ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/JSON-XS/Changes
Revision: 1.192
Committed: Wed Mar 6 07:32:06 2019 UTC (5 years, 2 months ago) by root
Branch: MAIN
CVS Tags: rel-4_02
Changes since 1.191: +5 -4 lines
Log Message:
4.02

File Contents

# Content
1 Revision history for Perl extension JSON::XS
2
3 TODO: maybe detect and croak on more invalid inputs (e.g. +-inf/nan)
4 TODO: maybe avoid the reblessing and better support readonly objects.
5 TODO: how to cope with tagged values and standard json decoders
6 TODO: investigate magic (Eric Brine)
7 TODO: [PATCH] Types::Serialiser: Inline true(), false() and error() functions
8
9 4.02 Wed Mar 6 08:31:24 CET 2019
10 - undo the fix from 4.01, it breaks more things than it fixes
11 (another tetscase by Wesley Schwengle).
12 - try a proper fix this time.
13
14 4.01 Sun Feb 24 05:03:30 CET 2019
15 - fix some stack corruption caused mostly when calling methods
16 in list context (testcase by Wesley Schwengle).
17
18 4.0 Fri Nov 16 00:06:54 CET 2018
19 - SECURITY IMPLICATION: this release enables allow_nonref by default
20 for compatibnility with RFC 7159 and newer. See "old" vs. "new"
21 JSON under SECURITY CONSIDERATIONS.
22 - reworked the "old" vs. "new" JSON section.
23 - add ->boolean_values to provide the values to which booleans
24 decode (requested by Aristotle Pagaltzis).
25 - decode would wrongly accept ASCII NUL characters instead of
26 reporting them as trailing garbage.
27 - work around what smells like a perl bug w.r.t. exceptions
28 thrown in callbacks.
29 - incremental parser now more or less respects allow_nonref.
30 - json_xs json-pretty now enables canonical mode.
31 - add documentation section about I-JSON.
32 - minor documentation fixes/updates.
33
34 3.04 Thu Aug 17 04:30:47 CEST 2017
35 - change exponential realloc algorithm on encoding and string decoding to be
36 really exponential (this helps slow/debugging allocators such as libumem)
37 (reported by Matthew Horsfall).
38 - string encoding would needlessly overallocate output space
39 (testcase by Matthew Horsfall).
40 - be very paranoid about extending buffer lengths and croak if buffers get too large,
41 which might (or might not) improve security.
42 - add cbor-packed type to json_xs.
43 - switch from YAML to YAML::XS in json_xs, as YAML is way too buggy and outdated.
44
45 3.03 Wed Nov 16 20:20:59 CET 2016
46 - fix a bug introduced by a perl bug workaround that would cause
47 incremental parsing to fail with a sv_chop panic.
48 - json_xs: toformat failure error message fixed.
49 - json_xs: allow cyclic data structures in CBOR.
50
51 3.02 Fri Feb 26 22:45:20 CET 2016
52 - allow_nonref now affects booleans (\1, $Types::Serialiser::Boolean)
53 as well (reported by Alex Efros).
54 - allow literal tabs in strings in relaxed mode (patch by
55 lubo.rintel@gooddata.com).
56 - support "cbor" format in json_xs tool.
57 - support (and fix) calling encode and decode in list context
58 (reported by Вадим Власов).
59 - work around a bug in older perls crashing when presented
60 with shared hash keys (Reini Urban).
61 - use stability canary.
62
63 3.01 Tue Oct 29 16:55:15 CET 2013
64 - backport to perls < 5.18 (reported by Paul Howarth).
65
66 3.0 Tue Oct 29 01:35:37 CET 2013
67 - implemented an object tagging extension (using the
68 Types::Serialiser serialisation protocol).
69 - reworked the documentation regarding object serialisation,
70 add a new OBJECT SERIALISATION section that explains th
71 whole process.
72 - new setting: allow_tags.
73 - switch to Types::Serialiser booleans.
74 - remove to_json/from_json.
75 - other minor improvements to the documentation.
76
77 2.34 Thu May 23 11:30:34 CEST 2013
78 - work around bugs in perl 5.18 breaking more than 100
79 widely used modules, without a fix in sight because
80 p5pers don't care about CPAN anymore.
81 - when canonicalising, only allocate up to 64 hash key
82 pointers on the stack. for larger hashes, use the heap,
83 to avoid using too much stackspace.
