third_party/simplejson/__init__.py

   1 r"""JSON (JavaScript Object Notation) <http://json.org> is a subset of
   2 JavaScript syntax (ECMA-262 3rd edition) used as a lightweight data
   3 interchange format.
   4
   5 :mod:`simplejson` exposes an API familiar to users of the standard library
   6 :mod:`marshal` and :mod:`pickle` modules. It is the externally maintained
   7 version of the :mod:`json` library contained in Python 2.6, but maintains
   8 compatibility with Python 2.4 and Python 2.5 and (currently) has
   9 significant performance advantages, even without using the optional C
  10 extension for speedups.
  11
  12 Encoding basic Python object hierarchies::
  13
  14     >>> import simplejson as json
  15     >>> json.dumps(['foo', {'bar': ('baz', None, 1.0, 2)}])
  16     '["foo", {"bar": ["baz", null, 1.0, 2]}]'
  17     >>> print json.dumps("\"foo\bar")
  18     "\"foo\bar"
  19     >>> print json.dumps(u'\u1234')
  20     "\u1234"
  21     >>> print json.dumps('\\')
  22     "\\"
  23     >>> print json.dumps({"c": 0, "b": 0, "a": 0}, sort_keys=True)
  24     {"a": 0, "b": 0, "c": 0}
  25     >>> from StringIO import StringIO
  26     >>> io = StringIO()
  27     >>> json.dump(['streaming API'], io)
  28     >>> io.getvalue()
  29     '["streaming API"]'
  30
  31 Compact encoding::
  32
  33     >>> import simplejson as json
  34     >>> json.dumps([1,2,3,{'4': 5, '6': 7}], separators=(',',':'))
  35     '[1,2,3,{"4":5,"6":7}]'
  36
  37 Pretty printing::
  38
  39     >>> import simplejson as json
  40     >>> s = json.dumps({'4': 5, '6': 7}, sort_keys=True, indent='    ')
  41     >>> print '\n'.join([l.rstrip() for l in  s.splitlines()])
  42     {
  43         "4": 5,
  44         "6": 7
  45     }
  46
  47 Decoding JSON::
  48
  49     >>> import simplejson as json
  50     >>> obj = [u'foo', {u'bar': [u'baz', None, 1.0, 2]}]
  51     >>> json.loads('["foo", {"bar":["baz", null, 1.0, 2]}]') == obj
  52     True
  53     >>> json.loads('"\\"foo\\bar"') == u'"foo\x08ar'
  54     True
  55     >>> from StringIO import StringIO
  56     >>> io = StringIO('["streaming API"]')
  57     >>> json.load(io)[0] == 'streaming API'
  58     True
  59
  60 Specializing JSON object decoding::
  61
  62     >>> import simplejson as json
  63     >>> def as_complex(dct):
  64     ...     if '__complex__' in dct:
  65     ...         return complex(dct['real'], dct['imag'])
  66     ...     return dct
  67     ...
  68     >>> json.loads('{"__complex__": true, "real": 1, "imag": 2}',
  69     ...     object_hook=as_complex)
  70     (1+2j)
  71     >>> from decimal import Decimal
  72     >>> json.loads('1.1', parse_float=Decimal) == Decimal('1.1')
  73     True
  74
  75 Specializing JSON object encoding::
  76
  77     >>> import simplejson as json
  78     >>> def encode_complex(obj):
  79     ...     if isinstance(obj, complex):
  80     ...         return [obj.real, obj.imag]
  81     ...     raise TypeError(repr(o) + " is not JSON serializable")
  82     ...
