OLD | NEW |
(Empty) | |
| 1 """Implementation of JSONDecoder |
| 2 """ |
| 3 import re |
| 4 import sys |
| 5 import struct |
| 6 |
| 7 from simplejson.scanner import make_scanner |
| 8 def _import_c_scanstring(): |
| 9 try: |
| 10 from simplejson._speedups import scanstring |
| 11 return scanstring |
| 12 except ImportError: |
| 13 return None |
| 14 c_scanstring = _import_c_scanstring() |
| 15 |
| 16 __all__ = ['JSONDecoder'] |
| 17 |
| 18 FLAGS = re.VERBOSE | re.MULTILINE | re.DOTALL |
| 19 |
| 20 def _floatconstants(): |
| 21 _BYTES = '7FF80000000000007FF0000000000000'.decode('hex') |
| 22 # The struct module in Python 2.4 would get frexp() out of range here |
| 23 # when an endian is specified in the format string. Fixed in Python 2.5+ |
| 24 if sys.byteorder != 'big': |
| 25 _BYTES = _BYTES[:8][::-1] + _BYTES[8:][::-1] |
| 26 nan, inf = struct.unpack('dd', _BYTES) |
| 27 return nan, inf, -inf |
| 28 |
| 29 NaN, PosInf, NegInf = _floatconstants() |
| 30 |
| 31 |
| 32 class JSONDecodeError(ValueError): |
| 33 """Subclass of ValueError with the following additional properties: |
| 34 |
| 35 msg: The unformatted error message |
| 36 doc: The JSON document being parsed |
| 37 pos: The start index of doc where parsing failed |
| 38 end: The end index of doc where parsing failed (may be None) |
| 39 lineno: The line corresponding to pos |
| 40 colno: The column corresponding to pos |
| 41 endlineno: The line corresponding to end (may be None) |
| 42 endcolno: The column corresponding to end (may be None) |
| 43 |
| 44 """ |
| 45 def __init__(self, msg, doc, pos, end=None): |
| 46 ValueError.__init__(self, errmsg(msg, doc, pos, end=end)) |
| 47 self.msg = msg |
| 48 self.doc = doc |
| 49 self.pos = pos |
| 50 self.end = end |
| 51 self.lineno, self.colno = linecol(doc, pos) |
| 52 if end is not None: |
| 53 self.endlineno, self.endcolno = linecol(doc, end) |
| 54 else: |
| 55 self.endlineno, self.endcolno = None, None |
| 56 |
| 57 |
| 58 def linecol(doc, pos): |
| 59 lineno = doc.count('\n', 0, pos) + 1 |
| 60 if lineno == 1: |
| 61 colno = pos |
| 62 else: |
| 63 colno = pos - doc.rindex('\n', 0, pos) |
| 64 return lineno, colno |
| 65 |
| 66 |
| 67 def errmsg(msg, doc, pos, end=None): |
| 68 # Note that this function is called from _speedups |
| 69 lineno, colno = linecol(doc, pos) |
| 70 if end is None: |
| 71 #fmt = '{0}: line {1} column {2} (char {3})' |
| 72 #return fmt.format(msg, lineno, colno, pos) |
| 73 fmt = '%s: line %d column %d (char %d)' |
| 74 return fmt % (msg, lineno, colno, pos) |
| 75 endlineno, endcolno = linecol(doc, end) |
| 76 #fmt = '{0}: line {1} column {2} - line {3} column {4} (char {5} - {6})' |
| 77 #return fmt.format(msg, lineno, colno, endlineno, endcolno, pos, end) |
| 78 fmt = '%s: line %d column %d - line %d column %d (char %d - %d)' |
| 79 return fmt % (msg, lineno, colno, endlineno, endcolno, pos, end) |
| 80 |
| 81 |
| 82 _CONSTANTS = { |
| 83 '-Infinity': NegInf, |
| 84 'Infinity': PosInf, |
| 85 'NaN': NaN, |
| 86 } |
| 87 |
| 88 STRINGCHUNK = re.compile(r'(.*?)(["\\\x00-\x1f])', FLAGS) |
| 89 BACKSLASH = { |
| 90 '"': u'"', '\\': u'\\', '/': u'/', |
| 91 'b': u'\b', 'f': u'\f', 'n': u'\n', 'r': u'\r', 't': u'\t', |
| 92 } |
| 93 |
| 94 DEFAULT_ENCODING = "utf-8" |
| 95 |
| 96 def py_scanstring(s, end, encoding=None, strict=True, |
| 97 _b=BACKSLASH, _m=STRINGCHUNK.match): |
| 98 """Scan the string s for a JSON string. End is the index of the |
| 99 character in s after the quote that started the JSON string. |
| 100 Unescapes all valid JSON string escape sequences and raises ValueError |
| 101 on attempt to decode an invalid string. If strict is False then literal |
| 102 control characters are allowed in the string. |
| 103 |
| 104 Returns a tuple of the decoded string and the index of the character in s |
| 105 after the end quote.""" |
