Jabref: EndNote .xml import to JabRef: PDF links are not imported correctly.

[AtaZadehgol]

JabRef version Windows 10 10.0 amd64
Java 13.0.2>

• [ x ] I have tested the latest development version from http://builds.jabref.org/master/ and the problem persists

When importing an EndNote library in .xml format into JabRef, the PDF attachment links are not imported into JabRef correctly.

Steps to reproduce the behavior:

1. I followed instructions under "Importing an EndNote library into JabRef" at https://www.mcgill.ca/library/files/library/jabref_guide_2016.pdf, and exported the EndNote library as an .xml file (attached here as file "My EndNote Library-Converted_xml.txt") . See below snapshot (attached here as file "EndNote Export settings = xml.PNG") showing the export configuration of EndNote version X9.3.1 (Bld 13758).

2. I then imported the EndNote .xml library file into JabRef version 5.0.

3. The EndNote .xml library is imported into JabRef; however, the PDF file attachments are not imported automatically.

4. It appears this is a bug because JabRef expects the pdf url to be encapsulated in a style span as this is the case for the other urls. As a temporary workaround, you can add the tag around the filename.pdf (i.e., filename.pdf --> ) yourself for the moment; this should work semi-automatically via a "search and replace" in your favorite text editor. Also, remember to replace the "internal-pdf://" before the filename.pdf with the correct path in your computer (e.g., "internal-pdf://path_folder1/filename.pdf" --> "C:/EndNote/PDFs_Folder/path_folder1/filename.pdf" in Windows, etc.).

My EndNote Library-Converted_xml.txt

Log File

Paste an excerpt of your log file here

[tobiasdiez]

Relevant code: https://github.com/JabRef/jabref/blob/06e3b7259098f344606c8580fa4cd8feb3476ee9/src/main/java/org/jabref/logic/importer/fileformat/EndnoteXmlImporter.java#L263-L268

[archeaopteryx]

@tobiasdiez , thank you for the hint for the relevant code. I am working on a fix but have an addition question. Should I post directly here, or can I send you an email?

[Siedlerchr]

@archeaopteryx Please ask here so we all can help. Best is to create a draft PR and then ask a specific question about the code

[archeaopteryx]

@Siedlerchr , thank you for the advice.
The problem is that I'm not sure how to to get the URL content. As stated by OP, it's fine to have an empty style span, but having no style span means that the unmarshaller returns a null object with no content.

My naive approach would be to change the behavior of the url class referenced when defining the unmarshaller context, but that isn't possible since the context is defined by generated files. Am I missing a better solution? And, if not, is it possible for me to change the generated files?

[Siedlerchr]

I came across this issue and it's inded a problem of the generated unmarshaller. I am not an expert in xml and those schema stuff. But from my opinion it should be necessary to adjust the xsd schema to add some kind of content attribute for the url attribute to represent the text without a style.

<urls>
                <related-urls>
                    <url>
                        <style face="normal" font="default" size="100%">&lt;Go to
                            ISI&gt;://WOS:000377368200039</style>
                    </url>
                </related-urls>
                <pdf-urls>
                    <url>/Users/xxxx/_JABREFTEP/Menezes2018 - Map2Check Using LLVM and KLEE.pdf</url>
                </pdf-urls>
            </urls>

The xsd file is here:
https://github.com/JabRef/jabref/blob/master/src/main/resources/xjc/endnote/endnote.xsd