middle + trail != word: |
| 197 | words[i] = lead + middle + trail |
| 198 | return u''.join(words) |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 199 | |
| 200 | |
Armin Ronacher | 4f7d2d5 | 2008-04-22 10:40:26 +0200 | [diff] [blame] | 201 | def generate_lorem_ipsum(n=5, html=True, min=20, max=100): |
| 202 | """Generate some lorem impsum for the template.""" |
| 203 | from jinja2.constants import LOREM_IPSUM_WORDS |
| 204 | from random import choice, random, randrange |
| 205 | words = LOREM_IPSUM_WORDS.split() |
| 206 | result = [] |
| 207 | |
| 208 | for _ in xrange(n): |
| 209 | next_capitalized = True |
| 210 | last_comma = last_fullstop = 0 |
| 211 | word = None |
| 212 | last = None |
| 213 | p = [] |
| 214 | |
| 215 | # each paragraph contains out of 20 to 100 words. |
| 216 | for idx, _ in enumerate(xrange(randrange(min, max))): |
| 217 | while True: |
| 218 | word = choice(words) |
| 219 | if word != last: |
| 220 | last = word |
| 221 | break |
| 222 | if next_capitalized: |
| 223 | word = word.capitalize() |
| 224 | next_capitalized = False |
| 225 | # add commas |
| 226 | if idx - randrange(3, 8) > last_comma: |
| 227 | last_comma = idx |
| 228 | last_fullstop += 2 |
| 229 | word += ',' |
| 230 | # add end of sentences |
| 231 | if idx - randrange(10, 20) > last_fullstop: |
| 232 | last_comma = last_fullstop = idx |
| 233 | word += '.' |
| 234 | next_capitalized = True |
| 235 | p.append(word) |
| 236 | |
| 237 | # ensure that the paragraph ends with a dot. |
| 238 | p = u' '.join(p) |
| 239 | if p.endswith(','): |
| 240 | p = p[:-1] + '.' |
| 241 | elif not p.endswith('.'): |
| 242 | p += '.' |
| 243 | result.append(p) |
| 244 | |
| 245 | if not html: |
| 246 | return u'\n\n'.join(result) |
| 247 | return Markup(u'\n'.join(u'<p>%s</p>' % escape(x) for x in result)) |
| 248 | |
| 249 | |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 250 | class Markup(unicode): |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 251 | r"""Marks a string as being safe for inclusion in HTML/XML output without |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 252 | needing to be escaped. This implements the `__html__` interface a couple |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 253 | of frameworks and web applications use. :class:`Markup` is a direct |
| 254 | subclass of `unicode` and provides all the methods of `unicode` just that |
| 255 | it escapes arguments passed and always returns `Markup`. |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 256 | |
| 257 | The `escape` function returns markup objects so that double escaping can't |
| 258 | happen. If you want to use autoescaping in Jinja just set the finalizer |
| 259 | of the environment to `escape`. |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 260 | |
| 261 | The constructor of the :class:`Markup` class can be used for three |
| 262 | different things: When passed an unicode object it's assumed to be safe, |
| 263 | when passed an object with an HTML representation (has an `__html__` |
| 264 | method) that representation is used, otherwise the object passed is |
| 265 | converted into a unicode string and then assumed to be safe: |
| 266 | |
| 267 | >>> Markup("Hello <em>World</em>!") |
| 268 | Markup(u'Hello <em>World</em>!') |
| 269 | >>> class Foo(object): |
| 270 | ... def __html__(self): |
| 271 | ... return '<a href="#">foo</a>' |
| 272 | ... |
| 273 | >>> Markup(Foo()) |
| 274 | Markup(u'<a href="#">foo</a>') |
| 275 | |
| 276 | If you want object passed being always treated as unsafe you can use the |
| 277 | :meth:`escape` classmethod to create a :class:`Markup` object: |
| 278 | |
| 279 | >>> Markup.escape("Hello <em>World</em>!") |
