1Revision history for Perl extension JSON::XS 2 32.27 Thu Jan 7 07:35:08 CET 2010 4 - support relaxed option inside the incremental parser 5 (testcase provided by IKEGAMI via Makamaka). 6 72.26 Sat Oct 10 03:26:19 CEST 2009 8 - big integers could become truncated (based on patch 9 by Strobl Anton). 10 - output format change: indent now adds a final newline, which is 11 more expected and more true to the documentation. 12 132.25 Sat Aug 8 12:04:41 CEST 2009 14 - the perl debugger completely breaks lvalue subs - try to work 15 around the issue. 16 - ignore RMAGICAL hashes w.r.t. CANONICAL. 17 - try to work around a possible char signedness issue on aix. 18 - require common sense. 19 202.24 Sat May 30 08:25:45 CEST 2009 21 - the incremental parser did not update its parse offset 22 pointer correctly when parsing utf8-strings (nicely 23 debugged by Martin Evans). 24 - appending a non-utf8-string to the incremental parser 25 in utf8 mode failed to upgrade the string. 26 - wording of parse error messages has been improved. 27 282.232 Sun Feb 22 11:12:25 CET 2009 29 - use an exponential algorithm to extend strings, to 30 help platforms with bad or abysmal==windows memory 31 allocater performance, at the expense of some memory 32 wastage (use shrink to recover this extra memory). 33 (nicely analysed by Dmitry Karasik). 34 352.2311 Thu Feb 19 02:12:54 CET 2009 36 - add a section "JSON and ECMAscript" to explain some 37 incompatibilities between the two (problem was noted by 38 various people). 39 - add t/20_faihu.t. 40 412.231 Thu Nov 20 04:59:08 CET 2008 42 - work around 5.10.0 magic bugs where manipulating magic values 43 (such as $1) would permanently damage them as perl would 44 ignore the magicalness, by making a full copy of the string, 45 reported by Dmitry Karasik. 46 - work around spurious warnings under older perl 5.8's. 47 482.23 Mon Sep 29 05:08:29 CEST 2008 49 - fix a compilation problem when perl is not using char * as, well, 50 char *. 51 - use PL_hexdigit in favour of rolling our own. 52 532.2222 Sun Jul 20 18:49:00 CEST 2008 54 - same game again, broken 5.10 finds yet another assertion 55 failure, and the workaround causes additional runtime warnings. 56 Work around the next assertion AND the warning. 5.10 seriously 57 needs to adjust it's attitude against working code. 58 592.222 Sat Jul 19 06:15:34 CEST 2008 60 - you work around one -DDEBUGGING assertion bug in perl 5.10 61 just to hit the next one. work around this one, too. 62 632.22 Tue Jul 15 13:26:51 CEST 2008 64 - allow higher nesting levels in incremental parser. 65 - error out earlier in some cases in the incremental parser 66 (as suggested by Yuval Kogman). 67 - improve incr-parser test (Yuval Kogman). 68 692.21 Tue Jun 3 08:43:23 CEST 2008 70 - (hopefully) work around a perl 5.10 bug with -DDEBUGGING. 71 - remove the experimental status of the incremental parser interface. 72 - move =encoding around again, to avoid bugs with search.cpan.org. 73 when can we finally have utf-8 in pod??? 74 - add ->incr_reset method. 75 762.2 Wed Apr 16 20:37:25 CEST 2008 77 - lifted the log2 rounding restriction of max_depth and max_size. 78 - make booleans mutable by creating a copy instead of handing out 79 the same scalar (reported by pasha sadri). 80 - added support for incremental json parsing (still EXPERIMENTAL). 81 - implemented and added a json_xs command line utility that can convert 82 from/to a number of serialisation formats - tell me if you need more. 83 - implement allow_unknown/get_allow_unknown methods. 84 - fixed documentation of max_depth w.r.t. higher and equal. 85 - moved down =encoding directive a bit, too much breaks if it's the first 86 pod directive :/. 87 - removed documentation section on other modules, it became somewhat 88 outdated and is nowadays mostly of historical interest. 89 902.1 Wed Mar 19 23:23:18 CET 2008 91 - update documentation here and there: add a large section 92 about utf8/latin1/ascii flags, add a security consideration 93 and extend and clarify the JSON and YAML section. 94 - medium speed enhancements when encoding/decoding non-ascii chars. 95 - minor speedup in number encoding case. 96 - extend and clarify the section on incompatibilities 97 between YAML and JSON. 98 - switch to static inline from just inline when using gcc. 99 - add =encoding utf-8 to the manpage, now that perl 5.10 supports it. 100 - fix some issues with UV to JSON conversion of unknown impact. 101 - published the yahoo locals search result used in benchmarks as the 102 original url changes so comparison is impossible. 103 1042.01 Wed Dec 5 11:40:28 CET 2007 105 - INCOMPATIBLE API CHANGE: to_json and from_json have been 106 renamed to encode_json/decode_json for JSON.pm compatibility. 107 The old functions croak and might be replaced by JSON.pm 108 comaptible versions in some later release. 