The Record Import Service (RIS) allows bibliographic, holdings or authority records to be automatically harvested or sent by users to the ANBD. RIS allows you to do all of your cataloguing on your local system then either let Libraries Australia automatically harvest the records or send the records in batch files for upload onto the ANBD. You can:
- add, edit and delete bibliographic records
- add, edit and delete your holdings
- add and edit authority records (although still being developed)
- add records in non-MARC formats (e.g. DB/TextWorks, XML)
- add records in non-Latin scripts including Chinese, Japanese, Korean, Cyrillic, Greek, Tamil, Thai, Arabic and Hebrew.
- do all cataloguing on your own system
Who should use the Record Import Service?
The Record Import Service can be used by any library, but is particularly suitable for libraries that prefer to catalogue in their own system and/or have large numbers of records/holdings to upload to the ANBD.
Prior to being authorised to use RIS, members must ensure their data conforms to the RIS specifications by supplying us with a sample file of records for test loading. The sample file should contain a range of record types (the size of the sample file will depend on the library). Members may be asked to amend and resubmit the sample file until their records meet the correct specifications.
Please send all sample files for test loading to the Help Desk.
Existing RIS users must advise us of any significant changes to their local system output prior to uploading affected records. This includes:
- character encoding changes in the local library management system e.g. from MARC8 to Unicode
- changes to control numbers or local system numbers
- the inclusion of OCLC control numbers in field 035
When you send a RIS file to Libraries Australia, you must ensure both the file name and record data are correctly formatted. File naming requirements are set out under RIS file specifications while data specifications are included in the information on each of the RIS formats (see below).
- Bibliographic - full MARC
- Bibliographic - non MARC
- Holdings - full MARC
- Holdings - abbreviated MARC
- Holdings - non MARC
- RIS for authorities (under development)
These formats enable libraries to add holdings, add new bibliographic and authority records and to add fields to existing bibliographic and authority records on the ANBD. For example, data from the incoming record may be added to the database record if it is not already present and by using the full MARC option, it is also possible (under some conditions) to substantially update database records. For more information on how records can be modified please refer to RIS processes & the match/merge algorithm.
Uploading RIS files
There are several ways in which files can be uploaded to the ANBD via the Record Import Service:
- Batch upload via http: RIS files can be uploaded via http through our Administration system. To upload files via http, members need to login to our Administration system using their RIS enabled account. Please note there is a size limit of 10 MB for files loaded in this way. Please contact the Help Desk if you need assistance with contributing via http.
- Batch upload via ftp: Files can be uploaded through our ftp server. Each RIS contributor is given their own directory on the Libraries Australia ftp server with a password to control access. To upload files via ftp (file transfer protocol), members need to login to our ftp server (ftp.librariesaustralia.nla.gov.au) and access their directory using the password they have been given. After changing to binary mode they upload their files and disconnect from the server. Please contact the Help Desk if you are interesting in contributing via ftp.
- Scheduled batch updates via OAI-PMH harvest: MARC files can be collected via the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH). Members make their MARCXML records available in an OAI-PMH accessible repository. Libraries Australia conducts an initial harvest to retrieve all available records and load them into Libraries Australia. Thereafter, update harvests are run weekly to capture changes from the intervening period. The returned new, updated and deleted records are then loaded or removed from Libraries Australia automatically. The benefits of this method of contributing records is that holdings are kept up-to-date and staff time is not required at member libraries to extract records from the local catalogue and upload RIS files. However, members need to provide the Base URL, Set name and Metadata Format of their OAI-PMH accessible repository. Please contact the Help Desk if you are interesting in contributing via OAI.
- Manual batch upload: Files can also be processed manually if they are in comma delimited ASCII format or Inmagic dump file. Note that the records in these files must have either ANBD or LC control numbers. We can assist you with adding these control numbers to your records before you sent them to us. Please contact the Help Desk for further information.
Loading RIS files onto the ANBD
Once the files have been uploaded via RIS, the data is converted and loaded onto the ANBD at regular intervals by Libraries Australia staff. Please note that we reserve the right to determine priorities for file loading based on the significance of the data for national resource sharing.
For more detailed information on what happens to the records once they are uploaded by contributors, please refer to RIS processes & the match/merge algorithm.
A processing report will be sent to the email account established for your RIS profile upon completion of each file load. The report will describe the file that was processed and when loading was completed.
Rejected records reports
Records rejected during loading will be placed on the Libraries Australia ftp server for retrieval, correction and resubmission. They can also be downloaded from the LA Admin. System. Reject files come in 2 formats:
- errors.mrc: file of MARC records without the error messages.
- errors.txt: text file of records (in full view) with the error messages.
A PDF providing a comprehensive explanation of error messages contained in the "errors.txt" report is available here.
If records contain both errors and warnings they will only be reported in the files that contain errors (e.g. ‘errors.mrc’ or ‘errors.txt’).
Contributors receive notification of these reject files via an email message. Rejected files will be automatically deleted from the Libraries Australia ftp server after 90 days.
A separate email alert is generated and sent to the contributors account when reject files are available and created.
Additional reports are available and include:
- review.txt: text file of records (in full view) that have tentatively matched database records and require further review.
- unprocessed.txt: text file of records (in full view) that could not be processed correctly through all matching and merging steps and were not added to the database or used for updating existing records.
- warnings.mrc: file of MARC records which generated warnings without the warning messages.
- warnings.txt: text file of records (in full view) which generated warnings with the warning messages. Note: The warnings do not prevent records from processing through RIS or being matched and merged.
Detailed statistical reports are available from the LA Admin. System.