| 280 | Markup(u'Hello <em>World</em>!') |
| 281 | |
| 282 | Operations on a markup string are markup aware which means that all |
| 283 | arguments are passed through the :func:`escape` function: |
| 284 | |
| 285 | >>> em = Markup("<em>%s</em>") |
| 286 | >>> em % "foo & bar" |
| 287 | Markup(u'<em>foo & bar</em>') |
| 288 | >>> strong = Markup("<strong>%(text)s</strong>") |
| 289 | >>> strong % {'text': '<blink>hacker here</blink>'} |
| 290 | Markup(u'<strong><blink>hacker here</blink></strong>') |
| 291 | >>> Markup("<em>Hello</em> ") + "<foo>" |
| 292 | Markup(u'<em>Hello</em> <foo>') |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 293 | """ |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 294 | __slots__ = () |
| 295 | |
Armin Ronacher | 4e6f9a2 | 2008-05-23 23:57:38 +0200 | [diff] [blame] | 296 | def __new__(cls, base=u''): |
| 297 | if hasattr(base, '__html__'): |
| 298 | base = base.__html__() |
| 299 | return unicode.__new__(cls, base) |
| 300 | |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 301 | def __html__(self): |
| 302 | return self |
| 303 | |
| 304 | def __add__(self, other): |
| 305 | if hasattr(other, '__html__') or isinstance(other, basestring): |
| 306 | return self.__class__(unicode(self) + unicode(escape(other))) |
| 307 | return NotImplemented |
| 308 | |
| 309 | def __radd__(self, other): |
| 310 | if hasattr(other, '__html__') or isinstance(other, basestring): |
| 311 | return self.__class__(unicode(escape(other)) + unicode(self)) |
| 312 | return NotImplemented |
| 313 | |
| 314 | def __mul__(self, num): |
Armin Ronacher | d71fff0 | 2008-05-26 23:57:07 +0200 | [diff] [blame] | 315 | if isinstance(num, (int, long)): |
| 316 | return self.__class__(unicode.__mul__(self, num)) |
| 317 | return NotImplemented |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 318 | __rmul__ = __mul__ |
| 319 | |
| 320 | def __mod__(self, arg): |
| 321 | if isinstance(arg, tuple): |
| 322 | arg = tuple(imap(_MarkupEscapeHelper, arg)) |
| 323 | else: |
| 324 | arg = _MarkupEscapeHelper(arg) |
| 325 | return self.__class__(unicode.__mod__(self, arg)) |
| 326 | |
| 327 | def __repr__(self): |
| 328 | return '%s(%s)' % ( |
| 329 | self.__class__.__name__, |
| 330 | unicode.__repr__(self) |
| 331 | ) |
| 332 | |
| 333 | def join(self, seq): |
| 334 | return self.__class__(unicode.join(self, imap(escape, seq))) |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 335 | join.__doc__ = unicode.join.__doc__ |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 336 | |
| 337 | def split(self, *args, **kwargs): |
| 338 | return map(self.__class__, unicode.split(self, *args, **kwargs)) |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 339 | split.__doc__ = unicode.split.__doc__ |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 340 | |
| 341 | def rsplit(self, *args, **kwargs): |
| 342 | return map(self.__class__, unicode.rsplit(self, *args, **kwargs)) |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 343 | rsplit.__doc__ = unicode.rsplit.__doc__ |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 344 | |
| 345 | def splitlines(self, *args, **kwargs): |
| 346 | return map(self.__class__, unicode.splitlines(self, *args, **kwargs)) |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 347 | splitlines.__doc__ = unicode.splitlines.__doc__ |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 348 | |
Armin Ronacher | 76c280b | 2008-05-04 12:31:48 +0200 | [diff] [blame] | 349 | def unescape(self): |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 350 | r"""Unescape markup again into an unicode string. This also resolves |
| 351 | known HTML4 and XHTML entities: |
| 352 | |
| 353 | >>> Markup("Main » <em>About</em>").unescape() |
