ProCite to Zotero Conversion: Translator, RIS, and Testing

maevepotter

Just wanted to make sure, since I posted several times, did you see the error above?

It seems that workform "Book Short Form" is not importing multiple authors correctly.

See screenshot: http://dl.dropbox.com/u/19141190/error with multauthors slashes.png

The // between names are not being registered as separating the authors. Perhaps because there are no spaces? Because it will only do the first authors "name//nextauthorslastname" and then nothing else because of the comma after the second authors name. So those names after the second authors last name are being lost completely. This is true across all multiple author records I have looked at so far. I am using the new ProCite.js file from above.

here is the examples RIS test:

TY - BOOK
N1 - Record ID: 2753
A1 - Bloomer, Kent//Moore, Charles
T1 - Body, Memory, and Architecture
CY - New Haven
PB - Yale University Press
PY - 1977
KW - spatial analysis
ER -

aurimas

I don't think we'll be able to support something like that. Are the authors entered properly in ProCite? If they are, then we may have to edit the export format for Book Short Form.

maevepotter

This should be fairly standard. I figured it might just be a missing bit of code. The authors are entered the same as elsewhere in the database:

see here:
http://dl.dropbox.com/u/19141190/bookshortform.png

They are simply not importing as separate people. Does that make sense?

It goes into Zotero
Author: Bloomer, Kent//Moore

Instead of
Author: Bloomer, Kent
Author: Moore, Charles

aurimas

You'll have to fix the output style. // between authors is not part of the RIS specification.

If you go to edit the style, Bibliography tab, Book Short Form. Click "Configure Fields" (top of the window, right under font selector). Select, Authors (07) and go to Separators. Put ^pA1 - in all the fields there. (Note the spaces, two before - and one after)

adamsmith

Just to be clear - almost everything aurimas' has done in this thread is because ProCite's export doesn't conform to the RIS standard.

There should most certainly not be multiple authors within one A1 field.
On the prior page of this thread you have one entry with
A1 - Dickson, Michael
A1 - Littleton, Judith
A1 - Frohlich, Bruno
A1 - Amgalantögs, Tsend
A1 - Karstens, Sarah
A1 - Floyd, Bruce

how are these entered in Procite? This is the way multiple authors should be exported to RIS. While it looks possible to me to divide authors on slashes, it's yet another hack and the more of those are in the translators, the less stable it will be in the end.

aurimas

@adamsmith
Actually, authors in ProCite are entered with // separating each author, but for most workforms ProCite exports authors with ^pA1 - separator, but for Book Short Form, the separator is set to //

maevepotter

Thanks for the clarifications Adam and Aurimas. Well thank god for you both helping us get out of ProCite. It would have been almost impossible to do otherwise with such a large database with so many workforms. ProCite in general sounds like it is a nightmare to code around. Thank you for your patience and assistance so far. It is really appreciated.

I just tried the edit to the RIS style, and it seems to have worked beautifully.

Thanks so much for your expertise. I would have never known to check for that separator.

aurimas

No problem at all. Let us know when you're done importing all your items and there are no more bugs that we can iron out. I'll try to incorporate this translator into the main RIS translator, so that everything is streamlined in the future

maevepotter

One question:

As I move these files to a separate account, can I trust the Zotero RDF export? I figured since it would go from Zotero to Zotero, it would be fine, but the documentation page says that it is not a reliable way to back-up/transfer.

Opinions?

Rintze

Do you need to transfer (a lot of) files as well, or just the item metadata? In general, you're better off with syncing or just making a copy of your Zotero data directory.

maevepotter

There are not very many PDFs attached to the records. Some things from newer imports seem to have files from the websites.

But you would recommend syncing or copying the entire Zotero data directory to the other computer/account?

Thanks..

EDIT: Is there a problem with direct copying the Zotero data directory of one account, ie. maevepotter, to a different account?

aurimas

EDIT: Is there a problem with direct copying the Zotero data directory of one account, ie. maevepotter, to a different account?

Should be safe
http://www.zotero.org/support/kb/transferring_a_library

maevepotter

Yep it worked well.

Thanks everyone! I have all of the citations into Zotero now, and it seems to be working well. I will let you know if anything crops up, but I think it is all looking pretty good! Thanks Aurimas! I know this will really help all of those ProCite users out there, it's amazing. Takes some work, but its great. Let me know if you want a tutorial on this anywhere that I could write up to contribute to the project.

aurimas

Let me know if you want a tutorial on this anywhere that I could write up to contribute to the project

I'll have to see how much of the modified translator we can merge into Zotero's RIS translator. Hopefully I can make all of the changes play along with non-ProCite RIS files, then the tutorial should be pretty short.

Takes some work, but its great.

I wonder what you mean by this. Do you mean you need to tweak some things (besides what's already been discussed) after importing with the latest version of the translator?

And thank you for not giving up on me. It's been a looooong journey :-)

maevepotter

I basically mean that there is a lot of prep work involved to make it go right. As far as separating all of the exports from Procite into specific workforms so you can check on what is happening on a macro scale to records of similar types... and then tweaking the RIS in procite to go to either BOOK or CHAP, or add the ^pA1 - separator to book short form. Or having to search for #4 = EMPTY etc, to separate their book chapters from books, if they used Book Long Form for both types, since the RIS has to be modified to either BOOK or CHAP in those cases, and can't be done to the workform en masse without errors.

Those would all be things someone would have to do on their end before using the translator. And they are not intuitive things, so there would have to be a guide on how to. I suppose they could skip exporting by workform, but if their library was really large, the export often fails.

Another issue would be that everyone uses the workforms differently, so they may have new issues that I did not encounter. Therefore, it would be smart for them to export by workform so they can more easily see if anything went wrong.

Also there are always errors, like one / instead of // which messes up the author's name. This is of course not the fault of the translator, but was still an error I often find and correct, which is due to human error in the past database. For this I have to compare the data between both programs.

Beyond that, there is just the downloading, and finding the right folders etc. I am very computer savvy and know computer programmers personally and how they work. I think I had a huge advantage there with being able to know the jargon, and understand how a computer works in general. Some less tech-savvy users will find this process a lot more difficult, I imagine. That was why I was thinking the tutorial might be helpful.

Although I will say, unless they are professional writers, they may not require the same level of fidelity to the original data that we did. So ::shrugs:: might be ok to just use the translator without as much detailed error checking as I did. Depends on what they use Zotero for.

And thank you for not giving up on me. It's been a looooong journey :-)

No problem! I could never have done it without you, thanks for hanging in there and helping!! I appreciate all the work you went through. :)Others definitely will too!