  83     >>> json.dumps(2 + 1j, default=encode_complex)
  84     '[2.0, 1.0]'
  85     >>> json.JSONEncoder(default=encode_complex).encode(2 + 1j)
  86     '[2.0, 1.0]'
  87     >>> ''.join(json.JSONEncoder(default=encode_complex).iterencode(2 + 1j))
  88     '[2.0, 1.0]'
  89
  90
  91 Using simplejson.tool from the shell to validate and pretty-print::
  92
  93     $ echo '{"json":"obj"}' | python -m simplejson.tool
  94     {
  95         "json": "obj"
  96     }
  97     $ echo '{ 1.2:3.4}' | python -m simplejson.tool
  98     Expecting property name: line 1 column 2 (char 2)
  99 """
 100 __version__ = '2.6.2'
 101 __all__ = [
 102     'dump', 'dumps', 'load', 'loads',
 103     'JSONDecoder', 'JSONDecodeError', 'JSONEncoder',
 104     'OrderedDict', 'simple_first',
 105 ]
 106
 107 __author__ = 'Bob Ippolito <bob@redivi.com>'
 108
 109 from decimal import Decimal
 110
 111 from decoder import JSONDecoder, JSONDecodeError
 112 from encoder import JSONEncoder, JSONEncoderForHTML
 113 def _import_OrderedDict():
 114     import collections
 115     try:
 116         return collections.OrderedDict
 117     except AttributeError:
 118         import ordered_dict
 119         return ordered_dict.OrderedDict
 120 OrderedDict = _import_OrderedDict()
 121
 122 def _import_c_make_encoder():
 123     try:
 124         from simplejson._speedups import make_encoder
 125         return make_encoder
 126     except ImportError:
 127         return None
 128
 129 _default_encoder = JSONEncoder(
 130     skipkeys=False,
 131     ensure_ascii=True,
 132     check_circular=True,
 133     allow_nan=True,
 134     indent=None,
 135     separators=None,
 136     encoding='utf-8',
 137     default=None,
 138     use_decimal=True,
 139     namedtuple_as_object=True,
 140     tuple_as_array=True,
 141     bigint_as_string=False,
 142     item_sort_key=None,
 143 )
 144
 145 def dump(obj, fp, skipkeys=False, ensure_ascii=True, check_circular=True,
 146         allow_nan=True, cls=None, indent=None, separators=None,
 147         encoding='utf-8', default=None, use_decimal=True,
 148         namedtuple_as_object=True, tuple_as_array=True,
 149         bigint_as_string=False, sort_keys=False, item_sort_key=None,
 150         **kw):
 151     """Serialize ``obj`` as a JSON formatted stream to ``fp`` (a
 152     ``.write()``-supporting file-like object).
 153
 154     If ``skipkeys`` is true then ``dict`` keys that are not basic types
 155     (``str``, ``unicode``, ``int``, ``long``, ``float``, ``bool``, ``None``)
 156     will be skipped instead of raising a ``TypeError``.
 157
 158     If ``ensure_ascii`` is false, then the some chunks written to ``fp``
 159     may be ``unicode`` instances, subject to normal Python ``str`` to
 160     ``unicode`` coercion rules. Unless ``fp.write()`` explicitly
 161     understands ``unicode`` (as in ``codecs.getwriter()``) this is likely
 162     to cause an error.
 163
 164     If ``check_circular`` is false, then the circular reference check
 165     for container types will be skipped and a circular reference will
 166     result in an ``OverflowError`` (or worse).
 167
 168     If ``allow_nan`` is false, then it will be a ``ValueError`` to
 169     serialize out of range ``float`` values (``nan``, ``inf``, ``-inf``)
 170     in strict compliance of the JSON specification, instead of using the
 171     JavaScript equivalents (``NaN``, ``Infinity``, ``-Infinity``).
 172
 173     If *indent* is a string, then JSON array elements and object members
 174     will be pretty-printed with a newline followed by that string repeated
 175     for each level of nesting. ``None`` (the default) selects the most compact
 176     representation without any newlines. For backwards compatibility with
 177     versions of simplejson earlier than 2.1.0, an integer is also accepted
 178     and is converted to a string with that many spaces.
