Incorrect bibkey creation for >4 authors

[umuen]

JabRef version

5.12 (latest release)

Operating system

macOS

Details on version and operating system

MacOS Ventura 13.6.5 (22G621)

Checked with the latest development build (copy version output from About dialog)

• I made a backup of my libraries before testing the latest development version.
• I have tested the latest development version and the problem persists

Steps to reproduce the behaviour

When using the Citation Key Generator specification: [authIni4][year][journal:abbr:lower] it works as it should for 1, 2, 3 or 4 authors. However, if there are more than 4 authors it only uses the first THREE initials but not the first FOUR as it should. I was not able to check with the newest build as Apple does not allow me to open it (and I had to tick the boxes to be able to submit the report - sort of a catch 22)

Appendix

...

Log File

Paste an excerpt of your log file here

[Siedlerchr]

@umuen To run the newest development build you have run it first then allow it in settings under data protection and security (Make sure to pick the right one for your OS M1/m1 - arm64 suffix

[umuen]

thanks for the info. I tested it and the issue is still there in the most recent build

[LoayGhreeb]

@Siedlerchr, it seems the issue is related to PR #9703.

We changed the handling of an "overflow" of authors at [authIniN]: JabRef uses + to indicate an overflow. For example, [authIni2] produces A+ (instead of AB) for Aachen and Berlin and Chemnitz.

However, I would like to know where this + appears and why this feature was added.

jabref/src/main/java/org/jabref/logic/citationkeypattern/BracketedPattern.java

Lines 1039 to 1050 in 4432dcf

	final boolean lastAuthorIsOthers = authorList.getAuthor(numberOfAuthors - 1).equals(Author.OTHERS);
	if ((n > 1) && ((n < numberOfAuthors) || lastAuthorIsOthers)) {
	final int limit = Math.min(n - 1, numberOfAuthors - 1);
	// special handling if the last author is "Others"
	// This gets the single char "+" only
	AuthorList allButOthers = AuthorList.of(
	authorList.getAuthors()
	.stream()
	.limit(limit)
	.toList());
	return authIniN(allButOthers, n - 1) + "+";
	}

[koppor]

@umuen Thank you so much for your report. I checked the documentation at https://docs.jabref.org/setup/citationkeypatterns#author-related-field-markers. It reads

The beginning of each author's last name, using at most N characters in total.

First, I thought, we should introduce [authIni4+] with the semantics to show the first 4 lastnames (and thereof the first letter), and then adding a + if there are more than four. However, this is maybe unneeded.

Thus, we should change the logic to contain 4 letters. The only "exception" is "and others" - this is rendered as "+". Meaning:

Aachen and Berlin and Chemnitz and others

Gets

ABC+

(and not ABCo or something else weird stuff)

[koppor]

@LoayGhreeb Typically + is used for saying "and more". In BibTeX terms and others. In BibTeX, one can explicitly write and others. This is a fixed term. Defined by BibTeX. Nothing we can change. Need to accept that. - JabRef could not handle "and others". It thought "others" be a name instead of an indicator for "and more". Therefore the issue JabRef#626. -- The fix for the isuse here is to distinguish and others with "more names" exist. Example:

Aachen and Berlin and Chemnitz and Düsseldorf and Essen

should lead to

ABCD

but

Aachen and Berlin and Chemnitz and others

should lead to

ABC+

[Siedlerchr]

@umuen Could you give us a sample entry where this occured?

[umuen]

Below is an example where it should give "GAHS2023jpcc" but gives "GAH2023jpcc". I would like to point out that the behaviour was correct up to some weeks/months ago (I am not sure up to when exactly), i.e., for greater equal 4 authors only the initials of the first authors are used (without a "+" or similar). This is how my large bibfile is structured and is used in citations in many documents. It would be most unfortunate if the behaviour remains changed. Thanks for looking into this.

@Article{GAHS2023jpcc,
  author   = {Golomb, M. and Arndt, N. B. and Honnigfort, C. and Shakhayeva, B. and Ravoo, B. J. and Braunschweig, B.},
  journal  = {J. Phys. Chem. C},
  title    = {Molecular kinetics and wetting dynamics of self-assembled monolayers with fluorinated arylazopyrazoles},
  year     = {2023},
  pages    = {15316--15325},
  volume   = {127},
	  doi      = {10.1021/acs.jpcc.3c02472},
}

[koppor]

@LoayGhreeb May I assign you? I can also put that as small university project.

[LoayGhreeb]

Typically + is used for saying "and more". In BibTeX terms and others. In BibTeX, one can explicitly write and others.

I understand this, but where does the + sign actually appear?
If I set the Citation Key Generator to use [authIni2] in the preferences, an entry with author = {Aachen and Berlin and others}
results in the generated citation key being A instead of A+.

My question is not about the "and others" being replaced with '+'. In PR #9703, there are two changelog entries. The first one is:

We changed the handling of an "overflow" of authors at [authIniN]: JabRef uses + to indicate an overflow. Example: [authIni2] produces A+ (instead of AB) for Aachen and Berlin and Chemnitz.

This is what I'm asking about. Why is this done?

May I assign you? I can also put that as small university project.

Sure.

[koppor]

I understand this, but where does the + sign actually appear? If I set the Citation Key Generator to use [authIni2] in the preferences, an entry with author = {Aachen and Berlin and others} results in the generated citation key being A instead of A+.

I see that, too.

Think, this is a bug.

We changed the handling of an "overflow" of authors at [authIniN]: JabRef uses + to indicate an overflow. Example: [authIni2] produces A+ (instead of AB) for Aachen and Berlin and Chemnitz.

As discussed above. This is wrong. The result has to be "AB".

Code needs to be fixed.