Module: OkJson
Overview
Some parts adapted from golang.org/src/pkg/json/decode.go and golang.org/src/pkg/utf8/utf8.go
Defined Under Namespace
Classes: ParserError
Constant Summary collapse
- Utagx =
1000 0000
0x80
- Utag2 =
1100 0000
0xc0
- Utag3 =
1110 0000
0xe0
- Utag4 =
1111 0000
0xf0
- Utag5 =
1111 1000
0xF8
- Umaskx =
0011 1111
0x3f
- Umask2 =
0001 1111
0x1f
- Umask3 =
0000 1111
0x0f
- Umask4 =
0000 0111
0x07
- Uchar1max =
(1 << 7) - 1
- Uchar2max =
(1 << 11) - 1
- Uchar3max =
(1 << 16) - 1
- Ucharerr =
unicode “replacement char”
0xFFFD
- Usurrself =
0x10000
- Usurr1 =
0xd800
- Usurr2 =
0xdc00
- Usurr3 =
0xe000
- Umax =
0x10ffff
- Spc =
- Unesc =
{ 'b' => "\b", 'f' => "\f", 'n' => "\n", 'r' => "\r", 't' => "\t" }.freeze
- Hex =
'0123456789abcdef'.freeze
Instance Method Summary collapse
- #abbrev(s) ⇒ Object
- #arrenc(a) ⇒ Object
-
#arrparse(ts) ⇒ Object
Parses an “array” in the sense of RFC 4627.
-
#decode(s) ⇒ Object
Decodes a json document in string s and returns the corresponding ruby value.
- #eat(typ, ts) ⇒ Object
-
#encode(x) ⇒ Object
Encodes x into a json text.
- #falsetok(s) ⇒ Object
- #hexdec4(s) ⇒ Object
- #hexenc4(t, u) ⇒ Object
-
#lex(s) ⇒ Object
Sans s and returns a list of json tokens, excluding white space (as defined in RFC 4627).
- #nibble(c) ⇒ Object
- #nulltok(s) ⇒ Object
- #numenc(x) ⇒ Object
- #numtok(s) ⇒ Object
- #objenc(x) ⇒ Object
-
#objparse(ts) ⇒ Object
Parses an “object” in the sense of RFC 4627.
-
#pairparse(ts) ⇒ Object
Parses a “member” in the sense of RFC 4627.
- #strenc(s) ⇒ Object
- #strtok(s) ⇒ Object
- #subst(u1, u2) ⇒ Object
- #surrogate?(u) ⇒ Boolean
-
#textparse(ts) ⇒ Object
Parses a “json text” in the sense of RFC 4627.
-
#tok(s) ⇒ Object
Scans the first token in s and returns a 3-element list, or nil if no such token exists.
- #truetok(s) ⇒ Object
-
#uchardec(s, i) ⇒ Object
Decodes unicode character u from UTF-8 bytes in string s at position i.
-
#ucharenc(a, i, u) ⇒ Object
Encodes unicode character u as UTF-8 bytes in string a at position i.
-
#unquote(q) ⇒ Object
Converts a quoted json string literal q into a UTF-8-encoded string.
- #unsubst(u) ⇒ Object
-
#valparse(ts) ⇒ Object
Parses a “value” in the sense of RFC 4627.
Instance Method Details
#abbrev(s) ⇒ Object
221 222 223 224 225 226 227 |
# File 'lib/vendor/okjson.rb', line 221 def abbrev(s) t = s[0, 10] p = t['`'] t = t[0, p] if p t += '...' if t.length < s.length '`' + t + '`' end |
#arrenc(a) ⇒ Object
348 349 350 |
# File 'lib/vendor/okjson.rb', line 348 def arrenc(a) '[' + a.map { |x| encode(x) }.join(',') + ']' end |
#arrparse(ts) ⇒ Object
Parses an “array” in the sense of RFC 4627. Returns the parsed value and any trailing tokens.