 179
 180     If ``separators`` is an ``(item_separator, dict_separator)`` tuple
 181     then it will be used instead of the default ``(', ', ': ')`` separators.
 182     ``(',', ':')`` is the most compact JSON representation.
 183
 184     ``encoding`` is the character encoding for str instances, default is UTF-8.
 185
 186     ``default(obj)`` is a function that should return a serializable version
 187     of obj or raise TypeError. The default simply raises TypeError.
 188
 189     If *use_decimal* is true (default: ``True``) then decimal.Decimal
 190     will be natively serialized to JSON with full precision.
 191
 192     If *namedtuple_as_object* is true (default: ``True``),
 193     :class:`tuple` subclasses with ``_asdict()`` methods will be encoded
 194     as JSON objects.
 195
 196     If *tuple_as_array* is true (default: ``True``),
 197     :class:`tuple` (and subclasses) will be encoded as JSON arrays.
 198
 199     If *bigint_as_string* is true (default: ``False``), ints 2**53 and higher
 200     or lower than -2**53 will be encoded as strings. This is to avoid the
 201     rounding that happens in Javascript otherwise. Note that this is still a
 202     lossy operation that will not round-trip correctly and should be used
 203     sparingly.
 204
 205     If specified, *item_sort_key* is a callable used to sort the items in
 206     each dictionary. This is useful if you want to sort items other than
 207     in alphabetical order by key. This option takes precedence over
 208     *sort_keys*.
 209
 210     If *sort_keys* is true (default: ``False``), the output of dictionaries
 211     will be sorted by item.
 212
 213     To use a custom ``JSONEncoder`` subclass (e.g. one that overrides the
 214     ``.default()`` method to serialize additional types), specify it with
 215     the ``cls`` kwarg.
 216
 217     """
 218     # cached encoder
 219     if (not skipkeys and ensure_ascii and
 220         check_circular and allow_nan and
 221         cls is None and indent is None and separators is None and
 222         encoding == 'utf-8' and default is None and use_decimal
 223         and namedtuple_as_object and tuple_as_array
 224         and not bigint_as_string and not item_sort_key and not kw):
 225         iterable = _default_encoder.iterencode(obj)
 226     else:
 227         if cls is None:
 228             cls = JSONEncoder
 229         iterable = cls(skipkeys=skipkeys, ensure_ascii=ensure_ascii,
 230             check_circular=check_circular, allow_nan=allow_nan, indent=indent,
 231             separators=separators, encoding=encoding,
 232             default=default, use_decimal=use_decimal,
 233             namedtuple_as_object=namedtuple_as_object,
 234             tuple_as_array=tuple_as_array,
 235             bigint_as_string=bigint_as_string,
 236             sort_keys=sort_keys,
 237             item_sort_key=item_sort_key,
 238             **kw).iterencode(obj)
 239     # could accelerate with writelines in some versions of Python, at
 240     # a debuggability cost
 241     for chunk in iterable:
 242         fp.write(chunk)
 243
 244
 245 def dumps(obj, skipkeys=False, ensure_ascii=True, check_circular=True,
 246         allow_nan=True, cls=None, indent=None, separators=None,
 247         encoding='utf-8', default=None, use_decimal=True,
 248         namedtuple_as_object=True, tuple_as_array=True,
 249         bigint_as_string=False, sort_keys=False, item_sort_key=None,
 250         **kw):
 251     """Serialize ``obj`` to a JSON formatted ``str``.
 252
 253     If ``skipkeys`` is false then ``dict`` keys that are not basic types
 254     (``str``, ``unicode``, ``int``, ``long``, ``float``, ``bool``, ``None``)
 255     will be skipped instead of raising a ``TypeError``.
 256
 257     If ``ensure_ascii`` is false, then the return value will be a
 258     ``unicode`` instance subject to normal Python ``str`` to ``unicode``
 259     coercion rules instead of being escaped to an ASCII ``str``.
