* Still respects "-Dtarget" unity/nonunity selection. Still defaults to
unity.
* Adds a new target (rippled_classic or rippled_unity) depending on
unity/nonunity selection.
* New target does not build by default.
eg. Use `cmake --build . -target rippled_classic`
* Copy all config variables (Debug/Release to DebugClassic/ReleaseClassic) for nonunity builds
* CI uses the more generic "cmake --build" command
* Unit test malformed/tem responses
* Update free trust line test to cover the case where creation is actually adding the opposite trust direction to an existing account lines ledger entry
* Add unit tests for setting and removing quality on trust lines.
* Put all the relevant calls together.
* Sensibly handle stage vs. stage64 lib directories.
* Unable to move target_link_libraries acur_project not defined in new location, so breaks non-win builds
Migrate tests in offer-test.js to cpp/jtx. Minimally reformat existing
test code. Augment self-cross test to include partner account as well as bridged and
direct crossing.
* Remove extraneous passing of transaction set hashes
* Remove recentPositions_. InboundTXs does the job now
* Move responsibility for sending "have TX set" out of consensus
* Simplify Travis APT config.
* Automatically retry Travis build and test script. Will result in fewer
false negatives.
* Travis install scripts use absolute paths.
* Build a library of cmake functions for reuse.
* Disallow cmake builds in project root.
* Disallow cmake default 32-bit Visual Studio builds.
* Add several missing nonunity / header files, including all unit tests to
cmake.
* Change gcc.debug.nounity Travis build to use cmake, instead of adding
builds.
* Change Appveyor build to cmake. Eliminates most spurious failures, which
are caused by python or scons failing to download.
When attempting to parse a BufferSequence as a JSON object,
if the sequence contained more than buffer, the JSON parser
would incorrectly attempt to decode each buffer as a separate
JSON object, instead of one complete object.
When attempting to parse a BufferSequence as a JSON object,
if the sequence contained more than buffer, the JSON parser
would incorrectly attempt to decode each buffer as a separate
JSON object, instead of one complete object.
If the mantissas of two non-native amounts differ by less than 10, then
subtracting them leaves a result of zero. This can cause situations
where `a>b`, yet `a-b == 0`.
One consequence of this is unfunded offers were incorrectly left in
order books. The code would check if the offer would be
consumed (`amount in offer > amount needed`), assume it wouldn't be,
yet when `amount needed` was subtracted from `amount in offer` the
result was zero and the offer was unfunded. This unfunded offer
incorrectly remained on the order book.
This patch fixes this bug.