109 1102.0 Tue Dec 4 11:30:46 CET 2007 111 - this is supposed to be the first version of JSON::XS 112 compatible with version 2.0+ of the JSON module. 113 Using the JSON module as frontend to JSON::XS should be 114 as fast as using JSON::XS directly, so consider using it 115 instead. 116 - added get_* methods for all "simple" options. 117 - make JSON::XS subclassable. 118 1191.53 Tue Nov 13 23:58:33 CET 2007 120 - minor doc clarifications. 121 - fixed many doc typos (patch by Thomas L. Shinnick). 122 1231.52 Mon Oct 15 03:22:06 CEST 2007 124 - remove =encoding pod directive again, it confuses too many pod 125 parsers :/. 126 1271.51 Sat Oct 13 03:55:56 CEST 2007 128 - encode empty arrays/hashes in a compact way when pretty is enabled. 129 - apparently JSON::XS was used to find some bugs in the 130 JSON_checker testsuite, so add (the corrected) JSON_checker tests to 131 the testsuite. 132 - quite a bit of doc updates/extension. 133 - require 5.8.2, as this seems to be the first unicode-stable version. 134 1351.5 Tue Aug 28 04:05:38 CEST 2007 136 - add support for tied hashes, based on ideas and testcase by 137 Marcus Holland-Moritz. 138 - implemented relaxed parsing mode where some extensions are being 139 accepted. generation is still JSON-only. 140 1411.44 Wed Aug 22 01:02:44 CEST 2007 142 - very experimental process-emulation support, slowing everything down. 143 the horribly broken perl threads are still not supported - YMMV. 144 1451.43 Thu Jul 26 13:26:37 CEST 2007 146 - convert big json numbers exclusively consisting of digits to NV 147 only when there is no loss of precision, otherwise to string. 148 1491.42 Tue Jul 24 00:51:18 CEST 2007 150 - fix a crash caused by not handling missing array elements 151 (report and testcase by Jay Kuri). 152 1531.41 Tue Jul 10 18:21:44 CEST 2007 154 - fix compilation with NDEBUG (assert side-effect), 155 affects convert_blessed only. 156 - fix a bug in decode filters calling ENTER; SAVETMPS; 157 one time too often. 158 - catch a typical error in TO_JSON methods. 159 - antique-ised XS.xs again to work with outdated 160 C compilers (windows...). 161 1621.4 Mon Jul 2 10:06:30 CEST 2007 163 - add convert_blessed setting. 164 - encode did not catch all blessed objects, encoding their 165 contents in most cases. This has been fixed by introducing 166 the allow_blessed setting. 167 - added filter_json_object and filter_json_single_key_object 168 settings that specify a callback to be called when 169 all/specific json objects are encountered. 170 - assume that most object keys are simple ascii words and 171 optimise this case, penalising the general case. This can 172 speed up decoding by 30% in typical cases and gives 173 a smaller and faster perl hash. 174 - implemented simpleminded, optional resource size checking 175 in decode_json. 176 - remove objToJson/jsonToObj aliases, as the next version 177 of JSON will not have them either. 178 - bit the bullet and converted the very simple json object 179 into a more complex one. 180 - work around a bug where perl wrongly claims an integer 181 is not an integer. 182 - unbundle JSON::XS::Boolean into own pm file so Storable 183 and similar modules can resolve the overloading when thawing. 184 1851.3 Sun Jun 24 01:55:02 CEST 2007 186 - make JSON::XS::true and false special overloaded objects 187 and return those instead of 1 and 0 for those json atoms 188 (JSON::PP compatibility is NOT achieved yet). 189 - add JSON::XS::is_bool predicate to test for those special 190 values. 191 - add a reference to 192 http://jpsykes.com/47/practical-csrf-and-json-security. 193 - removed require 5.8.8 again, it is just not very expert-friendly. 194 Also try to be more compatible with slightly older versions, 195 which are not recommended (because they are buggy). 196 1971.24 Mon Jun 11 05:40:49 CEST 2007 198 - added informative section on JSON-as-YAML. 199 - get rid of some c99-isms again. 200 - localise dec->cur in decode_str, speeding up 201 string decoding considerably (>15% on my amd64 + gcc). 202 - increased SHORT_STRING_LEN to 16kb: stack space is 203 usually plenty, and this actually saves memory 204 when !shrinking as short strings will fit perfectly. 205 2061.23 Wed Jun 6 20:13:06 CEST 2007 207 - greatly improved small integer encoding and decoding speed. 208 - implement a number of µ-optimisations. 209 - updated benchmarks. 210 2111.22 Thu May 24 00:07:25 CEST 2007 212 - require 5.8.8 explicitly as older perls do not seem to offer 213 the required macros. 214 - possibly made it compile on so-called C compilers by microsoft. 215 2161.21 Wed May 9 18:40:32 CEST 2007 217 - character offset reported for trailing garbage was random. 218 2191.2 Wed May 9 18:35:01 CEST 2007 220 - decode did not work with magical scalars (doh!). 221 - added latin1 flag to produce JSON texts in the latin1 subset 222 of unicode. 223 - flag trailing garbage as error. 