| 106 if encoding is None: |
| 107 encoding = DEFAULT_ENCODING |
| 108 chunks = [] |
| 109 _append = chunks.append |
| 110 begin = end - 1 |
| 111 while 1: |
| 112 chunk = _m(s, end) |
| 113 if chunk is None: |
| 114 raise JSONDecodeError( |
| 115 "Unterminated string starting at", s, begin) |
| 116 end = chunk.end() |
| 117 content, terminator = chunk.groups() |
| 118 # Content is contains zero or more unescaped string characters |
| 119 if content: |
| 120 if not isinstance(content, unicode): |
| 121 content = unicode(content, encoding) |
| 122 _append(content) |
| 123 # Terminator is the end of string, a literal control character, |
| 124 # or a backslash denoting that an escape sequence follows |
| 125 if terminator == '"': |
| 126 break |
| 127 elif terminator != '\\': |
| 128 if strict: |
| 129 msg = "Invalid control character %r at" % (terminator,) |
| 130 #msg = "Invalid control character {0!r} at".format(terminator) |
| 131 raise JSONDecodeError(msg, s, end) |
| 132 else: |
| 133 _append(terminator) |
| 134 continue |
| 135 try: |
| 136 esc = s[end] |
| 137 except IndexError: |
| 138 raise JSONDecodeError( |
| 139 "Unterminated string starting at", s, begin) |
| 140 # If not a unicode escape sequence, must be in the lookup table |
| 141 if esc != 'u': |
| 142 try: |
| 143 char = _b[esc] |
| 144 except KeyError: |
| 145 msg = "Invalid \\escape: " + repr(esc) |
| 146 raise JSONDecodeError(msg, s, end) |
| 147 end += 1 |
| 148 else: |
| 149 # Unicode escape sequence |
| 150 esc = s[end + 1:end + 5] |
| 151 next_end = end + 5 |
| 152 if len(esc) != 4: |
| 153 msg = "Invalid \\uXXXX escape" |
| 154 raise JSONDecodeError(msg, s, end) |
| 155 uni = int(esc, 16) |
| 156 # Check for surrogate pair on UCS-4 systems |
| 157 if 0xd800 <= uni <= 0xdbff and sys.maxunicode > 65535: |
| 158 msg = "Invalid \\uXXXX\\uXXXX surrogate pair" |
| 159 if not s[end + 5:end + 7] == '\\u': |
| 160 raise JSONDecodeError(msg, s, end) |
| 161 esc2 = s[end + 7:end + 11] |
| 162 if len(esc2) != 4: |
| 163 raise JSONDecodeError(msg, s, end) |
| 164 uni2 = int(esc2, 16) |
| 165 uni = 0x10000 + (((uni - 0xd800) << 10) | (uni2 - 0xdc00)) |
| 166 next_end += 6 |
| 167 char = unichr(uni) |
| 168 end = next_end |
| 169 # Append the unescaped character |
| 170 _append(char) |
| 171 return u''.join(chunks), end |
| 172 |
| 173 |
| 174 # Use speedup if available |
| 175 scanstring = c_scanstring or py_scanstring |
| 176 |
| 177 WHITESPACE = re.compile(r'[ \t\n\r]*', FLAGS) |
| 178 WHITESPACE_STR = ' \t\n\r' |
| 179 |
| 180 def JSONObject((s, end), encoding, strict, scan_once, object_hook, |
| 181 object_pairs_hook, memo=None, |
| 182 _w=WHITESPACE.match, _ws=WHITESPACE_STR): |
| 183 # Backwards compatibility |
| 184 if memo is None: |
| 185 memo = {} |
| 186 memo_get = memo.setdefault |
| 187 pairs = [] |
| 188 # Use a slice to prevent IndexError from being raised, the following |
| 189 # check will raise a more specific ValueError if the string is empty |
| 190 nextchar = s[end:end + 1] |
| 191 # Normally we expect nextchar == '"' |
| 192 if nextchar != '"': |
| 193 if nextchar in _ws: |
| 194 end = _w(s, end).end() |
| 195 nextchar = s[end:end + 1] |
| 196 # Trivial empty object |
| 197 if nextchar == '}': |
| 198 if object_pairs_hook is not None: |
| 199 result = object_pairs_hook(pairs) |
| 200 return result, end + 1 |
| 201 pairs = {} |
| 202 if object_hook is not None: |
| 203 pairs = object_hook(pairs) |
| 204 return pairs, end + 1 |
| 205 elif nextchar != '"': |
| 206 raise JSONDecodeError("Expecting property name", s, end) |
| 207 end += 1 |
| 208 while True: |
| 209 key, end = scanstring(s, end, encoding, strict) |
| 210 key = memo_get(key, key) |
| 211 |
| 212 # To skip some function call overhead we optimize the fast paths where |
| 213 # the JSON key separator is ": " or just ":". |
