michael@0: r"""
michael@0: A simple, fast, extensible JSON encoder and decoder
michael@0: 
michael@0: JSON (JavaScript Object Notation) <http://json.org> is a subset of
michael@0: JavaScript syntax (ECMA-262 3rd edition) used as a lightweight data
michael@0: interchange format.
michael@0: 
michael@0: simplejson exposes an API familiar to uses of the standard library
michael@0: marshal and pickle modules.
michael@0: 
michael@0: Encoding basic Python object hierarchies::
michael@0:     
michael@0:     >>> import simplejson
michael@0:     >>> simplejson.dumps(['foo', {'bar': ('baz', None, 1.0, 2)}])
michael@0:     '["foo", {"bar": ["baz", null, 1.0, 2]}]'
michael@0:     >>> print simplejson.dumps("\"foo\bar")
michael@0:     "\"foo\bar"
michael@0:     >>> print simplejson.dumps(u'\u1234')
michael@0:     "\u1234"
michael@0:     >>> print simplejson.dumps('\\')
michael@0:     "\\"
michael@0:     >>> print simplejson.dumps({"c": 0, "b": 0, "a": 0}, sort_keys=True)
michael@0:     {"a": 0, "b": 0, "c": 0}
michael@0:     >>> from StringIO import StringIO
michael@0:     >>> io = StringIO()
michael@0:     >>> simplejson.dump(['streaming API'], io)
michael@0:     >>> io.getvalue()
michael@0:     '["streaming API"]'
michael@0: 
michael@0: Compact encoding::
michael@0: 
michael@0:     >>> import simplejson
michael@0:     >>> simplejson.dumps([1,2,3,{'4': 5, '6': 7}], separators=(',',':'))
michael@0:     '[1,2,3,{"4":5,"6":7}]'
michael@0: 
michael@0: Pretty printing::
michael@0: 
michael@0:     >>> import simplejson
michael@0:     >>> print simplejson.dumps({'4': 5, '6': 7}, sort_keys=True, indent=4)
michael@0:     {
michael@0:         "4": 5, 
michael@0:         "6": 7
michael@0:     }
michael@0: 
michael@0: Decoding JSON::
michael@0:     
michael@0:     >>> import simplejson
michael@0:     >>> simplejson.loads('["foo", {"bar":["baz", null, 1.0, 2]}]')
michael@0:     [u'foo', {u'bar': [u'baz', None, 1.0, 2]}]
michael@0:     >>> simplejson.loads('"\\"foo\\bar"')
michael@0:     u'"foo\x08ar'
michael@0:     >>> from StringIO import StringIO
michael@0:     >>> io = StringIO('["streaming API"]')
michael@0:     >>> simplejson.load(io)
michael@0:     [u'streaming API']
michael@0: 
michael@0: Specializing JSON object decoding::
michael@0: 
michael@0:     >>> import simplejson
michael@0:     >>> def as_complex(dct):
michael@0:     ...     if '__complex__' in dct:
michael@0:     ...         return complex(dct['real'], dct['imag'])
michael@0:     ...     return dct
michael@0:     ... 
michael@0:     >>> simplejson.loads('{"__complex__": true, "real": 1, "imag": 2}',
michael@0:     ...     object_hook=as_complex)
michael@0:     (1+2j)
michael@0:     >>> import decimal
michael@0:     >>> simplejson.loads('1.1', parse_float=decimal.Decimal)
michael@0:     Decimal("1.1")
michael@0: 
michael@0: Extending JSONEncoder::
michael@0:     
michael@0:     >>> import simplejson
michael@0:     >>> class ComplexEncoder(simplejson.JSONEncoder):
michael@0:     ...     def default(self, obj):
michael@0:     ...         if isinstance(obj, complex):
michael@0:     ...             return [obj.real, obj.imag]
michael@0:     ...         return simplejson.JSONEncoder.default(self, obj)
michael@0:     ... 
michael@0:     >>> dumps(2 + 1j, cls=ComplexEncoder)
michael@0:     '[2.0, 1.0]'
michael@0:     >>> ComplexEncoder().encode(2 + 1j)
michael@0:     '[2.0, 1.0]'
michael@0:     >>> list(ComplexEncoder().iterencode(2 + 1j))
michael@0:     ['[', '2.0', ', ', '1.0', ']']
michael@0:     
michael@0: 
michael@0: Using simplejson from the shell to validate and
michael@0: pretty-print::
michael@0:     
michael@0:     $ echo '{"json":"obj"}' | python -msimplejson.tool
michael@0:     {
michael@0:         "json": "obj"
michael@0:     }
michael@0:     $ echo '{ 1.2:3.4}' | python -msimplejson.tool
michael@0:     Expecting property name: line 1 column 2 (char 2)
michael@0: 
michael@0: Note that the JSON produced by this module's default settings
michael@0: is a subset of YAML, so it may be used as a serializer for that as well.
