| 1 | Revision history for Perl extension JSON::XS |
1 | Revision history for Perl extension JSON::XS |
| 2 | |
2 | |
| 3 | TODO: maybe simply use \0 and \1 as true/false when encoding (and overloaded/blessed versions thereof on |
3 | TODO: maybe detetc and croak on more invalid inputs (e.g. +-inf/nan) |
| 4 | decoding?) |
4 | TODO: maybe avoid the reblessing and better support readonly objects. |
|
|
5 | TODO: http://stevehanov.ca/blog/index.php?id=104 compression |
|
|
6 | TODO: how to cope with tagged values and standard json decoders |
|
|
7 | TODO: investigate magic (Eric Brine) |
|
|
8 | TODO: rfc7464 |
|
|
9 | TODO: Victor Efimov |
|
|
10 | TODO: move json_xs to types::serialiser |
|
|
11 | TODO: [PATCH] Types::Serialiser: Inline true(), false() and error() functions |
| 5 | |
12 | |
| 6 | 0.9 |
13 | 3.04 Thu Aug 17 04:30:47 CEST 2017 |
|
|
14 | - change exponential realloc algorithm on encoding and string decoding to be |
|
|
15 | really exponential (this helps slow/debugging allocators such as libumem) |
|
|
16 | (reported by Matthew Horsfall). |
|
|
17 | - string encoding would needlessly overallocate output space |
|
|
18 | (testcase by Matthew Horsfall). |
|
|
19 | - be very paranoid about extending buffer lengths and croak if buffers get too large, |
|
|
20 | which might (or might not) improve security. |
|
|
21 | - add cbor-packed type to json_xs. |
|
|
22 | - switch from YAML to YAML::XS in json_xs, as YAML is way too buggy and outdated. |
|
|
23 | |
|
|
24 | 3.03 Wed Nov 16 20:20:59 CET 2016 |
|
|
25 | - fix a bug introduced by a perl bug workaround that would cause |
|
|
26 | incremental parsing to fail with a sv_chop panic. |
|
|
27 | - json_xs: toformat failure error message fixed. |
|
|
28 | - json_xs: allow cyclic data structures in CBOR. |
|
|
29 | |
|
|
30 | 3.02 Fri Feb 26 22:45:20 CET 2016 |
|
|
31 | - allow_nonref now affects booleans (\1, $Types::Serialiser::Boolean) |
|
|
32 | as well (reported by Alex Efros). |
|
|
33 | - allow literal tabs in strings in relaxed mode (patch by |
|
|
34 | lubo.rintel@gooddata.com). |
|
|
35 | - support "cbor" format in json_xs tool. |
|
|
36 | - support (and fix) calling encode and decode in list context |
|
|
37 | (reported by Вадим Власов). |
|
|
38 | - work around a bug in older perls crashing when presented |
|
|
39 | with shared hash keys (Reini Urban). |
|
|
40 | - use stability canary. |
|
|
41 | |
|
|
42 | 3.01 Tue Oct 29 16:55:15 CET 2013 |
|
|
43 | - backport to perls < 5.18 (reported by Paul Howarth). |
|
|
44 | |
|
|
45 | 3.0 Tue Oct 29 01:35:37 CET 2013 |
|
|
46 | - implemented an object tagging extension (using the |
|
|
47 | Types::Serialiser serialisation protocol). |
|
|
48 | - reworked the documentation regarding object serialisation, |
|
|
49 | add a new OBJECT SERIALISATION section that explains th |
|
|
50 | whole process. |
|
|
51 | - new setting: allow_tags. |
|
|
52 | - switch to Types::Serialiser booleans. |
|
|
53 | - remove to_json/from_json. |
|
|
54 | - other minor improvements to the documentation. |
|
|
55 | |
|
|
56 | 2.34 Thu May 23 11:30:34 CEST 2013 |
|
|
57 | - work around bugs in perl 5.18 breaking more than 100 |
|
|
58 | widely used modules, without a fix in sight because |
|
|
59 | p5pers don't care about CPAN anymore. |
|
|
60 | - when canonicalising, only allocate up to 64 hash key |
|
|
61 | pointers on the stack. for larger hashes, use the heap, |
|
|
62 | to avoid using too much stackspace. |
|
|