224 - new decode_prefix method that returns the number 225 of characters consumed by a decode. 226 - max octets/char in perls UTF-X is actually 13, not 11, 227 as pointed out by Glenn Linderman. 228 - fixed typoe reported by YAMASHINA Hio. 229 2301.11 Mon Apr 9 07:05:49 CEST 2007 231 - properly 0-terminate sv's returned by encode to help 232 C libraries that expect that 0 to be there. 233 - partially "port" JSON from C to microsofts fucking broken 234 pseudo-C. They should be burned to the ground for pissing 235 on standards. And I should be stoned for even trying to 236 support this filthy excuse for a c compiler. 237 2381.1 Wed Apr 4 01:45:00 CEST 2007 239 - clarify documentation (pointed out by Quinn Weaver). 240 - decode_utf8 sometimes did not correctly flag errors, 241 leading to segfaults. 242 - further reduced default nesting depth to 512 due to the test 243 failure by that anonymous "chris" whose e-mail address seems 244 to be impossible to get. Tests on other freebsd systems indicate 245 that this is likely a problem in his/her configuration and not this 246 module. 247 - renamed json => JSON in error messages. 248 - corrected the character offset in some error messages. 249 2501.01 Sat Mar 31 16:15:40 CEST 2007 251 - do not segfault when from_json/decode gets passed 252 a non-string object (reported by Florian Ragwitz). 253 This has no effect on normal operation. 254 2551.0 Thu Mar 29 04:43:34 CEST 2007 256 - the long awaited (by me) 1.0 version. 257 - add \0 (JSON::XS::false) and \1 (JSON::XS::true) mappings to JSON 258 true and false. 259 - add some more notes to shrink, as suggested by Alex Efros. 260 - improve testsuite. 261 - halve the default nesting depth limit, to hopefully make it 262 work on Freebsd (unfortunately, the cpan tester did not 263 send me his report, so I cannot ask about the stack limit on fbsd). 264 2650.8 Mon Mar 26 00:10:48 CEST 2007 266 - fix a memleak when decoding hashes. 267 - export jsonToBj and objToJson as aliases 268 to to_json and from_json, to reduce incompatibilities 269 between JSON/JSON::PC and JSON::XS. (experimental). 270 - implement a maximum nesting depth for both en- and de-coding. 271 - added a security considerations sections. 272 2730.7 Sun Mar 25 01:46:30 CET 2007 274 - code cleanup. 275 - fix a memory overflow bug when indenting. 276 - pretty-printing now up to 15% faster. 277 - improve decoding speed of strings by 278 up to 50% by specialcasing short strings. 279 - further decoding speedups for strings using 280 lots of \u escapes. 281 - improve utf8 decoding speed for U+80 .. U+7FF. 282 2830.5 Sat Mar 24 20:41:51 CET 2007 284 - added the UTF-16 encoding example hinted at in previous 285 versions. 286 - minor documentation fixes. 287 - fix a bug in and optimise canonicalising fastpath 288 (reported by Craig Manley). 289 - remove a subtest that breaks with bleadperl (reported 290 by Andreas König). 291 2920.31 Sat Mar 24 02:14:34 CET 2007 293 - documentation updates. 294 - do some casting to hopefully fix Andreas' problem. 295 - nuke bogus json rpc stuff. 296 2970.3 Fri Mar 23 19:33:21 CET 2007 298 - remove spurious PApp::Util reference (John McNamara). 299 - adapted lots of tests from other json modules 300 (idea by Chris Carline). 301 - documented mapping from json to perl and vice versa. 302 - improved the documentation by adding more examples. 303 - added short escaping forms, reducing the created 304 json texts a bit. 305 - added shrink flag. 306 - when flag methods are called without enable argument 307 they will by default enable their flag. 308 - considerably improved string encoding speed (at least 309 with gcc 4). 310 - added a test that covers lots of different characters. 311 - clarified some error messages. 312 - error messages now use correct character offset 313 with F_UTF8. 314 - improve the "no bytes" and "no warnings" hacks in 315 case the called functions do... stuff. 316 - croak when encoding to ascii and an out-of-range 317 (non-unicode) codepoint is encountered. 318 3190.2 Fri Mar 23 00:23:34 CET 2007 320 - the "could not sleep without debugging release". 321 it should basically work now, with many bugs as 322 no production tests have been run yet. 323 - added more testcases. 324 - the expected shitload of bugfixes. 325 - handle utf8 flag correctly in decode. 326 - fix segfault in decoder. 327 - utf8n_to_uvuni sets retlen to -1, but retlen is an 328 unsigned types (argh). 329 - fix decoding of utf-8 strings. 330 - improved error diagnostics. 331 - fix decoding of 'null'. 332 - fix parsing of empty array/hashes 333 - silence warnings when we prepare the croak message. 334 3350.1 Thu Mar 22 22:13:43 CET 2007 336 - first release, very untested, basically just to claim 337 the namespace. 338 3390.01 Thu Mar 22 06:08:12 CET 2007 340 - original version; cloned from Convert-Scalar 341 342