Baptiste Lepilleur
d496e044b1
Fixed unit tests execution on MSVC 6 by removing usage of std::numeric_limits. It was returning 0 value in some max cases. Fixed Value::asFloat() to use integerToDouble().
2011-05-27 08:12:41 +00:00
Baptiste Lepilleur
f587e6a420
Fixed compilation issues with MSVC 6: replace usage of ostringstream with valueToString to support 64 bits integer and high precision floating point conversion to string. Replace usage of ULL and LL literal with UInt64(expr) and Int64(expr). Introduced helper function uint64ToDouble() to work-around missing conversion. Unit tests do not pass yet.
2011-05-26 22:55:24 +00:00
Baptiste Lepilleur
f0b24e705f
Fixed MSVS 2003, 2005 and 2008 tests execution by normalizing floating-point string representation using helper normalizeFloatingPointStr().
2011-05-26 20:14:32 +00:00
Baptiste Lepilleur
e807a7640e
Fixed unit test failure on IBM AIX xlC by hard-coding the maxUInt64AsDouble as double constant instead of relying on double(Value::maxUInt64) which produces an incorrect value.
2011-05-26 17:14:26 +00:00
Baptiste Lepilleur
d3cd9a7fc5
- Fixed unit test compilation on MSVS 2003, 2005 and 2008.
...
- Worked-around unit test failure with MSVS* by "forcing" all floating-point numbers to be loaded from memory instead of FPU registers.
2011-05-26 07:32:36 +00:00
Aaron Jacobs
a2fb7fb918
Fixed some test bugs that show up when 64-bit mode is disabled.
2011-05-26 06:58:52 +00:00
Aaron Jacobs
c649badb95
Another round of attempting to fix VC++ errors...
2011-05-26 03:44:02 +00:00
Aaron Jacobs
a9eb1eccc0
Fixed more default cases.
2011-05-26 03:32:11 +00:00
Aaron Jacobs
6ffff91c54
Got rid of some unreachable code.
2011-05-26 03:27:44 +00:00
Aaron Jacobs
acdefb0869
Fixed a double -> float compilation warning/error.
2011-05-26 03:04:01 +00:00
Aaron Jacobs
c025697ea5
Reworked the type conversion system again, so that:A
...
* isFoo methods determine exact representability.
* asFoo methods cause casting when safe.
* isConvertibleTo indicates whether casting is safe.
See NEWS.txt for details.
2011-05-26 02:46:28 +00:00
Aaron Jacobs
b0ec41c3e3
Made the unit test's output more readable, adding to jsontest's
...
capabilities (and simplifying its implementation) in the process.
2011-05-26 00:30:39 +00:00
Aaron Jacobs
2a2b5cf3ad
Made jsontest work with 64-bit integers, and fixed an error.
2011-05-26 00:12:48 +00:00
Aaron Jacobs
b6620e2801
Removed some out of date TODOs.
2011-05-25 23:26:58 +00:00
Aaron Jacobs
ccde848fd1
Fixed test failures with 64-bit support disabled.
2011-05-25 05:53:59 +00:00
Aaron Jacobs
e082248001
Fixed a 'comparison between signed and unsigned' error.
2011-05-25 05:50:13 +00:00
Aaron Jacobs
e91a68cb9e
Fixed a compilation warning/error.
2011-05-25 04:34:57 +00:00
Aaron Jacobs
1b138e8544
Gave a more consistent behavior to the Value::isFoo methods. See
...
NEWS.txt for more details.
2011-05-25 04:19:17 +00:00
Aaron Jacobs
4f081b50e6
Fixed bugs in asInt64 and asUInt64.
2011-05-25 03:16:49 +00:00
Aaron Jacobs
3c9fdeb859
Added tests for default numeric values.
2011-05-25 02:54:11 +00:00
Aaron Jacobs
4b79fd1a00
Fixed a test bug.
2011-05-25 01:51:30 +00:00
Aaron Jacobs
e12d84ebaa
Made tests more comprehensive.
2011-05-25 01:46:50 +00:00
Aaron Jacobs
078e0d7c37
Gave tests more general names in preparation for making them much more
...
comprehensive.
2011-05-25 01:24:23 +00:00
Aaron Jacobs
fee49b1a37
Fixed some whitespace.
2011-05-25 01:23:47 +00:00
Aaron Jacobs
22eede44c1
Added tests for 64-bit integers.
2011-05-25 01:23:08 +00:00
Aaron Jacobs
d9ec234fc2
Greatly fleshed out numeric type tests.
2011-05-25 01:04:07 +00:00
Aaron Jacobs
3e5b347f75
Added some missing checks.
2011-05-25 01:03:29 +00:00
Aaron Jacobs
96408a30e1
Renamed test cases to make more sense with the upcoming new behavior of
...
isFoo methods.
2011-05-25 00:39:55 +00:00
Aaron Jacobs
1d648f089a
Fixed a whitespace problem.