| 354 | u'Main \xbb <em>About</em>' |
| 355 | """ |
Armin Ronacher | 76c280b | 2008-05-04 12:31:48 +0200 | [diff] [blame] | 356 | def handle_match(m): |
| 357 | name = m.group(1) |
| 358 | if name in _entities: |
| 359 | return unichr(_entities[name]) |
| 360 | try: |
| 361 | if name[:2] in ('#x', '#X'): |
| 362 | return unichr(int(name[2:], 16)) |
| 363 | elif name.startswith('#'): |
| 364 | return unichr(int(name[1:])) |
| 365 | except ValueError: |
| 366 | pass |
| 367 | return u'' |
| 368 | return _entity_re.sub(handle_match, unicode(self)) |
| 369 | |
| 370 | def striptags(self): |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 371 | r"""Unescape markup into an unicode string and strip all tags. This |
| 372 | also resolves known HTML4 and XHTML entities. Whitespace is |
| 373 | normalized to one: |
| 374 | |
| 375 | >>> Markup("Main » <em>About</em>").striptags() |
| 376 | u'Main \xbb About' |
| 377 | """ |
Armin Ronacher | 76c280b | 2008-05-04 12:31:48 +0200 | [diff] [blame] | 378 | stripped = u' '.join(_striptags_re.sub('', self).split()) |
| 379 | return Markup(stripped).unescape() |
| 380 | |
Armin Ronacher | f35e281 | 2008-05-06 16:04:10 +0200 | [diff] [blame] | 381 | @classmethod |
| 382 | def escape(cls, s): |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 383 | """Escape the string. Works like :func:`escape` with the difference |
| 384 | that for subclasses of :class:`Markup` this function would return the |
| 385 | correct subclass. |
| 386 | """ |
Armin Ronacher | f35e281 | 2008-05-06 16:04:10 +0200 | [diff] [blame] | 387 | rv = escape(s) |
| 388 | if rv.__class__ is not cls: |
| 389 | return cls(rv) |
| 390 | return rv |
| 391 | |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 392 | def make_wrapper(name): |
| 393 | orig = getattr(unicode, name) |
| 394 | def func(self, *args, **kwargs): |
Armin Ronacher | d71fff0 | 2008-05-26 23:57:07 +0200 | [diff] [blame] | 395 | args = _escape_argspec(list(args), enumerate(args)) |
| 396 | _escape_argspec(kwargs, kwargs.iteritems()) |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 397 | return self.__class__(orig(self, *args, **kwargs)) |
Armin Ronacher | 4f7d2d5 | 2008-04-22 10:40:26 +0200 | [diff] [blame] | 398 | func.__name__ = orig.__name__ |
| 399 | func.__doc__ = orig.__doc__ |
| 400 | return func |
Armin Ronacher | d71fff0 | 2008-05-26 23:57:07 +0200 | [diff] [blame] | 401 | |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 402 | for method in '__getitem__', '__getslice__', 'capitalize', \ |
| 403 | 'title', 'lower', 'upper', 'replace', 'ljust', \ |
Armin Ronacher | 709f6e5 | 2008-04-28 18:18:16 +0200 | [diff] [blame] | 404 | 'rjust', 'lstrip', 'rstrip', 'center', 'strip', \ |
Armin Ronacher | 316157d | 2008-04-28 18:30:27 +0200 | [diff] [blame] | 405 | 'translate', 'expandtabs', 'swapcase', 'zfill': |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 406 | locals()[method] = make_wrapper(method) |
Armin Ronacher | 709f6e5 | 2008-04-28 18:18:16 +0200 | [diff] [blame] | 407 | |
| 408 | # new in python 2.5 |
| 409 | if hasattr(unicode, 'partition'): |
Armin Ronacher | d71fff0 | 2008-05-26 23:57:07 +0200 | [diff] [blame] | 410 | partition = make_wrapper('partition'), |
| 411 | rpartition = make_wrapper('rpartition') |
| 412 | |
| 413 | # new in python 2.6 |
| 414 | if hasattr(unicode, 'format'): |
| 415 | format = make_wrapper('format') |
| 416 | |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 417 | del method, make_wrapper |
| 418 | |
| 419 | |
Armin Ronacher | d71fff0 | 2008-05-26 23:57:07 +0200 | [diff] [blame] | 420 | def _escape_argspec(obj, iterable): |
| 421 | """Helper for various string-wrapped functions.""" |
