]>
Commit | Line | Data |
---|---|---|
4d256d41 CP |
1 | r""" |
2 | A simple, fast, extensible JSON encoder and decoder | |
3 | ||
4 | JSON (JavaScript Object Notation) <http://json.org> is a subset of | |
5 | JavaScript syntax (ECMA-262 3rd edition) used as a lightweight data | |
6 | interchange format. | |
7 | ||
8 | simplejson exposes an API familiar to uses of the standard library | |
9 | marshal and pickle modules. | |
10 | ||
11 | Encoding basic Python object hierarchies:: | |
12 | ||
13 | >>> import simplejson | |
14 | >>> simplejson.dumps(['foo', {'bar': ('baz', None, 1.0, 2)}]) | |
15 | '["foo", {"bar": ["baz", null, 1.0, 2]}]' | |
16 | >>> print simplejson.dumps("\"foo\bar") | |
17 | "\"foo\bar" | |
18 | >>> print simplejson.dumps(u'\u1234') | |
19 | "\u1234" | |
20 | >>> print simplejson.dumps('\\') | |
21 | "\\" | |
22 | >>> print simplejson.dumps({"c": 0, "b": 0, "a": 0}, sort_keys=True) | |
23 | {"a": 0, "b": 0, "c": 0} | |
24 | >>> from StringIO import StringIO | |
25 | >>> io = StringIO() | |
26 | >>> simplejson.dump(['streaming API'], io) | |
27 | >>> io.getvalue() | |
28 | '["streaming API"]' | |
29 | ||
30 | Compact encoding:: | |
31 | ||
32 | >>> import simplejson | |
33 | >>> simplejson.dumps([1,2,3,{'4': 5, '6': 7}], separators=(',',':')) | |
34 | '[1,2,3,{"4":5,"6":7}]' | |
35 | ||
36 | Pretty printing:: | |
37 | ||
38 | >>> import simplejson | |
39 | >>> print simplejson.dumps({'4': 5, '6': 7}, sort_keys=True, indent=4) | |
40 | { | |
41 | "4": 5, | |
42 | "6": 7 | |
43 | } | |
44 | ||
45 | Decoding JSON:: | |
46 | ||
47 | >>> import simplejson | |
48 | >>> simplejson.loads('["foo", {"bar":["baz", null, 1.0, 2]}]') | |
49 | [u'foo', {u'bar': [u'baz', None, 1.0, 2]}] | |
50 | >>> simplejson.loads('"\\"foo\\bar"') | |
51 | u'"foo\x08ar' | |
52 | >>> from StringIO import StringIO | |
53 | >>> io = StringIO('["streaming API"]') | |
54 | >>> simplejson.load(io) | |
55 | [u'streaming API'] | |
56 | ||
57 | Specializing JSON object decoding:: | |
58 | ||
59 | >>> import simplejson | |
60 | >>> def as_complex(dct): | |
61 | ... if '__complex__' in dct: | |
62 | ... return complex(dct['real'], dct['imag']) | |
63 | ... return dct | |
64 | ... | |
65 | >>> simplejson.loads('{"__complex__": true, "real": 1, "imag": 2}', | |
66 | ... object_hook=as_complex) | |
67 | (1+2j) | |
68 | >>> import decimal | |
69 | >>> simplejson.loads('1.1', parse_float=decimal.Decimal) | |
70 | Decimal("1.1") | |
71 | ||
72 | Extending JSONEncoder:: | |
73 | ||
74 | >>> import simplejson | |
75 | >>> class ComplexEncoder(simplejson.JSONEncoder): | |
76 | ... def default(self, obj): | |
77 | ... if isinstance(obj, complex): | |
78 | ... return [obj.real, obj.imag] | |
79 | ... return simplejson.JSONEncoder.default(self, obj) | |
80 | ... | |
81 | >>> dumps(2 + 1j, cls=ComplexEncoder) | |
82 | '[2.0, 1.0]' | |
83 | >>> ComplexEncoder().encode(2 + 1j) | |
84 | '[2.0, 1.0]' | |
85 | >>> list(ComplexEncoder().iterencode(2 + 1j)) | |
86 | ['[', '2.0', ', ', '1.0', ']'] | |
87 | ||
88 | ||
89 | Using simplejson from the shell to validate and | |
90 | pretty-print:: | |
91 | ||
92 | $ echo '{"json":"obj"}' | python -msimplejson | |
93 | { | |
94 | "json": "obj" | |
95 | } | |
96 | $ echo '{ 1.2:3.4}' | python -msimplejson | |
97 | Expecting property name: line 1 column 2 (char 2) | |