84 - discuss the problem with setlocale (reported by a few victims).
85
86 2.33 Wed Aug 1 21:03:52 CEST 2012
87 - internal encode/decode XS wrappers did not expect stack
88 moves caused by callbacks (analyzed and testcase by Jesse Luehrs).
89 - add bencode as to/from option in bin/json_xs.
90 - add -e option to json_xs, and none and string in/out formats.
91
92 2.32 Thu Aug 11 19:06:38 CEST 2011
93 - fix a bug in the initial whitespace accumulation.
94
95 2.31 Wed Jul 27 17:53:05 CEST 2011
96 - don't accumulate initial whitespace in the incremental buffer
97 (this can be useful to allow whitespace-keepalive on a tcp
98 connection without triggering the max_size limit).
99 - properly croak on some invalid inputs that are not strings
100 (e.g. undef) when trying to decode a json text (reported
101 and analyzed by Goro Fuji).
102
103 2.3 Wed Aug 18 01:26:47 CEST 2010
104 - make sure decoder doesn't change the decoding in the incremental
105 parser (testcase provided by Hendrik Schumacher).
106 - applied patch by DaTa for Data::Dumper support in json_xs.
107 - added -t dump support to json_xs, using Data::Dump.
108 - added -f eval support to json_xs.
109
110 2.29 Wed Mar 17 02:39:12 CET 2010
111 - fix a memory leak when callbacks set using filter_json_object
112 or filter_json_single_key_object were called (great testcase
113 by Eric Wilhelm).
114
115 2.28 Thu Mar 11 20:30:46 CET 2010
116 - implement our own atof function - perl's can be orders of
117 magnitudes slower than even the system one. on the positive
118 side, ours seems to be more exact in general than perl's.
119 (testcase provided by Tim Meadowcroft).
120 - clarify floating point conversion issues a bit.
121 - update jpsykes csrf article url.
122 - updated benchmark section - JSON::PP became much faster!
123
124 2.27 Thu Jan 7 07:35:08 CET 2010
125 - support relaxed option inside the incremental parser
126 (testcase provided by IKEGAMI via Makamaka).
127
128 2.26 Sat Oct 10 03:26:19 CEST 2009
129 - big integers could become truncated (based on patch
130 by Strobl Anton).
131 - output format change: indent now adds a final newline, which is
132 more expected and more true to the documentation.
133
134 2.25 Sat Aug 8 12:04:41 CEST 2009
135 - the perl debugger completely breaks lvalue subs - try to work
136 around the issue.
137 - ignore RMAGICAL hashes w.r.t. CANONICAL.
138 - try to work around a possible char signedness issue on aix.
139 - require common sense.
140
141 2.24 Sat May 30 08:25:45 CEST 2009
142 - the incremental parser did not update its parse offset
143 pointer correctly when parsing utf8-strings (nicely
144 debugged by Martin Evans).
145 - appending a non-utf8-string to the incremental parser
146 in utf8 mode failed to upgrade the string.
147 - wording of parse error messages has been improved.
148
149 2.232 Sun Feb 22 11:12:25 CET 2009
150 - use an exponential algorithm to extend strings, to
151 help platforms with bad or abysmal==windows memory
152 allocater performance, at the expense of some memory
153 wastage (use shrink to recover this extra memory).
154 (nicely analysed by Dmitry Karasik).
155
156 2.2311 Thu Feb 19 02:12:54 CET 2009
157 - add a section "JSON and ECMAscript" to explain some
158 incompatibilities between the two (problem was noted by
159 various people).
160 - add t/20_faihu.t.
161
162 2.231 Thu Nov 20 04:59:08 CET 2008
163 - work around 5.10.0 magic bugs where manipulating magic values
164 (such as $1) would permanently damage them as perl would
165 ignore the magicalness, by making a full copy of the string,
166 reported by Dmitry Karasik.
167 - work around spurious warnings under older perl 5.8's.
168
169 2.23 Mon Sep 29 05:08:29 CEST 2008
170 - fix a compilation problem when perl is not using char * as, well,
171 char *.
172 - use PL_hexdigit in favour of rolling our own.
173
174 2.2222 Sun Jul 20 18:49:00 CEST 2008
175 - same game again, broken 5.10 finds yet another assertion
176 failure, and the workaround causes additional runtime warnings.