63 | - discuss the problem with setlocale (reported by a few victims). |
|
|
64 | |
|
|
65 | 2.33 Wed Aug 1 21:03:52 CEST 2012 |
|
|
66 | - internal encode/decode XS wrappers did not expect stack |
|
|
67 | moves caused by callbacks (analyzed and testcase by Jesse Luehrs). |
|
|
68 | - add bencode as to/from option in bin/json_xs. |
|
|
69 | - add -e option to json_xs, and none and string in/out formats. |
|
|
70 | |
|
|
71 | 2.32 Thu Aug 11 19:06:38 CEST 2011 |
|
|
72 | - fix a bug in the initial whitespace accumulation. |
|
|
73 | |
|
|
74 | 2.31 Wed Jul 27 17:53:05 CEST 2011 |
|
|
75 | - don't accumulate initial whitespace in the incremental buffer |
|
|
76 | (this can be useful to allow whitespace-keepalive on a tcp |
|
|
77 | connection without triggering the max_size limit). |
|
|
78 | - properly croak on some invalid inputs that are not strings |
|
|
79 | (e.g. undef) when trying to decode a json text (reported |
|
|
80 | and analyzed by Goro Fuji). |
|
|
81 | |
|
|
82 | 2.3 Wed Aug 18 01:26:47 CEST 2010 |
|
|
83 | - make sure decoder doesn't change the decoding in the incremental |
|
|
84 | parser (testcase provided by Hendrik Schumacher). |
|
|
85 | - applied patch by DaTa for Data::Dumper support in json_xs. |
|
|
86 | - added -t dump support to json_xs, using Data::Dump. |
|
|
87 | - added -f eval support to json_xs. |
|
|
88 | |
|
|
89 | 2.29 Wed Mar 17 02:39:12 CET 2010 |
|
|
90 | - fix a memory leak when callbacks set using filter_json_object |
|
|
91 | or filter_json_single_key_object were called (great testcase |
|
|
92 | by Eric Wilhelm). |
|
|
93 | |
|
|
94 | 2.28 Thu Mar 11 20:30:46 CET 2010 |
|
|
95 | - implement our own atof function - perl's can be orders of |
|
|
96 | magnitudes slower than even the system one. on the positive |
|
|
97 | side, ours seems to be more exact in general than perl's. |
|
|
98 | (testcase provided by Tim Meadowcroft). |
|
|
99 | - clarify floating point conversion issues a bit. |
|
|
100 | - update jpsykes csrf article url. |
|
|
101 | - updated benchmark section - JSON::PP became much faster! |
|
|
102 | |
|
|
103 | 2.27 Thu Jan 7 07:35:08 CET 2010 |
|
|
104 | - support relaxed option inside the incremental parser |
|
|
105 | (testcase provided by IKEGAMI via Makamaka). |
|
|
106 | |
|
|
107 | 2.26 Sat Oct 10 03:26:19 CEST 2009 |
|
|
108 | - big integers could become truncated (based on patch |
|
|
109 | by Strobl Anton). |
|
|
110 | - output format change: indent now adds a final newline, which is |
|
|
111 | more expected and more true to the documentation. |
|
|
112 | |
|
|
113 | 2.25 Sat Aug 8 12:04:41 CEST 2009 |
|
|
114 | - the perl debugger completely breaks lvalue subs - try to work |
|
|
115 | around the issue. |
|
|
116 | - ignore RMAGICAL hashes w.r.t. CANONICAL. |
|
|
117 | - try to work around a possible char signedness issue on aix. |
|
|
118 | - require common sense. |
|
|
119 | |
|
|
120 | 2.24 Sat May 30 08:25:45 CEST 2009 |
|
|
121 | - the incremental parser did not update its parse offset |
|
|
122 | pointer correctly when parsing utf8-strings (nicely |
|
|
123 | debugged by Martin Evans). |
|
|
124 | - appending a non-utf8-string to the incremental parser |
|
|
125 | in utf8 mode failed to upgrade the string. |
|
|
126 | - wording of parse error messages has been improved. |
|
|
127 | |
|
|
128 | 2.232 Sun Feb 22 11:12:25 CET 2009 |
|
|
129 | - use an exponential algorithm to extend strings, to |
|
|
130 | help platforms with bad or abysmal==windows memory |
