Escape underscores on template labels

[Rdornier]

Fix part of issue #486, related to

I didn't change much because I think the markdown reading need to be preserved (see PR #481 (comment)).

For each label added from a template (tags, channel name and key-values only ; image and dataset name are correctly handled), the underscore is escaped with a backslah replaceAll("_", "\\_"). This enables a correct display of the label in omero.figure and it is compatible with Figure-to-pdf.py (_ and \_ are identically interpretated in the script).

I've also integrated this feature in the PR #528.

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-plugins-push#311. See the console output for more details.
Possible conflicts:

• PR Labels from ALL KVP  #528 Rdornier 'Labels from ALL KVP '
  • src/js/views/labels_from_maps_modal.js

--conflicts

[snoopycrimecop]

Conflicting PR. Removed from build OMERO-plugins-push#312. See the console output for more details.
Possible conflicts:

• PR Labels from ALL KVP  #528 Rdornier 'Labels from ALL KVP '
  • src/js/views/labels_from_maps_modal.js

--conflicts Conflict resolved in build OMERO-plugins-push#319. See the console output for more details.

[will-moore]

@Rdornier can you now rebase or merge in the master branch to fix the merge conflicts here (now that #528 is merged)? thx

[Rdornier]

I have a question regarding the documentation in the "Tips". It is written that to make a text italic, you need to use one start *italic* but it is also working with underscores _italic_. Should we add the two options in the tips ?
And should we also inicate how to escape the underscores to not be interpreted as markdown formatting ?

[will-moore]

Yes, tips improvements sounds good. 👍

[will-moore]

Tips look good and underscore escaping is working as expected.

[will-moore]

I just had a final thought/question before I merge this...
I wonder if we only need to escape underscores if there are more than 1 in the label?
E.g. If my tag is dead_cell then this won't be understood as markdown.
But if I have KV pair as my_key: my_value then this would need to be escaped.

I'm undecided by this:

• Pro: we don't clutter the labels unnecessarily. For most users without multiple underscores, this stays clean and simple.
• Cons: This is slightly "magic" behaviour that may be hard to understand as it's not consistent. Also, if you want to add markdown to an existing Tag dead_cell then you'd have to manually add escaping to the existing underscores.

So, I'm happy to stick with the existing behaviour, but just thought I'd ask the question?

Also, I think once this is merged, we are close to getting all this work released as https://github.com/ome/omero-figure/milestone/16.
Just a couple of my PRs to get reviewed and merged...

We could consider #524 too (needs merge conflicts fixed).

[Rdornier]

I see your point...
Preferentially, I would keep the current implementation because it is quite simple and consistant. However, if you really see an issue with escaping all underscores, then I can implement the counting. We can also ask the question to @jburel

I'm also thinking about a different solution to use the markdown. By default, nothing is interpretated as markdown. If the user introduce md: at the beginning, then the label is markdown compatible.
I don't know if this solution could be good / feasible ; it just comes out of the mind. It will however introduce a major change and we should open a need issue / PR to discuss about it.

[will-moore]

OK, agreed. 👍