| 1 |
Revision history for Perl extension JSON::XS |
| 2 |
|
| 3 |
TODO: maybe detetc and croak on more invalid inputs (e.g. +-inf/nan) |
| 4 |
TODO: maybe avoid the reblessing and better support readonly objects. |
| 5 |
TODO: http://stevehanov.ca/blog/index.php?id=104 compression |
| 6 |
|
| 7 |
2.34 Thu May 23 11:30:34 CEST 2013 |
| 8 |
- work around bugs in perl 5.18 breaking more than 100 |
| 9 |
widely used modules, without a fix in sight because |
| 10 |
p5pers don't care about CPAN anymore. |
| 11 |
- when canonicalising, only allocate up to 64 hash key |
| 12 |
pointers on the stack. for larger hashes, use the heap, |
| 13 |
to avoid using too much stackspace. |
| 14 |
- discuss the problem with setlocale (reported by a few victims). |
| 15 |
|
| 16 |
2.33 Wed Aug 1 21:03:52 CEST 2012 |
| 17 |
- internal encode/decode XS wrappers did not expect stack |
| 18 |
moves caused by callbacks (analyzed and testcase by Jesse Luehrs). |
| 19 |
- add bencode as to/from option in bin/json_xs. |
| 20 |
- add -e option to json_xs, and none and string in/out formats. |
| 21 |
|
| 22 |
2.32 Thu Aug 11 19:06:38 CEST 2011 |
| 23 |
- fix a bug in the initial whitespace accumulation. |
| 24 |
|
| 25 |
2.31 Wed Jul 27 17:53:05 CEST 2011 |
| 26 |
- don't accumulate initial whitespace in the incremental buffer |
| 27 |
(this can be useful to allow whitespace-keepalive on a tcp |
| 28 |
connection without triggering the max_size limit). |
| 29 |
- properly croak on some invalid inputs that are not strings |
| 30 |
(e.g. undef) when trying to decode a json text (reported |
| 31 |
and analyzed by Goro Fuji). |
| 32 |
|
| 33 |
2.3 Wed Aug 18 01:26:47 CEST 2010 |
| 34 |
- make sure decoder doesn't change the decoding in the incremental |
| 35 |
parser (testcase provided by Hendrik Schumacher). |
| 36 |
- applied patch by DaTa for Data::Dumper support in json_xs. |
| 37 |
- added -t dump support to json_xs, using Data::Dump. |
| 38 |
- added -f eval support to json_xs. |
| 39 |
|
| 40 |
2.29 Wed Mar 17 02:39:12 CET 2010 |
| 41 |
- fix a memory leak when callbacks set using filter_json_object |
| 42 |
or filter_json_single_key_object were called (great testcase |
| 43 |
by Eric Wilhelm). |
| 44 |
|
| 45 |
2.28 Thu Mar 11 20:30:46 CET 2010 |
| 46 |
- implement our own atof function - perl's can be orders of |
| 47 |
magnitudes slower than even the system one. on the positive |
| 48 |
side, ours seems to be more exact in general than perl's. |
| 49 |
(testcase provided by Tim Meadowcroft). |
| 50 |
- clarify floating point conversion issues a bit. |
| 51 |
- update jpsykes csrf article url. |
| 52 |
- updated benchmark section - JSON::PP became much faster! |
| 53 |
|
| 54 |
2.27 Thu Jan 7 07:35:08 CET 2010 |
| 55 |
- support relaxed option inside the incremental parser |
| 56 |
(testcase provided by IKEGAMI via Makamaka). |
| 57 |
|
| 58 |
2.26 Sat Oct 10 03:26:19 CEST 2009 |
| 59 |
- big integers could become truncated (based on patch |
| 60 |
by Strobl Anton). |
| 61 |
- output format change: indent now adds a final newline, which is |
| 62 |
more expected and more true to the documentation. |
| 63 |
|
| 64 |
2.25 Sat Aug 8 12:04:41 CEST 2009 |
| 65 |
- the perl debugger completely breaks lvalue subs - try to work |
| 66 |
around the issue. |
| 67 |
- ignore RMAGICAL hashes w.r.t. CANONICAL. |
| 68 |
- try to work around a possible char signedness issue on aix. |
| 69 |
- require common sense. |
| 70 |
|
| 71 |
2.24 Sat May 30 08:25:45 CEST 2009 |
| 72 |
- the incremental parser did not update its parse offset |
| 73 |
pointer correctly when parsing utf8-strings (nicely |
| 74 |
debugged by Martin Evans). |
| 75 |
- appending a non-utf8-string to the incremental parser |
| 76 |
in utf8 mode failed to upgrade the string. |
| 77 |
- wording of parse error messages has been improved. |
| 78 |
|
| 79 |
2.232 Sun Feb 22 11:12:25 CET 2009 |