|
|
131 | allocater performance, at the expense of some memory |
|
|
132 | wastage (use shrink to recover this extra memory). |
|
|
133 | (nicely analysed by Dmitry Karasik). |
|
|
134 | |
|
|
135 | 2.2311 Thu Feb 19 02:12:54 CET 2009 |
|
|
136 | - add a section "JSON and ECMAscript" to explain some |
|
|
137 | incompatibilities between the two (problem was noted by |
|
|
138 | various people). |
|
|
139 | - add t/20_faihu.t. |
|
|
140 | |
|
|
141 | 2.231 Thu Nov 20 04:59:08 CET 2008 |
|
|
142 | - work around 5.10.0 magic bugs where manipulating magic values |
|
|
143 | (such as $1) would permanently damage them as perl would |
|
|
144 | ignore the magicalness, by making a full copy of the string, |
|
|
145 | reported by Dmitry Karasik. |
|
|
146 | - work around spurious warnings under older perl 5.8's. |
|
|
147 | |
|
|
148 | 2.23 Mon Sep 29 05:08:29 CEST 2008 |
|
|
149 | - fix a compilation problem when perl is not using char * as, well, |
|
|
150 | char *. |
|
|
151 | - use PL_hexdigit in favour of rolling our own. |
|
|
152 | |
|
|
153 | 2.2222 Sun Jul 20 18:49:00 CEST 2008 |
|
|
154 | - same game again, broken 5.10 finds yet another assertion |
|
|
155 | failure, and the workaround causes additional runtime warnings. |
|
|
156 | Work around the next assertion AND the warning. 5.10 seriously |
|
|
157 | needs to adjust it's attitude against working code. |
|
|
158 | |
|
|
159 | 2.222 Sat Jul 19 06:15:34 CEST 2008 |
|
|
160 | - you work around one -DDEBUGGING assertion bug in perl 5.10 |
|
|
161 | just to hit the next one. work around this one, too. |
|
|
162 | |
|
|
163 | 2.22 Tue Jul 15 13:26:51 CEST 2008 |
|
|
164 | - allow higher nesting levels in incremental parser. |
|
|
165 | - error out earlier in some cases in the incremental parser |
|
|
166 | (as suggested by Yuval Kogman). |
|
|
167 | - improve incr-parser test (Yuval Kogman). |
|
|
168 | |
|
|
169 | 2.21 Tue Jun 3 08:43:23 CEST 2008 |
|
|
170 | - (hopefully) work around a perl 5.10 bug with -DDEBUGGING. |
|
|
171 | - remove the experimental status of the incremental parser interface. |
|
|
172 | - move =encoding around again, to avoid bugs with search.cpan.org. |
|
|
173 | when can we finally have utf-8 in pod??? |
|
|
174 | - add ->incr_reset method. |
|
|
175 | |
|
|
176 | 2.2 Wed Apr 16 20:37:25 CEST 2008 |
|
|
177 | - lifted the log2 rounding restriction of max_depth and max_size. |
|
|
178 | - make booleans mutable by creating a copy instead of handing out |
|
|
179 | the same scalar (reported by pasha sadri). |
|
|
180 | - added support for incremental json parsing (still EXPERIMENTAL). |
|
|
181 | - implemented and added a json_xs command line utility that can convert |
|
|
182 | from/to a number of serialisation formats - tell me if you need more. |
|
|
183 | - implement allow_unknown/get_allow_unknown methods. |
|
|
184 | - fixed documentation of max_depth w.r.t. higher and equal. |
|
|
185 | - moved down =encoding directive a bit, too much breaks if it's the first |
|
|
186 | pod directive :/. |
|
|
187 | - removed documentation section on other modules, it became somewhat |
|
|
188 | outdated and is nowadays mostly of historical interest. |
|
|
189 | |
|
|
190 | 2.1 Wed Mar 19 23:23:18 CET 2008 |
|
|
191 | - update documentation here and there: add a large section |
|
|
192 | about utf8/latin1/ascii flags, add a security consideration |
|
|
193 | and extend and clarify the JSON and YAML section. |
|
|
