ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/JSON-XS/Changes
(Generate patch)

Comparing JSON-XS/Changes (file contents):
Revision 1.36 by root, Wed May 9 15:34:04 2007 UTC vs.
Revision 1.175 by root, Wed Nov 16 19:21:53 2016 UTC

1Revision history for Perl extension JSON::XS 1Revision history for Perl extension JSON::XS
2 2
3TODO: maybe use some overload magic to decode true and false into something 3TODO: maybe detetc and croak on more invalid inputs (e.g. +-inf/nan)
4 not 0 or 1 in perl. or use attached magic? 4TODO: maybe avoid the reblessing and better support readonly objects.
5TODO: get rid of dynamic arrays to support so-called C compilers 5TODO: http://stevehanov.ca/blog/index.php?id=104 compression
6 by fucking microsoft that do not do C. argl, snprintf is also 6TODO: how to cope with tagged values and standard json decoders
7 missing. 7TODO: investigate magic (Eric Brine)
8TODO: latin1 flag working similar to ascii. 8TODO: rfc7464
9TODO: Victor Efimov
9 10
101.12 113.03 Wed Nov 16 20:20:59 CET 2016
11 - fixed typoe reported by YAMASHINA Hio. 12 - fix a bug introduced by a perl bug workaround that would cause
13 incremental parsing to fail with a sv_chop panic.
14 - json_xs: toformat failure error message fixed.
15 - json_xs: allow cyclic data structures in CBOR.
16
173.02 Fri Feb 26 22:45:20 CET 2016
18 - allow_nonref now affects booleans (\1, $Types::Serialiser::Boolean)
19 as well (reported by Alex Efros).
20 - allow literal tabs in strings in relaxed mode (patch by
21 lubo.rintel@gooddata.com).
22 - support "cbor" format in json_xs tool.
23 - support (and fix) calling encode and decode in list context
24 (reported by Вадим Власов).
25 - work around a bug in older perls crashing when presented
26 with shared hash keys (Reini Urban).
27 - use stability canary.
28
293.01 Tue Oct 29 16:55:15 CET 2013
30 - backport to perls < 5.18 (reported by Paul Howarth).
31
323.0 Tue Oct 29 01:35:37 CET 2013
33 - implemented an object tagging extension (using the
34 Types::Serialiser serialisation protocol).
35 - reworked the documentation regarding object serialisation,
36 add a new OBJECT SERIALISATION section that explains th
37 whole process.
38 - new setting: allow_tags.
39 - switch to Types::Serialiser booleans.
40 - remove to_json/from_json.
41 - other minor improvements to the documentation.
42
432.34 Thu May 23 11:30:34 CEST 2013
44 - work around bugs in perl 5.18 breaking more than 100
45 widely used modules, without a fix in sight because
46 p5pers don't care about CPAN anymore.
47 - when canonicalising, only allocate up to 64 hash key
48 pointers on the stack. for larger hashes, use the heap,
49 to avoid using too much stackspace.
50 - discuss the problem with setlocale (reported by a few victims).
51
522.33 Wed Aug 1 21:03:52 CEST 2012
53 - internal encode/decode XS wrappers did not expect stack
54 moves caused by callbacks (analyzed and testcase by Jesse Luehrs).
55 - add bencode as to/from option in bin/json_xs.
56 - add -e option to json_xs, and none and string in/out formats.
57
582.32 Thu Aug 11 19:06:38 CEST 2011
59 - fix a bug in the initial whitespace accumulation.
60
612.31 Wed Jul 27 17:53:05 CEST 2011
62 - don't accumulate initial whitespace in the incremental buffer
63 (this can be useful to allow whitespace-keepalive on a tcp
64 connection without triggering the max_size limit).
65 - properly croak on some invalid inputs that are not strings
66 (e.g. undef) when trying to decode a json text (reported
67 and analyzed by Goro Fuji).
68
692.3 Wed Aug 18 01:26:47 CEST 2010
70 - make sure decoder doesn't change the decoding in the incremental
71 parser (testcase provided by Hendrik Schumacher).
72 - applied patch by DaTa for Data::Dumper support in json_xs.
73 - added -t dump support to json_xs, using Data::Dump.
74 - added -f eval support to json_xs.
