How extracting only one kind of bib entry - misc

Help
1

For the bibliography of my book, I have about 1000 entries. Since it is difficult to program in Latex the style for all kind of entry types, I want to implement only one general type - misc.
This type should contain all import fields like author, title, year, etc. Not supportet information should only generate blank fields.
Although I can define those universal nececessary entry fields within the JabRef preferences, I need the merging and unification also by the fetchers when they get the information online.

It seems that this is not user definable. So, I have to program it myself.

My question: Which module should I select for this kind of adaption? Or does the architecture prohibit this? Is this task too complicated and better manually resolved?

What do you think?

2

I don’t get the question. Do you say that none (!) of the existing bibtex or BibLaTeX styles matches your needs? :sweat_smile:

3

In case you can read German books, I would recommend to start with Bibliografien | DANTE e.V..

4

Hi Rudi, i suspect you have misunderstood latex a little bit. Latex is all about separating style from content: the content of your book as well as the citations from their style in the bibliography.
The basic idea is: in an ideal world you describe in the most precise way the type of the citation you have and you put a reference (“citationkey”) in you text, configure latex and then you let latex and bibtex/biblatex do the magic. And in the end a perfectly formatted document is created. Of course, there are always glitches and bugs that have to be fixed “manually”. You don’t have to program the style for 1000 kind of entry types, because there exists already a lot of styles created and provided from nice people aroung the world. E.g. if you working in humanities in germany I would suggest you take a look at biblatex-dw – Dominik Waßenhoven ❧ Historiker .
Also note that JabRef is essentially an editor for bib files, that provide the pure citation information, which is processed by latex and biblatex.

5

I understand the question to mean “How do I make all entries have the same output format, where the output contains all the fields I want and ignores any others?”. See: Universal Import Filter

  • Option1: Apply the same output format to all entry types. Output styles typically contain different fields for different entry types, so I think this option would require the creation of a style that includes a default output format and no entry-type-specific formats.
  • Option 2: Convert all entries to the same entry type to force the same output style apply to all entries.
6

Hi Carl, thank you for arguing.

Nevertheless, I want to evaluate my idea more precisely. Until now, all efforts in this community is targeting to an accurate and most rich citation possible. But today, this is no longer necessary. For all my ca. 1000 citations, an internet search revealed the data and sources in a motsly better way than the existing bibliography references.
In former days, I had to write a post card in order to obtain an article from an author. This is obsolete now.

For all my citations I found that a small number of important data, i.e. the author, the title, the year and the organization are sufficient. But of all data, the most important one is the URL of the source.
Thus, is does not matter what kind of source we have, either an article, an audio file or a law citation, most of the special citation data (BibTex Format) is unnecessary. The essential data including the URL are necessary and suffient. A simple, but powerfull universal citation BibTex format is possible.

After changing all existing citations manually, my current approach is to normalize all new citations of articles which I obtain from the internet. Finally, I found the way to obtain this: changing the Firefox extension, and here, the Zotero BibTex translator.
Unfortunaly, after 3 years of no activity, the JabRef Browser extension throws errors when being compiled. It is not clear, whether this is due to many deprecated modules or other related issues, but this is another problem I have to deal with.

In any case: It would be very helpful if Tobias Diez could make a successful compilation of his code.

7

I have been working in a library and I can tell you that there are a lot of works out there that do not have links yet, because they were written in an age when the internet did not yet exist. There are other edge cases, like citing a page in an article or an article contained within a book from the same author in the same year and similar title (don’t ask why they would have the same title. People can do dumb things) or a specific passage within a video. You clearly don’t want to cite a “page” in a video, right? Instead you would want to cite the time (hour, minute, second etc.). Also, citationstyles usually serve the purpose to differentiate certain types of works from other types of works. Having the information available that the referenced work is a video and not a written article holds value to certain people. E.g. a teacher that told their students to only use scientific articles can see immediately, if they did it. Another problem is that links tend to vanish on the internet, because the organisations and servers hosting them go offline eventually. (Even well established identifiers like DOI or ISBN are dependent on their organisation not going bankrupt, but I digress).

For the sake of going along with your use case, I think coding this in latex could be better, because usually it is the citationstyle that renders the citation a particular way and entry data is dependent on its author and publishers. Instead of having to call every single publisher of a work to change their entry data or you doing the labour of changing the original data or modify JabRef to modify the original data automatically, it is more sustainable, if you write a citationstyle that ignores entry types altogether (and sets them all to misc (e.g. via a RegEx find and replace)), if at one point in time you decide you want to go back to the original data. After all, you already have high quality data that apparently lets you find the entries on the internet, so why degrade it? Your changes may eventually lead to a distinction between your data and the original published data, which makes it HARDER to find things on the internet, because eventually, you will find two things out there: Your new entries and the original entries by the author and publisher and then you have to decide, if you now have two different works or if it is the same thing. I wonder how you are going to make that distinction if there is only partial overlap.

8

image
image526Ă—534 28.7 KB

9

Thank you Thilo, for your insight thoughts.

Nevertheless, there are some arguments lacking:

URL: An internet reference is not only practical for literature, but mandatory in many cases. Assume that you have a reference like this:
@Misc{Breasted1930a,
author = {James Henry Breasted},
organization = {The University of Chicago Press},
title = {The Edwin Smith Surgical Papyrus},
date = {1930},
}
How do you check this source? Certainly, your library does not supply this. You have to ask the New York library for the original or a copy, which you will not get because it is an antique.
Therefore, if you do not need this source really badly, you will not check it.

However, if you also get the URL
url = {https://web.archive.org/web/20141014155705/https://ceb.nlm.nih.gov/proj/ttp/smith_home.html}

By this, the case becomes completely different. In general, this is the case for all sources. Nowadays, if you do not find the primary literature in your library and you are only interested in new developments, without an URL you will only reference the old sources, but you will not read them.

Better citations by details: Certaily, the more details you have, the better is the citation. Nevertheless, redundand details are not necessary. If you want some specific details like a page number or a minute in a video, you better do not add it as a special entry type, but use it in an option of the \cite command.

Use all details by special entry types: Using many special entries is not the best answer to the problem of accurate citation. Often, the meta information of web sites contain erronious information already. This can not be automated without an AI approach. Therefore, I prefer to check the information extracted by JabRef manually, correct it and normalize it to the Misc Entry type.

10

Have you checked with the publisher of your book what citationstyle they prefer?

11

(Also, did you really just respond with an AI generated message?)

12

No. I think, as author I am responsible for the content, including the references.

1 Like
13

??? AI generated? Why?

14

Having short summarizer sentences in front of text-sections and format them in fat is a lot of work that many humans are too lazy to do, whereas many of the LLMs that I have been using do it by default. If it’s indeed all original, then thank you for the work you did to format your text.

E.g.

“Use all details by special entry types:”
“Better citations by details:”