194 | - medium speed enhancements when encoding/decoding non-ascii chars. |
|
|
195 | - minor speedup in number encoding case. |
|
|
196 | - extend and clarify the section on incompatibilities |
|
|
197 | between YAML and JSON. |
|
|
198 | - switch to static inline from just inline when using gcc. |
|
|
199 | - add =encoding utf-8 to the manpage, now that perl 5.10 supports it. |
|
|
200 | - fix some issues with UV to JSON conversion of unknown impact. |
|
|
201 | - published the yahoo locals search result used in benchmarks as the |
|
|
202 | original url changes so comparison is impossible. |
|
|
203 | |
|
|
204 | 2.01 Wed Dec 5 11:40:28 CET 2007 |
|
|
205 | - INCOMPATIBLE API CHANGE: to_json and from_json have been |
|
|
206 | renamed to encode_json/decode_json for JSON.pm compatibility. |
|
|
207 | The old functions croak and might be replaced by JSON.pm |
|
|
208 | comaptible versions in some later release. |
|
|
209 | |
|
|
210 | 2.0 Tue Dec 4 11:30:46 CET 2007 |
|
|
211 | - this is supposed to be the first version of JSON::XS |
|
|
212 | compatible with version 2.0+ of the JSON module. |
|
|
213 | Using the JSON module as frontend to JSON::XS should be |
|
|
214 | as fast as using JSON::XS directly, so consider using it |
|
|
215 | instead. |
|
|
216 | - added get_* methods for all "simple" options. |
|
|
217 | - make JSON::XS subclassable. |
|
|
218 | |
|
|
219 | 1.53 Tue Nov 13 23:58:33 CET 2007 |
|
|
220 | - minor doc clarifications. |
|
|
221 | - fixed many doc typos (patch by Thomas L. Shinnick). |
|
|
222 | |
|
|
223 | 1.52 Mon Oct 15 03:22:06 CEST 2007 |
|
|
224 | - remove =encoding pod directive again, it confuses too many pod |
|
|
225 | parsers :/. |
|
|
226 | |
|
|
227 | 1.51 Sat Oct 13 03:55:56 CEST 2007 |
|
|
228 | - encode empty arrays/hashes in a compact way when pretty is enabled. |
|
|
229 | - apparently JSON::XS was used to find some bugs in the |
|
|
230 | JSON_checker testsuite, so add (the corrected) JSON_checker tests to |
|
|
231 | the testsuite. |
|
|
232 | - quite a bit of doc updates/extension. |
|
|
233 | - require 5.8.2, as this seems to be the first unicode-stable version. |
|
|
234 | |
|
|
235 | 1.5 Tue Aug 28 04:05:38 CEST 2007 |
|
|
236 | - add support for tied hashes, based on ideas and testcase by |
|
|
237 | Marcus Holland-Moritz. |
|
|
238 | - implemented relaxed parsing mode where some extensions are being |
|
|
239 | accepted. generation is still JSON-only. |
|
|
240 | |
|
|
241 | 1.44 Wed Aug 22 01:02:44 CEST 2007 |
|
|
242 | - very experimental process-emulation support, slowing everything down. |
|
|
243 | the horribly broken perl threads are still not supported - YMMV. |
|
|
244 | |
|
|
245 | 1.43 Thu Jul 26 13:26:37 CEST 2007 |
|
|
246 | - convert big json numbers exclusively consisting of digits to NV |
|
|
247 | only when there is no loss of precision, otherwise to string. |
|
|
248 | |
|
|
249 | 1.42 Tue Jul 24 00:51:18 CEST 2007 |
|
|
250 | - fix a crash caused by not handling missing array elements |
|
|
251 | (report and testcase by Jay Kuri). |
|
|
252 | |
|
|
253 | 1.41 Tue Jul 10 18:21:44 CEST 2007 |
|
|
254 | - fix compilation with NDEBUG (assert side-effect), |
|
|
255 | affects convert_blessed only. |
|
|
256 | - fix a bug in decode filters calling ENTER; SAVETMPS; |
|
|
257 | one time too often. |
|
|
258 | - catch a typical error in TO_JSON methods. |
|
|
259 | - antique-ised XS.xs again to work with outdated |
|
|
260 | C compilers (windows...). |
|
|
261 | |
|
|
