Reimplementing virtual functions with
"const &" arguments wasn't behaving as
expected because these arguments were
copied.
Now, "const &" for arguments (in virtual
function reimplementation) is not implemented
as a copy.
In addition, now it's possible to declare
results as references always (also if const &).
See gsiTest.cc:1078 for example:
// gsi::arg_make_reference makes the function's return value
// always being taken as a reference
gsi::method<C_P, const CopyDetector &, const CopyDetector &, gsi::arg_make_reference> ("pass_cd_cref_as_ref", &C_P::pass_cd_cref)
Reason for the problem: the interaction test
has to keep separate "inside" records for both
below and above the scanline, not just once.
With the single record, the step in the
left polygon erased the "inside" condition.
The cause for the problem was that the layout got updated
while iterating causing the mess within the iterator.
This solution is to lock the layout while an iterator
is present. This happens for various Cell and Shapes
iterator, so it's a major enhancement.
This will include the normal Python tests also
in pymod, hence aligning the test base for pymod
and pya.
This feature requires the pya compatibility module.
Python 2 is a bit more picky with respect to
the correct name of the module. If passing "db"
to a module called dbcore, it will exit with
a SystemError.
- Unify structure of Python module sources so the pymod.pri can be used for all modules
- switch to tlcore, dbcore and rdbcore + distutils_src wrapper modules
With these fixes, unit tests pass with MacOS
(set DYLD_LIBRARY_PATH!)
1.) MacOS takes popen vs. pclose seriously
By using fopen we basically spoil the system and other popen
won't work.
2.) For system integrity, MacOS does not propagate DYLD_LIBRARY_PATH
to child processes such as sh. This has to be done explictly.
3.) Search the klayout binary in the right path (klayout.app/...)
4.) Reset KLAYOUT_HOME for less intervention by installed macros