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

Comparing JSON-XS/Changes (file contents):
Revision 1.45 by root, Wed May 23 22:07:43 2007 UTC vs.
Revision 1.138 by root, Sun Jan 10 10:07:13 2010 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 3 - update jpsykes csrf article url.
4 not 0 or 1 in perl. or use attached magic? 4
52.27 Thu Jan 7 07:35:08 CET 2010
6 - support relaxed option inside the incremental parser
7 (testcase provided by IKEGAMI via Makamaka).
8
92.26 Sat Oct 10 03:26:19 CEST 2009
10 - big integers could become truncated (based on patch
11 by Strobl Anton).
12 - output format change: indent now adds a final newline, which is
13 more expected and more true to the documentation.
14
152.25 Sat Aug 8 12:04:41 CEST 2009
16 - the perl debugger completely breaks lvalue subs - try to work
17 around the issue.
18 - ignore RMAGICAL hashes w.r.t. CANONICAL.
19 - try to work around a possible char signedness issue on aix.
20 - require common sense.
21
222.24 Sat May 30 08:25:45 CEST 2009
23 - the incremental parser did not update its parse offset
24 pointer correctly when parsing utf8-strings (nicely
25 debugged by Martin Evans).
26 - appending a non-utf8-string to the incremental parser
27 in utf8 mode failed to upgrade the string.
28 - wording of parse error messages has been improved.
29
302.232 Sun Feb 22 11:12:25 CET 2009
31 - use an exponential algorithm to extend strings, to
32 help platforms with bad or abysmal==windows memory
33 allocater performance, at the expense of some memory
34 wastage (use shrink to recover this extra memory).
35 (nicely analysed by Dmitry Karasik).
36
372.2311 Thu Feb 19 02:12:54 CET 2009
38 - add a section "JSON and ECMAscript" to explain some
39 incompatibilities between the two (problem was noted by
40 various people).
41 - add t/20_faihu.t.
42
432.231 Thu Nov 20 04:59:08 CET 2008
44 - work around 5.10.0 magic bugs where manipulating magic values
45 (such as $1) would permanently damage them as perl would
46 ignore the magicalness, by making a full copy of the string,
47 reported by Dmitry Karasik.
48 - work around spurious warnings under older perl 5.8's.
49
502.23 Mon Sep 29 05:08:29 CEST 2008
51 - fix a compilation problem when perl is not using char * as, well,
52 char *.
53 - use PL_hexdigit in favour of rolling our own.
54
552.2222 Sun Jul 20 18:49:00 CEST 2008
56 - same game again, broken 5.10 finds yet another assertion
57 failure, and the workaround causes additional runtime warnings.
58 Work around the next assertion AND the warning. 5.10 seriously
59 needs to adjust it's attitude against working code.
60
612.222 Sat Jul 19 06:15:34 CEST 2008
62 - you work around one -DDEBUGGING assertion bug in perl 5.10
63 just to hit the next one. work around this one, too.
64
652.22 Tue Jul 15 13:26:51 CEST 2008
66 - allow higher nesting levels in incremental parser.
67 - error out earlier in some cases in the incremental parser
68 (as suggested by Yuval Kogman).
69 - improve incr-parser test (Yuval Kogman).
70
712.21 Tue Jun 3 08:43:23 CEST 2008
72 - (hopefully) work around a perl 5.10 bug with -DDEBUGGING.
73 - remove the experimental status of the incremental parser interface.
74 - move =encoding around again, to avoid bugs with search.cpan.org.
75 when can we finally have utf-8 in pod???
76 - add ->incr_reset method.
77
782.2 Wed Apr 16 20:37:25 CEST 2008
79 - lifted the log2 rounding restriction of max_depth and max_size.
80 - make booleans mutable by creating a copy instead of handing out
81 the same scalar (reported by pasha sadri).
82 - added support for incremental json parsing (still EXPERIMENTAL).
83 - implemented and added a json_xs command line utility that can convert
84 from/to a number of serialisation formats - tell me if you need more.
