json_object_get_int/int64/uint64() now sets errno to ERANGE when the source value
can't be represented in the target type.

Json objects of type double with the value NaN could cause undefined
behavior when casting double to int in `json_object_get_int`.

The latter is the proper macro defined by Windows headers.

Fixes compilation under at least clang-cl which mandates function
declarations.

Signed-off-by: Rosen Penev <rosenp@gmail.com>

Enable all disabled tests, add a few more including some with null documents.

Essentially, this change adds the test cases from this repo:
   https://github.com/json-patch/json-patch-tests

Specifically:
   https://github.com/json-patch/json-patch-tests/blob/master/spec_tests.json
   https://github.com/json-patch/json-patch-tests/blob/master/tests.json

The files were taken at the date of this commit, at git hash
  ea3af85790

Some tests may not have an 'expected' or 'error' field. Those are ignored.
One test was disabled manually via "disabled_in_json_c", because it tries
an impossible test, i.e. to add 2 "op" fields in the same patch entry,
something which is impossible in a JSON object.

For the 'error' cases, right now we only test that an error happens.
Later, we can extend this to check the error codes.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>

The out-of-bounds check is useful when trying to index/obtain a value from
an array.
However, when we set a value to a specific JSON pointer, we can allow
values that are outside the length of the current array.
The RFC6901 doc isn't clear on that aspect, and doing so is a bit more
in-line with how json_object_array_{put,insert}_idx() functions behave.

This changes the behavior of json_pointer_set{f}() because now a value can
be set anywhere in the array.

Also, added a test-case for this behavior change.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>

This change adds a few test cases to test the behavior of the new
json_object_array_insert_idx() function, to make sure it behaves according
to specification in doc-string.

This test uses assert() vs the old method of comparing outputs.
This will cause the test to fail because the outputs won't match, since the
assert() will kick in.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>

That hasn't been needed since since commit 6068d3f, which changed that code to
 check an env var instead ("_JSON_C_STRERROR_ENABLE").
Fixes issue #812, about dup symbols in static builds with clang.

Fixes
json_util.c:63:35: error: a function declaration without a prototype is deprecated in all versions of C [-We
rror,-Wstrict-prototypes]
const char *json_util_get_last_err()
                                  ^
                                   void

Signed-off-by: Khem Raj <raj.khem@gmail.com>

Do not fix typos from past commits found in release notes.

cmake rules matters, and fix compile warnings now that we're building
those sources with all the regular flags.

Mostly found with codespell and during code review.

Some users may not want to included it in their build/system. So allow a
cmake symbol to disable it.

A user can do 'cmake -DDISABLE_JSON_POINTER=ON <json_c_root_dir>' and
disable the json_pointer functionality. That saves about 17 KB (on an
x86_64) machine. This may be useful on smaller embedded systems; even
though the saving would be fewer kilobytes.

One thing that also needs to change a bit, is that the 'json.h' be
autogenerated via cmake, in order to conditionally include that
"json_pointer.h" file.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>

The intent is to be able to disable some features that get built into the
library. When we do that, we also need to disable some tests.

It's easier when adjusting a variable that contains the list of test names,
versus modifying the list in the foreach() statement.

Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>

SIZEOF_SIZE_T might be only defined in config.h.

Include config.h for these systems to pass tests which are only
supposed to be run on 32 bit systems.

Fixes issue #666.

Validate size arguments in arraylist functions.

The array_list_new2 function, which is externally reachable through
json_object_new_array_ext, does not check if specified initial size
actually fits into memory on 32 bit architectures.

It also allows negative values, which could lead to an overflow on these
architectures as well. I have added test cases for these situations.

While at it, also protect array_list_shrink against too large
empty_slots argument. No test added because it takes a huge length
value, therefore a lot of items within the array, to overflow the
calculation. In theory this affects 64 bit sytems as well, but since the
arraylist API is not supposed to be used by external applications
according to its header file, the call is protected due to int
limitation of json_object_array_shrink.

The function _json_c_strerror does not properly format unknown errnos.
The int to ascii loop ignores the leading digit if the number can be
divided by 10 and if an errno has been formatted, shorter errnos would
not properly terminate the newly created string, showing the ending
numbers of the previous output.

A test case has been added to show these effects.

Since this function has been introduced for tests, the effect of this on
real life code is basically non-existing. First an environment variable
has to be set to activate this strerror code and second an unknown errno
would have to be encountered.

See also issue #658.

This causes previously failing strings like "123-456" to return a valid json_object with the appropriate value.  If you care about the trailing content, call json_tokener_parse_ex() and check the parse end point with json_tokener_get_parse_end().

Deprecate the json_parse_double() function from json_util.h