| 422 | for key, value in iterable: |
| 423 | if hasattr(value, '__html__') or isinstance(value, basestring): |
| 424 | obj[key] = escape(value) |
| 425 | return obj |
| 426 | |
| 427 | |
Armin Ronacher | 18c6ca0 | 2008-04-17 10:03:29 +0200 | [diff] [blame] | 428 | class _MarkupEscapeHelper(object): |
| 429 | """Helper for Markup.__mod__""" |
| 430 | |
| 431 | def __init__(self, obj): |
| 432 | self.obj = obj |
| 433 | |
| 434 | __getitem__ = lambda s, x: _MarkupEscapeHelper(s.obj[x]) |
| 435 | __unicode__ = lambda s: unicode(escape(s.obj)) |
| 436 | __str__ = lambda s: str(escape(s.obj)) |
| 437 | __repr__ = lambda s: str(repr(escape(s.obj))) |
| 438 | __int__ = lambda s: int(s.obj) |
| 439 | __float__ = lambda s: float(s.obj) |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 440 | |
| 441 | |
| 442 | class LRUCache(object): |
| 443 | """A simple LRU Cache implementation.""" |
Armin Ronacher | 58f351d | 2008-05-28 21:30:14 +0200 | [diff] [blame^] | 444 | |
| 445 | # this is fast for small capacities (something below 1000) but doesn't |
| 446 | # scale. But as long as it's only used as storage for templates this |
| 447 | # won't do any harm. |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 448 | |
| 449 | def __init__(self, capacity): |
| 450 | self.capacity = capacity |
| 451 | self._mapping = {} |
| 452 | self._queue = deque() |
Armin Ronacher | 7962ce7 | 2008-05-20 17:52:52 +0200 | [diff] [blame] | 453 | self._postinit() |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 454 | |
Armin Ronacher | 7962ce7 | 2008-05-20 17:52:52 +0200 | [diff] [blame] | 455 | def _postinit(self): |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 456 | # alias all queue methods for faster lookup |
| 457 | self._popleft = self._queue.popleft |
| 458 | self._pop = self._queue.pop |
| 459 | if hasattr(self._queue, 'remove'): |
| 460 | self._remove = self._queue.remove |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 461 | self._wlock = allocate_lock() |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 462 | self._append = self._queue.append |
| 463 | |
| 464 | def _remove(self, obj): |
| 465 | """Python 2.4 compatibility.""" |
| 466 | for idx, item in enumerate(self._queue): |
| 467 | if item == obj: |
| 468 | del self._queue[idx] |
| 469 | break |
| 470 | |
Armin Ronacher | 7962ce7 | 2008-05-20 17:52:52 +0200 | [diff] [blame] | 471 | def __getstate__(self): |
| 472 | return { |
| 473 | 'capacity': self.capacity, |
| 474 | '_mapping': self._mapping, |
| 475 | '_queue': self._queue |
| 476 | } |
| 477 | |
| 478 | def __setstate__(self, d): |
| 479 | self.__dict__.update(d) |
| 480 | self._postinit() |
| 481 | |
| 482 | def __getnewargs__(self): |
| 483 | return (self.capacity,) |
| 484 | |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 485 | def copy(self): |
| 486 | """Return an shallow copy of the instance.""" |
Armin Ronacher | be4ae24 | 2008-04-18 09:49:08 +0200 | [diff] [blame] | 487 | rv = self.__class__(self.capacity) |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 488 | rv._mapping.update(self._mapping) |
Armin Ronacher | be4ae24 | 2008-04-18 09:49:08 +0200 | [diff] [blame] | 489 | rv._queue = deque(self._queue) |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 490 | return rv |
| 491 | |
| 492 | def get(self, key, default=None): |
| 493 | """Return an item from the cache dict or `default`""" |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 494 | try: |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 495 | return self[key] |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 496 | except KeyError: |
| 497 | return default |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 498 | |
