2015-03-15 12:03:52 -07:00
v2.5.2
* fix regression parsing << "-0e...">> (thanks @c -bik)
2015-03-03 10:44:36 -08:00
v2.5.1
* assume all datetimes are UTC time and add `Z` designator to indicate
* fix parsing issue with datetimes in arrays
2015-02-25 20:40:17 -08:00
v2.5.0
* `consult/2` function for reading a file directly to a json term
* `maps_always` build flag for always returning maps on platforms
that support them
* dialyzer fixes
2014-12-09 16:24:09 -08:00
v2.4.0
* enough performance improvements to justify a new version. 2-3x
speedup depending on mode of operation
2014-12-07 15:38:59 -08:00
v2.3.1
* fixes an issue where astral plane json escape sequences were
inadvertently being converted to the unicode replacement
character
2014-12-03 23:16:41 -08:00
v2.3
* switched to a faster implementation of string parsing in both
the decoder and encoder
* expand `uescape` option to the decoder
* allow control codes in json passed to decoder (contrary to the spec,
yes)
2014-11-24 17:27:06 -08:00
v2.2
* `return_maps` option
2014-11-24 18:16:54 -08:00
* `uescape` option for 7-bit clean output
2014-11-24 19:18:10 -08:00
* add `Makefile` for slightly better `erlang.mk` compatibility
* add `maps_support/0` call to determine whether `jsx` was compiled
with support for maps or not
2014-11-24 17:27:06 -08:00
2014-08-27 05:45:12 -07:00
v2.1.1
2014-08-26 22:38:46 -07:00
* faster generation of json via iolists
* `repeat_keys` option
2014-08-13 23:00:35 -07:00
v2.1
* force the end of streams with `end_json` in addition to `end_stream`
2014-08-13 23:32:19 -07:00
* support for encoding erlang datetime tuples to iso8601 format
2014-08-19 17:53:22 -07:00
* allow a single trailing comma in objects and arrays
2014-08-13 23:00:35 -07:00
2014-06-03 21:43:43 +00:00
v2.0.4
* more typespec adjustments
2014-05-18 07:16:44 +00:00
v2.0.3
* update some typespecs to make them more comprehensive
v2.0.2
* fixes travis-ci spec
2014-04-27 22:48:48 +00:00
v2.0.1
* fix regression in output of empty objects/arrays
2013-12-18 02:39:43 +00:00
v2.0
* jsx is much more pragmatic by default; common json errors are silently
ignored (and fixed). stricter parsing must be enabled with options
2014-04-19 00:35:56 +00:00
* add support for encoding otp 17.0's new maps data type
2013-12-18 02:39:43 +00:00
* removed `pre_encode` and `post_decode` options in favour of making jsx
functions easier to wrap and customize
* streaming behavior is now disabled by default and must be requested explicitly
* removed deprecated function names (`to_json` , `to_term` , `term_to_json` , etc)
2014-01-13 07:59:59 +00:00
* expanded test coverage
2013-12-18 02:39:43 +00:00
2013-12-12 11:20:30 -08:00
v1.4.5
* various fixes to typespecs uncovered by dialyzer
* allow integer keys during encoding
2013-12-12 11:30:26 -08:00
* convert atoms (other than `true` , `false` and `null` ) to strings during encoding
2013-12-12 11:20:30 -08:00
2013-10-29 19:47:30 +00:00
v1.4.4
* typespec for `json_term/0` fixed
* incorrect boolean shortcircuiting fixed in multibyte escape processing
2013-08-27 05:09:19 +00:00
v1.4.3
* add empty rebar.config for mix build tool
* add `attempt_atom` option for decoding json objects
* fix a bug related to multibyte codepoints and streaming input
* add a missing error state in the encoder
2013-06-04 01:19:39 +00:00
v1.4.2
* build apparatus cleaned up and streamlined
* new `{raw, <<"json goes here">>}` intermediate form to support direct generation of json
* bugfixes involving inappropriate exceptions from jsx functions
2013-03-11 19:10:08 -07:00
v1.4.1
* fixes a bug with interaction between `dirty_strings` and even numbers of escape characters
* performance enhancements
2013-03-04 16:01:45 -08:00
v1.4
* radically refactored decoder
* `dirty_strings` now behaves intuitively in decoding. bad codepoints, bad utf8, illegal characters and escapes (except `"` and `'` if `single_quoted_strings` is enabled) are ignored completely
2013-03-04 23:49:01 -08:00
* `incomplete_handler` & `error_handler` are now available for use, see documentation in README
2013-03-04 16:01:45 -08:00
2012-09-03 21:28:32 -07:00
v1.3.3
* `pre_encode` now orders input in the order you'd expect
2012-09-03 21:17:06 -07:00
v1.3.2
* `pre_encode` is now able to handle tuples *correctly*
2012-07-01 23:42:57 -07:00
v1.3.1
* `pre_encode` is now able to handle tuples
2012-05-23 22:16:44 -07:00
v1.3
* introduces `prettify/1` and `minify/1` , shortcuts for `format/2`
* introduce `encode/1,2` and `decode/1,2` as primary interface to built in tokenizers. `to_json/1,2` and `to_term/1,2` remain accessible but not advertised
* new `parser/3` function exposes syntactic analysis stage for use with user defined tokenizers
* improved documentation
v1.2.1
* fixes incorrect handling of escaped forward slashes, thanks bob ippolito
v1.2
* rewritten handling of string escaping to improve performance
* `pre_encode` and `post_decode` hooks, see README
* `relax` option
v1.1.2
* add `dirty_strings` option
* more fixes for invalid unicode in strings
v1.1.1
* fixes bug regarding handling of invalid unicode in R14Bxx
v1.1
* improvements to string escaping and json generation performance
v1.0.2
* fixes to function specs
* rewritten README
* `comments` option
v1.0.1
2013-12-12 11:20:30 -08:00
* rebar fix