98 | ||
99 | Note that the JSON produced by this module's default settings | |
100 | is a subset of YAML, so it may be used as a serializer for that as well. | |
101 | """ | |
102 | __version__ = '1.9.1' | |
103 | __all__ = [ | |
104 | 'dump', 'dumps', 'load', 'loads', | |
105 | 'JSONDecoder', 'JSONEncoder', | |
106 | ] | |
107 | ||
108 | if __name__ == '__main__': | |
109 | from simplejson.decoder import JSONDecoder | |
110 | from simplejson.encoder import JSONEncoder | |
111 | else: | |
112 | from decoder import JSONDecoder | |
113 | from encoder import JSONEncoder | |
114 | ||
115 | _default_encoder = JSONEncoder( | |
116 | skipkeys=False, | |
117 | ensure_ascii=True, | |
118 | check_circular=True, | |
119 | allow_nan=True, | |
120 | indent=None, | |
121 | separators=None, | |
122 | encoding='utf-8', | |
123 | default=None, | |
124 | ) | |
125 | ||
126 | def dump(obj, fp, skipkeys=False, ensure_ascii=True, check_circular=True, | |
127 | allow_nan=True, cls=None, indent=None, separators=None, | |
128 | encoding='utf-8', default=None, **kw): | |
129 | """ | |
130 | Serialize ``obj`` as a JSON formatted stream to ``fp`` (a | |
131 | ``.write()``-supporting file-like object). | |
132 | ||
133 | If ``skipkeys`` is ``True`` then ``dict`` keys that are not basic types | |
134 | (``str``, ``unicode``, ``int``, ``long``, ``float``, ``bool``, ``None``) | |
135 | will be skipped instead of raising a ``TypeError``. | |
136 | ||
137 | If ``ensure_ascii`` is ``False``, then the some chunks written to ``fp`` | |
138 | may be ``unicode`` instances, subject to normal Python ``str`` to | |
139 | ``unicode`` coercion rules. Unless ``fp.write()`` explicitly | |
140 | understands ``unicode`` (as in ``codecs.getwriter()``) this is likely | |
141 | to cause an error. | |
142 | ||
143 | If ``check_circular`` is ``False``, then the circular reference check | |
144 | for container types will be skipped and a circular reference will | |
145 | result in an ``OverflowError`` (or worse). | |
146 | ||
147 | If ``allow_nan`` is ``False``, then it will be a ``ValueError`` to | |
148 | serialize out of range ``float`` values (``nan``, ``inf``, ``-inf``) | |
149 | in strict compliance of the JSON specification, instead of using the | |
150 | JavaScript equivalents (``NaN``, ``Infinity``, ``-Infinity``). | |
151 | ||
152 | If ``indent`` is a non-negative integer, then JSON array elements and object | |
153 | members will be pretty-printed with that indent level. An indent level | |
154 | of 0 will only insert newlines. ``None`` is the most compact representation. | |
155 | ||
156 | If ``separators`` is an ``(item_separator, dict_separator)`` tuple | |
157 | then it will be used instead of the default ``(', ', ': ')`` separators. | |
158 | ``(',', ':')`` is the most compact JSON representation. | |
159 | ||
160 | ``encoding`` is the character encoding for str instances, default is UTF-8. | |
161 | ||
162 | ``default(obj)`` is a function that should return a serializable version | |
163 | of obj or raise TypeError. The default simply raises TypeError. | |
164 | ||
165 | To use a custom ``JSONEncoder`` subclass (e.g. one that overrides the | |
166 | ``.default()`` method to serialize additional types), specify it with | |
167 | the ``cls`` kwarg. | |
168 | """ | |
169 | # cached encoder | |