| 214 if s[end:end + 1] != ':': |
| 215 end = _w(s, end).end() |
| 216 if s[end:end + 1] != ':': |
| 217 raise JSONDecodeError("Expecting : delimiter", s, end) |
| 218 |
| 219 end += 1 |
| 220 |
| 221 try: |
| 222 if s[end] in _ws: |
| 223 end += 1 |
| 224 if s[end] in _ws: |
| 225 end = _w(s, end + 1).end() |
| 226 except IndexError: |
| 227 pass |
| 228 |
| 229 try: |
| 230 value, end = scan_once(s, end) |
| 231 except StopIteration: |
| 232 raise JSONDecodeError("Expecting object", s, end) |
| 233 pairs.append((key, value)) |
| 234 |
| 235 try: |
| 236 nextchar = s[end] |
| 237 if nextchar in _ws: |
| 238 end = _w(s, end + 1).end() |
| 239 nextchar = s[end] |
| 240 except IndexError: |
| 241 nextchar = '' |
| 242 end += 1 |
| 243 |
| 244 if nextchar == '}': |
| 245 break |
| 246 elif nextchar != ',': |
| 247 raise JSONDecodeError("Expecting , delimiter", s, end - 1) |
| 248 |
| 249 try: |
| 250 nextchar = s[end] |
| 251 if nextchar in _ws: |
| 252 end += 1 |
| 253 nextchar = s[end] |
| 254 if nextchar in _ws: |
| 255 end = _w(s, end + 1).end() |
| 256 nextchar = s[end] |
| 257 except IndexError: |
| 258 nextchar = '' |
| 259 |
| 260 end += 1 |
| 261 if nextchar != '"': |
| 262 raise JSONDecodeError("Expecting property name", s, end - 1) |
| 263 |
| 264 if object_pairs_hook is not None: |
| 265 result = object_pairs_hook(pairs) |
| 266 return result, end |
| 267 pairs = dict(pairs) |
| 268 if object_hook is not None: |
| 269 pairs = object_hook(pairs) |
| 270 return pairs, end |
| 271 |
| 272 def JSONArray((s, end), scan_once, _w=WHITESPACE.match, _ws=WHITESPACE_STR): |
| 273 values = [] |
| 274 nextchar = s[end:end + 1] |
| 275 if nextchar in _ws: |
| 276 end = _w(s, end + 1).end() |
| 277 nextchar = s[end:end + 1] |
| 278 # Look-ahead for trivial empty array |
| 279 if nextchar == ']': |
| 280 return values, end + 1 |
| 281 _append = values.append |
| 282 while True: |
| 283 try: |
| 284 value, end = scan_once(s, end) |
| 285 except StopIteration: |
| 286 raise JSONDecodeError("Expecting object", s, end) |
| 287 _append(value) |
| 288 nextchar = s[end:end + 1] |
| 289 if nextchar in _ws: |
| 290 end = _w(s, end + 1).end() |
| 291 nextchar = s[end:end + 1] |
| 292 end += 1 |
| 293 if nextchar == ']': |
| 294 break |
| 295 elif nextchar != ',': |
| 296 raise JSONDecodeError("Expecting , delimiter", s, end) |
| 297 |
| 298 try: |
| 299 if s[end] in _ws: |
| 300 end += 1 |
| 301 if s[end] in _ws: |
| 302 end = _w(s, end + 1).end() |
| 303 except IndexError: |
| 304 pass |
| 305 |
| 306 return values, end |
| 307 |
| 308 class JSONDecoder(object): |
| 309 """Simple JSON <http://json.org> decoder |
| 310 |
| 311 Performs the following translations in decoding by default: |
| 312 |
| 313 +---------------+-------------------+ |
| 314 | JSON | Python | |
| 315 +===============+===================+ |
| 316 | object | dict | |
| 317 +---------------+-------------------+ |
| 318 | array | list | |
| 319 +---------------+-------------------+ |
| 320 | string | unicode | |
| 321 +---------------+-------------------+ |
| 322 | number (int) | int, long | |
| 323 +---------------+-------------------+ |
| 324 | number (real) | float | |
| 325 +---------------+-------------------+ |
| 326 | true | True | |
| 327 +---------------+-------------------+ |
| 328 | false | False | |
| 329 +---------------+-------------------+ |
| 330 | null | None | |
| 331 +---------------+-------------------+ |
| 332 |
| 333 It also understands ``NaN``, ``Infinity``, and ``-Infinity`` as |
| 334 their corresponding ``float`` values, which is outside the JSON spec. |
| 335 |
| 336 """ |
| 337 |
| 338 def __init__(self, encoding=None, object_hook=None, parse_float=None, |
| 339 parse_int=None, parse_constant=None, strict=True, |
| 340 object_pairs_hook=None): |
| 341 """ |
