Yury Matveev (ea2624e1) at 20 Mar 17:01
Attribute::set_value_date_quality: allow accepting nullptr value is...
Yury Matveev (1876b82b) at 20 Mar 15:19
Attribute::set_value_date_quality: allow accepting nullptr value is...
Yury Matveev (d2722b78) at 19 Mar 15:55
Attribute::set_value_date_quality: allow accepting nullptr value is...
Yury Matveev (12dfc627) at 19 Mar 15:50
Attribute::set_value_date_quality: allow accepting nullptr value is...
Comment on the coverage:
We do have two implementations, doing the same, but for time_t and TangoTimestamp as the time type. My catch2 tests calling only time_t overload. Instead of making new tests, I just made time_t overload to delegate to the TangoTimestamp overload, with all logic living in one place.
Yury Matveev (cc03d488) at 19 Mar 12:43
Attribute::set_value_date_quality: allow accepting nullptr value is...
After cppTango was modified to accept nullptr in case of INVALID quality (cppTango!1614), we adapt PyTango implementation to also allow doing it.
Yury Matveev (d344fd93) at 19 Mar 11:42
push_<...>_event: allow returning None for invalid quality
Yury Matveev (54822cd7) at 19 Mar 11:05
Attribute.set_value_date_quality: allow returning None for invalid ...
Yury Matveev (48876c94) at 19 Mar 11:01
Attribute::set_value_date_quality: allow accepting nullptr value is...
I do not recall any conclusion, actually. Let's disscuss it today
Closes #1631
I reworked logic of Attribute::set_value_date_quality:
before we did not accept nullptr value for invalid quality, since we always tried to set_value, where we check if value is not nullptr. This, actually, does not make any sense, since in case of invalid quality the value was deleted just after we set it. Now we just skiping set value in this case at all.
Added this scenario to tests
This is PyTango`s sister MR: pytango!990
Yury Matveev (1746d5f6) at 19 Mar 10:32
Attribute::set_value_date_quality: allow accepting nullptr value is...
Yury Matveev (0275896c) at 19 Mar 10:26
Yury Matveev (1cddc3cb) at 16 Mar 13:23
Pimpl Tango::Group: fix asan and ubsan issues
... and 4 more commits
Yury Matveev (061b3ba2) at 16 Mar 12:02
Yury Matveev (71a8405c) at 16 Mar 12:02
Merge branch 'improve_command_history' into 'develop'
... and 1 more commit
We do not need to provide our overload for DeviceProxy::command_history - pybind11 does everything by itslef PYBIND11_MAKE_OPAQUE. In addition, if we let pybind11 to use its caster - we get proper type hints for DeviceProxy.command_history
DeviceProxy::attribute_history: improve type hints a bit. Now it is at least typing.list, not just typing.Any
We do not need to provide our overload for DeviceProxy::command_history - pybind11 does everything by itslef PYBIND11_MAKE_OPAQUE. In addition, if we let pybind11 to use its caster - we get proper type hints for DeviceProxy.command_history
DeviceProxy::attribute_history: improve type hints a bit. Now it is at least typing.list, not just typing.Any
Yury Matveev (061b3ba2) at 16 Mar 10:26
DeviceProxy::command_history: usie pybind11 caster instead of our o...