| 499 | def setdefault(self, key, default=None): |
Armin Ronacher | be4ae24 | 2008-04-18 09:49:08 +0200 | [diff] [blame] | 500 | """Set `default` if the key is not in the cache otherwise |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 501 | leave unchanged. Return the value of this key. |
| 502 | """ |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 503 | try: |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 504 | return self[key] |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 505 | except KeyError: |
| 506 | self[key] = default |
| 507 | return default |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 508 | |
| 509 | def clear(self): |
| 510 | """Clear the cache.""" |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 511 | self._wlock.acquire() |
| 512 | try: |
| 513 | self._mapping.clear() |
| 514 | self._queue.clear() |
| 515 | finally: |
| 516 | self._wlock.release() |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 517 | |
| 518 | def __contains__(self, key): |
| 519 | """Check if a key exists in this cache.""" |
| 520 | return key in self._mapping |
| 521 | |
| 522 | def __len__(self): |
| 523 | """Return the current size of the cache.""" |
| 524 | return len(self._mapping) |
| 525 | |
| 526 | def __repr__(self): |
| 527 | return '<%s %r>' % ( |
| 528 | self.__class__.__name__, |
| 529 | self._mapping |
| 530 | ) |
| 531 | |
| 532 | def __getitem__(self, key): |
| 533 | """Get an item from the cache. Moves the item up so that it has the |
| 534 | highest priority then. |
| 535 | |
| 536 | Raise an `KeyError` if it does not exist. |
| 537 | """ |
| 538 | rv = self._mapping[key] |
| 539 | if self._queue[-1] != key: |
| 540 | self._remove(key) |
| 541 | self._append(key) |
| 542 | return rv |
| 543 | |
| 544 | def __setitem__(self, key, value): |
| 545 | """Sets the value for an item. Moves the item up so that it |
| 546 | has the highest priority then. |
| 547 | """ |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 548 | self._wlock.acquire() |
| 549 | try: |
| 550 | if key in self._mapping: |
| 551 | self._remove(key) |
| 552 | elif len(self._mapping) == self.capacity: |
| 553 | del self._mapping[self._popleft()] |
| 554 | self._append(key) |
| 555 | self._mapping[key] = value |
| 556 | finally: |
| 557 | self._wlock.release() |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 558 | |
| 559 | def __delitem__(self, key): |
| 560 | """Remove an item from the cache dict. |
| 561 | Raise an `KeyError` if it does not exist. |
| 562 | """ |
Armin Ronacher | 000b491 | 2008-05-01 18:40:15 +0200 | [diff] [blame] | 563 | self._wlock.acquire() |
| 564 | try: |
| 565 | del self._mapping[key] |
| 566 | self._remove(key) |
| 567 | finally: |
| 568 | self._wlock.release() |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 569 | |
Armin Ronacher | e25f24d | 2008-05-19 11:20:41 +0200 | [diff] [blame] | 570 | def items(self): |
| 571 | """Return a list of items.""" |
| 572 | result = [(key, self._mapping[key]) for key in list(self._queue)] |
| 573 | result.reverse() |
| 574 | return result |
| 575 | |
| 576 | def iteritems(self): |
| 577 | """Iterate over all items.""" |
| 578 | return iter(self.items()) |
| 579 | |
| 580 | def values(self): |
| 581 | """Return a list of all values.""" |
| 582 | return [x[1] for x in self.items()] |
| 583 | |
| 584 | def itervalue(self): |
| 585 | """Iterate over all values.""" |
| 586 | return iter(self.values()) |
| 587 | |
| 588 | def keys(self): |
| 589 | """Return a list of all keys ordered by most recent usage.""" |
| 590 | return list(self) |
| 591 | |
| 592 | def iterkeys(self): |
| 593 | """Iterate over all keys in the cache dict, ordered by |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 594 | the most recent usage. |