michael@0: """
michael@0: __version__ = '1.9.2'
michael@0: __all__ = [
michael@0:     'dump', 'dumps', 'load', 'loads',
michael@0:     'JSONDecoder', 'JSONEncoder',
michael@0: ]
michael@0: 
michael@0: if __name__ == '__main__':
michael@0:     import warnings
michael@0:     warnings.warn('python -msimplejson is deprecated, use python -msiplejson.tool', DeprecationWarning)
michael@0:     from simplejson.decoder import JSONDecoder
michael@0:     from simplejson.encoder import JSONEncoder
michael@0: else:
michael@0:     from decoder import JSONDecoder
michael@0:     from encoder import JSONEncoder
michael@0: 
michael@0: _default_encoder = JSONEncoder(
michael@0:     skipkeys=False,
michael@0:     ensure_ascii=True,
michael@0:     check_circular=True,
michael@0:     allow_nan=True,
michael@0:     indent=None,
michael@0:     separators=None,
michael@0:     encoding='utf-8',
michael@0:     default=None,
michael@0: )
michael@0: 
michael@0: def dump(obj, fp, skipkeys=False, ensure_ascii=True, check_circular=True,
michael@0:         allow_nan=True, cls=None, indent=None, separators=None,
michael@0:         encoding='utf-8', default=None, **kw):
michael@0:     """
michael@0:     Serialize ``obj`` as a JSON formatted stream to ``fp`` (a
michael@0:     ``.write()``-supporting file-like object).
michael@0: 
michael@0:     If ``skipkeys`` is ``True`` then ``dict`` keys that are not basic types
michael@0:     (``str``, ``unicode``, ``int``, ``long``, ``float``, ``bool``, ``None``) 
michael@0:     will be skipped instead of raising a ``TypeError``.
michael@0: 
michael@0:     If ``ensure_ascii`` is ``False``, then the some chunks written to ``fp``
michael@0:     may be ``unicode`` instances, subject to normal Python ``str`` to
michael@0:     ``unicode`` coercion rules. Unless ``fp.write()`` explicitly
michael@0:     understands ``unicode`` (as in ``codecs.getwriter()``) this is likely
michael@0:     to cause an error.
michael@0: 
michael@0:     If ``check_circular`` is ``False``, then the circular reference check
michael@0:     for container types will be skipped and a circular reference will
michael@0:     result in an ``OverflowError`` (or worse).
michael@0: 
michael@0:     If ``allow_nan`` is ``False``, then it will be a ``ValueError`` to
michael@0:     serialize out of range ``float`` values (``nan``, ``inf``, ``-inf``)
michael@0:     in strict compliance of the JSON specification, instead of using the
michael@0:     JavaScript equivalents (``NaN``, ``Infinity``, ``-Infinity``).
michael@0: 
michael@0:     If ``indent`` is a non-negative integer, then JSON array elements and object
michael@0:     members will be pretty-printed with that indent level. An indent level
michael@0:     of 0 will only insert newlines. ``None`` is the most compact representation.
michael@0: 
michael@0:     If ``separators`` is an ``(item_separator, dict_separator)`` tuple
michael@0:     then it will be used instead of the default ``(', ', ': ')`` separators.
michael@0:     ``(',', ':')`` is the most compact JSON representation.
michael@0: 
michael@0:     ``encoding`` is the character encoding for str instances, default is UTF-8.
michael@0: 
michael@0:     ``default(obj)`` is a function that should return a serializable version
michael@0:     of obj or raise TypeError. The default simply raises TypeError.
michael@0: 
michael@0:     To use a custom ``JSONEncoder`` subclass (e.g. one that overrides the
michael@0:     ``.default()`` method to serialize additional types), specify it with
michael@0:     the ``cls`` kwarg.
michael@0:     """
michael@0:     # cached encoder
michael@0:     if (skipkeys is False and ensure_ascii is True and
michael@0:         check_circular is True and allow_nan is True and
michael@0:         cls is None and indent is None and separators is None and
michael@0:         encoding == 'utf-8' and default is None and not kw):
michael@0:         iterable = _default_encoder.iterencode(obj)
michael@0:     else:
michael@0:         if cls is None:
michael@0:             cls = JSONEncoder
michael@0:         iterable = cls(skipkeys=skipkeys, ensure_ascii=ensure_ascii,
michael@0:             check_circular=check_circular, allow_nan=allow_nan, indent=indent,
michael@0:             separators=separators, encoding=encoding,
michael@0:             default=default, **kw).iterencode(obj)
michael@0:     # could accelerate with writelines in some versions of Python, at
michael@0:     # a debuggability cost
michael@0:     for chunk in iterable:
michael@0:         fp.write(chunk)
michael@0: 
michael@0: 
michael@0: def dumps(obj, skipkeys=False, ensure_ascii=True, check_circular=True,
michael@0:         allow_nan=True, cls=None, indent=None, separators=None,
michael@0:         encoding='utf-8', default=None, **kw):
michael@0:     """
michael@0:     Serialize ``obj`` to a JSON formatted ``str``.