177 Work around the next assertion AND the warning. 5.10 seriously
178 needs to adjust it's attitude against working code.
179
180 2.222 Sat Jul 19 06:15:34 CEST 2008
181 - you work around one -DDEBUGGING assertion bug in perl 5.10
182 just to hit the next one. work around this one, too.
183
184 2.22 Tue Jul 15 13:26:51 CEST 2008
185 - allow higher nesting levels in incremental parser.
186 - error out earlier in some cases in the incremental parser
187 (as suggested by Yuval Kogman).
188 - improve incr-parser test (Yuval Kogman).
189
190 2.21 Tue Jun 3 08:43:23 CEST 2008
191 - (hopefully) work around a perl 5.10 bug with -DDEBUGGING.
192 - remove the experimental status of the incremental parser interface.
193 - move =encoding around again, to avoid bugs with search.cpan.org.
194 when can we finally have utf-8 in pod???
195 - add ->incr_reset method.
196
197 2.2 Wed Apr 16 20:37:25 CEST 2008
198 - lifted the log2 rounding restriction of max_depth and max_size.
199 - make booleans mutable by creating a copy instead of handing out
200 the same scalar (reported by pasha sadri).
201 - added support for incremental json parsing (still EXPERIMENTAL).
202 - implemented and added a json_xs command line utility that can convert
203 from/to a number of serialisation formats - tell me if you need more.
204 - implement allow_unknown/get_allow_unknown methods.
205 - fixed documentation of max_depth w.r.t. higher and equal.
206 - moved down =encoding directive a bit, too much breaks if it's the first
207 pod directive :/.
208 - removed documentation section on other modules, it became somewhat
209 outdated and is nowadays mostly of historical interest.
210
211 2.1 Wed Mar 19 23:23:18 CET 2008
212 - update documentation here and there: add a large section
213 about utf8/latin1/ascii flags, add a security consideration
214 and extend and clarify the JSON and YAML section.
215 - medium speed enhancements when encoding/decoding non-ascii chars.
216 - minor speedup in number encoding case.
217 - extend and clarify the section on incompatibilities
218 between YAML and JSON.
219 - switch to static inline from just inline when using gcc.
220 - add =encoding utf-8 to the manpage, now that perl 5.10 supports it.
221 - fix some issues with UV to JSON conversion of unknown impact.
222 - published the yahoo locals search result used in benchmarks as the
223 original url changes so comparison is impossible.
224
225 2.01 Wed Dec 5 11:40:28 CET 2007
226 - INCOMPATIBLE API CHANGE: to_json and from_json have been
227 renamed to encode_json/decode_json for JSON.pm compatibility.
228 The old functions croak and might be replaced by JSON.pm
229 comaptible versions in some later release.
230
231 2.0 Tue Dec 4 11:30:46 CET 2007
232 - this is supposed to be the first version of JSON::XS
233 compatible with version 2.0+ of the JSON module.
234 Using the JSON module as frontend to JSON::XS should be
235 as fast as using JSON::XS directly, so consider using it
236 instead.
237 - added get_* methods for all "simple" options.
238 - make JSON::XS subclassable.
239
240 1.53 Tue Nov 13 23:58:33 CET 2007
241 - minor doc clarifications.
242 - fixed many doc typos (patch by Thomas L. Shinnick).
243
244 1.52 Mon Oct 15 03:22:06 CEST 2007
245 - remove =encoding pod directive again, it confuses too many pod
246 parsers :/.
247
248 1.51 Sat Oct 13 03:55:56 CEST 2007
249 - encode empty arrays/hashes in a compact way when pretty is enabled.
250 - apparently JSON::XS was used to find some bugs in the
251 JSON_checker testsuite, so add (the corrected) JSON_checker tests to
252 the testsuite.
253 - quite a bit of doc updates/extension.
254 - require 5.8.2, as this seems to be the first unicode-stable version.
255
256 1.5 Tue Aug 28 04:05:38 CEST 2007
257 - add support for tied hashes, based on ideas and testcase by
258 Marcus Holland-Moritz.
259 - implemented relaxed parsing mode where some extensions are being
260 accepted. generation is still JSON-only.
261
262 1.44 Wed Aug 22 01:02:44 CEST 2007
263 - very experimental process-emulation support, slowing everything down.
264 the horribly broken perl threads are still not supported - YMMV.