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 |
# File 'lib/vendor/okjson.rb', line 114 def arrparse(ts) ts = eat('[', ts) arr = [] return arr, ts[1..-1] if ts[0][0] == ']' v, ts = valparse(ts) arr << v return arr, ts[1..-1] if ts[0][0] == ']' loop do ts = eat(',', ts) v, ts = valparse(ts) arr << v return arr, ts[1..-1] if ts[0][0] == ']' end end |
#decode(s) ⇒ Object
Decodes a json document in string s and returns the corresponding ruby value. String s must be valid UTF-8. If you have a string in some other encoding, convert it first.
String values in the resulting structure will be UTF-8.
41 42 43 44 45 46 |
# File 'lib/vendor/okjson.rb', line 41 def decode(s) ts = lex(s) v, ts = textparse(ts) raise OkJson::ParserError, 'trailing garbage' unless ts.empty? v end |
#eat(typ, ts) ⇒ Object
135 136 137 138 139 140 |
# File 'lib/vendor/okjson.rb', line 135 def eat(typ, ts) if ts[0][0] != typ raise OkJson::ParserError, "expected #{typ} (got #{ts[0].inspect})" end ts[1..-1] end |
#encode(x) ⇒ Object
Encodes x into a json text. It may contain only Array, Hash, String, Numeric, true, false, nil. (Note, this list excludes Symbol.) Strings contained in x must be valid UTF-8. Values that cannot be represented, such as Nan, Infinity, Symbol, and Proc, are encoded as null, in accordance with ECMA-262, 5th ed.
330 331 332 333 334 335 336 337 338 339 340 341 342 |
# File 'lib/vendor/okjson.rb', line 330 def encode(x) case x when Hash then objenc(x) when Array then arrenc(x) when String then strenc(x) when Numeric then numenc(x) when Symbol then strenc(x.to_s) when true then 'true' when false then 'false' when nil then 'null' else 'null' end end |
#falsetok(s) ⇒ Object
198 199 200 |
# File 'lib/vendor/okjson.rb', line 198 def falsetok(s) s[0, 5] == 'false' && [:val, 'false', false] end |
#hexdec4(s) ⇒ Object
291 292 293 294 |
# File 'lib/vendor/okjson.rb', line 291 def hexdec4(s) raise OkJson::ParserError, 'short' if s.length != 4 (nibble(s[0]) << 12) | (nibble(s[1]) << 8) | (nibble(s[2]) << 4) | nibble(s[3]) end |
#hexenc4(t, u) ⇒ Object
391 392 393 394 395 396 |
# File 'lib/vendor/okjson.rb', line 391 def hexenc4(t, u) t.putc(Hex[(u >> 12) & 0xf]) t.putc(Hex[(u >> 8) & 0xf]) t.putc(Hex[(u >> 4) & 0xf]) t.putc(Hex[u & 0xf]) end |
#lex(s) ⇒ Object
Sans s and returns a list of json tokens, excluding white space (as defined in RFC 4627).