| 80 |
- use an exponential algorithm to extend strings, to |
| 81 |
help platforms with bad or abysmal==windows memory |
| 82 |
allocater performance, at the expense of some memory |
| 83 |
wastage (use shrink to recover this extra memory). |
| 84 |
(nicely analysed by Dmitry Karasik). |
| 85 |
|
| 86 |
2.2311 Thu Feb 19 02:12:54 CET 2009 |
| 87 |
- add a section "JSON and ECMAscript" to explain some |
| 88 |
incompatibilities between the two (problem was noted by |
| 89 |
various people). |
| 90 |
- add t/20_faihu.t. |
| 91 |
|
| 92 |
2.231 Thu Nov 20 04:59:08 CET 2008 |
| 93 |
- work around 5.10.0 magic bugs where manipulating magic values |
| 94 |
(such as $1) would permanently damage them as perl would |
| 95 |
ignore the magicalness, by making a full copy of the string, |
| 96 |
reported by Dmitry Karasik. |
| 97 |
- work around spurious warnings under older perl 5.8's. |
| 98 |
|
| 99 |
2.23 Mon Sep 29 05:08:29 CEST 2008 |
| 100 |
- fix a compilation problem when perl is not using char * as, well, |
| 101 |
char *. |
| 102 |
- use PL_hexdigit in favour of rolling our own. |
| 103 |
|
| 104 |
2.2222 Sun Jul 20 18:49:00 CEST 2008 |
| 105 |
- same game again, broken 5.10 finds yet another assertion |
| 106 |
failure, and the workaround causes additional runtime warnings. |
| 107 |
Work around the next assertion AND the warning. 5.10 seriously |
| 108 |
needs to adjust it's attitude against working code. |
| 109 |
|
| 110 |
2.222 Sat Jul 19 06:15:34 CEST 2008 |
| 111 |
- you work around one -DDEBUGGING assertion bug in perl 5.10 |
| 112 |
just to hit the next one. work around this one, too. |
| 113 |
|
| 114 |
2.22 Tue Jul 15 13:26:51 CEST 2008 |
| 115 |
- allow higher nesting levels in incremental parser. |
| 116 |
- error out earlier in some cases in the incremental parser |
| 117 |
(as suggested by Yuval Kogman). |
| 118 |
- improve incr-parser test (Yuval Kogman). |
| 119 |
|
| 120 |
2.21 Tue Jun 3 08:43:23 CEST 2008 |
| 121 |
- (hopefully) work around a perl 5.10 bug with -DDEBUGGING. |
| 122 |
- remove the experimental status of the incremental parser interface. |
| 123 |
- move =encoding around again, to avoid bugs with search.cpan.org. |
| 124 |
when can we finally have utf-8 in pod??? |
| 125 |
- add ->incr_reset method. |
| 126 |
|
| 127 |
2.2 Wed Apr 16 20:37:25 CEST 2008 |
| 128 |
- lifted the log2 rounding restriction of max_depth and max_size. |
| 129 |
- make booleans mutable by creating a copy instead of handing out |
| 130 |
the same scalar (reported by pasha sadri). |
| 131 |
- added support for incremental json parsing (still EXPERIMENTAL). |
| 132 |
- implemented and added a json_xs command line utility that can convert |
| 133 |
from/to a number of serialisation formats - tell me if you need more. |
| 134 |
- implement allow_unknown/get_allow_unknown methods. |
| 135 |
- fixed documentation of max_depth w.r.t. higher and equal. |
| 136 |
- moved down =encoding directive a bit, too much breaks if it's the first |
| 137 |
pod directive :/. |
| 138 |
- removed documentation section on other modules, it became somewhat |
| 139 |
outdated and is nowadays mostly of historical interest. |
| 140 |
|
| 141 |
2.1 Wed Mar 19 23:23:18 CET 2008 |
| 142 |
- update documentation here and there: add a large section |
| 143 |
about utf8/latin1/ascii flags, add a security consideration |
| 144 |
and extend and clarify the JSON and YAML section. |
| 145 |
- medium speed enhancements when encoding/decoding non-ascii chars. |
| 146 |
- minor speedup in number encoding case. |
| 147 |
- extend and clarify the section on incompatibilities |
| 148 |
between YAML and JSON. |
| 149 |
- switch to static inline from just inline when using gcc. |
| 150 |
- add =encoding utf-8 to the manpage, now that perl 5.10 supports it. |
| 151 |
- fix some issues with UV to JSON conversion of unknown impact. |
