ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/JSON-XS/Changes
Revision: 1.73
Committed: Mon Aug 13 16:05:42 2007 UTC (16 years, 9 months ago) by root
Branch: MAIN
Changes since 1.72: +4 -0 lines
Log Message:
bleh

File Contents

# Content
1 Revision history for Perl extension JSON::XS
2
3 1.44
4 - very experimental process-emulation support, slowing everything down.
5 the horribly broken perl threads are still not supported - YMMV.
6
7 1.43 Thu Jul 26 13:26:37 CEST 2007
8 - convert big json numbers exclusively consisting of digits to NV
9 only when there is no loss of precision, otherwise to string.
10
11 1.42 Tue Jul 24 00:51:18 CEST 2007
12 - fix a crash caused by not handling missing array elements
13 (report and testcase by Jay Kuri).
14
15 1.41 Tue Jul 10 18:21:44 CEST 2007
16 - fix compilation with NDEBUG (assert side-effect),
17 affects convert_blessed only.
18 - fix a bug in decode filters calling ENTER; SAVETMPS;
19 one time too often.
20 - catch a typical error in TO_JSON methods.
21 - antique-ised XS.xs again to work with outdated
22 C compilers (windows...).
23
24 1.4 Mon Jul 2 10:06:30 CEST 2007
25 - add convert_blessed setting.
26 - encode did not catch all blessed objects, encoding their
27 contents in most cases. This has been fixed by introducing
28 the allow_blessed setting.
29 - added filter_json_object and filter_json_single_key_object
30 settings that specify a callback to be called when
31 all/specific json objects are encountered.
32 - assume that most object keys are simple ascii words and
33 optimise this case, penalising the general case. This can
34 speed up decoding by 30% in typical cases and gives
35 a smaller and faster perl hash.
36 - implemented simpleminded, optional resource size checking
37 in decode_json.
38 - remove objToJson/jsonToObj aliases, as the next version
39 of JSON will not have them either.
40 - bit the bullet and converted the very simple json object
41 into a more complex one.
42 - work around a bug where perl wrongly claims an integer
43 is not an integer.
44 - unbundle JSON::XS::Boolean into own pm file so Storable
45 and similar modules can resolve the overloading when thawing.
46
47 1.3 Sun Jun 24 01:55:02 CEST 2007
48 - make JSON::XS::true and false special overloaded objects
49 and return those instead of 1 and 0 for those json atoms
50 (JSON::PP compatibility is NOT achieved yet).
51 - add JSON::XS::is_bool predicate to test for those special
52 values.
53 - add a reference to
54 http://jpsykes.com/47/practical-csrf-and-json-security.
55 - removed require 5.8.8 again, it is just not very expert-friendly.
56 Also try to be more compatible with slightly older versions,
57 which are not recommended (because they are buggy).
58
59 1.24 Mon Jun 11 05:40:49 CEST 2007
60 - added informative section on JSON-as-YAML.
61 - get rid of some c99-isms again.
62 - localise dec->cur in decode_str, speeding up
63 string decoding considerably (>15% on my amd64 + gcc).
64 - increased SHORT_STRING_LEN to 16kb: stack space is
65 usually plenty, and this actually saves memory
66 when !shrinking as short strings will fit perfectly.
67
68 1.23 Wed Jun 6 20:13:06 CEST 2007
69 - greatly improved small integer encoding and decoding speed.
70 - implement a number of µ-optimisations.
71 - updated benchmarks.
72
73 1.22 Thu May 24 00:07:25 CEST 2007
74 - require 5.8.8 explicitly as older perls do not seem to offer
75 the required macros.
76 - possibly made it compile on so-called C compilers by microsoft.
77
78 1.21 Wed May 9 18:40:32 CEST 2007
79 - character offset reported for trailing garbage was random.
80
81 1.2 Wed May 9 18:35:01 CEST 2007
82 - decode did not work with magical scalars (doh!).
83 - added latin1 flag to produce JSON texts in the latin1 subset
84 of unicode.
85 - flag trailing garbage as error.
86 - new decode_prefix method that returns the number
87 of characters consumed by a decode.
88 - max octets/char in perls UTF-X is actually 13, not 11,
89 as pointed out by Glenn Linderman.
90 - fixed typoe reported by YAMASHINA Hio.
91
92 1.11 Mon Apr 9 07:05:49 CEST 2007
93 - properly 0-terminate sv's returned by encode to help
94 C libraries that expect that 0 to be there.
95 - partially "port" JSON from C to microsofts fucking broken
96 pseudo-C. They should be burned to the ground for pissing
97 on standards. And I should be stoned for even trying to
98 support this filthy excuse for a c compiler.