144 145 146 147 148 149 150 151 152 153 154 155 |
# File 'lib/vendor/okjson.rb', line 144 def lex(s) ts = [] until s.empty? typ, lexeme, val = tok(s) if typ.nil? raise OkJson::ParserError, "invalid character at #{s[0, 10].inspect}" end ts << [typ, lexeme, val] if typ != :space s = s[lexeme.length..-1] end ts end |
#nibble(c) ⇒ Object
313 314 315 316 317 318 319 320 321 |
# File 'lib/vendor/okjson.rb', line 313 def nibble(c) case true when c >= '0' && c <= '9' then c.ord - '0'.ord when c >= 'a' && c <= 'z' then c.ord - 'a'.ord + 10 when c >= 'A' && c <= 'Z' then c.ord - 'A'.ord + 10 else raise OkJson::ParserError, "invalid hex code #{c}" end end |
#nulltok(s) ⇒ Object
190 191 192 |
# File 'lib/vendor/okjson.rb', line 190 def nulltok(s) s[0, 4] == 'null' && [:val, 'null', nil] end |
#numenc(x) ⇒ Object
398 399 400 401 402 403 404 405 |
# File 'lib/vendor/okjson.rb', line 398 def numenc(x) begin return 'null' if x.nan? || x.infinite? rescue nil end x.to_s end |
#numtok(s) ⇒ Object
202 203 204 205 206 207 208 209 210 211 212 213 |
# File 'lib/vendor/okjson.rb', line 202 def numtok(s) m = /-?([1-9][0-9]+|[0-9])([.][0-9]+)?([eE][+-]?[0-9]+)?/.match(s) if m && m.begin(0) == 0 if m[3] && !m[2] [:val, m[0], Integer(m[1]) * (10**Integer(m[3][1..-1]))] elsif m[2] [:val, m[0], Float(m[0])] else [:val, m[0], Integer(m[0])] end end end |
#objenc(x) ⇒ Object
344 345 346 |
# File 'lib/vendor/okjson.rb', line 344 def objenc(x) '{' + x.map { |k, v| encode(k) + ':' + encode(v) }.join(',') + '}' end |
#objparse(ts) ⇒ Object
Parses an “object” in the sense of RFC 4627. Returns the parsed value and any trailing tokens.
80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 |
# File 'lib/vendor/okjson.rb', line 80 def objparse(ts) ts = eat('{', ts) obj = {} return obj, ts[1..-1] if ts[0][0] == '}' k, v, ts = pairparse(ts) obj[k] = v return obj, ts[1..-1] if ts[0][0] == '}' loop do ts = eat(',', ts) k, v, ts = pairparse(ts) obj[k] = v return obj, ts[1..-1] if ts[0][0] == '}' end end |
#pairparse(ts) ⇒ Object
Parses a “member” in the sense of RFC 4627. Returns the parsed value and any trailing tokens.
103 104 105 106 107 108 109 110 |
# File 'lib/vendor/okjson.rb', line 103 def pairparse(ts) (typ, _, k) = ts[0] ts = ts[1..-1] raise OkJson::ParserError, "unexpected #{k.inspect}" if typ != :str ts = eat(':', ts) v, ts = valparse(ts) [k, v, ts] end |
#strenc(s) ⇒ Object
352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 |
# File 'lib/vendor/okjson.rb', line 352 def strenc(s) t = StringIO.new t.putc('"') r = 0 while r < s.length case s[r] when '"' then t.print('\\"') when '\\' then t.print('\\\\') when "\b" then t.print('\\b') when "\f" then t.print('\\f') when "\n" then t.print('\\n') when "\r" then t.print('\\r') when "\t" then t.print('\\t') else c = s[r] case true when Spc <= c && c <= '~' t.putc(c) when true u, size = uchardec(s, r) r += size - 1 # we add one more at the bottom of the loop if u < 0x10000 t.print('\\u') hexenc4(t, u) else u1, u2 = unsubst(u) t.print('\\u') hexenc4(t, u1) t.print('\\u') hexenc4(t, u2) end end end r += 1 end t.putc('"') t.string end |
#strtok(s) ⇒ Object
215 216 217 218 219 |
# File 'lib/vendor/okjson.rb', line 215 def strtok(s) m = /"([^"\\]|\\["\/\\bfnrt]|\\u[0-9a-fA-F]{4})*"/.match(s) raise OkJson::ParserError, "invalid string literal at #{abbrev(s)}" unless m [:str, m[0], unquote(m[0])] end |
#subst(u1, u2) ⇒ Object
296 297 298 299 300 301 |
# File 'lib/vendor/okjson.rb', line 296 def subst(u1, u2) if Usurr1 <= u1 && u1 < Usurr2 && Usurr2 <= u2 && u2 < Usurr3 return ((u1 - Usurr1) << 10) | (u2 - Usurr2) + Usurrself end Ucharerr end |
#surrogate?(u) ⇒ Boolean
309 310 311 |
# File 'lib/vendor/okjson.rb', line 309 def surrogate?(u) Usurr1 <= u && u < Usurr3 end |
#textparse(ts) ⇒ Object
Parses a “json text” in the sense of RFC 4627. Returns the parsed value and any trailing tokens. Note: this is almost the same as valparse, except that it does not accept atomic values.