85 - implement allow_unknown/get_allow_unknown methods.
86 - fixed documentation of max_depth w.r.t. higher and equal.
87 - moved down =encoding directive a bit, too much breaks if it's the first
88 pod directive :/.
89 - removed documentation section on other modules, it became somewhat
90 outdated and is nowadays mostly of historical interest.
91
922.1 Wed Mar 19 23:23:18 CET 2008
93 - update documentation here and there: add a large section
94 about utf8/latin1/ascii flags, add a security consideration
95 and extend and clarify the JSON and YAML section.
96 - medium speed enhancements when encoding/decoding non-ascii chars.
97 - minor speedup in number encoding case.
98 - extend and clarify the section on incompatibilities
99 between YAML and JSON.
100 - switch to static inline from just inline when using gcc.
101 - add =encoding utf-8 to the manpage, now that perl 5.10 supports it.
102 - fix some issues with UV to JSON conversion of unknown impact.
103 - published the yahoo locals search result used in benchmarks as the
104 original url changes so comparison is impossible.
105
1062.01 Wed Dec 5 11:40:28 CET 2007
107 - INCOMPATIBLE API CHANGE: to_json and from_json have been
108 renamed to encode_json/decode_json for JSON.pm compatibility.
109 The old functions croak and might be replaced by JSON.pm
110 comaptible versions in some later release.
111
1122.0 Tue Dec 4 11:30:46 CET 2007
113 - this is supposed to be the first version of JSON::XS
114 compatible with version 2.0+ of the JSON module.
115 Using the JSON module as frontend to JSON::XS should be
116 as fast as using JSON::XS directly, so consider using it
117 instead.
118 - added get_* methods for all "simple" options.
119 - make JSON::XS subclassable.
120
1211.53 Tue Nov 13 23:58:33 CET 2007
122 - minor doc clarifications.
123 - fixed many doc typos (patch by Thomas L. Shinnick).
124
1251.52 Mon Oct 15 03:22:06 CEST 2007
126 - remove =encoding pod directive again, it confuses too many pod
127 parsers :/.
128
1291.51 Sat Oct 13 03:55:56 CEST 2007
130 - encode empty arrays/hashes in a compact way when pretty is enabled.
131 - apparently JSON::XS was used to find some bugs in the
132 JSON_checker testsuite, so add (the corrected) JSON_checker tests to
133 the testsuite.
134 - quite a bit of doc updates/extension.
135 - require 5.8.2, as this seems to be the first unicode-stable version.
136
1371.5 Tue Aug 28 04:05:38 CEST 2007
138 - add support for tied hashes, based on ideas and testcase by
139 Marcus Holland-Moritz.
140 - implemented relaxed parsing mode where some extensions are being
141 accepted. generation is still JSON-only.
142
1431.44 Wed Aug 22 01:02:44 CEST 2007
144 - very experimental process-emulation support, slowing everything down.
145 the horribly broken perl threads are still not supported - YMMV.
146
1471.43 Thu Jul 26 13:26:37 CEST 2007
148 - convert big json numbers exclusively consisting of digits to NV
149 only when there is no loss of precision, otherwise to string.
150
1511.42 Tue Jul 24 00:51:18 CEST 2007
152 - fix a crash caused by not handling missing array elements
153 (report and testcase by Jay Kuri).
154
1551.41 Tue Jul 10 18:21:44 CEST 2007
156 - fix compilation with NDEBUG (assert side-effect),
157 affects convert_blessed only.
158 - fix a bug in decode filters calling ENTER; SAVETMPS;
159 one time too often.
160 - catch a typical error in TO_JSON methods.
161 - antique-ised XS.xs again to work with outdated
162 C compilers (windows...).
163
1641.4 Mon Jul 2 10:06:30 CEST 2007
165 - add convert_blessed setting.
166 - encode did not catch all blessed objects, encoding their
167 contents in most cases. This has been fixed by introducing
168 the allow_blessed setting.
