Work on plugins

EPrints import and Export Plug-ins

We tested most of the import and export plug-ins available from EPrints although not rigorously. The majority that were tested worked well. However an issue with all import plug-in is that we are unable to get a copy of the abstracts; this is partly due to the copyright/licensing restrictions.

BibTex

Import

Imported successfully however had to top and tail the file to get rid of the researcher's comments. Import via cut and paste from text file into Eprints import box

Export

Exporting the hits from a simple search worked fine.

DOI via CrossRef

Import

Metadata have been successfully obtained using the EPrints DOI plug-in however the dataset is not always complete. This is not because of the plug-in not working but the CrossRef database does not contain complete citation data sets. DOIs can be cut and pasted (individually or as a list) into the import box.

Assuming the DOI has been registered (not always the case – particularly for recent works), using the DOI will populate:

• Article Title
• Journal title
• Volume
• Number
• Start page (not complete range)
• Year (not full date)
• DOI (in EPrints “Identification Number” field)
• Official URL field (link to the published copy using http://dx.doi.org/doi)
• ISSN (no spaces or hyphen)

The major drawback with CrossRef is lack of author data.
Some of our test imports have pulled in very little data from CrossRef– e.g. the DOI field is populated but all others are empty. At the moment, EPrints records a “successful” import from DOI, even where most fields are blank. As far as we can tell from our testing, the EPrints DOI plug-in assumes the imported item is a journal paper; this may not always be the case (e.g. papers in Lectures Notes in Computer Science may be better described as book chapters).

Export

No export plug-in available.

EndNote

Import

Used this plug-in to import citation metadata in bulk. The import worked well. The issues we had were with the EndNote software. Importing files into EndNote does not work consistently.

Sample import from Philosophy. Rejected some entries but accepted others – ratio about 1:4. Unable to discover differences between the accepted and rejected records.
Have imported metadata only. Have not tested full text attachments.

Export

Successfully exported EndNote format outputs from EPrints.

PubMedID

Import

PubMedID works with one ID at a time cannot batch upload. Batch import of PubMedIDs work fine if importing from the text box.
There are issues with importing using PubMedID in that it has not worked consistently. When it works has a rich and good quality metadata.

It’s worth being aware of the distinction between ids used in PubMed and ids used in PubMed Central (PMC)/ UK PubMed Central (UKPMC); they are not the same. E.g. these two ids refer to the same paper:
PMID: 17210079
PMCID: PMC1774569
(It’s possible to find out the relevant PMCID from PMID and vice versa from the PMID : PMCID Converter site at http://www.ncbi.nlm.nih.gov/sites/pmctopmid ).

The current EPrints plugin is set up to import metadata from PubMed using PMID. A PMC/UKPMC plug could be useful; particularly as PMC holds open access content whereas PubMed holds metadata..

Other Import Plug-ins

We have not tested either PubMed XML or XML import plug-ins.

Export Plug-in from EPrints

Currently the export plug-ins available from EPrints are the following:

We tested exporting the outputs from searches for all the different formats (listed above) and all worked apart from MODS. Tested exporting in the MODS format in two internet browsers Explorer and Mozilla FireFox both had XML parsing problems.

Exporting in the Reference Manger format opened up EndNote however the file did not appear so could not be read. This could because Reference Manger was not installed on the computer being used.