52 53 54 55 56 57 58 59 60 61 |
# File 'lib/vendor/okjson.rb', line 52 def textparse(ts) raise OkJson::ParserError, 'empty' if ts.length < 0 typ, _, val = ts[0] case typ when '{' then objparse(ts) when '[' then arrparse(ts) else valparse(ts) end end |
#tok(s) ⇒ Object
Scans the first token in s and returns a 3-element list, or nil if no such token exists.
The first list element is one of ‘‘’, ‘:’, ‘,’, ‘[’, ‘]’, :val, :str, and :space.
The second element is the lexeme.
The third element is the value of the token for :val and :str, otherwise it is the lexeme.
170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 |
# File 'lib/vendor/okjson.rb', line 170 def tok(s) case s[0] when '{' then ['{', s[0, 1], s[0, 1]] when '}' then ['}', s[0, 1], s[0, 1]] when ':' then [':', s[0, 1], s[0, 1]] when ',' then [',', s[0, 1], s[0, 1]] when '[' then ['[', s[0, 1], s[0, 1]] when ']' then [']', s[0, 1], s[0, 1]] when 'n' then nulltok(s) when 't' then truetok(s) when 'f' then falsetok(s) when '"' then strtok(s) when Spc then [:space, s[0, 1], s[0, 1]] when "\t" then [:space, s[0, 1], s[0, 1]] when "\n" then [:space, s[0, 1], s[0, 1]] when "\r" then [:space, s[0, 1], s[0, 1]] else numtok(s) end end |
#truetok(s) ⇒ Object
194 195 196 |
# File 'lib/vendor/okjson.rb', line 194 def truetok(s) s[0, 4] == 'true' && [:val, 'true', true] end |
#uchardec(s, i) ⇒ Object
Decodes unicode character u from UTF-8 bytes in string s at position i. Returns u and the number of bytes read.
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 |
# File 'lib/vendor/okjson.rb', line 410 def uchardec(s, i) n = s.length - i return [Ucharerr, 1] if n < 1 c0 = s[i].ord # 1-byte, 7-bit sequence? return [c0, 1] if c0 < Utagx # unexpected continuation byte? return [Ucharerr, 1] if c0 < Utag2 # need continuation byte return [Ucharerr, 1] if n < 2 c1 = s[i + 1].ord return [Ucharerr, 1] if c1 < Utagx || Utag2 <= c1 # 2-byte, 11-bit sequence? if c0 < Utag3 u = (c0 & Umask2) << 6 | (c1 & Umaskx) return [Ucharerr, 1] if u <= Uchar1max return [u, 2] end # need second continuation byte return [Ucharerr, 1] if n < 3 c2 = s[i + 2].ord return [Ucharerr, 1] if c2 < Utagx || Utag2 <= c2 # 3-byte, 16-bit sequence? if c0 < Utag4 u = (c0 & Umask3) << 12 | (c1 & Umaskx) << 6 | (c2 & Umaskx) return [Ucharerr, 1] if u <= Uchar2max return [u, 3] end # need third continuation byte return [Ucharerr, 1] if n < 4 c3 = s[i + 3].ord return [Ucharerr, 1] if c3 < Utagx || Utag2 <= c3 # 4-byte, 21-bit sequence? if c0 < Utag5 u = (c0 & Umask4) << 18 | (c1 & Umaskx) << 12 | (c2 & Umaskx) << 6 | (c3 & Umaskx) return [Ucharerr, 1] if u <= Uchar3max return [u, 4] end [Ucharerr, 1] end |
#ucharenc(a, i, u) ⇒ Object
Encodes unicode character u as UTF-8 bytes in string a at position i. Returns the number of bytes written.