169 - added filter_json_object and filter_json_single_key_object
170 settings that specify a callback to be called when
171 all/specific json objects are encountered.
172 - assume that most object keys are simple ascii words and
173 optimise this case, penalising the general case. This can
174 speed up decoding by 30% in typical cases and gives
175 a smaller and faster perl hash.
176 - implemented simpleminded, optional resource size checking
177 in decode_json.
178 - remove objToJson/jsonToObj aliases, as the next version
179 of JSON will not have them either.
180 - bit the bullet and converted the very simple json object
181 into a more complex one.
182 - work around a bug where perl wrongly claims an integer
183 is not an integer.
184 - unbundle JSON::XS::Boolean into own pm file so Storable
185 and similar modules can resolve the overloading when thawing.
186
1871.3 Sun Jun 24 01:55:02 CEST 2007
188 - make JSON::XS::true and false special overloaded objects
189 and return those instead of 1 and 0 for those json atoms
190 (JSON::PP compatibility is NOT achieved yet).
191 - add JSON::XS::is_bool predicate to test for those special
192 values.
193 - add a reference to
194 http://jpsykes.com/47/practical-csrf-and-json-security.
195 - removed require 5.8.8 again, it is just not very expert-friendly.
196 Also try to be more compatible with slightly older versions,
197 which are not recommended (because they are buggy).
198
1991.24 Mon Jun 11 05:40:49 CEST 2007
200 - added informative section on JSON-as-YAML.
201 - get rid of some c99-isms again.
202 - localise dec->cur in decode_str, speeding up
203 string decoding considerably (>15% on my amd64 + gcc).
204 - increased SHORT_STRING_LEN to 16kb: stack space is
205 usually plenty, and this actually saves memory
206 when !shrinking as short strings will fit perfectly.
207
2081.23 Wed Jun 6 20:13:06 CEST 2007
209 - greatly improved small integer encoding and decoding speed.
210 - implement a number of ยต-optimisations.
211 - updated benchmarks.
5 212
61.22 Thu May 24 00:07:25 CEST 2007 2131.22 Thu May 24 00:07:25 CEST 2007
7 - require 5.8.8 explicitly as older perls do not seem to offer 214 - require 5.8.8 explicitly as older perls do not seem to offer
8 the required macros. 215 the required macros.
9 - possibly made it compile on so-called C compilers by microsoft. 216 - possibly made it compile on so-called C compilers by microsoft.
10 217
111.21 Wed May 9 18:40:32 CEST 2007 2181.21 Wed May 9 18:40:32 CEST 2007
12 - character offset reported for trailing garbage is random. 219 - character offset reported for trailing garbage was random.
13 220
141.2 Wed May 9 18:35:01 CEST 2007 2211.2 Wed May 9 18:35:01 CEST 2007
15 - decode did not work with magical scalars (doh!). 222 - decode did not work with magical scalars (doh!).
16 - added latin1 flag to produce JSON texts in the latin1 subset 223 - added latin1 flag to produce JSON texts in the latin1 subset
17 of unicode. 224 of unicode.
… …
110 case the called functions do... stuff. 317 case the called functions do... stuff.
111 - croak when encoding to ascii and an out-of-range 318 - croak when encoding to ascii and an out-of-range
112 (non-unicode) codepoint is encountered. 319 (non-unicode) codepoint is encountered.
113 320
1140.2 Fri Mar 23 00:23:34 CET 2007 3210.2 Fri Mar 23 00:23:34 CET 2007
115 - the "could not sleep without debuggign release". 322 - the "could not sleep without debugging release".
116 it should basically work now, with many bugs as 323 it should basically work now, with many bugs as
117 no production tests have been run yet. 324 no production tests have been run yet.
118 - added more testcases. 325 - added more testcases.
119 - the expected shitload of bugfixes. 326 - the expected shitload of bugfixes.
120 - handle utf8 flag correctly in decode. 327 - handle utf8 flag correctly in decode.

Diff Legend

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