SNS ADARA files: Load the instrument from the Nexus file

[Russell Taylor]

The nexus files from ADARA-enabled instruments include the full instrument definition XML. We should load that if it's available.

This will allow HYSPECA (an alias for HYSPEC while files were being generated by both the ADARA & legacy systems) to go away (but the files to still be loadable). In time, it will also mean that an updated instrument definition doesn't necessitate a patch release.

[Russell Taylor]

Re #8108. Generalise where to look for instrument.

Remove the hard-coding pointing to the ISIS place in favour of the variable that's used everywhere else.

Changeset: 095164a382d6f0e92320d446fe27541043b606fe

[Russell Taylor]

Re #8108. Use the LoadEventNexus loadInstrument static method.

As well as reducing code duplication, this method tries to load the instrument from the nexus file if possible.

Changeset: 594aff25cbea0eb0a16bf96edb6cb46f1ed60020

[Russell Taylor]

Re #8108. Add a test that the instrument is read in correctly.

This is for a nexus file that contains the full instrument XML. Right now, it's still reading the external IDF - this test will make sure things still work after the refactoring.

Changeset: 1cfc94e1d659f1d0cf59a6d8656ec6d564e4d3f1

[Russell Taylor]

Re #8108. Generalise where to look for instrument.

Remove the hard-coding pointing to the ISIS place in favour of the variable that's used everywhere else.

Changeset: ce1e29a1ecfac4a64e1b5f70b0864061b4bd3629

[Russell Taylor]

Re #8108. Use the LoadEventNexus loadInstrument static method.

As well as reducing code duplication, this method tries to load the instrument from the nexus file if possible.

Changeset: f2f0bdaed27029784f6a46b78730276a42debd9f

[Russell Taylor]

Re #8108. Refactor to correctly load from SNS ADARA files.

While also continuing to work for the other files that have the instrument XML inside the nexus file (processed, ISIS files).

Changeset: b424c810dcc41ac136b57923a64e96ec73ba8b92

[Russell Taylor]

Re #8108. Add .nxs.h5 to list of file extensions.

Also fix some formatting.

Changeset: 8d799faeeacb57b59155306f272972788ed37118

[Russell Taylor]

Re #8108. Add loading of the parameter file to LoadIDFFromNexus.

However, the algorithm doesn't fail if the parameter file is missing.

Changeset: f36487fb8724f264dcac25f765ef7d5433f1ab24

[Russell Taylor]

Re #8108. Fix and enable the last test.

As far as I can tell, setting the filename at this point in ExperimentInfo is not necessary.

Changeset: 029f69253d3a4b53495badb0b629183068f5ee66

[Russell Taylor]

Re #8108. Throw if unable to load instrument from IDF

Previously, this failure would just log and the calling algorithm thought it had succeeded when it hadn't.

Changeset: c7234148297d4b7958728aab0ebc0288fa99d9c7

[Russell Taylor]

Re #8108. Remove redundant code.

The strings will be empty on construction anyway.

Changeset: ccd65b8378a1e34afd0312e47dfe3854a1f44d68

[Russell Taylor]

Re #8108. Check early for presence of IDF in nexus file.

It seemed rather heavyweight to launch and plough through most of LoadIDFFromNexus & ExperimentInfo::loadNexus when the thing we want isn't there anyway.

Changeset: 2984b2aecfc245d14eff5a7492d797255cbbe8a7

[Russell Taylor]

Tester: Branch is feature/8108_load_inst_from_adara_nexus_files

This should probably be tested at the SNS on SEQUOIA or HYSPEC files. It should be checked on an analysis that would be affected if the instrument and the parameters were not loaded correctly.

All the unit and system tests are passing so I'm fairly confident that I haven't broken anything else. There's a rudimentary unit test, but really it's not too easy to tell from the 'outside' where the instrument was loaded from. The best test is that in #8100 HYSPECA_Definition.xml was removed (this is on develop) which would mean that trying to load a HYSA file (as the unit test and one system test - StepScan - do) would fail if this had not been done correctly.

[Andrei Savici]

Does not work as expected. I loaded HYS_28371.nxs.h5 (IPTS-9477) and it looks fine. Then I deleted HYSPEC_Definition.xml, HYSPEC_Definition.vtp, and HYSPEC_Parameters.xml from the instrument directory, and restarted MantidPlot. When I load the file, it shows some instrument, but the tank is not rotated at ~70 degrees, and monitor positions are wrong. Hyspec geometry depends on some log values, and it seems like the instrument is defined before the logs are read (or not using the logs). Strangely, if the IDF is there, it says that instrument definition is read from the nexus file, but obviously something is different

[Russell Taylor]

Re #8108. Add a magic line to tie logfile params to components.

Changeset: 119d6b080add1c7118d43cfde80feddc35d5485e

[Russell Taylor]

Re #8108. Fix cut-and-paste mistake.

Changeset: 564bf4ad466542c85177902a3fa12a9217ba8f99

[Russell Taylor]

The problem raised in comment:15 should now be fixed.

[Andrei Savici]

Merge remote-tracking branch 'origin/feature/8108_load_inst_from_adara_nexus_files'

Full changeset: ​084c85a5dc8098a8993afbb7bac63724086b50c7

[Andrei Savici]

Tested by removing HYSPEC definition and parameter files. Instrument shows up correctly

[Stuart Campbell]

This ticket has been transferred to github issue ​8953