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