2011-05-25 00:39:17 +00:00
Aaron Jacobs
f40c880585
Fixed a "comparison between signed and unsigned" warning/error.
2011-05-24 23:08:59 +00:00
Aaron Jacobs
ae9ffb5443
Fixed a parsing bug in decodeNumber, updating the failing test cases to be
...
correct in the process. (The test cases incorrectly used exact integers instead
of scientific notation.)
2011-05-24 03:59:24 +00:00
Aaron Jacobs
f1053e7acb
Fixed a bunch of compilation errors when JSON_HAS_INT64 is set.
2011-05-24 03:18:02 +00:00
Aaron Jacobs
e3d0eca9f4
Centralized assertion macros and made them obey JSON_USE_EXCEPTION.
2011-05-24 01:03:22 +00:00
Aaron Jacobs
a77a803c85
Made two security fixes.
2011-05-24 00:43:59 +00:00
Aaron Jacobs
785ba2675d
Updated a cast to use a more appropriate type.
2011-05-24 00:43:30 +00:00
Aaron Jacobs
3b556ec633
Fixed constructor initializer list order warnings/errors.
2011-05-24 00:42:58 +00:00
Aaron Jacobs
5fb0f09cbb
Removed an unused typedef.
2011-05-24 00:42:15 +00:00
Aaron Jacobs
73911f2e33
Fixed a hard to debug crash on OS X related to sscanf format strings.
...
See here for more info:
http://developer.apple.com/library/mac/#DOCUMENTATION/DeveloperTools/gcc-4.0.1/gcc/Incompatibilities.html
2011-05-24 00:41:12 +00:00
Baptiste Lepilleur
eadc478e50
Fixed typo: amalga*ma*te. Replaced macro JSON_IS_AMALGATED with JSON_IS_AMALGAMATION
2011-05-02 21:09:30 +00:00
Baptiste Lepilleur
1837a1c508
Value::compare() is now const and has an actual implementation with unit tests.
2011-05-02 20:11:48 +00:00
Baptiste Lepilleur
e3cc0f004b
Untabified some sources
2011-05-02 18:41:01 +00:00
Baptiste Lepilleur
fb17080142
- Added unit tests for comparison operators (except compare())
...
- Fixed Value::operator <= implementation (had the semantic of operator >=). Found when addigin unit tests for comparison operators.
2011-05-02 16:53:10 +00:00
Baptiste Lepilleur
e0e1fd37cd
- Bug #3200841 : removed "warning C4127: conditional expression is constant" concerning infinite loop by replacing while (true) with for (;;). Added new JSON_FAIL macro. Commented unused parameters.
2011-05-02 16:51:48 +00:00
Baptiste Lepilleur
d0a9f3d98d
Bug #3200841 : removed "warning C4127: conditional expression is constant" concerning infinite loop by replacing while (true) with for (;;).
2011-05-02 09:54:49 +00:00
Baptiste Lepilleur
64e40aafe5
Added support for amalgated source and header generation (a la sqlite). Refer to README.txt section "Generating amalgated source and header" for detail.
...
The amalgated sources are generated by concatenating JsonCpp source in the correct order and defining macro JSON_IS_AMALGATED to prevent inclusion of other headers. Sources and header has been modified to prevent any inclusion when this macro is defined.
The script amalgate.py handle the generation.
2011-05-01 20:13:40 +00:00
Baptiste Lepilleur
b2e8cccbc6
Renamed Reader::getFormatedErrorMessages() to getFormattedErrorMessages. Bug #3023708 (Formatted has 2 't'). The old member function is deprecated but still present for backward compatibility.
2011-05-01 16:27:55 +00:00
Baptiste Lepilleur
99043b32b5
Fixed bug #3139678 : stack buffer overflow when parsing a double with a length of 32 characters.
2011-05-01 15:47:38 +00:00
Baptiste Lepilleur
9c98f2277b
Fixed bug #3139677 : JSON [1 2 3] was incorrectly parsed as [1, 3]. Error is now correctly detected.
...
Modified runjsontests.py to allow test that expect failure in jsoncpp test suite.
2011-05-01 15:40:47 +00:00
Baptiste Lepilleur
842e9ac54b
Major rework of 64 integer support: 64 bits integer are only returned when explicitly request via Json::Value::asInt64(), unlike previous implementation where Json::Value::asInt() returned a 64 bits integer.
...
This eases porting portable code and does not break compatibility with the previous release.
Json::Value::asLargestInt() has also be added to ease writing portable code independent of 64 bits integer support. It is typically used to implement writers.
2010-12-27 17:45:23 +00:00
Baptiste Lepilleur
b96aed0f3e
Added float Json::Value::asFloat() to obtain a floating point value as a float (avoid lost of precision warning caused by used of asDouble() to initialize a float).
2010-12-24 19:30:06 +00:00