| 152 |
- published the yahoo locals search result used in benchmarks as the |
| 153 |
original url changes so comparison is impossible. |
| 154 |
|
| 155 |
2.01 Wed Dec 5 11:40:28 CET 2007 |
| 156 |
- INCOMPATIBLE API CHANGE: to_json and from_json have been |
| 157 |
renamed to encode_json/decode_json for JSON.pm compatibility. |
| 158 |
The old functions croak and might be replaced by JSON.pm |
| 159 |
comaptible versions in some later release. |
| 160 |
|
| 161 |
2.0 Tue Dec 4 11:30:46 CET 2007 |
| 162 |
- this is supposed to be the first version of JSON::XS |
| 163 |
compatible with version 2.0+ of the JSON module. |
| 164 |
Using the JSON module as frontend to JSON::XS should be |
| 165 |
as fast as using JSON::XS directly, so consider using it |
| 166 |
instead. |
| 167 |
- added get_* methods for all "simple" options. |
| 168 |
- make JSON::XS subclassable. |
| 169 |
|
| 170 |
1.53 Tue Nov 13 23:58:33 CET 2007 |
| 171 |
- minor doc clarifications. |
| 172 |
- fixed many doc typos (patch by Thomas L. Shinnick). |
| 173 |
|
| 174 |
1.52 Mon Oct 15 03:22:06 CEST 2007 |
| 175 |
- remove =encoding pod directive again, it confuses too many pod |
| 176 |
parsers :/. |
| 177 |
|
| 178 |
1.51 Sat Oct 13 03:55:56 CEST 2007 |
| 179 |
- encode empty arrays/hashes in a compact way when pretty is enabled. |
| 180 |
- apparently JSON::XS was used to find some bugs in the |
| 181 |
JSON_checker testsuite, so add (the corrected) JSON_checker tests to |
| 182 |
the testsuite. |
| 183 |
- quite a bit of doc updates/extension. |
| 184 |
- require 5.8.2, as this seems to be the first unicode-stable version. |
| 185 |
|
| 186 |
1.5 Tue Aug 28 04:05:38 CEST 2007 |
| 187 |
- add support for tied hashes, based on ideas and testcase by |
| 188 |
Marcus Holland-Moritz. |
| 189 |
- implemented relaxed parsing mode where some extensions are being |
| 190 |
accepted. generation is still JSON-only. |
| 191 |
|
| 192 |
1.44 Wed Aug 22 01:02:44 CEST 2007 |
| 193 |
- very experimental process-emulation support, slowing everything down. |
| 194 |
the horribly broken perl threads are still not supported - YMMV. |
| 195 |
|
| 196 |
1.43 Thu Jul 26 13:26:37 CEST 2007 |
| 197 |
- convert big json numbers exclusively consisting of digits to NV |
| 198 |
only when there is no loss of precision, otherwise to string. |
| 199 |
|
| 200 |
1.42 Tue Jul 24 00:51:18 CEST 2007 |
| 201 |
- fix a crash caused by not handling missing array elements |
| 202 |
(report and testcase by Jay Kuri). |
| 203 |
|
| 204 |
1.41 Tue Jul 10 18:21:44 CEST 2007 |
| 205 |
- fix compilation with NDEBUG (assert side-effect), |
| 206 |
affects convert_blessed only. |
| 207 |
- fix a bug in decode filters calling ENTER; SAVETMPS; |
| 208 |
one time too often. |
| 209 |
- catch a typical error in TO_JSON methods. |
| 210 |
- antique-ised XS.xs again to work with outdated |
| 211 |
C compilers (windows...). |
| 212 |
|
| 213 |
1.4 Mon Jul 2 10:06:30 CEST 2007 |
| 214 |
- add convert_blessed setting. |
| 215 |
- encode did not catch all blessed objects, encoding their |
| 216 |
contents in most cases. This has been fixed by introducing |
| 217 |
the allow_blessed setting. |
| 218 |
- added filter_json_object and filter_json_single_key_object |
| 219 |
settings that specify a callback to be called when |
| 220 |
all/specific json objects are encountered. |
| 221 |
- assume that most object keys are simple ascii words and |
| 222 |
optimise this case, penalising the general case. This can |
| 223 |
speed up decoding by 30% in typical cases and gives |
| 224 |
a smaller and faster perl hash. |
| 225 |
- implemented simpleminded, optional resource size checking |
| 226 |
in decode_json. |
| 227 |
- remove objToJson/jsonToObj aliases, as the next version |
| 228 |
of JSON will not have them either. |
| 229 |
- bit the bullet and converted the very simple json object |
| 230 |