michael@0: 
michael@0:     If ``skipkeys`` is ``True`` then ``dict`` keys that are not basic types
michael@0:     (``str``, ``unicode``, ``int``, ``long``, ``float``, ``bool``, ``None``) 
michael@0:     will be skipped instead of raising a ``TypeError``.
michael@0: 
michael@0:     If ``ensure_ascii`` is ``False``, then the return value will be a
michael@0:     ``unicode`` instance subject to normal Python ``str`` to ``unicode``
michael@0:     coercion rules instead of being escaped to an ASCII ``str``.
michael@0: 
michael@0:     If ``check_circular`` is ``False``, then the circular reference check
michael@0:     for container types will be skipped and a circular reference will
michael@0:     result in an ``OverflowError`` (or worse).
michael@0: 
michael@0:     If ``allow_nan`` is ``False``, then it will be a ``ValueError`` to
michael@0:     serialize out of range ``float`` values (``nan``, ``inf``, ``-inf``) in
michael@0:     strict compliance of the JSON specification, instead of using the
michael@0:     JavaScript equivalents (``NaN``, ``Infinity``, ``-Infinity``).
michael@0: 
michael@0:     If ``indent`` is a non-negative integer, then JSON array elements and
michael@0:     object members will be pretty-printed with that indent level. An indent
michael@0:     level of 0 will only insert newlines. ``None`` is the most compact
michael@0:     representation.
michael@0: 
michael@0:     If ``separators`` is an ``(item_separator, dict_separator)`` tuple
michael@0:     then it will be used instead of the default ``(', ', ': ')`` separators.
michael@0:     ``(',', ':')`` is the most compact JSON representation.
michael@0: 
michael@0:     ``encoding`` is the character encoding for str instances, default is UTF-8.
michael@0: 
michael@0:     ``default(obj)`` is a function that should return a serializable version
michael@0:     of obj or raise TypeError. The default simply raises TypeError.
michael@0: 
michael@0:     To use a custom ``JSONEncoder`` subclass (e.g. one that overrides the
michael@0:     ``.default()`` method to serialize additional types), specify it with
michael@0:     the ``cls`` kwarg.
michael@0:     """
michael@0:     # cached encoder
michael@0:     if (skipkeys is False and ensure_ascii is True and
michael@0:         check_circular is True and allow_nan is True and
michael@0:         cls is None and indent is None and separators is None and
michael@0:         encoding == 'utf-8' and default is None and not kw):
michael@0:         return _default_encoder.encode(obj)
michael@0:     if cls is None:
michael@0:         cls = JSONEncoder
michael@0:     return cls(
michael@0:         skipkeys=skipkeys, ensure_ascii=ensure_ascii,
michael@0:         check_circular=check_circular, allow_nan=allow_nan, indent=indent,
michael@0:         separators=separators, encoding=encoding, default=default,
michael@0:         **kw).encode(obj)
michael@0: 
michael@0: 
michael@0: _default_decoder = JSONDecoder(encoding=None, object_hook=None)
michael@0: 
michael@0: 
michael@0: def load(fp, encoding=None, cls=None, object_hook=None, parse_float=None,
michael@0:         parse_int=None, parse_constant=None, **kw):
michael@0:     """
michael@0:     Deserialize ``fp`` (a ``.read()``-supporting file-like object containing
michael@0:     a JSON document) to a Python object.
michael@0: 
michael@0:     If the contents of ``fp`` is encoded with an ASCII based encoding other
michael@0:     than utf-8 (e.g. latin-1), then an appropriate ``encoding`` name must
michael@0:     be specified. Encodings that are not ASCII based (such as UCS-2) are
michael@0:     not allowed, and should be wrapped with
michael@0:     ``codecs.getreader(fp)(encoding)``, or simply decoded to a ``unicode``
michael@0:     object and passed to ``loads()``
michael@0: 
michael@0:     ``object_hook`` is an optional function that will be called with the
michael@0:     result of any object literal decode (a ``dict``). The return value of
michael@0:     ``object_hook`` will be used instead of the ``dict``. This feature
michael@0:     can be used to implement custom decoders (e.g. JSON-RPC class hinting).
michael@0:     
michael@0:     To use a custom ``JSONDecoder`` subclass, specify it with the ``cls``
michael@0:     kwarg.