75
762.29 Wed Mar 17 02:39:12 CET 2010
77 - fix a memory leak when callbacks set using filter_json_object
78 or filter_json_single_key_object were called (great testcase
79 by Eric Wilhelm).
80
812.28 Thu Mar 11 20:30:46 CET 2010
82 - implement our own atof function - perl's can be orders of
83 magnitudes slower than even the system one. on the positive
84 side, ours seems to be more exact in general than perl's.
85 (testcase provided by Tim Meadowcroft).
86 - clarify floating point conversion issues a bit.
87 - update jpsykes csrf article url.
88 - updated benchmark section - JSON::PP became much faster!
89
902.27 Thu Jan 7 07:35:08 CET 2010
91 - support relaxed option inside the incremental parser
92 (testcase provided by IKEGAMI via Makamaka).
93
942.26 Sat Oct 10 03:26:19 CEST 2009
95 - big integers could become truncated (based on patch
96 by Strobl Anton).
97 - output format change: indent now adds a final newline, which is
98 more expected and more true to the documentation.
99
1002.25 Sat Aug 8 12:04:41 CEST 2009
101 - the perl debugger completely breaks lvalue subs - try to work
102 around the issue.
103 - ignore RMAGICAL hashes w.r.t. CANONICAL.
104 - try to work around a possible char signedness issue on aix.
105 - require common sense.
106
1072.24 Sat May 30 08:25:45 CEST 2009
108 - the incremental parser did not update its parse offset
109 pointer correctly when parsing utf8-strings (nicely
110 debugged by Martin Evans).
111 - appending a non-utf8-string to the incremental parser
112 in utf8 mode failed to upgrade the string.
113 - wording of parse error messages has been improved.
114
1152.232 Sun Feb 22 11:12:25 CET 2009
116 - use an exponential algorithm to extend strings, to
117 help platforms with bad or abysmal==windows memory
118 allocater performance, at the expense of some memory
119 wastage (use shrink to recover this extra memory).
120 (nicely analysed by Dmitry Karasik).
121
1222.2311 Thu Feb 19 02:12:54 CET 2009
123 - add a section "JSON and ECMAscript" to explain some
124 incompatibilities between the two (problem was noted by
125 various people).
126 - add t/20_faihu.t.
127
1282.231 Thu Nov 20 04:59:08 CET 2008
129 - work around 5.10.0 magic bugs where manipulating magic values
130 (such as $1) would permanently damage them as perl would
131 ignore the magicalness, by making a full copy of the string,
132 reported by Dmitry Karasik.
133 - work around spurious warnings under older perl 5.8's.
134
1352.23 Mon Sep 29 05:08:29 CEST 2008
136 - fix a compilation problem when perl is not using char * as, well,
137 char *.
138 - use PL_hexdigit in favour of rolling our own.
139
1402.2222 Sun Jul 20 18:49:00 CEST 2008
141 - same game again, broken 5.10 finds yet another assertion
142 failure, and the workaround causes additional runtime warnings.
143 Work around the next assertion AND the warning. 5.10 seriously
144 needs to adjust it's attitude against working code.
145
1462.222 Sat Jul 19 06:15:34 CEST 2008
147 - you work around one -DDEBUGGING assertion bug in perl 5.10
148 just to hit the next one. work around this one, too.
149
1502.22 Tue Jul 15 13:26:51 CEST 2008
151 - allow higher nesting levels in incremental parser.
152 - error out earlier in some cases in the incremental parser
153 (as suggested by Yuval Kogman).
154 - improve incr-parser test (Yuval Kogman).
155
1562.21 Tue Jun 3 08:43:23 CEST 2008
157 - (hopefully) work around a perl 5.10 bug with -DDEBUGGING.
158 - remove the experimental status of the incremental parser interface.
159 - move =encoding around again, to avoid bugs with search.cpan.org.
160 when can we finally have utf-8 in pod???
161 - add ->incr_reset method.
162
1632.2 Wed Apr 16 20:37:25 CEST 2008
164 - lifted the log2 rounding restriction of max_depth and max_size.
165 - make booleans mutable by creating a copy instead of handing out
166 the same scalar (reported by pasha sadri).