170 | if (skipkeys is False and ensure_ascii is True and | |
171 | check_circular is True and allow_nan is True and | |
172 | cls is None and indent is None and separators is None and | |
173 | encoding == 'utf-8' and default is None and not kw): | |
174 | iterable = _default_encoder.iterencode(obj) | |
175 | else: | |
176 | if cls is None: | |
177 | cls = JSONEncoder | |
178 | iterable = cls(skipkeys=skipkeys, ensure_ascii=ensure_ascii, | |
179 | check_circular=check_circular, allow_nan=allow_nan, indent=indent, | |
180 | separators=separators, encoding=encoding, | |
181 | default=default, **kw).iterencode(obj) | |
182 | # could accelerate with writelines in some versions of Python, at | |
183 | # a debuggability cost | |
184 | for chunk in iterable: | |
185 | fp.write(chunk) | |
186 | ||
187 | ||
188 | def dumps(obj, skipkeys=False, ensure_ascii=True, check_circular=True, | |
189 | allow_nan=True, cls=None, indent=None, separators=None, | |
190 | encoding='utf-8', default=None, **kw): | |
191 | """ | |
192 | Serialize ``obj`` to a JSON formatted ``str``. | |
193 | ||
194 | If ``skipkeys`` is ``True`` then ``dict`` keys that are not basic types | |
195 | (``str``, ``unicode``, ``int``, ``long``, ``float``, ``bool``, ``None``) | |
196 | will be skipped instead of raising a ``TypeError``. | |
197 | ||
198 | If ``ensure_ascii`` is ``False``, then the return value will be a | |
199 | ``unicode`` instance subject to normal Python ``str`` to ``unicode`` | |
200 | coercion rules instead of being escaped to an ASCII ``str``. | |
201 | ||
202 | If ``check_circular`` is ``False``, then the circular reference check | |
203 | for container types will be skipped and a circular reference will | |
204 | result in an ``OverflowError`` (or worse). | |
205 | ||
206 | If ``allow_nan`` is ``False``, then it will be a ``ValueError`` to | |
207 | serialize out of range ``float`` values (``nan``, ``inf``, ``-inf``) in | |
208 | strict compliance of the JSON specification, instead of using the | |
209 | JavaScript equivalents (``NaN``, ``Infinity``, ``-Infinity``). | |
210 | ||
211 | If ``indent`` is a non-negative integer, then JSON array elements and | |
212 | object members will be pretty-printed with that indent level. An indent | |
213 | level of 0 will only insert newlines. ``None`` is the most compact | |
214 | representation. | |
215 | ||
216 | If ``separators`` is an ``(item_separator, dict_separator)`` tuple | |
217 | then it will be used instead of the default ``(', ', ': ')`` separators. | |
218 | ``(',', ':')`` is the most compact JSON representation. | |
219 | ||
220 | ``encoding`` is the character encoding for str instances, default is UTF-8. | |
221 | ||
222 | ``default(obj)`` is a function that should return a serializable version | |
223 | of obj or raise TypeError. The default simply raises TypeError. | |
224 | ||
225 | To use a custom ``JSONEncoder`` subclass (e.g. one that overrides the | |
226 | ``.default()`` method to serialize additional types), specify it with | |
227 | the ``cls`` kwarg. | |
228 | """ | |
229 | # cached encoder | |
230 | if (skipkeys is False and ensure_ascii is True and | |
231 | check_circular is True and allow_nan is True and | |
232 | cls is None and indent is None and separators is None and | |
233 | encoding == 'utf-8' and default is None and not kw): | |
