ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/JSON-XS/Changes
Revision: 1.172
Committed: Fri Sep 23 07:45:58 2016 UTC (7 years, 7 months ago) by root
Branch: MAIN
Changes since 1.171: +1 -0 lines
Log Message:
*** empty log message ***

File Contents

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