michael@0:     """
michael@0:     return loads(fp.read(),
michael@0:         encoding=encoding, cls=cls, object_hook=object_hook,
michael@0:         parse_float=parse_float, parse_int=parse_int,
michael@0:         parse_constant=parse_constant, **kw)
michael@0: 
michael@0: 
michael@0: def loads(s, encoding=None, cls=None, object_hook=None, parse_float=None,
michael@0:         parse_int=None, parse_constant=None, **kw):
michael@0:     """
michael@0:     Deserialize ``s`` (a ``str`` or ``unicode`` instance containing a JSON
michael@0:     document) to a Python object.
michael@0: 
michael@0:     If ``s`` is a ``str`` instance and is encoded with an ASCII based encoding
michael@0:     other than utf-8 (e.g. latin-1) then an appropriate ``encoding`` name
michael@0:     must be specified. Encodings that are not ASCII based (such as UCS-2)
michael@0:     are not allowed and should be decoded to ``unicode`` first.
michael@0: 
michael@0:     ``object_hook`` is an optional function that will be called with the
michael@0:     result of any object literal decode (a ``dict``). The return value of
michael@0:     ``object_hook`` will be used instead of the ``dict``. This feature
michael@0:     can be used to implement custom decoders (e.g. JSON-RPC class hinting).
michael@0: 
michael@0:     ``parse_float``, if specified, will be called with the string
michael@0:     of every JSON float to be decoded. By default this is equivalent to
michael@0:     float(num_str). This can be used to use another datatype or parser
michael@0:     for JSON floats (e.g. decimal.Decimal).
michael@0: 
michael@0:     ``parse_int``, if specified, will be called with the string
michael@0:     of every JSON int to be decoded. By default this is equivalent to
michael@0:     int(num_str). This can be used to use another datatype or parser
michael@0:     for JSON integers (e.g. float).
michael@0: 
michael@0:     ``parse_constant``, if specified, will be called with one of the
michael@0:     following strings: -Infinity, Infinity, NaN, null, true, false.
michael@0:     This can be used to raise an exception if invalid JSON numbers
michael@0:     are encountered.
michael@0: 
michael@0:     To use a custom ``JSONDecoder`` subclass, specify it with the ``cls``
michael@0:     kwarg.
michael@0:     """
michael@0:     if (cls is None and encoding is None and object_hook is None and
michael@0:             parse_int is None and parse_float is None and
michael@0:             parse_constant is None and not kw):
michael@0:         return _default_decoder.decode(s)
michael@0:     if cls is None:
michael@0:         cls = JSONDecoder
michael@0:     if object_hook is not None:
michael@0:         kw['object_hook'] = object_hook
michael@0:     if parse_float is not None:
michael@0:         kw['parse_float'] = parse_float
michael@0:     if parse_int is not None:
michael@0:         kw['parse_int'] = parse_int
michael@0:     if parse_constant is not None:
michael@0:         kw['parse_constant'] = parse_constant
michael@0:     return cls(encoding=encoding, **kw).decode(s)
michael@0: 
michael@0: 
michael@0: #
michael@0: # Compatibility cruft from other libraries
michael@0: #
michael@0: 
michael@0: 
michael@0: def decode(s):
michael@0:     """
michael@0:     demjson, python-cjson API compatibility hook. Use loads(s) instead.
michael@0:     """
michael@0:     import warnings
michael@0:     warnings.warn("simplejson.loads(s) should be used instead of decode(s)",
michael@0:         DeprecationWarning)
michael@0:     return loads(s)
michael@0: 
michael@0: 
michael@0: def encode(obj):
michael@0:     """
michael@0:     demjson, python-cjson compatibility hook. Use dumps(s) instead.
michael@0:     """
michael@0:     import warnings
michael@0:     warnings.warn("simplejson.dumps(s) should be used instead of encode(s)",
michael@0:         DeprecationWarning)
michael@0:     return dumps(obj)
michael@0: 
michael@0: 
michael@0: def read(s):
michael@0:     """
michael@0:     jsonlib, JsonUtils, python-json, json-py API compatibility hook.
michael@0:     Use loads(s) instead.
michael@0:     """
michael@0:     import warnings
michael@0:     warnings.warn("simplejson.loads(s) should be used instead of read(s)",
michael@0:         DeprecationWarning)
michael@0:     return loads(s)
michael@0: 
michael@0: 
michael@0: def write(obj):
michael@0:     """
michael@0:     jsonlib, JsonUtils, python-json, json-py API compatibility hook.
michael@0:     Use dumps(s) instead.
michael@0:     """
michael@0:     import warnings
michael@0:     warnings.warn("simplejson.dumps(s) should be used instead of write(s)",
michael@0:         DeprecationWarning)
michael@0:     return dumps(obj)
michael@0: 
michael@0: 
michael@0: if __name__ == '__main__':
michael@0:     import simplejson.tool
michael@0:     simplejson.tool.main()