 260
 261     If ``check_circular`` is false, then the circular reference check
 262     for container types will be skipped and a circular reference will
 263     result in an ``OverflowError`` (or worse).
 264
 265     If ``allow_nan`` is false, then it will be a ``ValueError`` to
 266     serialize out of range ``float`` values (``nan``, ``inf``, ``-inf``) in
 267     strict compliance of the JSON specification, instead of using the
 268     JavaScript equivalents (``NaN``, ``Infinity``, ``-Infinity``).
 269
 270     If ``indent`` is a string, then JSON array elements and object members
 271     will be pretty-printed with a newline followed by that string repeated
 272     for each level of nesting. ``None`` (the default) selects the most compact
 273     representation without any newlines. For backwards compatibility with
 274     versions of simplejson earlier than 2.1.0, an integer is also accepted
 275     and is converted to a string with that many spaces.
 276
 277     If ``separators`` is an ``(item_separator, dict_separator)`` tuple
 278     then it will be used instead of the default ``(', ', ': ')`` separators.
 279     ``(',', ':')`` is the most compact JSON representation.
 280
 281     ``encoding`` is the character encoding for str instances, default is UTF-8.
 282
 283     ``default(obj)`` is a function that should return a serializable version
 284     of obj or raise TypeError. The default simply raises TypeError.
 285
 286     If *use_decimal* is true (default: ``True``) then decimal.Decimal
 287     will be natively serialized to JSON with full precision.
 288
 289     If *namedtuple_as_object* is true (default: ``True``),
 290     :class:`tuple` subclasses with ``_asdict()`` methods will be encoded
 291     as JSON objects.
 292
 293     If *tuple_as_array* is true (default: ``True``),
 294     :class:`tuple` (and subclasses) will be encoded as JSON arrays.
 295
 296     If *bigint_as_string* is true (not the default), ints 2**53 and higher
 297     or lower than -2**53 will be encoded as strings. This is to avoid the
 298     rounding that happens in Javascript otherwise.
 299
 300     If specified, *item_sort_key* is a callable used to sort the items in
 301     each dictionary. This is useful if you want to sort items other than
 302     in alphabetical order by key. This option takes precendence over
 303     *sort_keys*.
 304
 305     If *sort_keys* is true (default: ``False``), the output of dictionaries
 306     will be sorted by item.
 307
 308     To use a custom ``JSONEncoder`` subclass (e.g. one that overrides the
 309     ``.default()`` method to serialize additional types), specify it with
 310     the ``cls`` kwarg.
 311
 312     """
 313     # cached encoder
 314     if (not skipkeys and ensure_ascii and
 315         check_circular and allow_nan and
 316         cls is None and indent is None and separators is None and
 317         encoding == 'utf-8' and default is None and use_decimal
 318         and namedtuple_as_object and tuple_as_array
 319         and not bigint_as_string and not sort_keys
 320         and not item_sort_key and not kw):
 321         return _default_encoder.encode(obj)
 322     if cls is None:
 323         cls = JSONEncoder
 324     return cls(
 325         skipkeys=skipkeys, ensure_ascii=ensure_ascii,
 326         check_circular=check_circular, allow_nan=allow_nan, indent=indent,
 327         separators=separators, encoding=encoding, default=default,
 328         use_decimal=use_decimal,
 329         namedtuple_as_object=namedtuple_as_object,
 330         tuple_as_array=tuple_as_array,
 331         bigint_as_string=bigint_as_string,
 332         sort_keys=sort_keys,
 333         item_sort_key=item_sort_key,
 334         **kw).encode(obj)
 335
 336
 337 _default_decoder = JSONDecoder(encoding=None, object_hook=None,
 338                                object_pairs_hook=None)
 339
 340
 341 def load(fp, encoding=None, cls=None, object_hook=None, parse_float=None,
 342         parse_int=None, parse_constant=None, object_pairs_hook=None,
 343         use_decimal=False, namedtuple_as_object=True, tuple_as_array=True,
 344         **kw):
 345     """Deserialize ``fp`` (a ``.read()``-supporting file-like object containing
 346     a JSON document) to a Python object.