167 - added support for incremental json parsing (still EXPERIMENTAL).
168 - implemented and added a json_xs command line utility that can convert
169 from/to a number of serialisation formats - tell me if you need more.
170 - implement allow_unknown/get_allow_unknown methods.
171 - fixed documentation of max_depth w.r.t. higher and equal.
172 - moved down =encoding directive a bit, too much breaks if it's the first
173 pod directive :/.
174 - removed documentation section on other modules, it became somewhat
175 outdated and is nowadays mostly of historical interest.
176
1772.1 Wed Mar 19 23:23:18 CET 2008
178 - update documentation here and there: add a large section
179 about utf8/latin1/ascii flags, add a security consideration
180 and extend and clarify the JSON and YAML section.
181 - medium speed enhancements when encoding/decoding non-ascii chars.
182 - minor speedup in number encoding case.
183 - extend and clarify the section on incompatibilities
184 between YAML and JSON.
185 - switch to static inline from just inline when using gcc.
186 - add =encoding utf-8 to the manpage, now that perl 5.10 supports it.
187 - fix some issues with UV to JSON conversion of unknown impact.
188 - published the yahoo locals search result used in benchmarks as the
189 original url changes so comparison is impossible.
190
1912.01 Wed Dec 5 11:40:28 CET 2007
192 - INCOMPATIBLE API CHANGE: to_json and from_json have been
193 renamed to encode_json/decode_json for JSON.pm compatibility.
194 The old functions croak and might be replaced by JSON.pm
195 comaptible versions in some later release.
196
1972.0 Tue Dec 4 11:30:46 CET 2007
198 - this is supposed to be the first version of JSON::XS
199 compatible with version 2.0+ of the JSON module.
200 Using the JSON module as frontend to JSON::XS should be
201 as fast as using JSON::XS directly, so consider using it
202 instead.
203 - added get_* methods for all "simple" options.
204 - make JSON::XS subclassable.
205
2061.53 Tue Nov 13 23:58:33 CET 2007
207 - minor doc clarifications.
208 - fixed many doc typos (patch by Thomas L. Shinnick).
209
2101.52 Mon Oct 15 03:22:06 CEST 2007
211 - remove =encoding pod directive again, it confuses too many pod
212 parsers :/.
213
2141.51 Sat Oct 13 03:55:56 CEST 2007
215 - encode empty arrays/hashes in a compact way when pretty is enabled.
216 - apparently JSON::XS was used to find some bugs in the
217 JSON_checker testsuite, so add (the corrected) JSON_checker tests to
218 the testsuite.
219 - quite a bit of doc updates/extension.
220 - require 5.8.2, as this seems to be the first unicode-stable version.
221
2221.5 Tue Aug 28 04:05:38 CEST 2007
223 - add support for tied hashes, based on ideas and testcase by
224 Marcus Holland-Moritz.
225 - implemented relaxed parsing mode where some extensions are being
226 accepted. generation is still JSON-only.
227
2281.44 Wed Aug 22 01:02:44 CEST 2007
229 - very experimental process-emulation support, slowing everything down.
230 the horribly broken perl threads are still not supported - YMMV.
231
2321.43 Thu Jul 26 13:26:37 CEST 2007
233 - convert big json numbers exclusively consisting of digits to NV
234 only when there is no loss of precision, otherwise to string.
235
2361.42 Tue Jul 24 00:51:18 CEST 2007
237 - fix a crash caused by not handling missing array elements
238 (report and testcase by Jay Kuri).
239
2401.41 Tue Jul 10 18:21:44 CEST 2007
241 - fix compilation with NDEBUG (assert side-effect),
242 affects convert_blessed only.
243 - fix a bug in decode filters calling ENTER; SAVETMPS;
244 one time too often.
245 - catch a typical error in TO_JSON methods.
246 - antique-ised XS.xs again to work with outdated
247 C compilers (windows...).
248
2491.4 Mon Jul 2 10:06:30 CEST 2007
250 - add convert_blessed setting.
251 - encode did not catch all blessed objects, encoding their
252 contents in most cases. This has been fixed by introducing
253 the allow_blessed setting.
254 - added filter_json_object and filter_json_single_key_object
255 settings that specify a callback to be called when
256 all/specific json objects are encountered.