265
266 1.43 Thu Jul 26 13:26:37 CEST 2007
267 - convert big json numbers exclusively consisting of digits to NV
268 only when there is no loss of precision, otherwise to string.
269
270 1.42 Tue Jul 24 00:51:18 CEST 2007
271 - fix a crash caused by not handling missing array elements
272 (report and testcase by Jay Kuri).
273
274 1.41 Tue Jul 10 18:21:44 CEST 2007
275 - fix compilation with NDEBUG (assert side-effect),
276 affects convert_blessed only.
277 - fix a bug in decode filters calling ENTER; SAVETMPS;
278 one time too often.
279 - catch a typical error in TO_JSON methods.
280 - antique-ised XS.xs again to work with outdated
281 C compilers (windows...).
282
283 1.4 Mon Jul 2 10:06:30 CEST 2007
284 - add convert_blessed setting.
285 - encode did not catch all blessed objects, encoding their
286 contents in most cases. This has been fixed by introducing
287 the allow_blessed setting.
288 - added filter_json_object and filter_json_single_key_object
289 settings that specify a callback to be called when
290 all/specific json objects are encountered.
291 - assume that most object keys are simple ascii words and
292 optimise this case, penalising the general case. This can
293 speed up decoding by 30% in typical cases and gives
294 a smaller and faster perl hash.
295 - implemented simpleminded, optional resource size checking
296 in decode_json.
297 - remove objToJson/jsonToObj aliases, as the next version
298 of JSON will not have them either.
299 - bit the bullet and converted the very simple json object
300 into a more complex one.
301 - work around a bug where perl wrongly claims an integer
302 is not an integer.
303 - unbundle JSON::XS::Boolean into own pm file so Storable
304 and similar modules can resolve the overloading when thawing.
305
306 1.3 Sun Jun 24 01:55:02 CEST 2007
307 - make JSON::XS::true and false special overloaded objects
308 and return those instead of 1 and 0 for those json atoms
309 (JSON::PP compatibility is NOT achieved yet).
310 - add JSON::XS::is_bool predicate to test for those special
311 values.
312 - add a reference to
313 http://jpsykes.com/47/practical-csrf-and-json-security.
314 - removed require 5.8.8 again, it is just not very expert-friendly.
315 Also try to be more compatible with slightly older versions,
316 which are not recommended (because they are buggy).
317
318 1.24 Mon Jun 11 05:40:49 CEST 2007
319 - added informative section on JSON-as-YAML.
320 - get rid of some c99-isms again.
321 - localise dec->cur in decode_str, speeding up
322 string decoding considerably (>15% on my amd64 + gcc).
323 - increased SHORT_STRING_LEN to 16kb: stack space is
324 usually plenty, and this actually saves memory
325 when !shrinking as short strings will fit perfectly.
326
327 1.23 Wed Jun 6 20:13:06 CEST 2007
328 - greatly improved small integer encoding and decoding speed.
329 - implement a number of µ-optimisations.
330 - updated benchmarks.
331
332 1.22 Thu May 24 00:07:25 CEST 2007
333 - require 5.8.8 explicitly as older perls do not seem to offer
334 the required macros.
335 - possibly made it compile on so-called C compilers by microsoft.
336
337 1.21 Wed May 9 18:40:32 CEST 2007
338 - character offset reported for trailing garbage was random.
339
340 1.2 Wed May 9 18:35:01 CEST 2007
341 - decode did not work with magical scalars (doh!).
342 - added latin1 flag to produce JSON texts in the latin1 subset
343 of unicode.
344 - flag trailing garbage as error.
345 - new decode_prefix method that returns the number
346 of characters consumed by a decode.
347 - max octets/char in perls UTF-X is actually 13, not 11,
348 as pointed out by Glenn Linderman.
349 - fixed typoe reported by YAMASHINA Hio.
350
351 1.11 Mon Apr 9 07:05:49 CEST 2007
352 - properly 0-terminate sv's returned by encode to help
353 C libraries that expect that 0 to be there.
354 - partially "port" JSON from C to microsofts fucking broken
355 pseudo-C. They should be burned to the ground for pissing
356 on standards. And I should be stoned for even trying to
357 support this filthy excuse for a c compiler.
358
359 1.1 Wed Apr 4 01:45:00 CEST 2007
360 - clarify documentation (pointed out by Quinn Weaver).