 347
 348     *encoding* determines the encoding used to interpret any
 349     :class:`str` objects decoded by this instance (``'utf-8'`` by
 350     default).  It has no effect when decoding :class:`unicode` objects.
 351
 352     Note that currently only encodings that are a superset of ASCII work,
 353     strings of other encodings should be passed in as :class:`unicode`.
 354
 355     *object_hook*, if specified, will be called with the result of every
 356     JSON object decoded and its return value will be used in place of the
 357     given :class:`dict`.  This can be used to provide custom
 358     deserializations (e.g. to support JSON-RPC class hinting).
 359
 360     *object_pairs_hook* is an optional function that will be called with
 361     the result of any object literal decode with an ordered list of pairs.
 362     The return value of *object_pairs_hook* will be used instead of the
 363     :class:`dict`.  This feature can be used to implement custom decoders
 364     that rely on the order that the key and value pairs are decoded (for
 365     example, :func:`collections.OrderedDict` will remember the order of
 366     insertion). If *object_hook* is also defined, the *object_pairs_hook*
 367     takes priority.
 368
 369     *parse_float*, if specified, will be called with the string of every
 370     JSON float to be decoded.  By default, this is equivalent to
 371     ``float(num_str)``. This can be used to use another datatype or parser
 372     for JSON floats (e.g. :class:`decimal.Decimal`).
 373
 374     *parse_int*, if specified, will be called with the string of every
 375     JSON int to be decoded.  By default, this is equivalent to
 376     ``int(num_str)``.  This can be used to use another datatype or parser
 377     for JSON integers (e.g. :class:`float`).
 378
 379     *parse_constant*, if specified, will be called with one of the
 380     following strings: ``'-Infinity'``, ``'Infinity'``, ``'NaN'``.  This
 381     can be used to raise an exception if invalid JSON numbers are
 382     encountered.
 383
 384     If *use_decimal* is true (default: ``False``) then it implies
 385     parse_float=decimal.Decimal for parity with ``dump``.
 386
 387     To use a custom ``JSONDecoder`` subclass, specify it with the ``cls``
 388     kwarg.
 389
 390     """
 391     return loads(fp.read(),
 392         encoding=encoding, cls=cls, object_hook=object_hook,
 393         parse_float=parse_float, parse_int=parse_int,
 394         parse_constant=parse_constant, object_pairs_hook=object_pairs_hook,
 395         use_decimal=use_decimal, **kw)
 396
 397
 398 def loads(s, encoding=None, cls=None, object_hook=None, parse_float=None,
 399         parse_int=None, parse_constant=None, object_pairs_hook=None,
 400         use_decimal=False, **kw):
 401     """Deserialize ``s`` (a ``str`` or ``unicode`` instance containing a JSON
 402     document) to a Python object.
 403
 404     *encoding* determines the encoding used to interpret any
 405     :class:`str` objects decoded by this instance (``'utf-8'`` by
 406     default).  It has no effect when decoding :class:`unicode` objects.
 407
 408     Note that currently only encodings that are a superset of ASCII work,
 409     strings of other encodings should be passed in as :class:`unicode`.
 410
 411     *object_hook*, if specified, will be called with the result of every
 412     JSON object decoded and its return value will be used in place of the
 413     given :class:`dict`.  This can be used to provide custom
 414     deserializations (e.g. to support JSON-RPC class hinting).
 415
 416     *object_pairs_hook* is an optional function that will be called with
 417     the result of any object literal decode with an ordered list of pairs.
