Group names can now be rendered using LaTeXs commands

[oscargus]

Inspired by #1681

• Change in CHANGELOG.md described
• Manually tested changed features in running JabRef

[tobiasdiez]

Mhh, I think this is the wrong place to add this functionality. If I name my group \relax (or \beta) then I probably want to keep this name.
But I think it is a good idea to add the Latex-to-Unicode-formatter to determine the name when groups are automatically created.

[oscargus]

Really!? If you name a group \beta, in a program dealing with LaTeX, would you be upset if it shows the beta properly? (The \relax use case is, ehh, far fetched?)

[stefan-kolb]

Do we really need to allow latex code for group names? Why is normal text not enough? I'd rather not have this option and keep it simple.

[oscargus]

See #1681. Names containing accents is the main reason. Already now, the names can contain LaTeX code (except that the reading is flawed, again see #1681). It's just that they are properly rendered now. I do not really see the problem.

[tobiasdiez]

For me the problem is that the Latex -> Unicode converter is not good enough to handle arbitrary text and I think we may run into problems where names of groups are not properly rendered (i.e. "A or B" in the form "A\B", or similar things with backslash). As a user, I kind of expect that JabRef takes what I give it as a name (especially since I can provide unicode strings, I don't see the reason to support latex accents). Moreover, we might run into big performance problems with groups again.

The idea is nonetheless good. Just put the conversion code in the automatic generation of groups (since there the user has no control over the names of the groups).

[oscargus]

I see your point, although I'm not fully agreeing with it (I'd rather write $\Delta\Sigma$ than look up the Unicode key and writing \ in anything related to LaTeX expecting it to stay a backslash seems questionable... ;-)), but you solution is fair enough and leads to more or less the same end result. Only question is if BibTeX is bothered by Unicode characters in fields that are ignored anyway? (Yes, there are good reasons to use BibTeX for many people.)

Could you have a look at why the backslashes are removed when reading back the group definitions? (Since you have better insight into the group package.)

[oscargus]

I changed it to do the conversion upon generation of automatic groups (and correspondingly changed the CHANGELOG entry).

[tobiasdiez]

I would make the formatter a normal variable (initialized directly before the for loop below) instead of static final.

[oscargus]

I had it like that first. ;-)

[tobiasdiez]

Ok, just make the fomatter a normal variable and then this can be merged. I will have a look at the serialization / parsing problem as soon as I'm back in Germany (i.e. in a month)