Commit graph

2638 commits

Author SHA1 Message Date
Niels Lohmann
afe4571309
🔨 cleanup + some noexcept 2018-01-16 20:41:04 +01:00
Niels Lohmann
b182308eff
🔨 cleanup 2018-01-16 20:40:24 +01:00
abolz
9b9919d460 Use max_digits10 in dump_float for float->text->float round-trip 2018-01-15 21:26:03 +01:00
abolz
810f81bbd9 Regenerate src/json.hpp 2018-01-15 21:24:59 +01:00
abolz
787204f076 Add unit-tests for detail::to_chars 2018-01-15 21:13:17 +01:00
abolz
9f7c2c04c8 Use the Grisu2 algorithm for formatting 'float's and 'double's 2018-01-15 21:07:26 +01:00
abolz
332f352033 Add an implementation of the Grisu2 algorithm for binary to decimal floating-point conversion
This is an attempt to fix #360. The algorithm produces
decimal representations which are guaranteed to roundtrip
and in ~99.8% actually produces the shortest possible
representation. So this is a nice compromise between using
a precision of digits10 and max_digits10.

Note 1:

The implementation only works for IEEE single/double precision
numbers. So the old implementation is kept for compatibility
with non-IEEE implementations and 'long double'.

Note 2:

If number_float_t is 'float', not all serialized numbers can
be recovered using strtod (strtof works, though). (There is
exactly one such number and the result is off by 1 ulp.)
This can be avoided by changing the implementation (the fix
is trivial), but then the resulting decimal numbers are not
exactly short.
2018-01-15 21:07:26 +01:00
abolz
0695ac4001 Add tests for #360 2018-01-15 21:07:25 +01:00
abolz
68a9736738 Tests: Re-enable some round-trip tests 2018-01-15 21:07:25 +01:00
abolz
6e2e466c27 Tests: Don't rely on the format used for floating-point formatting 2018-01-15 21:07:25 +01:00
abolz
107c21a488 Tests: Exponents are formatted with a leading '+' sign 2018-01-15 21:07:25 +01:00
abolz
3ae82d91a2 Tests: Floating-point formatting uses lower case 'e' 2018-01-15 21:07:24 +01:00
Niels Lohmann
92f7295063
🔨 cleanup 2018-01-15 13:21:33 +01:00
Niels Lohmann
3ac674562f
🔨 clean up 2018-01-14 22:26:20 +01:00
Niels Lohmann
d9446b0e6e
Merge pull request #911 from theodelrieu/fix/cmake_install
Fix/cmake install
2018-01-14 20:08:49 +01:00
Niels Lohmann
9d6b3731b9
improved test coverage 2018-01-14 20:07:38 +01:00
Niels Lohmann
06cddd371b
🔨 removed failing amalgamation test 2018-01-14 17:22:42 +01:00
Niels Lohmann
f85f4967fe
improved test coverage 2018-01-14 17:22:06 +01:00
Niels Lohmann
6965ff00c8
Merge branch 'develop' into feature/ubjson 2018-01-14 13:58:06 +01:00
Niels Lohmann
411c16cbb2
📝 overworked documentation wrt. amalgamation #906 2018-01-14 13:08:28 +01:00
Niels Lohmann
fea5f3792b
📝 it's 2018 2018-01-14 10:41:59 +01:00
Niels Lohmann
541b46132d
♻️ adjusted code to split headers 2018-01-14 10:34:43 +01:00
Niels Lohmann
0e8f01a963
Merge branch 'develop' into feature/rfc7396 2018-01-14 10:28:09 +01:00
Niels Lohmann
c772c01a48
♻️ refactored code to split headers 2018-01-14 10:27:30 +01:00
Niels Lohmann
1b54d4a5aa
Merge branch 'develop' into feature/ubjson 2018-01-14 10:10:56 +01:00
Niels Lohmann
420dcf1f25
🚧 added check whether code is amalgamated 2018-01-14 10:10:23 +01:00
Théo DELRIEU
5775084ffc
cmake: add option to use/install the non-amalgamated version 2018-01-13 13:07:03 +01:00
Théo DELRIEU
84bffd5d36
move amalgamate tool to third_party folder 2018-01-13 13:05:42 +01:00
Théo DELRIEU
922b56e492
cmake: add back trailing slash to NLOHMANN_JSON_SOURCE_DIR 2018-01-13 12:27:33 +01:00
Niels Lohmann
a66b2d20c6
🚨 removed linter warnings for Python code 2018-01-13 11:15:23 +01:00
Niels Lohmann
f4a55f26b0
added amalgamate Python script 2018-01-13 10:59:49 +01:00
Niels Lohmann
241eea0c0e
📝 documentation 2018-01-12 09:23:33 +01:00
Niels Lohmann
10bad9381d
🚧 added size benchmark for binary formats 2018-01-10 23:13:48 +01:00
Niels Lohmann
3a7585e738
added more tests 2018-01-10 22:47:07 +01:00
Niels Lohmann
31bfabc4c0
🔨 optimized input format 2018-01-10 16:14:43 +01:00
Niels Lohmann
965a70e38d
🔨 optimized output format 2018-01-10 11:22:19 +01:00
Niels Lohmann
85173f5627
🔨 some clean up 2018-01-10 10:18:31 +01:00
Niels Lohmann
ce53537ba2
replaces amalgamation tool
The tool used before created a lot of duplicates inside the generated amalgamation. The new tool is a single Python file which seems to do the same job.
2018-01-09 23:15:06 +01:00
Niels Lohmann
0a2920e0fd
♻️ reorganized code 2018-01-09 18:30:02 +01:00
Niels Lohmann
b67e00b9b5
Merge pull request #700 from theodelrieu/refactor/split_it
Refactor/split it
2018-01-09 07:57:01 +01:00
Niels Lohmann
fd04967676
🐛 fixed copy-paste error 2018-01-08 21:37:51 +01:00
Niels Lohmann
fb1154c237
🚧 debug 2018-01-08 21:26:41 +01:00
Niels Lohmann
97e0d20ce9
🚧 debug 2018-01-08 21:16:20 +01:00
Niels Lohmann
ebf28a26ee
🚧 another try 2018-01-08 20:49:02 +01:00
Théo DELRIEU
7e4ee23f40
add single_header CMake target 2018-01-08 20:09:43 +01:00
Théo DELRIEU
57d822b6e2
add missing includes, put back include comments 2018-01-08 20:09:41 +01:00
Théo DELRIEU
9cab30cfce
add adl_serializer.hpp 2018-01-08 20:09:38 +01:00
Théo DELRIEU
8e9714fe3d
add detail/json_ref.hpp 2018-01-08 20:09:35 +01:00
Théo DELRIEU
a3473fda6a
add detail/serializer.hpp 2018-01-08 20:09:33 +01:00
Théo DELRIEU
c117515e31
add detail/parsing/binary_writer.hpp 2018-01-08 20:09:29 +01:00