| 342 *encoding* determines the encoding used to interpret any |
| 343 :class:`str` objects decoded by this instance (``'utf-8'`` by |
| 344 default). It has no effect when decoding :class:`unicode` objects. |
| 345 |
| 346 Note that currently only encodings that are a superset of ASCII work, |
| 347 strings of other encodings should be passed in as :class:`unicode`. |
| 348 |
| 349 *object_hook*, if specified, will be called with the result of every |
| 350 JSON object decoded and its return value will be used in place of the |
| 351 given :class:`dict`. This can be used to provide custom |
| 352 deserializations (e.g. to support JSON-RPC class hinting). |
| 353 |
| 354 *object_pairs_hook* is an optional function that will be called with |
| 355 the result of any object literal decode with an ordered list of pairs. |
| 356 The return value of *object_pairs_hook* will be used instead of the |
| 357 :class:`dict`. This feature can be used to implement custom decoders |
| 358 that rely on the order that the key and value pairs are decoded (for |
| 359 example, :func:`collections.OrderedDict` will remember the order of |
| 360 insertion). If *object_hook* is also defined, the *object_pairs_hook* |
| 361 takes priority. |
| 362 |
| 363 *parse_float*, if specified, will be called with the string of every |
| 364 JSON float to be decoded. By default, this is equivalent to |
| 365 ``float(num_str)``. This can be used to use another datatype or parser |
| 366 for JSON floats (e.g. :class:`decimal.Decimal`). |
| 367 |
| 368 *parse_int*, if specified, will be called with the string of every |
| 369 JSON int to be decoded. By default, this is equivalent to |
| 370 ``int(num_str)``. This can be used to use another datatype or parser |
| 371 for JSON integers (e.g. :class:`float`). |
| 372 |
| 373 *parse_constant*, if specified, will be called with one of the |
| 374 following strings: ``'-Infinity'``, ``'Infinity'``, ``'NaN'``. This |
| 375 can be used to raise an exception if invalid JSON numbers are |
| 376 encountered. |
| 377 |
| 378 *strict* controls the parser's behavior when it encounters an |
| 379 invalid control character in a string. The default setting of |
| 380 ``True`` means that unescaped control characters are parse errors, if |
| 381 ``False`` then control characters will be allowed in strings. |
| 382 |
| 383 """ |
| 384 self.encoding = encoding |
| 385 self.object_hook = object_hook |
| 386 self.object_pairs_hook = object_pairs_hook |
| 387 self.parse_float = parse_float or float |
| 388 self.parse_int = parse_int or int |
| 389 self.parse_constant = parse_constant or _CONSTANTS.__getitem__ |
| 390 self.strict = strict |
| 391 self.parse_object = JSONObject |
| 392 self.parse_array = JSONArray |
| 393 self.parse_string = scanstring |
| 394 self.memo = {} |
| 395 self.scan_once = make_scanner(self) |
| 396 |
| 397 def decode(self, s, _w=WHITESPACE.match): |
| 398 """Return the Python representation of ``s`` (a ``str`` or ``unicode`` |
| 399 instance containing a JSON document) |
| 400 |
| 401 """ |
| 402 obj, end = self.raw_decode(s, idx=_w(s, 0).end()) |
| 403 end = _w(s, end).end() |
| 404 if end != len(s): |
| 405 raise JSONDecodeError("Extra data", s, end, len(s)) |
| 406 return obj |
| 407 |
| 408 def raw_decode(self, s, idx=0): |
| 409 """Decode a JSON document from ``s`` (a ``str`` or ``unicode`` |
| 410 beginning with a JSON document) and return a 2-tuple of the Python |
| 411 representation and the index in ``s`` where the document ended. |
| 412 |
| 413 This can be used to decode a JSON document from a string that may |
| 414 have extraneous data at the end. |
| 415 |
| 416 """ |
| 417 try: |
| 418 obj, end = self.scan_once(s, idx) |
| 419 except StopIteration: |
| 420 raise JSONDecodeError("No JSON object could be decoded", s, idx) |
| 421 return obj, end |
OLD | NEW |