Preview rdef start end ints

[will-moore]

What this PR does

Fixes an error loading Preview panel when the rendering setting channel start/end values are floats, not integers. Although we only see this error on web-dev-integration (doesn't fail locally, maybe different Python version).

Internal Server Error: /webclient/metadata_preview/image/5595/
Traceback (most recent call last):
...
 File "/opt/hudson/workspace/WEB-DEV-integration-deploy/OMERO.py/lib/python/omeroweb/webclient/views.py", line 1585, in load_metadata_preview
   % (act, i+1, c['start'], c['end'], reverse, color))
TypeError: %d format: a number is required, not float

However, this also highlighted that we shouldn't be casting start/end to Integer because some float images have very small ranges and it's important to use float start/end values.
Fixed another example of casting these to Integer in the JavaScript viewer.

Testing this PR

1. Find/import a floating point image with small dynamic range, e.g. /test_images_good/EM-data/EMD-2225.map
2. Adjust rendering settings, save them.
3. Re-load the Preview panel, adjust settings again then click the thumbnail showing your rendering settings to re-load the saved settings.
4. These actions should all support floating values for start/end.

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-DEV-merge-push#790. See the console output for more details.
Possible conflicts:

• PR Api experimenters groups #5315 will-moore 'Api experimenters groups'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• PR Reverse invert webclient #5470 will-moore 'Reverse invert webclient'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• Upstream changes
  • components/tools/OmeroWeb/omeroweb/webclient/views.py

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-DEV-merge-push#791. See the console output for more details.
Possible conflicts:

• PR Api experimenters groups #5315 will-moore 'Api experimenters groups'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• PR Reverse invert webclient #5470 will-moore 'Reverse invert webclient'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• Upstream changes
  • components/tools/OmeroWeb/omeroweb/webclient/views.py

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-DEV-merge-push#792. See the console output for more details.
Possible conflicts:

• PR Reverse invert webclient #5470 will-moore 'Reverse invert webclient'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• Upstream changes
  • components/tools/OmeroWeb/omeroweb/webclient/views.py

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-DEV-merge-push#793. See the console output for more details.
Possible conflicts:

• PR Reverse invert webclient #5470 will-moore 'Reverse invert webclient'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• Upstream changes
  • components/tools/OmeroWeb/omeroweb/webclient/views.py

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-DEV-merge-push#794. See the console output for more details.
Possible conflicts:

• PR Reverse invert webclient #5470 will-moore 'Reverse invert webclient'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• Upstream changes
  • components/tools/OmeroWeb/omeroweb/webclient/views.py

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-DEV-merge-push#796. See the console output for more details.
Possible conflicts:

• PR Reverse invert webclient #5470 will-moore 'Reverse invert webclient'
  • components/tools/OmeroWeb/omeroweb/webclient/views.py
• Upstream changes
  • components/tools/OmeroWeb/omeroweb/webclient/views.py

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-DEV-merge-push#797. See the console output for more details.
Possible conflicts:

• Upstream changes
  • components/tools/OmeroWeb/omeroweb/webclient/views.py

[will-moore]

@rgozim Did you get a chance to look at this?

[rgozim]

Tested on web-dev-merge. Works as described.

[jburel]

Could a robot test be doable in that case?
i.e. import a float image and open the preview panel

[will-moore]

Updated robot test to check that rdef settings (floats) are copied, pasted & applied correctly.

[jburel]

I think you should keep a mix of int and floats not all floats in the tests

[will-moore]

Why?

[will-moore]

They are both testing different actions (Copy -> 'Paste & Save' and click rdef thumbnail)

[jburel]

To be sure handle both correctly

[will-moore]

Added testing of copy & paste of start/end Integers into another test.

[jburel]

@will-moore thanks

[dominikl]

PR works as described, the Test Rdef Copy Paste Save test passed, too. But Test Owners Rdef failed today. Can't really rule out that this is not related to this PR. Rerun the robot tests maybe?

[dominikl]

Argh, that was yesterday's robot tests, today's not finished yet. Will check again later.

[dominikl]

Finally dev-merge-robot job ran again. There are a couple of Rdef related failed tests. I guess that might have been caused by this PR. https://ci.openmicroscopy.org/view/Failing/job/OMERO-DEV-merge-robotframework/718/robot/

[will-moore]

Hmmm - All Rdef tests are passing for me locally on this branch... Let's see what test result we get today...

[will-moore]

Rdef tests all passing today: https://ci.openmicroscopy.org/view/Failing/job/OMERO-DEV-merge-robotframework/719/robot/

[dominikl]

Good to merge then. Manual testing of the PR worked as described. 👍