 418     The return value of *object_pairs_hook* will be used instead of the
 419     :class:`dict`.  This feature can be used to implement custom decoders
 420     that rely on the order that the key and value pairs are decoded (for
 421     example, :func:`collections.OrderedDict` will remember the order of
 422     insertion). If *object_hook* is also defined, the *object_pairs_hook*
 423     takes priority.
 424
 425     *parse_float*, if specified, will be called with the string of every
 426     JSON float to be decoded.  By default, this is equivalent to
 427     ``float(num_str)``. This can be used to use another datatype or parser
 428     for JSON floats (e.g. :class:`decimal.Decimal`).
 429
 430     *parse_int*, if specified, will be called with the string of every
 431     JSON int to be decoded.  By default, this is equivalent to
 432     ``int(num_str)``.  This can be used to use another datatype or parser
 433     for JSON integers (e.g. :class:`float`).
 434
 435     *parse_constant*, if specified, will be called with one of the
 436     following strings: ``'-Infinity'``, ``'Infinity'``, ``'NaN'``.  This
 437     can be used to raise an exception if invalid JSON numbers are
 438     encountered.
 439
 440     If *use_decimal* is true (default: ``False``) then it implies
 441     parse_float=decimal.Decimal for parity with ``dump``.
 442
 443     To use a custom ``JSONDecoder`` subclass, specify it with the ``cls``
 444     kwarg.
 445
 446     """
 447     if (cls is None and encoding is None and object_hook is None and
 448             parse_int is None and parse_float is None and
 449             parse_constant is None and object_pairs_hook is None
 450             and not use_decimal and not kw):
 451         return _default_decoder.decode(s)
 452     if cls is None:
 453         cls = JSONDecoder
 454     if object_hook is not None:
 455         kw['object_hook'] = object_hook
 456     if object_pairs_hook is not None:
 457         kw['object_pairs_hook'] = object_pairs_hook
 458     if parse_float is not None:
 459         kw['parse_float'] = parse_float
 460     if parse_int is not None:
 461         kw['parse_int'] = parse_int
 462     if parse_constant is not None:
 463         kw['parse_constant'] = parse_constant
 464     if use_decimal:
 465         if parse_float is not None:
 466             raise TypeError("use_decimal=True implies parse_float=Decimal")
 467         kw['parse_float'] = Decimal
 468     return cls(encoding=encoding, **kw).decode(s)
 469
 470
 471 def _toggle_speedups(enabled):
 472     import simplejson.decoder as dec
 473     import simplejson.encoder as enc
 474     import simplejson.scanner as scan
 475     c_make_encoder = _import_c_make_encoder()
 476     if enabled:
 477         dec.scanstring = dec.c_scanstring or dec.py_scanstring
 478         enc.c_make_encoder = c_make_encoder
 479         enc.encode_basestring_ascii = (enc.c_encode_basestring_ascii or
 480             enc.py_encode_basestring_ascii)
 481         scan.make_scanner = scan.c_make_scanner or scan.py_make_scanner
 482     else:
 483         dec.scanstring = dec.py_scanstring
 484         enc.c_make_encoder = None
 485         enc.encode_basestring_ascii = enc.py_encode_basestring_ascii
 486         scan.make_scanner = scan.py_make_scanner
 487     dec.make_scanner = scan.make_scanner
 488     global _default_decoder
 489     _default_decoder = JSONDecoder(
 490         encoding=None,
 491         object_hook=None,
 492         object_pairs_hook=None,
 493     )
 494     global _default_encoder
 495     _default_encoder = JSONEncoder(
 496        skipkeys=False,
 497        ensure_ascii=True,
 498        check_circular=True,
 499        allow_nan=True,
 500        indent=None,
 501        separators=None,
 502        encoding='utf-8',
 503        default=None,
 504    )
 505
 506 def simple_first(kv):
 507     """Helper function to pass to item_sort_key to sort simple
 508     elements to the top, then container elements.
 509     """
 510     return (isinstance(kv[1], (list, dict, tuple)), kv[0])