I added our institute affiliations

Anything else from you @jburel? Sorry for being pushy, I'm trying to close that topic.

One heads up in terms of versioning: the removal or renaming of scripts in my mind would count as a breaking change.

So it would mean a v5.9.0 tag once this is merged?

One heads up in terms of versioning: the removal or renaming of scripts in my mind would count as a breaking change.

Seconding that. In my mind that would almost justify calling the next release of omero-scripts as v6.0.0.

Semi-related, is there any reference documentation about the existing scripts that would need to be updated? And for end-users who would have been using the previous key/value scripts in production, are there simple upgrade recommendations?

About documentation, I have made documentation for the scripts, that lives on its own for now. The plan was to merge it with your doc after the merge of this branch is completed, if I recall correctly what JM said.

The repo:
https://github.com/German-BioImaging/guide-KVpairs-scripts

The doc:
https://guide-kvpairs-scripts.readthedocs.io/en/latest/index.html

Hello,

Apologize for entering the game so late. I've tested the scripts and I found them very good !

I have a suggestion regarding the csv file that is generated. Would it be possible to add the separator used when it is generated (by adding in the first line sep=%separator% ) ?

This will ease the csv file opening and saving with Excel. For example, I have an Excel configured to recognize commas as separator. If the image names don't contain any comma, that's fine but if they contain some, it becomes a mess. I already give those scripts to some users and I made them aware of such issues but it is always better to force reading the file with the correct separator.

I hope it's not too late for that.

Rémy.

Hi @Rdornier,
I was not aware that such magic was possible for Excel. I simply thought it was poorly written because with Open Office (on my home computer) I can open CSV files without this problem.

The annoying trick I used with Excel is to open a blank sheet and in Data -> From Text/CSV select your CSV file. Only then Excel is capable of detecting the separator. So I agree to have it easier.

Two issues I can fix:

• writing the CSV better (quotes around the text so commas are fine with the export of comma as separator)
• using that metadata in the first line. Should be minor changes

Thanks a lot for the feedback. That shouldn't be too late

I was not aware that such magic was possible for Excel

I didn't know neither until today... I found this on stackoverflow

Hi @Tom-TBT

I have another feedback for you. When I try to convert a namespace from default_namspace to A_namespace, with the checkbox to create new and merge activated, it converts the KVP namespace correctly but doesn't merge the two KVPs together (before running the script, some KVPs were already existing under A_namespace).

To be able to merge the two KVPs in one, I had to run again the script, convert from A_namespace to A_namespace and check the box.
I think it is better if we can remove the extra run.

Thanks,
Rémy.

Oh, I see. The extra run could currently be avoided by giving to "Old Namespace" the two namespaces as input:
openmicroscopy.org/omero/client/mapAnnotation, A_namespace

But ticking "create new and merge" seems like an explicit agreement to merge all key-value pairs into a single A_namespace. When ticking that box, I can automatically add the "New namespace" to the "Old namespace" list.

The extra run could currently be avoided by giving to "Old Namespace" the two namespaces as input:
openmicroscopy.org/omero/client/mapAnnotation, A_namespace

Ok , it works !

But ticking "create new and merge" seems like an explicit agreement to merge all key-value pairs into a single A_namespace. When ticking that box, I can automatically add the "New namespace" to the "Old namespace" list.

Yes, checking the box to bypass the need of writing both namespaces would be very good to have.
Thanks !

@Rdornier I included both of your suggestions. You can go ahead and try them out with your users. Let me know if that solves your issues.

@Tom-TBT Thank you very much for your changes and for your reactivity ; I've just tested them and it works perfectly on my side ! I'm pretty sure it will also work for the users.

This pull request has been mentioned on Image.sc Forum. There might be relevant details there:

https://forum.image.sc/t/omero-export-list-of-file-names-from-a-project-dataset/102118/2

This pull request has been mentioned on Image.sc Forum. There might be relevant details there:

https://forum.image.sc/t/omero-which-keys-are-used/103853/3

@Tom-TBT see #216 (comment) please - not sure if you want to change the script at this stage, but the use case mentioned is clearly asking for more robust options. Such as, imho, to not to have select the KVPs on Images or Datasets only in the options of the Export_to_csv.py script, but to have also an option which would export both types of KVPs into one CSV in one run of the script. In any case, it just shows how useful this set of script really is, so thank you.

This pull request has been mentioned on Image.sc Forum. There might be relevant details there:

https://forum.image.sc/t/omero-which-keys-are-used/103853/6

Hi @Tom-TBT ,

I recently had a small issue with the export script when trying to export from an entire plate.
This was coming from the fact that there were no PlateAcquisition object for the plate (see the first screen)

The error I got was NoneType has no attribute OMERO_CLASS . I fixed it by adding a None check on the PlateAcquitision object (l226-227)

 if o is not None:
      ancestry.append((o.OMERO_CLASS, get_obj_name(o)))

Rémy