0

I have been for decades using the standard BibTeX with citations choices that place initials, dates, number sequences, etc in the text like:

... by the work of Fraenkel and Zelmanov [23] ...

By 1980, good citations systems in Mathematics were using large descriptive keys in the text, and by the 90's almost all bibliographical citation where using either number-sequence or short keys made up by abbreviation of last names and dates. We owe that to Bibtex since other branches of STEM and the Humanities developed much better citation system than ours.

Nowadays, it is desirable to have a larger target for hyperlink inserted in these citations -- since it is easier to click on a larger target... so I decide to change to Biblatex on a citation system that includes a link into the whole Name+[Date] target in the text.

Re-processing existing files, one is then left with a repetition of names like 

... by the work of Fraenkel and Zelmanov Fraenkel and Zelmanov [1996]...

whose duplicity can be eliminate by hand, but one if frequently left with a source reading:

... by the work of~\cite{ZF96}...

meaning, less readable because of the long-time use of small unreadable keys.

My question is: Is it possible to use Biblatex to eliminate the repetition in the text?

Paulo Ney
  • 2,405
  • 3
    The format of citation call-outs is not determined by BibTeX. Instead, it is determined by (a) the bibliography style that's in use and (b) the facilities of any citation management package (cite, natbib, apacite, etc) that the user may have loaded. Your claim that we "We owe [this sorry situation] to Bibtex" may be satisfyingly polemical, but it's without factual base. The claim that "other branches of STEM and the Humanities developed much better citation system than ours" is not directly verifiable. But since many of these fields do use BibTeX, the claim is likely incorrect as well. – Mico Mar 31 '18 at 02:00
  • Since you haven't specified how you use BibTeX and how you generate citation call-outs, it's not possible to evaluate your claim that changing to biblatex (and biber, right?) will provide the remedies you need. For sure, if what you need is to make the citation call-outs into hyperlinks to the corresponding formatted bibliographic entries, you can achieved that goal simply by loading the hyperref package. This is true whether or not you employ biblatex/biber instead of BibTeX. – Mico Mar 31 '18 at 02:05
  • @Mico I would be glad to provide you with plenty of examples on how citations in Mathematics got stuck into this "display of keys" while other areas went into a nicer display during the 80's - that is not debatable - you just have to look at the nice books of the era. ... but we should move our discussion elsewhere on SX probably. The question here is if Biblatex can curb the repetition. – Paulo Ney Mar 31 '18 at 03:17
  • 2
    You misunderstood my comments. I wasn't questioning whether "citations in Mathematics got stuck" in a less-than-appealing system in the 1980s. My point was that it cannot be correct to blame it on BibTeX. After all, many other fields which also employ LaTeX and BibTeX did not appear to share this experience. Blame it, rather, on some kind of unhealthy and/or stultifying conservatism that's specific to Mathematics? I think it would be helpful if you edited your query to (a) eliminate the needlessly distracting polemics and (b) provide more information about what you're actually trying to do. – Mico Mar 31 '18 at 04:14
  • 1
    I'm going to stick my neck out and say that this is not possible with TeX. You want to delete text that comes before the actual command that would have to do the deletion. That is the first issue. The closest command I know that deletes something before it is \unskip and that only removes spaces, not words. Secondly the text that needs to be deleted is only known when the citation is actually processed. But at the point where the citations are processed, the words before will already have been processed and are long gone. A Lua solution could be possible but is complicated by that fact. – moewe Mar 31 '18 at 05:46
  • 1
    Finally, at least with standard biblatex styles the hyperlink won't be that much larger since only the year is hyperlinked and not the author. Longer hyperlinks need significant modifications (https://tex.stackexchange.com/q/15951/35864). – moewe Mar 31 '18 at 05:48
  • @moewe First let me say that I think the links should be (by default) larger and encompass the whole citation instead of just the year. I have modified my styles to do that. It makes it for a much more dexterous citation when using a tablet, kindle, phone ... for reading. If one makes it easier to create a good citation -- people will do it. – Paulo Ney Mar 31 '18 at 06:31
  • @moewe I am not buying so easily - that it cannot be done! I believe one can write a TAG to the .aux file and on the second pass eliminate the duplication, preserving the linked names. – Paulo Ney Mar 31 '18 at 06:34
  • There are some challenges to hyperref labels that span more than just the year in biblatex. In particular compressed and multiple citations as well as punctuation pose challenges. The asynchronous punctuation tracker could lead to punctuation or brackets being included in the wrong link. I think this is one of the reasons why Audrey decided not to include the code into the biblatex core. But they may well be other reasons as well. See at least https://github.com/plk/biblatex/issues/428 and the comments in the question already linked. – moewe Mar 31 '18 at 06:40
  • Writing to the .aux might help, but then there is still the challenge of removing arbitrary text from a document without additional markup. https://tex.stackexchange.com/q/248632/35864 is related, the solutions there either use LuaTeX, external scripts or need a macro that is fed the entire text. – moewe Mar 31 '18 at 06:42
  • @Mico Indeed I misunderstood some of your comments, sorry for that. Indeed at about 1985 citations in Mathematics where either the simple number-sequence or very nice, then it came the "[EKG83]" tags that took over the entire field and make all publication look crap. It may be related to other things like sharing of files, the Internet and even some institutions like the AMS that used to (and even now) pass style files along. Right now it is just one in a sea of styles -- back them they were ALL the styles it was available. It is hard to not notice that this is coincidental with BibTeX. – Paulo Ney Mar 31 '18 at 06:53
  • 3
    You ask about changing the input file. If the input file is faulty, the reason is on the user side. – Johannes_B Mar 31 '18 at 07:11
  • 2
    natbib has been available since the mid-1990s, and one of its highlights is the author-year style. many mathematicians do use this style, and it's not discouraged by ams. making really long links acceptably breakable does pose problems, but they are potentially solvable. so the real problem is probably that many (most?) mathematicians are innately conservative when it comes to their publishing style. – barbara beeton Mar 31 '18 at 16:47
  • @barbarabeeton - I'm not a mathematician, but I do have friends who are, and one of them is of the opinion that when it comes to fine typography, "most mathematicians are Neanderthals". In his view, Don Knuth is nothing but a fine example of the exception that proves the rule... :-) – Mico Apr 01 '18 at 03:06
  • 1
    @Mico -- while i don't disagree at all with that assessment, i've had the very good fortune to meet a number of mathematicians who are also exceptions to that rule, though seldom to the degree that knuth is. that's why it's so important to have skilled and conscientious editors -- another, sadly, vanishing species. – barbara beeton Apr 01 '18 at 13:58

0 Answers0