Giovanni Di Siena authored 4 years ago and Gigg, Martyn Anthony committed 4 years ago

In places other substitutions have been made, e.g
Clang does not yet specialize std::shared_ptr for T[]. Vector
has been used instead. The operator[] methods were incorrectly
marked const but returning a non-const reference - this has been fixed.
Refs #25842

Also do some minor tidying of existing tests and add a missing polarization correction test.

Re #27826

Expose the background subtraction options in ReflectometryISISLoadAndProcess in the ISIS Reflectometry interface. The background processing instructions need to be specified on a per-angle basis so these are added to the Per-Angle Defaults table. The other options can be specified as defaults for all runs so are added as their own widgets on the Experiment Settings tab.

Re #27826

Clicking restore-defaults on an Experiment or Instrument tab in a batch should only affect that tab and batch. This commit fixes a bug where a notification was sent to all batches, and every Experiment and Instrument tab was being reset. The fix is to separate updating the instrument (which is required to ensure we have up-to-date settings loaded from the instrument) from the subsequent notification which updates all batches. Instead, the particular Experiment/Instrument tab where the request was made restores its own defaults itself.

Re #28429

The problem here was that when the instrument is set on the new batch, a notification is sent that the instrument has changed and all batches were being reset. However, although the instrument has been set, we can just check whether it is different to the current instrument to check whether it really has changed, and if not, do not do the updates.

Re #28429

The original test json string contained largely default/empty values so it is hard to tell whether they are really being loaded. Add a new string containing fully populated values as well as an "empty" batch representation containing defaults. Use these for the main window containing two batches.

Add a new unit test to test the case where the instrument has changed.

Re #28429

Previously this tooltip was misleading because it was referencing the transmission stitch params property of the reduction algorithm. This input is actually for all properties for the Stitch1DMany algorithm so we cannot take the documentation from a single algorithm property as we do for the other tooltips. Therefore, this commit adds a function to be able to set a tooltip independently of any algorithm and sets this tooltip manually.

Tidy up existing tests to improve readability

Re #28429

Re #28429

Runs clang-tidy pass by ref across the codebase to fix-up various
warnings we are seeing in cppcheck

First checking, more to save state over the weekend

Also updated class_maker.py

This commit replaces all header guards in qt/scientific_interfaces with #pragma once.

Bumps the version of GTest to 1.10.0, removes the previous patching
steps, which are no longer needed. Removes the extra handling which is
provided in upstreams CMake, such as managing include directories

These exceptions were not handled further up. Upon investigation there is no handling necessary so I have changed them to logger warnings instead.

This commit disables the context menu in the fit plot widgets in Indirect Data Analysis, this is because the plot is not properly communicating with the rest of the gui that can result in crashes.

re: #28104

This is a quick fix to to replace the fact that previously the fits were always updated using the zero table index. In the longer term there is a lot of tidying up to be done in the multi-input code.

Previously changeing the selected data was emitting spectrum changed rather than data changed meaning that not everything which needed to be updated was being updated.

Multiple input can be either width or EISF data. Therefore in line with the old functionality  in this mode the full range of fitting functions are enabled.

Updates single input so that it only fits one spectrum at a time. Updates multiple inputs so that you can load spectrum from more than one workspace at a time.

Attempting to use an improperly defined spectra range (e.g. 5-3)
causes appendValue in PropertyHelper.h to throw a logic error.  The
gui alerts the user to runtime errors, but fails over on logic errors.
We now catch the logic error and demote it to a runtime error.

This commit fixes a bug that resulted in a crash when a spectra was removed from the selected spectra if there was no fit function selected. This bug resulted from the plot attempting to remove the fit function but not finding one.

Re: #28086