into a more complex one. |
| 231 |
- work around a bug where perl wrongly claims an integer |
| 232 |
is not an integer. |
| 233 |
- unbundle JSON::XS::Boolean into own pm file so Storable |
| 234 |
and similar modules can resolve the overloading when thawing. |
| 235 |
|
| 236 |
1.3 Sun Jun 24 01:55:02 CEST 2007 |
| 237 |
- make JSON::XS::true and false special overloaded objects |
| 238 |
and return those instead of 1 and 0 for those json atoms |
| 239 |
(JSON::PP compatibility is NOT achieved yet). |
| 240 |
- add JSON::XS::is_bool predicate to test for those special |
| 241 |
values. |
| 242 |
- add a reference to |
| 243 |
http://jpsykes.com/47/practical-csrf-and-json-security. |
| 244 |
- removed require 5.8.8 again, it is just not very expert-friendly. |
| 245 |
Also try to be more compatible with slightly older versions, |
| 246 |
which are not recommended (because they are buggy). |
| 247 |
|
| 248 |
1.24 Mon Jun 11 05:40:49 CEST 2007 |
| 249 |
- added informative section on JSON-as-YAML. |
| 250 |
- get rid of some c99-isms again. |
| 251 |
- localise dec->cur in decode_str, speeding up |
| 252 |
string decoding considerably (>15% on my amd64 + gcc). |
| 253 |
- increased SHORT_STRING_LEN to 16kb: stack space is |
| 254 |
usually plenty, and this actually saves memory |
| 255 |
when !shrinking as short strings will fit perfectly. |
| 256 |
|
| 257 |
1.23 Wed Jun 6 20:13:06 CEST 2007 |
| 258 |
- greatly improved small integer encoding and decoding speed. |
| 259 |
- implement a number of µ-optimisations. |
| 260 |
- updated benchmarks. |
| 261 |
|
| 262 |
1.22 Thu May 24 00:07:25 CEST 2007 |
| 263 |
- require 5.8.8 explicitly as older perls do not seem to offer |
| 264 |
the required macros. |
| 265 |
- possibly made it compile on so-called C compilers by microsoft. |
| 266 |
|
| 267 |
1.21 Wed May 9 18:40:32 CEST 2007 |
| 268 |
- character offset reported for trailing garbage was random. |
| 269 |
|
| 270 |
1.2 Wed May 9 18:35:01 CEST 2007 |
| 271 |
- decode did not work with magical scalars (doh!). |
| 272 |
- added latin1 flag to produce JSON texts in the latin1 subset |
| 273 |
of unicode. |
| 274 |
- flag trailing garbage as error. |
| 275 |
- new decode_prefix method that returns the number |
| 276 |
of characters consumed by a decode. |
| 277 |
- max octets/char in perls UTF-X is actually 13, not 11, |
| 278 |
as pointed out by Glenn Linderman. |
| 279 |
- fixed typoe reported by YAMASHINA Hio. |
| 280 |
|
| 281 |
1.11 Mon Apr 9 07:05:49 CEST 2007 |
| 282 |
- properly 0-terminate sv's returned by encode to help |
| 283 |
C libraries that expect that 0 to be there. |
| 284 |
- partially "port" JSON from C to microsofts fucking broken |
| 285 |
pseudo-C. They should be burned to the ground for pissing |
| 286 |
on standards. And I should be stoned for even trying to |
| 287 |
support this filthy excuse for a c compiler. |
| 288 |
|
| 289 |
1.1 Wed Apr 4 01:45:00 CEST 2007 |
| 290 |
- clarify documentation (pointed out by Quinn Weaver). |
| 291 |
- decode_utf8 sometimes did not correctly flag errors, |
| 292 |
leading to segfaults. |
| 293 |
- further reduced default nesting depth to 512 due to the test |
| 294 |
failure by that anonymous "chris" whose e-mail address seems |
| 295 |
to be impossible to get. Tests on other freebsd systems indicate |
| 296 |
that this is likely a problem in his/her configuration and not this |
| 297 |
module. |
| 298 |
- renamed json => JSON in error messages. |
| 299 |
- corrected the character offset in some error messages. |
| 300 |
|
| 301 |
1.01 Sat Mar 31 16:15:40 CEST 2007 |
| 302 |
- do not segfault when from_json/decode gets passed |
| 303 |
a non-string object (reported by Florian Ragwitz). |
| 304 |
This has no effect on normal operation. |
| 305 |
|
| 306 |
1.0 Thu Mar 29 04:43:34 CEST 2007 |
| 307 |
- the long awaited (by me) 1.0 version. |
| 308 |