361 - decode_utf8 sometimes did not correctly flag errors,
362 leading to segfaults.
363 - further reduced default nesting depth to 512 due to the test
364 failure by that anonymous "chris" whose e-mail address seems
365 to be impossible to get. Tests on other freebsd systems indicate
366 that this is likely a problem in his/her configuration and not this
367 module.
368 - renamed json => JSON in error messages.
369 - corrected the character offset in some error messages.
370
371 1.01 Sat Mar 31 16:15:40 CEST 2007
372 - do not segfault when from_json/decode gets passed
373 a non-string object (reported by Florian Ragwitz).
374 This has no effect on normal operation.
375
376 1.0 Thu Mar 29 04:43:34 CEST 2007
377 - the long awaited (by me) 1.0 version.
378 - add \0 (JSON::XS::false) and \1 (JSON::XS::true) mappings to JSON
379 true and false.
380 - add some more notes to shrink, as suggested by Alex Efros.
381 - improve testsuite.
382 - halve the default nesting depth limit, to hopefully make it
383 work on Freebsd (unfortunately, the cpan tester did not
384 send me his report, so I cannot ask about the stack limit on fbsd).
385
386 0.8 Mon Mar 26 00:10:48 CEST 2007
387 - fix a memleak when decoding hashes.
388 - export jsonToBj and objToJson as aliases
389 to to_json and from_json, to reduce incompatibilities
390 between JSON/JSON::PC and JSON::XS. (experimental).
391 - implement a maximum nesting depth for both en- and de-coding.
392 - added a security considerations sections.
393
394 0.7 Sun Mar 25 01:46:30 CET 2007
395 - code cleanup.
396 - fix a memory overflow bug when indenting.
397 - pretty-printing now up to 15% faster.
398 - improve decoding speed of strings by
399 up to 50% by specialcasing short strings.
400 - further decoding speedups for strings using
401 lots of \u escapes.
402 - improve utf8 decoding speed for U+80 .. U+7FF.
403
404 0.5 Sat Mar 24 20:41:51 CET 2007
405 - added the UTF-16 encoding example hinted at in previous
406 versions.
407 - minor documentation fixes.
408 - fix a bug in and optimise canonicalising fastpath
409 (reported by Craig Manley).
410 - remove a subtest that breaks with bleadperl (reported
411 by Andreas König).
412
413 0.31 Sat Mar 24 02:14:34 CET 2007
414 - documentation updates.
415 - do some casting to hopefully fix Andreas' problem.
416 - nuke bogus json rpc stuff.
417
418 0.3 Fri Mar 23 19:33:21 CET 2007
419 - remove spurious PApp::Util reference (John McNamara).
420 - adapted lots of tests from other json modules
421 (idea by Chris Carline).
422 - documented mapping from json to perl and vice versa.
423 - improved the documentation by adding more examples.
424 - added short escaping forms, reducing the created
425 json texts a bit.
426 - added shrink flag.
427 - when flag methods are called without enable argument
428 they will by default enable their flag.
429 - considerably improved string encoding speed (at least
430 with gcc 4).
431 - added a test that covers lots of different characters.
432 - clarified some error messages.
433 - error messages now use correct character offset
434 with F_UTF8.
435 - improve the "no bytes" and "no warnings" hacks in
436 case the called functions do... stuff.
437 - croak when encoding to ascii and an out-of-range
438 (non-unicode) codepoint is encountered.
439
440 0.2 Fri Mar 23 00:23:34 CET 2007
441 - the "could not sleep without debugging release".
442 it should basically work now, with many bugs as
443 no production tests have been run yet.
444 - added more testcases.
445 - the expected shitload of bugfixes.
446 - handle utf8 flag correctly in decode.
447 - fix segfault in decoder.
448 - utf8n_to_uvuni sets retlen to -1, but retlen is an
449 unsigned types (argh).
450 - fix decoding of utf-8 strings.
451 - improved error diagnostics.
452 - fix decoding of 'null'.
453 - fix parsing of empty array/hashes
454 - silence warnings when we prepare the croak message.
455
456 0.1 Thu Mar 22 22:13:43 CET 2007
457 - first release, very untested, basically just to claim
458 the namespace.
459
460 0.01 Thu Mar 22 06:08:12 CET 2007
461 - original version; cloned from Convert-Scalar
462