Stitch1D now looks at the min and max possible. If StartOverlap or EndOverlap are outside the min-max range, then they are forced reassigned StartOverlap=min, EndOverlap=max within the algorithm.

Also correct spelling in CopyLogs.cpp .

Signed-off-by: Karl Palmen <karl.palmen@stfc.ac.uk>

Signed-off-by: Karl Palmen <karl.palmen@stfc.ac.uk>

rename the run function and always have it return the OutputWorkspace value. This saves lots of lines of code and makes it clearer.

This prevents orphaned workspaces from outliving the algorithm.

Also update unit tests for Stitch1DMany.

Fix parameter input creation/selection and update unit tests.

Constrain the input workspace types and introduce tests for this behaviour. Also add behaviour to documentation.

Unit tests have been added for this too.

Get manual scaling factor aspects of the algorithm working.

Tests updated. Minor changes required to algorithm as a result of problems found during testing. More tests to come.

This is stitching very nicely now with the test data. The most important change here is that I've limited the zero-masking range for the rhs and lhs rebinned workspace, so that the boundaries can be handled properly. 

This should bring the API to be more in line with combine2 from combineMulti. These changes will break the unit tests.

This is more in line with how the new functions work and how the C++
side works too. Refs #970

There is now a single directory called plugins that contains sub-
directories for algorithms & functions. These get installed into the
package under ROOT/plugins/python. This requires new keys to find the
directories and deprecates pythonalgorithms.directories key.

Users will receive warnings about the old key being deprecated and are
instructed about what to do.

Refs #970

This has been applied to both APIs as there are still quite a few old API
algorithms.

Firstly, I made it more robust to having either 1d or 2d md workspaces as inputs. This allows us to iterate over the algorithm in order to recursively combine many 1D segments.

 I also fixed a bug in the calculation of the target index for the overlapping weighted-mean sections.

Tests have been extended to cover both of these.

 The algorithm has been executed on Tims test data and seems close to the benchmark outputs.