B3Kat - Open Data

The joint union catalogue of the Bavarian Library Network (Bibliotheksverbund Bayern - BVB) and the Cooperative Library Network Berlin-Brandenburg (KOBV) contains more than 27 million bibliographic records from university libraries, academic libraries and many other libraries of Bavaria, Berlin and Brandenburg.
The data is published under the Creative Commons License CC0 in MARCXML data format in compliance with the conventions made for Germany.

Regionally defined MARC21 fields, specifically assigned fields:

001 B3Kat-ID (example: BV018229257)
049 holdings information (ISIL), see the List of B3Kat libraries
940 Ind.1 = 1 Regional and local codes (MAB 078)
  Subfield f: selection code language area
  Subfield n: selection code interlibrary bibliographies and projects
  Subfield q: library-specific selection codes
940 Ind.1 = 2 Subfield r: ISIL of revising library (in the context of VD18, MAB 088 subfield r)
999 OAI identifier of the record (example: oai:aleph.bib-bvb.de:BVB01-011000001)

 

Download and OAI-Repository

The catalog (as of May 10th, 2018) is provided as a download in 31 packages:

Open Data B3Kat Export Part 01
Open Data B3Kat Export Part 02
Open Data B3Kat Export Part 03
Open Data B3Kat Export Part 04
Open Data B3Kat Export Part 05
Open Data B3Kat Export Part 06
Open Data B3Kat Export Part 07
Open Data B3Kat Export Part 08
Open Data B3Kat Export Part 09
Open Data B3Kat Export Part 10
Open Data B3Kat Export Part 11
Open Data B3Kat Export Part 12
Open Data B3Kat Export Part 13
Open Data B3Kat Export Part 14
Open Data B3Kat Export Part 15
Open Data B3Kat Export Part 16
Open Data B3Kat Export Part 17
Open Data B3Kat Export Part 18
Open Data B3Kat Export Part 19
Open Data B3Kat Export Part 20
Open Data B3Kat Export Part 21
Open Data B3Kat Export Part 22
Open Data B3Kat Export Part 23
Open Data B3Kat Export Part 24
Open Data B3Kat Export Part 25
Open Data B3Kat Export Part 26
Open Data B3Kat Export Part 27
Open Data B3Kat Export Part 28
Open Data B3Kat Export Part 29
Open Data B3Kat Export Part 30
Open Data B3Kat Export Part 31

You can harvest new data and data modifications since May 10th, 2018, 11:00 UTC+01:00 from our OAI repository. The request

http://bvbr.bib-bvb.de:8991/aleph-cgi/oai/oai_opendata.pl?verb=ListRecords&metadataPrefix=marc21&from=2018-05-10T10:00:00Z

would return all records that have been added or updated since the creation of the latest full export. Please note that we strongly recommend to split the request with from and until parameters into shorter timestamp area sections to recall the data consecutively.

Harvesting individual records

Individual records can be retrieved from the OAI repository by using the verb GetRecord. To specify the desired record you can chose either the OAI-identifier or the B3Kat-ID as the value of the accompanying parameter identifier.

http://bvbr.bib-bvb.de:8991/aleph-cgi/oai/oai_opendata.pl?verb=GetRecord&metadataPrefix=marc21&identifier=oai:aleph.bib-bvb.de:BVB01-011000001

http://bvbr.bib-bvb.de:8991/aleph-cgi/oai/oai_opendata.pl?verb=GetRecord&metadataPrefix=marc21&identifier=BV018229257

Please note that the syntax of the second option expands the OAI-PMH standard to provide you more comfort. A consistent synchronization with B3Kat (in particular keeping track of deletions) can only be accomplished by using the OAI-identifier instead of the B3Kat-ID!

Set-specific harvesting

In addition to the options described above, you can use an appropriate set parameter to restrict harvesting to data records with specific contents in MARC fields 024, 049 or 940.

For example you may want to request

Please note that this use of the "set" parameter for a type of free text search in certain data fields goes beyond the OAI-PMH standard functionally, but is also subject to the restrictions described in the following section.

Restrictions for set-specific harvesting

  • The (potentially infinitely large) list of all "sets" available in this way can not be requested using the OAI verb ListSets.
  • Result sets exceeding 30 records are delivered in portions (with a resumption token at the end of each portion, which can be used to request the next one if necessary). If you specify a "set" parameter with your query, the repository ignores it internally in the frist instance. Only when it comes to delivery the more general result set is split up into portions and records that do not qualify for the specified set are discarded. Therefore, apart from the resumtion token, delivered portions may well be empty - even repeatedly. This, however, is not in compliance with the OAI-PMH standard (according to the OAI-PMH XML schema definition OAI-PMH.xsd both a ListIdentifiers response and a ListRecords response must contain at least one <header/> or at least one <record/> element) and could thus confuse some of the ready-made OAI harvesters out there.
  • You can try to work around this problem by shortening the timestamp range queried per request, because this procedure tends to yield result sets small enough so they never get split up into portions at all. Doing so increases the likelyhood for but cannot guarantee getting 100 % valid OAI-PMH-XML.
  • In case you are dependent on 100% valid OAI-PMH-XML the only option remaining at the moment is to harvest without any "set" parameter and to filter out the records relevant to you on the client side yourself.

Please keep these restrictions in mind when using set specific harvesting!

Nach oben