- add \0 (JSON::XS::false) and \1 (JSON::XS::true) mappings to JSON |
| 309 |
true and false. |
| 310 |
- add some more notes to shrink, as suggested by Alex Efros. |
| 311 |
- improve testsuite. |
| 312 |
- halve the default nesting depth limit, to hopefully make it |
| 313 |
work on Freebsd (unfortunately, the cpan tester did not |
| 314 |
send me his report, so I cannot ask about the stack limit on fbsd). |
| 315 |
|
| 316 |
0.8 Mon Mar 26 00:10:48 CEST 2007 |
| 317 |
- fix a memleak when decoding hashes. |
| 318 |
- export jsonToBj and objToJson as aliases |
| 319 |
to to_json and from_json, to reduce incompatibilities |
| 320 |
between JSON/JSON::PC and JSON::XS. (experimental). |
| 321 |
- implement a maximum nesting depth for both en- and de-coding. |
| 322 |
- added a security considerations sections. |
| 323 |
|
| 324 |
0.7 Sun Mar 25 01:46:30 CET 2007 |
| 325 |
- code cleanup. |
| 326 |
- fix a memory overflow bug when indenting. |
| 327 |
- pretty-printing now up to 15% faster. |
| 328 |
- improve decoding speed of strings by |
| 329 |
up to 50% by specialcasing short strings. |
| 330 |
- further decoding speedups for strings using |
| 331 |
lots of \u escapes. |
| 332 |
- improve utf8 decoding speed for U+80 .. U+7FF. |
| 333 |
|
| 334 |
0.5 Sat Mar 24 20:41:51 CET 2007 |
| 335 |
- added the UTF-16 encoding example hinted at in previous |
| 336 |
versions. |
| 337 |
- minor documentation fixes. |
| 338 |
- fix a bug in and optimise canonicalising fastpath |
| 339 |
(reported by Craig Manley). |
| 340 |
- remove a subtest that breaks with bleadperl (reported |
| 341 |
by Andreas König). |
| 342 |
|
| 343 |
0.31 Sat Mar 24 02:14:34 CET 2007 |
| 344 |
- documentation updates. |
| 345 |
- do some casting to hopefully fix Andreas' problem. |
| 346 |
- nuke bogus json rpc stuff. |
| 347 |
|
| 348 |
0.3 Fri Mar 23 19:33:21 CET 2007 |
| 349 |
- remove spurious PApp::Util reference (John McNamara). |
| 350 |
- adapted lots of tests from other json modules |
| 351 |
(idea by Chris Carline). |
| 352 |
- documented mapping from json to perl and vice versa. |
| 353 |
- improved the documentation by adding more examples. |
| 354 |
- added short escaping forms, reducing the created |
| 355 |
json texts a bit. |
| 356 |
- added shrink flag. |
| 357 |
- when flag methods are called without enable argument |
| 358 |
they will by default enable their flag. |
| 359 |
- considerably improved string encoding speed (at least |
| 360 |
with gcc 4). |
| 361 |
- added a test that covers lots of different characters. |
| 362 |
- clarified some error messages. |
| 363 |
- error messages now use correct character offset |
| 364 |
with F_UTF8. |
| 365 |
- improve the "no bytes" and "no warnings" hacks in |
| 366 |
case the called functions do... stuff. |
| 367 |
- croak when encoding to ascii and an out-of-range |
| 368 |
(non-unicode) codepoint is encountered. |
| 369 |
|
| 370 |
0.2 Fri Mar 23 00:23:34 CET 2007 |
| 371 |
- the "could not sleep without debugging release". |
| 372 |
it should basically work now, with many bugs as |
| 373 |
no production tests have been run yet. |
| 374 |
- added more testcases. |
| 375 |
- the expected shitload of bugfixes. |
| 376 |
- handle utf8 flag correctly in decode. |
| 377 |
- fix segfault in decoder. |
| 378 |
- utf8n_to_uvuni sets retlen to -1, but retlen is an |
| 379 |
unsigned types (argh). |
| 380 |
- fix decoding of utf-8 strings. |
| 381 |
- improved error diagnostics. |
| 382 |
- fix decoding of 'null'. |
| 383 |
- fix parsing of empty array/hashes |
| 384 |
- silence warnings when we prepare the croak message. |
| 385 |
|
| 386 |
0.1 Thu Mar 22 22:13:43 CET 2007 |
| 387 |
- first release, very untested, basically just to claim |
| 388 |
the namespace. |
| 389 |
|
| 390 |
0.01 Thu Mar 22 06:08:12 CET 2007 |
| 391 |
- original version; cloned from Convert-Scalar |
| 392 |
|
