## Full first names? & 2 other wishes

Joe E.
2009-01-23
2013-05-28

• Joe E.
2009-01-23

I yesterday downloaded Refbase and got it working on my desktop windows xampp PC -- testing whether we could use it for a philosophy journal articles database. I'm impressed by the work done so far: good job!

(This forum is very informative but difficult to navigate -- is there a search function I didn't find?) We want to share the metadata extracted of TOC of journals by scanning and OCRing them. Until now we are collecting them in a desktop bibliographic application called Citavi. But I want to share it through the web so I'm looking for a webbased database. We could export data from Citavi in RIS or Bibtex style and import it in Refbase.

It took me some time to get Bibutils working: in Windows you have to change the direction of the slashes. So your advice on the page http://www.refbase.net/index.php/Bibutils

<blockquote>Alternatively, if you'd like to enable Bibutils support for an existing refbase installation, you can set the path to Bibutils directly by issuing following MySQL command via a tool such as phpMyAdmin or from within your MySQL command line interpreter:

UPDATE depends SET depends_path = "/path/to/bibutils/" WHERE depends_external = "bibutils";
</blockquote>
<blockquote>UPDATE depends SET depends_path = "\path\to\bibutils\&quot; WHERE depends_external = "bibutils";
</blockquote>

After that I was able to import Bibtex without difficulties.
To my astonishment I noticed two things. 1. Christian names were shortened to their beginning letter. Is there any possibility to change this? I want author names to be as informative as possible, and I don't want to lose information already collected.
2. There's a problem with the RIS import.

Look at this RIS-File:
<blockquote>
TY  - JOUR
T1  - Erwiderung auf den Aufsatz von R. Schmidt: "Kant - Vaihinger - Adickes"
KW  -
PY  - 1928
JF  - Annalen der Philosophie und philosophischen Kritik
VL  - 7
SP  - 165
EP  - 168
ER  -
</blockquote>
The T1 field is imported (via the textfield or via the file import feature) as
"Erwiderung auf den Aufsatz von R. Schmidt: "Kant - Vaihinger - Adickes" KW  -"
This is true even of your official demo testbase, so it must be a bug. I noticed something interesting though. If my RIS File contains an ISSN of the journal the RIS import has another failure.
Look at this file
<blockquote>
TY  - JOUR
T1  - On the Dimensionality of Surfaces, Solids, and Spaces
SN  - 0165-0106
KW  -
PY  - 1986
IS  - 2
JF  - Erkenntnis
VL  - 24
SP  - 137
EP  - 203
ER  -
</blockquote>
Now after the import the ISSN reads "0165-0106 Kw-"
Does this mean that a RIS File cannot have fields without content? I tested this by deleting the empty field and importing again: voilà: no failure.
But I would like to have the possibility to import via RIS without deleting all empty fields first.

Oh and a third idea. I would very much like to have the possibility to enter different versions of an author name to make searching easier. That means: I would like to have for each author an author file connected to the database that specifies the main name of the author that is shown in the records and some alternative spellings. This would be useful for authors with pseudonyms or authors that married and changed their last name.

• Hi Joe,

> I yesterday downloaded Refbase and got it working on my desktop
> windows xampp PC

Great, and welcome to refbase!

> I'm impressed by the work done so far: good job!

>  (This forum is very informative but difficult to navigate -- is
>  there a search function I didn't find?)

In the project-specific menu bar, hover your mouse over the "Forums" menu, then click on the "Search" link that is at the bottom of the dropdown menu. This should allow you to search across all of the refbase forums. Btw, A similar search facilty is available for the refbase mailing lists.

Btw, in the future, please separate different topics into individual forum posts (with appropriate subject lines). This helps other refbase users having similar problems to more easily find the discussion. Thanks!

> We want to share the metadata extracted of TOC of journals by
> scanning and OCRing them.

If this task isn't done yet, and if some if your journal/article info is available online, you might want to try Zotero (zotero.org) for scraping the metadata. Then export as BibTeX, RIS or MODS XML and import into refbase.

> It took me some time to get Bibutils working: in Windows you have to
> change the direction of the slashes.

Thanks for the note, and glad you've got Bibutils working! I've added a short note about using backslashes as path separators on Windows to the Bibutils wiki page:

http://bibutils.refbase.net/

> I noticed two things. 1. Christian names were shortened to their
> beginning letter. Is there any possibility to change this?

Yes, but there isn't currently a global setting. To avoid the shortening of given names
to initials, do the following:

1. Open file 'includes/import.inc.php'

2. Find all occurrences of "$shortenGivenNames = true;" and change them to "$shortenGivenNames = false;"

If variable '\$shortenGivenNames' is set to 'false', given names (or any initials) available in the import data are taken as is.

> I want author names to be as informative as possible, and I don't
> want to lose information already collected.

Yes, I fully understand this. However, it is important to understand the implications of the above change: In your database, you should stick to either fully written given names OR initials. If you mix these, records will NOT get sorted correctly upon citation output.

This is also the reason why the shortening of given names is the default behaviour -- it's always possible for refbase to shorten full names, no matter whether full names or initials where given in the import data. But refbase cannot explode initials in the import data to full names.

Ideally, there would be a setting in 'initialize/ini.inc.php' which would allow the admin to choose globally whether given names should be reduced to initials on import. I've noted this on my ToDo list and I'll try to look into this. However, I cannot promise if or when this will be implemented.

And if refbase would feature a dedicated author database (see below), a smarter behaviour w.r.t. name handling and sorting might be possible. But we're not there yet.

> 2. There's a problem with the RIS import.
>
> Look at this RIS-File:
> <blockquote>
> TY  - JOUR
> T1  - Erwiderung auf den Aufsatz von R. Schmidt: "Kant - Vaihinger - Adickes"
> KW  -
> PY  - 1928
> JF  - Annalen der Philosophie und philosophischen Kritik
> VL  - 7
> SP  - 165
> EP  - 168
> ER  -
> </blockquote>
> The T1 field is imported (via the textfield or via the file import
> feature) as "Erwiderung auf den Aufsatz von R. Schmidt: "Kant -
> Vaihinger - Adickes" KW  -" This is true even of your official demo
> testbase, so it must be a bug.

Your RIS file is not valid. According to the RIS spec at:

http://www.refman.com/support/risformat_fields_01.asp

the RIS tag format must be as follows:

Each tag consists of six characters, and must be in the following format:
* Character 1: Uppercase alphabetic character
* Character 2: Uppercase alphabetic or numeric character
* Characters 3 and 4: Spaces (ANSI 32)
* Character 5: Dash (ANSI 45)
* Character 6: Space (ANSI 32)

Unless your RIS sample got mangled by the SF forum software, your RIS file is lacking the trailing space character for the KW tag. I.e., instead of "KW  -" it should be "KW  - ". If you make this change refbase should correctly import your data.

> Oh and a third idea. I would very much like to have the possibility
> to enter different versions of an author name to make searching
> easier. That means: I would like to have for each author an author
> file connected to the database that specifies the main name of the
> author that is shown in the records and some alternative spellings.
> This would be useful for authors with pseudonyms or authors that
> married and changed their last name.

Yes, I agree that this would be very useful. The same is true for journals, and we have been thinking about a dedicated journals database for a long time. However, they would require BIG changes under the hood. Note that this is simply a matter of time available for development. Refactoring of the entire database logic and all the SQL queries would be possible, but there's only so much time, and we still have other much wanted features (such as public groups/tags) on our ToDo list. So, as much as I'd like to see these features myself, this probably won't happen soon. Thanks for your understanding!

Best, Matthias