234 | return _default_encoder.encode(obj) | |
235 | if cls is None: | |
236 | cls = JSONEncoder | |
237 | return cls( | |
238 | skipkeys=skipkeys, ensure_ascii=ensure_ascii, | |
239 | check_circular=check_circular, allow_nan=allow_nan, indent=indent, | |
240 | separators=separators, encoding=encoding, default=default, | |
241 | **kw).encode(obj) | |
242 | ||
243 | ||
244 | _default_decoder = JSONDecoder(encoding=None, object_hook=None) | |
245 | ||
246 | ||
247 | def load(fp, encoding=None, cls=None, object_hook=None, parse_float=None, | |
248 | parse_int=None, parse_constant=None, **kw): | |
249 | """ | |
250 | Deserialize ``fp`` (a ``.read()``-supporting file-like object containing | |
251 | a JSON document) to a Python object. | |
252 | ||
253 | If the contents of ``fp`` is encoded with an ASCII based encoding other | |
254 | than utf-8 (e.g. latin-1), then an appropriate ``encoding`` name must | |
255 | be specified. Encodings that are not ASCII based (such as UCS-2) are | |
256 | not allowed, and should be wrapped with | |
257 | ``codecs.getreader(fp)(encoding)``, or simply decoded to a ``unicode`` | |
258 | object and passed to ``loads()`` | |
259 | ||
260 | ``object_hook`` is an optional function that will be called with the | |
261 | result of any object literal decode (a ``dict``). The return value of | |
262 | ``object_hook`` will be used instead of the ``dict``. This feature | |
263 | can be used to implement custom decoders (e.g. JSON-RPC class hinting). | |
264 | ||
265 | To use a custom ``JSONDecoder`` subclass, specify it with the ``cls`` | |
266 | kwarg. | |
267 | """ | |
268 | return loads(fp.read(), | |
269 | encoding=encoding, cls=cls, object_hook=object_hook, | |
270 | parse_float=parse_float, parse_int=parse_int, | |
271 | parse_constant=parse_constant, **kw) | |
272 | ||
273 | ||
274 | def loads(s, encoding=None, cls=None, object_hook=None, parse_float=None, | |
275 | parse_int=None, parse_constant=None, **kw): | |
276 | """ | |
277 | Deserialize ``s`` (a ``str`` or ``unicode`` instance containing a JSON | |
278 | document) to a Python object. | |
279 | ||
280 | If ``s`` is a ``str`` instance and is encoded with an ASCII based encoding | |
281 | other than utf-8 (e.g. latin-1) then an appropriate ``encoding`` name | |
282 | must be specified. Encodings that are not ASCII based (such as UCS-2) | |
283 | are not allowed and should be decoded to ``unicode`` first. | |
284 | ||
285 | ``object_hook`` is an optional function that will be called with the | |
286 | result of any object literal decode (a ``dict``). The return value of | |
287 | ``object_hook`` will be used instead of the ``dict``. This feature | |
288 | can be used to implement custom decoders (e.g. JSON-RPC class hinting). | |
289 | ||
290 | ``parse_float``, if specified, will be called with the string | |
291 | of every JSON float to be decoded. By default this is equivalent to | |
292 | float(num_str). This can be used to use another datatype or parser | |
293 | for JSON floats (e.g. decimal.Decimal). | |
294 | ||
295 | ``parse_int``, if specified, will be called with the string | |
296 | of every JSON int to be decoded. By default this is equivalent to | |
297 | int(num_str). This can be used to use another datatype or parser | |
298 | for JSON integers (e.g. float). | |
299 | ||
