Allow for a choice of how data is named in the Data Explorer

[smk78]

See also #1701.

Up to and including release 5.0.3, the Data Explorer displayed the name of the file that had been loaded. Since the release, several improvements and bug fixes have been made to the data loader. One improvement being that if it can recognise the name of the data in the file, it will display this in the Data Explorer instead of the file name. The CanSAS1D and NXcanSAS formats will benefit most from this.

User Adrian R makes the following comment on this switch:

Perhaps I can make a suggestion about this enhancement in version 5: while I realise that often people like the sample 
description as provided in the title of a data file to appear in plots or in the table when selecting data, there are 
equally cases where one does not find that appropriate.   A choice would be useful.

For example, there can be repeated runs for the same sample, either as a time sequence with varying properties/sample 
conditions or simply made as checks that data are reproduced.  In this case the title may be identical for several data 
sets but the measurements are distinguished by file name or run/sequence number.  It could then be convenient to 
choose what is presented both in the data panel and on the plot/fit output.

The data files that I sent [see #1701 ] happened to be some where the displaying by title could be appropriate but with 
part of the information missing, this became much less helpful.  There are quite a number of data sets where I have 
several measurements with the same title that would only be distinguished by other metadata (instrument configuration, 
wavelength, ....).