262 | 1.4 Mon Jul 2 10:06:30 CEST 2007 |
|
|
263 | - add convert_blessed setting. |
|
|
264 | - encode did not catch all blessed objects, encoding their |
|
|
265 | contents in most cases. This has been fixed by introducing |
|
|
266 | the allow_blessed setting. |
|
|
267 | - added filter_json_object and filter_json_single_key_object |
|
|
268 | settings that specify a callback to be called when |
|
|
269 | all/specific json objects are encountered. |
|
|
270 | - assume that most object keys are simple ascii words and |
|
|
271 | optimise this case, penalising the general case. This can |
|
|
272 | speed up decoding by 30% in typical cases and gives |
|
|
273 | a smaller and faster perl hash. |
|
|
274 | - implemented simpleminded, optional resource size checking |
|
|
275 | in decode_json. |
|
|
276 | - remove objToJson/jsonToObj aliases, as the next version |
|
|
277 | of JSON will not have them either. |
|
|
278 | - bit the bullet and converted the very simple json object |
|
|
279 | into a more complex one. |
|
|
280 | - work around a bug where perl wrongly claims an integer |
|
|
281 | is not an integer. |
|
|
282 | - unbundle JSON::XS::Boolean into own pm file so Storable |
|
|
283 | and similar modules can resolve the overloading when thawing. |
|
|
284 | |
|
|
285 | 1.3 Sun Jun 24 01:55:02 CEST 2007 |
|
|
286 | - make JSON::XS::true and false special overloaded objects |
|
|
287 | and return those instead of 1 and 0 for those json atoms |
|
|
288 | (JSON::PP compatibility is NOT achieved yet). |
|
|
289 | - add JSON::XS::is_bool predicate to test for those special |
|
|
290 | values. |
|
|
291 | - add a reference to |
|
|
292 | http://jpsykes.com/47/practical-csrf-and-json-security. |
|
|
293 | - removed require 5.8.8 again, it is just not very expert-friendly. |
|
|
294 | Also try to be more compatible with slightly older versions, |
|
|
295 | which are not recommended (because they are buggy). |
|
|
296 | |
|
|
297 | 1.24 Mon Jun 11 05:40:49 CEST 2007 |
|
|
298 | - added informative section on JSON-as-YAML. |
|
|
299 | - get rid of some c99-isms again. |
|
|
300 | - localise dec->cur in decode_str, speeding up |
|
|
301 | string decoding considerably (>15% on my amd64 + gcc). |
|
|
302 | - increased SHORT_STRING_LEN to 16kb: stack space is |
|
|
303 | usually plenty, and this actually saves memory |
|
|
304 | when !shrinking as short strings will fit perfectly. |
|
|
305 | |
|
|
306 | 1.23 Wed Jun 6 20:13:06 CEST 2007 |
|
|
307 | - greatly improved small integer encoding and decoding speed. |
|
|
308 | - implement a number of µ-optimisations. |
|
|
309 | - updated benchmarks. |
|
|
310 | |
|
|
311 | 1.22 Thu May 24 00:07:25 CEST 2007 |
|
|
312 | - require 5.8.8 explicitly as older perls do not seem to offer |
|
|
313 | the required macros. |
|
|
314 | - possibly made it compile on so-called C compilers by microsoft. |
|
|
315 | |
|
|
316 | 1.21 Wed May 9 18:40:32 CEST 2007 |
|
|
317 | - character offset reported for trailing garbage was random. |
|
|
318 | |
|
|
319 | 1.2 Wed May 9 18:35:01 CEST 2007 |
|
|
320 | - decode did not work with magical scalars (doh!). |
|
|
321 | - added latin1 flag to produce JSON texts in the latin1 subset |
|
|
322 | of unicode. |
|
|
323 | - flag trailing garbage as error. |
|
|
324 | - new decode_prefix method that returns the number |
|
|
325 | of characters consumed by a decode. |
|
|
326 | - max octets/char in perls UTF-X is actually 13, not 11, |
|
|
