- Added Mac Binary
- Fixed missing returns in PosiTests.h
- Fixed bug in DataManager where FormatLine was being sent a string
instead of a c string
- Increased mac timeout
I'm not sure why this is necessary. Should look into object lifetimes in MATLAB to see if there's a valid reason for the second & subsequent test runs to fail when using automatic socket management.
The C client tests were getting a bit unwieldy. Moved each block of tests to a separate file based on the functionality being tested. I've just put all of the test code directly in the headers because there doesn't seem to be any reason for the test headers to be included anywhere except in main.c.
- Swaped names of openUDP and aopenUDP.
- openUDP is now the "automatic" version, requiring only the X-Plane IP address.
- aopenUDP is now the "advanced" version, requiring both the X-Plane port and a client port.
- Added support for passing NULL to sendTEXT.
- openTest now validates that "localhost" resolves to "127.0.0.1".
- sendTEXTTest now sets and clears text. Note that since there is no way to externally verify the result of a TEXT command, the user must now step through this test using the debugger to fully verify functionality.
- requestDREFTest renamed to getDREFTest.
- getDREFTest now gets one dataref of each type except multidimensional arrays.
- sendDREFTest now sets one dataref of each type except mulidimensional arrays.
- sendDREFTest now attempts to set a read-only dataref. Need to work out how to test the result of this effectively.
- psendCTRLTest & sendCTRLTest now test multiple values, default values and sending only some parameters.
- Added psendPOSITest.
- psendPOSITest & sendPOSIest now test multiple values, default values and sending only some parameters.
- sendWYPTTest now clears waypoints before setting them.
- These tests create a great deal of duplicated work.
- The only errors caught by these tests would be errors that occur in the C client only when compiled by a C++ compiler.
- This error case is deemed unlikely enough that it is not worth the extra work to maintain two sets of tests.