99
100 1.1 Wed Apr 4 01:45:00 CEST 2007
101 - clarify documentation (pointed out by Quinn Weaver).
102 - decode_utf8 sometimes did not correctly flag errors,
103 leading to segfaults.
104 - further reduced default nesting depth to 512 due to the test
105 failure by that anonymous "chris" whose e-mail address seems
106 to be impossible to get. Tests on other freebsd systems indicate
107 that this is likely a problem in his/her configuration and not this
108 module.
109 - renamed json => JSON in error messages.
110 - corrected the character offset in some error messages.
111
112 1.01 Sat Mar 31 16:15:40 CEST 2007
113 - do not segfault when from_json/decode gets passed
114 a non-string object (reported by Florian Ragwitz).
115 This has no effect on normal operation.
116
117 1.0 Thu Mar 29 04:43:34 CEST 2007
118 - the long awaited (by me) 1.0 version.
119 - add \0 (JSON::XS::false) and \1 (JSON::XS::true) mappings to JSON
120 true and false.
121 - add some more notes to shrink, as suggested by Alex Efros.
122 - improve testsuite.
123 - halve the default nesting depth limit, to hopefully make it
124 work on Freebsd (unfortunately, the cpan tester did not
125 send me his report, so I cannot ask about the stack limit on fbsd).
126
127 0.8 Mon Mar 26 00:10:48 CEST 2007
128 - fix a memleak when decoding hashes.
129 - export jsonToBj and objToJson as aliases
130 to to_json and from_json, to reduce incompatibilities
131 between JSON/JSON::PC and JSON::XS. (experimental).
132 - implement a maximum nesting depth for both en- and de-coding.
133 - added a security considerations sections.
134
135 0.7 Sun Mar 25 01:46:30 CET 2007
136 - code cleanup.
137 - fix a memory overflow bug when indenting.
138 - pretty-printing now up to 15% faster.
139 - improve decoding speed of strings by
140 up to 50% by specialcasing short strings.
141 - further decoding speedups for strings using
142 lots of \u escapes.
143 - improve utf8 decoding speed for U+80 .. U+7FF.
144
145 0.5 Sat Mar 24 20:41:51 CET 2007
146 - added the UTF-16 encoding example hinted at in previous
147 versions.
148 - minor documentation fixes.
149 - fix a bug in and optimise canonicalising fastpath
150 (reported by Craig Manley).
151 - remove a subtest that breaks with bleadperl (reported
152 by Andreas König).
153
154 0.31 Sat Mar 24 02:14:34 CET 2007
155 - documentation updates.
156 - do some casting to hopefully fix Andreas' problem.
157 - nuke bogus json rpc stuff.
158
159 0.3 Fri Mar 23 19:33:21 CET 2007
160 - remove spurious PApp::Util reference (John McNamara).
161 - adapted lots of tests from other json modules
162 (idea by Chris Carline).
163 - documented mapping from json to perl and vice versa.
164 - improved the documentation by adding more examples.
165 - added short escaping forms, reducing the created
166 json texts a bit.
167 - added shrink flag.
168 - when flag methods are called without enable argument
169 they will by default enable their flag.
170 - considerably improved string encoding speed (at least
171 with gcc 4).
172 - added a test that covers lots of different characters.
173 - clarified some error messages.
174 - error messages now use correct character offset
175 with F_UTF8.
176 - improve the "no bytes" and "no warnings" hacks in
177 case the called functions do... stuff.
178 - croak when encoding to ascii and an out-of-range
179 (non-unicode) codepoint is encountered.
180
181 0.2 Fri Mar 23 00:23:34 CET 2007
182 - the "could not sleep without debuggign release".
183 it should basically work now, with many bugs as
184 no production tests have been run yet.
185 - added more testcases.
186 - the expected shitload of bugfixes.
187 - handle utf8 flag correctly in decode.
188 - fix segfault in decoder.
189 - utf8n_to_uvuni sets retlen to -1, but retlen is an
190 unsigned types (argh).
191 - fix decoding of utf-8 strings.
192 - improved error diagnostics.
193 - fix decoding of 'null'.
194 - fix parsing of empty array/hashes
195 - silence warnings when we prepare the croak message.
196
197 0.1 Thu Mar 22 22:13:43 CET 2007
198 - first release, very untested, basically just to claim
199 the namespace.
200
201 0.01 Thu Mar 22 06:08:12 CET 2007
202 - original version; cloned from Convert-Scalar
203