Refl_gui Improvements - Live Data fix

[Keith Brown]

Before Owen dealt with it in #8372, quick did something bizarre and bad.

In quick.coadd it used to call startlivedata internally inside quick if the run number was 0 then used the output for processing. Owen has since removed this functionality and it needs put back, but in the right way.

What should happen is that the GUI should be the one to call startlivedata and then pass the resulting workspace to quick as quick can now accept a workspace as a parameter

[Keith Brown]

Refs #8590. Working on alternate loading functionality

I've created a script that wraps Load and StartLiveData which will invoke the one apropriate for the run number supplied

Changeset: f05ee9e28de6c0f794178b56eda00bec53df48c9

[Keith Brown]

Refs #8590. Changed the way the check happens

The LoadInstrumentData script has been repurposed as load_live_data and doen't do a zero check when you ask for live data.

Changeset: afd9570efd251437db31c4c05cab26a25c83f255

[Keith Brown]

Refs #8590 continuing work on integrating startlivedata

The user is now prompted for what Accumulation method they want to use for the live data.

However there is a problem where quick crashed after it tries to get the workspace defualts on the live data workspace. I will need to investigate this.

I plan on merging master into this branch in order to gain the benefit of some of the newer edits to quick, in order to see if they fix this.

Changeset: eaa56fb78c6dfb5c57790bdcd6998abdb22b805f

[Keith Brown]

Refs #8590. Merge remote-tracking branch 'origin' into feature/8590_Refl_gui_Live_Data_fix

Changeset: 1e9ee5233df2cffdbd33173e81cc39ba3a7225a8

[Keith Brown]

Refs #8590. Remove debug code

FakeEventDataListener wasn't doing the job as the workspace it generated didn't have an accociated paremeter file so it caused an error later in quick when it tried to get those parameters.

Changeset: 4370b23e1b1e1f972f4ab2835077e7e27ddb865d

[Keith Brown]

This should only be tested by Owen.

This will likely need to be tested in debug as the fake event listeners don't appear in a release build. The wrapper script get_live_data() will also need modifying slightly so that you can force the fake event listeners, otherwise it'll try and connect to an instrument and fail.

Run FakeISISEventDAE and FakeISISHistoDAE to initialise the listeners

Load a tbl file and change the run number of one to 0, then hit process (processing all rows). It should then ask what accumulation mode you require. At this point the Live Data Listener should start and create a workspace. Quick.py will then fail as the fake event listener won't have the appropriate parameters

[Owen Arnold]

No checkbuild! Cannot test it.

[Owen Arnold]

I can see what you've done here, and I think what you've done should work. However, the following should be addressed within this ticket.

• Dead code line 371 etc. Dead code should be deleted.
• Menu option for accumulation method rather than pop-up. Store choice in qtsettings or similar so that it is remembered. Users will definitely get annoyed if they have to select this each time.
• Add menu option above to documentation (screenshot with help text)
• Line 520. Replace logic seems unnecessary. Should be able to load via the run number.

Once these things are done, this should be fine to go into master.

[Owen Arnold]

The decision logic does seem to be working as I have checked that. The other issues will still need to be resolved.

[Keith Brown]

Update from owen:

Remove dead code

The accumulation method can be saved for another ticket

Ignore the logic change at 520, owen is changing that area anyway

[Keith Brown]

Refs #8590 Merge remote-tracking branch 'origin' into feature/8590

Conflicts:

Code/Mantid/scripts/Interface/ui/reflectometer/refl_gui.py

Changeset: 0e46b72dfffb13f906db7c51ebca79979c88eb35

[Keith Brown]

Refs #8590 Dead Code removed

Dead code has been removed

Changeset: fa73b4d289e73bd5a6bc7ecf98595480d8c07929

[Keith Brown]

Owen should really be the one to test this.

To Tester

Make sure the existing functionality works: http://www.mantidproject.org/Testing_ISIS_Reflectometry_GUI

Make sure an attempt to start a live data service for the selected instrument is actually attempted. This attempt should fail as the live data listeners won't be able to connect properly.

The logic is there, and should work provided a connection is available, but we have no way of making sure.

[Owen Arnold]

Hasn't been fully merged into develop. Once you've done this, I can test it.

[Keith Brown]

Refs #8590 Merge branch 'feature/8590' into develop

Conflicts:

Code/Mantid/scripts/Interface/ui/reflectometer/refl_gui.py

Changeset: f09050426d6c23bf9ca4b45cdf9548446399281c

[Owen Arnold]

Had to reopen load_live_runs is undefined. Can you please fix this and retest it before adding it back to the testing pool. Thanks, Owen.

[Keith Brown]

I think that was caused by the movement of modules in your other ticket as the module won't have moved on my branch. I just need to put it in the proper place

[Owen Arnold]

I've also seen that you're using uppercase parameter names in some of the new functions. Our standard is to only use names starting with upper case to indicate a type. See ​http://google-styleguide.googlecode.com/svn/trunk/pyguide.html. You can write a separate ticket to clean this up at a later date.

[Keith Brown]

refs #8590 Replaced Import Statement

The import had dissapeared, persumably from when owen cleaned some stuff up and git got confused

Changeset: c9b51bc9f9b6e7e201898193399b6e455a532456

[Owen Arnold]

Merge remote-tracking branch 'origin/feature/8590_Refl_gui_Live_Data_fix'

Full changeset: ​975a4d59d35ed2f9024b349dfb14c7fa127ecbf6

[Stuart Campbell]

This ticket has been transferred to github issue ​9434