464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 |
# File 'lib/vendor/okjson.rb', line 464 def ucharenc(a, i, u) case true when u <= Uchar1max a[i] = (u & 0xff).chr 1 when u <= Uchar2max a[i + 0] = (Utag2 | ((u >> 6) & 0xff)).chr a[i + 1] = (Utagx | (u & Umaskx)).chr 2 when u <= Uchar3max a[i + 0] = (Utag3 | ((u >> 12) & 0xff)).chr a[i + 1] = (Utagx | ((u >> 6) & Umaskx)).chr a[i + 2] = (Utagx | (u & Umaskx)).chr 3 else a[i + 0] = (Utag4 | ((u >> 18) & 0xff)).chr a[i + 1] = (Utagx | ((u >> 12) & Umaskx)).chr a[i + 2] = (Utagx | ((u >> 6) & Umaskx)).chr a[i + 3] = (Utagx | (u & Umaskx)).chr 4 end end |
#unquote(q) ⇒ Object
Converts a quoted json string literal q into a UTF-8-encoded string. The rules are different than for Ruby, so we cannot use eval. Unquote will raise OkJson::ParserError, an error if q contains control characters.
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 |
# File 'lib/vendor/okjson.rb', line 232 def unquote(q) q = q[1...-1] a = q.dup # allocate a big enough string r = 0 w = 0 while r < q.length c = q[r] case true when c == '\\' r += 1 if r >= q.length raise OkJson::ParserError, "string literal ends with a \"\\\": \"#{q}\"" end case q[r] when '"', '\\', '/', "'" a[w] = q[r] r += 1 w += 1 when 'b', 'f', 'n', 'r', 't' a[w] = Unesc[q[r]] r += 1 w += 1 when 'u' r += 1 uchar = begin hexdec4(q[r, 4]) rescue RuntimeError => e raise OkJson::ParserError, "invalid escape sequence \\u#{q[r, 4]}: #{e}" end r += 4 if surrogate? uchar if q.length >= r + 6 uchar1 = hexdec4(q[r + 2, 4]) uchar = subst(uchar, uchar1) if uchar != Ucharerr # A valid pair; consume. r += 6 end end end w += ucharenc(a, w, uchar) else raise OkJson::ParserError, "invalid escape char #{q[r]} in \"#{q}\"" end when c == '"', c < Spc raise OkJson::ParserError, "invalid character in string literal \"#{q}\"" else # Copy anything else byte-for-byte. # Valid UTF-8 will remain valid UTF-8. # Invalid UTF-8 will remain invalid UTF-8. a[w] = c r += 1 w += 1 end end a[0, w] end |
#unsubst(u) ⇒ Object
303 304 305 306 307 |
# File 'lib/vendor/okjson.rb', line 303 def unsubst(u) return Ucharerr, Ucharerr if u < Usurrself || u > Umax || surrogate?(u) u -= Usurrself [Usurr1 + ((u >> 10) & 0x3ff), Usurr2 + (u & 0x3ff)] end |
#valparse(ts) ⇒ Object
Parses a “value” in the sense of RFC 4627. Returns the parsed value and any trailing tokens.
65 66 67 68 69 70 71 72 73 74 75 76 |
# File 'lib/vendor/okjson.rb', line 65 def valparse(ts) raise OkJson::ParserError, 'empty' if ts.length < 0 typ, _, val = ts[0] case typ when '{' then objparse(ts) when '[' then arrparse(ts) when :val, :str then [val, ts[1..-1]] else raise OkJson::ParserError, "unexpected #{val.inspect}" end end |