Opened 8 years ago

Closed 4 years ago

#1202 closed defect (fixed)

Need Translator For Primo Library OPAC

Reported by: tjowens Owned by: ajlyon
Priority: critical Milestone:
Component: translators Version: 2.1
Keywords: Cc:

Description

Primo is a relatively new EX Libris Product, at this point at least NYU and UMN are using the catalog. In both of these catalogs users can click on a "Full Record view" from individual item pages. These full record pages work fine through the Aleph translator. See the following links.

University of Minnesota
http://prime2.oit.umn.edu/primo_library/libweb/action/search.do

Individual record:
http://prime2.oit.umn.edu/primo_library/libweb/action/display.do?ct=display&doc=umn_aleph001903500&indx=1&srt=rank&tab=default_tab&mode=Basic&dum=true&fn=search&frbg=&vl(13244016UI1)=all_items&vl(185109809UI0)=any&ct=search&vid=TWINCITIES&indx=1&vl(freeText0)=robot&vl(1UI0)=contains

Full Record View of same item:
http://umnlib.oit.umn.edu/F/3QBTFTSQAXNBR3IBC2MYVDLM5JD3L9XLF95QD2RPSNV1VKMBDL-93274?func=direct&=&=&local_base=umn01pub&doc_number=001903500&pds_handle=GUEST

New York University
http://www.bobcat.nyu.edu:1701/primo_library/libweb/action/search.do?vid=NYU&fromLogin=true

Detail view of individual item:
http://www.bobcat.nyu.edu:1701/primo_library/libweb/action/display.do?ct=display&doc=nyu_aleph000824282&indx=4&srt=rank&tab=all&mode=Basic&dum=true&fn=search&frbg=&vl(177077488UI0)=any&ct=search&vid=NYU&vl(34909979UI1)=all_items&indx=1&vl(freeText0)=robot&vl(1UI0)=contains

Full Record view of same item:
http://hopkins1.bobst.nyu.edu:8991/F/1S4DINCP33D5LHYTH75VA8K2BARI8EANVHM4MN8B3SYEFHCBDB-03491?func=direct&local_base=PRIMOCOMMON&doc_number=000824282&pds_handle=GUEST

Attachments (2)

Primo.js (3.1 KB) - added by ajlyon 6 years ago.
Primo translator using RIS output
Primo.2.js (16.4 KB) - added by ajlyon 6 years ago.
Version using PNX again, but fetching more data.

Download all attachments as: .zip

Change History (10)

comment:1 Changed 8 years ago by connie

To add: Emory University's site is: http://www.library.emory.edu:32888/DB=discovere
Do you need something more?

comment:2 Changed 8 years ago by tjowens

  • Owner changed from stakats to mcburton
  • Priority changed from major to critical
  • Status changed from new to assigned

comment:3 Changed 8 years ago by connie

Just a couple of added notes -- the Primo site is not very scraper/translator friendly.

If you append &showPnx=true to the end of the URL, you can see the XML for the "books and more" section. Primo does not have appear to have anything for articles, so ignore that. An example from Emory is:

http://discovere.emory.edu/primo_library/libweb/action/display.do?ct=display&doc=emory_unicorn2453437&indx=1&frbg=&dum=true&vl(1UI0)=contains&vid=discovere&srt=rank&indx=1&tab=default_tab&doc=emory_unicorn1669464&ct=search&scp.scps=&vl(freeText0)=marilynne%20robinson%20gilead&fn=search&vl(40482334UI1)=all_items&mode=Basic&vl(40482333UI0)=any&showPnx=true

Note that the PNX/XML structure is customized to campus systems -- for example: You'll note from the examples from NYU and UMN that when you look up the PnX, they note for the tags sourceformat and sourcesystem "0" and "Aleph" while an Emory record will allow MARC21 and DC (not sure if there are more). For a MARC21 example, take the one above and you'll see that sourceformat and sourcesystem are instead "MARC21" and "Unicorn."

We've made a request to see if it's possible to get any more information on the PNX structure.

We've also made the strong suggestion that the software should embed COiNS or implement a standard citation structure for its data available to the public, as many university library sites do.

comment:4 Changed 8 years ago by mcburton

(In [4204]) Addresses #1202, First cut at the Primo translator, probably needs some wider testing on different primo installations (currently tested with NYU, Emory and UMN). Also probably some further data cleaning required.

comment:5 Changed 6 years ago by ajlyon

I've written a new version of the Primo translator that uses the carefully-hidden RIS output. This sometimes is less detailed than the proprietary PNX, but it should be more universal. In cases where it is altogether insufficient, perhaps the fact that RIS improvements will improve both Zotero and Endnote support will make the improvement of export data more likely. New version attached.

Changed 6 years ago by ajlyon

Primo translator using RIS output

comment:6 Changed 6 years ago by ajlyon

  • Owner changed from mcburton to ajlyon

Apparently the RIS output is unreliable as well; in the Bibliotheque Sainte-Barbe I ran into this RIS record for the item http://www.bsb.univ-paris3.fr/primo_library/libweb/action/display.do?ct=display&doc=primo_bsb000079921 :

TY  - BOOK
PB  - New York : Oxford University Press
JA  - New York
AU  - 19970822d1998----k--y0frey50------ba
T1  - Why we watch : the attractions of violent entertainment
Y1  - 1998
KW  - Violence dans les médias -- États-Unis
KW  - Violence -- Aspect social -- États-Unis
KW  - Culture populaire -- États-Unis
ID  - primo_bsb000079921
ER  -

The RIS approach is also not good enough for the British Library implementation of Primo, which also spits out bad RIS, with author names appended to the title and other loose ends:

TY  - BOOK
N1  - Includes bibliography: p. 176-[179].
PB  - Kazanʹ : Izdatelʹstvo Kazanskogo universiteta, 1968.
AU  - Tumasheva, D
T1  - I︠A︡zyk sibirskikh tatar / D. G. Tumasheva.
CY  - Kazanʹ
ID  - BLL01011919904
KW  - Tatar language
ER  -

Unfortunately, PNX isn't always very good either, giving us data like:
Jeffrey H Goldstein (Jeffrey Haskell). Why we watch : the attractions of violent entertainment / edited by Jeffrey Goldstein. New York ; Oxford: Oxford University Press, 1998., n.d.
for http://searchbeta.bl.uk/primo_library/libweb/action/display.do?ct=display&doc=BLL01012590758

I'm leaning again towards using PNX, and I'm attaching another version of the translator that gets language right more often (using a full mapping table for ISO 639-2) and gets names right more often, but neither PNX nor RIS exposes the richness of the underlying data in a reliable fashion.

Changed 6 years ago by ajlyon

Version using PNX again, but fetching more data.

comment:8 Changed 4 years ago by adamsmith

  • Resolution set to fixed
  • Status changed from assigned to closed

all samples above work now. Closing as fixed.

Note: See TracTickets for help on using tickets.