300 | ``parse_constant``, if specified, will be called with one of the | |
301 | following strings: -Infinity, Infinity, NaN, null, true, false. | |
302 | This can be used to raise an exception if invalid JSON numbers | |
303 | are encountered. | |
304 | ||
305 | To use a custom ``JSONDecoder`` subclass, specify it with the ``cls`` | |
306 | kwarg. | |
307 | """ | |
308 | if (cls is None and encoding is None and object_hook is None and | |
309 | parse_int is None and parse_float is None and | |
310 | parse_constant is None and not kw): | |
311 | return _default_decoder.decode(s) | |
312 | if cls is None: | |
313 | cls = JSONDecoder | |
314 | if object_hook is not None: | |
315 | kw['object_hook'] = object_hook | |
316 | if parse_float is not None: | |
317 | kw['parse_float'] = parse_float | |
318 | if parse_int is not None: | |
319 | kw['parse_int'] = parse_int | |
320 | if parse_constant is not None: | |
321 | kw['parse_constant'] = parse_constant | |
322 | return cls(encoding=encoding, **kw).decode(s) | |
323 | ||
324 | ||
325 | # | |
326 | # Compatibility cruft from other libraries | |
327 | # | |
328 | ||
329 | ||
330 | def decode(s): | |
331 | """ | |
332 | demjson, python-cjson API compatibility hook. Use loads(s) instead. | |
333 | """ | |
334 | import warnings | |
335 | warnings.warn("simplejson.loads(s) should be used instead of decode(s)", | |
336 | DeprecationWarning) | |
337 | return loads(s) | |
338 | ||
339 | ||
340 | def encode(obj): | |
341 | """ | |
342 | demjson, python-cjson compatibility hook. Use dumps(s) instead. | |
343 | """ | |
344 | import warnings | |
345 | warnings.warn("simplejson.dumps(s) should be used instead of encode(s)", | |
346 | DeprecationWarning) | |
347 | return dumps(obj) | |
348 | ||
349 | ||
350 | def read(s): | |
351 | """ | |
352 | jsonlib, JsonUtils, python-json, json-py API compatibility hook. | |
353 | Use loads(s) instead. | |
354 | """ | |
355 | import warnings | |
356 | warnings.warn("simplejson.loads(s) should be used instead of read(s)", | |
357 | DeprecationWarning) | |
358 | return loads(s) | |
359 | ||
360 | ||
361 | def write(obj): | |
362 | """ | |
363 | jsonlib, JsonUtils, python-json, json-py API compatibility hook. | |
364 | Use dumps(s) instead. | |
365 | """ | |
366 | import warnings | |
367 | warnings.warn("simplejson.dumps(s) should be used instead of write(s)", | |
368 | DeprecationWarning) | |
369 | return dumps(obj) | |
370 | ||
371 | ||
372 | # | |
373 | # Pretty printer: | |
374 | # curl http://mochikit.com/examples/ajax_tables/domains.json | python -msimplejson | |
375 | # | |
376 | ||
377 | ||
378 | def main(): | |
379 | import sys | |
380 | if len(sys.argv) == 1: | |
381 | infile = sys.stdin | |
382 | outfile = sys.stdout | |
383 | elif len(sys.argv) == 2: | |
384 | infile = open(sys.argv[1], 'rb') | |
385 | outfile = sys.stdout | |
386 | elif len(sys.argv) == 3: | |
387 | infile = open(sys.argv[1], 'rb') | |
388 | outfile = open(sys.argv[2], 'wb') | |
389 | else: | |
390 | raise SystemExit("%s [infile [outfile]]" % (sys.argv[0],)) | |
391 | try: | |
392 | obj = load(infile) | |
393 | except ValueError, e: | |
394 | raise SystemExit(e) | |
395 | dump(obj, outfile, sort_keys=True, indent=4) | |
396 | outfile.write('\n') | |
397 | ||
398 | ||
399 | if __name__ == '__main__': | |
400 | main() |