327 | as pointed out by Glenn Linderman. |
|
|
328 | - fixed typoe reported by YAMASHINA Hio. |
|
|
329 | |
|
|
330 | 1.11 Mon Apr 9 07:05:49 CEST 2007 |
|
|
331 | - properly 0-terminate sv's returned by encode to help |
|
|
332 | C libraries that expect that 0 to be there. |
|
|
333 | - partially "port" JSON from C to microsofts fucking broken |
|
|
334 | pseudo-C. They should be burned to the ground for pissing |
|
|
335 | on standards. And I should be stoned for even trying to |
|
|
336 | support this filthy excuse for a c compiler. |
|
|
337 | |
|
|
338 | 1.1 Wed Apr 4 01:45:00 CEST 2007 |
|
|
339 | - clarify documentation (pointed out by Quinn Weaver). |
|
|
340 | - decode_utf8 sometimes did not correctly flag errors, |
|
|
341 | leading to segfaults. |
|
|
342 | - further reduced default nesting depth to 512 due to the test |
|
|
343 | failure by that anonymous "chris" whose e-mail address seems |
|
|
344 | to be impossible to get. Tests on other freebsd systems indicate |
|
|
345 | that this is likely a problem in his/her configuration and not this |
|
|
346 | module. |
|
|
347 | - renamed json => JSON in error messages. |
|
|
348 | - corrected the character offset in some error messages. |
|
|
349 | |
|
|
350 | 1.01 Sat Mar 31 16:15:40 CEST 2007 |
|
|
351 | - do not segfault when from_json/decode gets passed |
|
|
352 | a non-string object (reported by Florian Ragwitz). |
|
|
353 | This has no effect on normal operation. |
|
|
354 | |
|
|
355 | 1.0 Thu Mar 29 04:43:34 CEST 2007 |
|
|
356 | - the long awaited (by me) 1.0 version. |
|
|
357 | - add \0 (JSON::XS::false) and \1 (JSON::XS::true) mappings to JSON |
|
|
358 | true and false. |
| 7 | - add some more notes to shrink, as suggested by Alex Efros. |
359 | - add some more notes to shrink, as suggested by Alex Efros. |
|
|
360 | - improve testsuite. |
|
|
361 | - halve the default nesting depth limit, to hopefully make it |
|
|
362 | work on Freebsd (unfortunately, the cpan tester did not |
|
|
363 | send me his report, so I cannot ask about the stack limit on fbsd). |
| 8 | |
364 | |
| 9 | 0.8 Mon Mar 26 00:10:48 CEST 2007 |
365 | 0.8 Mon Mar 26 00:10:48 CEST 2007 |
| 10 | - fix a memleak when decoding hashes. |
366 | - fix a memleak when decoding hashes. |
| 11 | - export jsonToBj and objToJson as aliases |
367 | - export jsonToBj and objToJson as aliases |
| 12 | to to_json and from_json, to reduce incompatibilities |
368 | to to_json and from_json, to reduce incompatibilities |
| … | |
… | |
| 59 | case the called functions do... stuff. |
415 | case the called functions do... stuff. |
| 60 | - croak when encoding to ascii and an out-of-range |
416 | - croak when encoding to ascii and an out-of-range |
| 61 | (non-unicode) codepoint is encountered. |
417 | (non-unicode) codepoint is encountered. |
| 62 | |
418 | |
| 63 | 0.2 Fri Mar 23 00:23:34 CET 2007 |
419 | 0.2 Fri Mar 23 00:23:34 CET 2007 |
| 64 | - the "could not sleep without debuggign release". |
420 | - the "could not sleep without debugging release". |
| 65 | it should basically work now, with many bugs as |
421 | it should basically work now, with many bugs as |
| 66 | no production tests have been run yet. |
422 | no production tests have been run yet. |
| 67 | - added more testcases. |
423 | - added more testcases. |
| 68 | - the expected shitload of bugfixes. |
424 | - the expected shitload of bugfixes. |
| 69 | - handle utf8 flag correctly in decode. |
425 | - handle utf8 flag correctly in decode. |