| 595 | """ |
Armin Ronacher | e224488 | 2008-05-19 09:25:57 +0200 | [diff] [blame] | 596 | return reversed(tuple(self._queue)) |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 597 | |
Armin Ronacher | e25f24d | 2008-05-19 11:20:41 +0200 | [diff] [blame] | 598 | __iter__ = iterkeys |
| 599 | |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 600 | def __reversed__(self): |
| 601 | """Iterate over the values in the cache dict, oldest items |
| 602 | coming first. |
| 603 | """ |
Armin Ronacher | e224488 | 2008-05-19 09:25:57 +0200 | [diff] [blame] | 604 | return iter(tuple(self._queue)) |
Armin Ronacher | 814f6c2 | 2008-04-17 15:52:23 +0200 | [diff] [blame] | 605 | |
| 606 | __copy__ = copy |
| 607 | |
Armin Ronacher | bd33f11 | 2008-04-18 09:17:32 +0200 | [diff] [blame] | 608 | |
Armin Ronacher | 9bb7e47 | 2008-05-28 11:26:59 +0200 | [diff] [blame] | 609 | # register the LRU cache as mutable mapping if possible |
| 610 | try: |
| 611 | from collections import MutableMapping |
| 612 | MutableMapping.register(LRUCache) |
| 613 | except ImportError: |
| 614 | pass |
| 615 | |
| 616 | |
Armin Ronacher | bd33f11 | 2008-04-18 09:17:32 +0200 | [diff] [blame] | 617 | # we have to import it down here as the speedups module imports the |
| 618 | # markup type which is define above. |
| 619 | try: |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 620 | from jinja2._speedups import escape, soft_unicode |
Armin Ronacher | bd33f11 | 2008-04-18 09:17:32 +0200 | [diff] [blame] | 621 | except ImportError: |
Lukas Meuser | ad48a2e | 2008-05-01 18:19:57 +0200 | [diff] [blame] | 622 | def escape(s): |
Armin Ronacher | f35e281 | 2008-05-06 16:04:10 +0200 | [diff] [blame] | 623 | """Convert the characters &, <, >, ' and " in string s to HTML-safe |
| 624 | sequences. Use this if you need to display text that might contain |
Armin Ronacher | 9a1e33c | 2008-05-05 22:00:46 +0200 | [diff] [blame] | 625 | such characters in HTML. Marks return value as markup string. |
Armin Ronacher | bd33f11 | 2008-04-18 09:17:32 +0200 | [diff] [blame] | 626 | """ |
Lukas Meuser | ad48a2e | 2008-05-01 18:19:57 +0200 | [diff] [blame] | 627 | if hasattr(s, '__html__'): |
| 628 | return s.__html__() |
| 629 | return Markup(unicode(s) |
Armin Ronacher | bd33f11 | 2008-04-18 09:17:32 +0200 | [diff] [blame] | 630 | .replace('&', '&') |
| 631 | .replace('>', '>') |
| 632 | .replace('<', '<') |
Armin Ronacher | f35e281 | 2008-05-06 16:04:10 +0200 | [diff] [blame] | 633 | .replace("'", ''') |
Armin Ronacher | 9d42abf | 2008-05-14 18:10:41 +0200 | [diff] [blame] | 634 | .replace('"', '"') |
Armin Ronacher | bd33f11 | 2008-04-18 09:17:32 +0200 | [diff] [blame] | 635 | ) |
Armin Ronacher | f59bac2 | 2008-04-20 13:11:43 +0200 | [diff] [blame] | 636 | |
| 637 | def soft_unicode(s): |
| 638 | """Make a string unicode if it isn't already. That way a markup |
| 639 | string is not converted back to unicode. |
| 640 | """ |
| 641 | if not isinstance(s, unicode): |
| 642 | s = unicode(s) |
| 643 | return s |
Armin Ronacher | 4f7d2d5 | 2008-04-22 10:40:26 +0200 | [diff] [blame] | 644 | |
| 645 | |
| 646 | # partials |
| 647 | try: |
| 648 | from functools import partial |
| 649 | except ImportError: |
| 650 | class partial(object): |
| 651 | def __init__(self, _func, *args, **kwargs): |
Benjamin Wiegand | 228c183 | 2008-04-28 18:09:27 +0200 | [diff] [blame] | 652 | self._func = _func |
Armin Ronacher | 4f7d2d5 | 2008-04-22 10:40:26 +0200 | [diff] [blame] | 653 | self._args = args |
| 654 | self._kwargs = kwargs |
| 655 | def __call__(self, *args, **kwargs): |
| 656 | kwargs.update(self._kwargs) |
| 657 | return self._func(*(self._args + args), **kwargs) |