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