257 - assume that most object keys are simple ascii words and
258 optimise this case, penalising the general case. This can
259 speed up decoding by 30% in typical cases and gives
260 a smaller and faster perl hash.
261 - implemented simpleminded, optional resource size checking
262 in decode_json.
263 - remove objToJson/jsonToObj aliases, as the next version
264 of JSON will not have them either.
265 - bit the bullet and converted the very simple json object
266 into a more complex one.
267 - work around a bug where perl wrongly claims an integer
268 is not an integer.
269 - unbundle JSON::XS::Boolean into own pm file so Storable
270 and similar modules can resolve the overloading when thawing.
271
2721.3 Sun Jun 24 01:55:02 CEST 2007
273 - make JSON::XS::true and false special overloaded objects
274 and return those instead of 1 and 0 for those json atoms
275 (JSON::PP compatibility is NOT achieved yet).
276 - add JSON::XS::is_bool predicate to test for those special
277 values.
278 - add a reference to
279 http://jpsykes.com/47/practical-csrf-and-json-security.
280 - removed require 5.8.8 again, it is just not very expert-friendly.
281 Also try to be more compatible with slightly older versions,
282 which are not recommended (because they are buggy).
283
2841.24 Mon Jun 11 05:40:49 CEST 2007
285 - added informative section on JSON-as-YAML.
286 - get rid of some c99-isms again.
287 - localise dec->cur in decode_str, speeding up
288 string decoding considerably (>15% on my amd64 + gcc).
289 - increased SHORT_STRING_LEN to 16kb: stack space is
290 usually plenty, and this actually saves memory
291 when !shrinking as short strings will fit perfectly.
292
2931.23 Wed Jun 6 20:13:06 CEST 2007
294 - greatly improved small integer encoding and decoding speed.
295 - implement a number of µ-optimisations.
296 - updated benchmarks.
297
2981.22 Thu May 24 00:07:25 CEST 2007
299 - require 5.8.8 explicitly as older perls do not seem to offer
300 the required macros.
301 - possibly made it compile on so-called C compilers by microsoft.
302
3031.21 Wed May 9 18:40:32 CEST 2007
304 - character offset reported for trailing garbage was random.
305
3061.2 Wed May 9 18:35:01 CEST 2007
307 - decode did not work with magical scalars (doh!).
308 - added latin1 flag to produce JSON texts in the latin1 subset
309 of unicode.
310 - flag trailing garbage as error.
311 - new decode_prefix method that returns the number
312 of characters consumed by a decode.
12 - max octets/char in perls UTF-X is actually 13, not 11, 313 - max octets/char in perls UTF-X is actually 13, not 11,
13 as pointed out by Glenn Linderman. 314 as pointed out by Glenn Linderman.
14 - decode did not work with magicical scalars (doh!). 315 - fixed typoe reported by YAMASHINA Hio.
15 316
161.11 Mon Apr 9 07:05:49 CEST 2007 3171.11 Mon Apr 9 07:05:49 CEST 2007
17 - properly 0-terminate sv's returned by encode to help 318 - properly 0-terminate sv's returned by encode to help
18 C libraries that expect that 0 to be there. 319 C libraries that expect that 0 to be there.
19 - partially "port" JSON from C to microsofts fucking broken 320 - partially "port" JSON from C to microsofts fucking broken
… …
101 case the called functions do... stuff. 402 case the called functions do... stuff.
102 - croak when encoding to ascii and an out-of-range 403 - croak when encoding to ascii and an out-of-range
103 (non-unicode) codepoint is encountered. 404 (non-unicode) codepoint is encountered.
104 405
1050.2 Fri Mar 23 00:23:34 CET 2007 4060.2 Fri Mar 23 00:23:34 CET 2007
106 - the "could not sleep without debuggign release". 407 - the "could not sleep without debugging release".
107 it should basically work now, with many bugs as 408 it should basically work now, with many bugs as
108 no production tests have been run yet. 409 no production tests have been run yet.
109 - added more testcases. 410 - added more testcases.
110 - the expected shitload of bugfixes. 411 - the expected shitload of bugfixes.
111 - handle utf8 flag correctly in decode. 412 - handle utf8 flag correctly in decode.

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines