SimpleWebLinkParser does not work for Zenodo records

[Steffengreiner]

What happened?

What has happened
I wanted to try out the FAIR signposting example outlined in the compass library which includes linksmith for weblink parsing with an example Zenodo record.
When trying to parse the record with the following code snippet:

HttpClient httpClient = HttpClient.newHttpClient();


 URI uri = URI.create("https://zenodo.org/records/17179862");


 HttpRequest request = HttpRequest.newBuilder()
     .uri(uri)
     .GET()
     .build();
 HttpResponse<String> response = httpClient.send(request, HttpResponse.BodyHandlers.ofString());
 HttpHeaders header = response.headers();
 String rawHeader = header.toString();
 // Raw header of an HTTP response with link attribute;

 WebLinkProcessor webLinkProcessor = new WebLinkProcessor.Builder().build();
 ValidationResult result = webLinkProcessor.process(rawHeader);

Linksmiths WeblinkProcessor.process() validation method returns the following error:

Exception in thread "main" life.qbic.linksmith.spi.WebLinkParser$StructureException: Expected LT but found IDENT('java.net.http.HttpHeaders@c7873be7') at position 0
	at life.qbic.linksmith.internal.parsing.SimpleWebLinkParser.expectCurrent(SimpleWebLinkParser.java:240)

I checked the zenodo documentation and found out that they enabled FAIR Signposting 2 which i can also see in the Header section if i curl the record curl -I https://zenodo.org/records/17179862

Can you help me?

link: ; rel="author" ,  ; rel="author" ,  ; rel="author" ,  ; rel="cite-as" ,  ; rel="describedby" ; type="application/dcat+xml" ,  ; rel="describedby" ; type="application/json" ,  ; rel="describedby" ; type="application/ld+json" ,  ; rel="describedby" ; type="application/ld+json;profile="https://datapackage.org/profiles/2.0/datapackage.json"" ,  ; rel="describedby" ; type="application/marcxml+xml" ,  ; rel="describedby" ; type="application/vnd.citationstyles.csl+json" ,  ; rel="describedby" ; type="application/vnd.datacite.datacite+json" ,  ; rel="describedby" ; type="application/vnd.datacite.datacite+xml" ,  ; rel="describedby" ; type="application/vnd.geo+json" ,  ; rel="describedby" ; type="application/vnd.inveniordm.v1+json" ,  ; rel="describedby" ; type="application/vnd.inveniordm.v1.full+csv" ,  ; rel="describedby" ; type="application/vnd.inveniordm.v1.simple+csv" ,  ; rel="describedby" ; type="application/x-bibtex" ,  ; rel="describedby" ; type="application/x-dc+xml" ,  ; rel="describedby" ; type="text/x-bibliography" ,  ; rel="item" ; type="application/pdf" ,  ; rel="item" ; type="application/octet-stream" ,  ; rel="item" ; type="application/octet-stream" ,  ; rel="license" ,  ; rel="type" ,  ; rel="type" ,  ; rel="linkset" ; type="application/linkset+json"

What browsers are you seeing the problem on?

Firefox

Relevant log output

Code of Conduct

• I agree to follow this project's Code of Conduct

[devluiscavalcante]

HttpHeaders.toString() is not a valid Link header. It returns a Java object string, so the Linksmith parser fails.

Pass only the Link header value:

String linkHeader = response.headers().firstValue("Link").orElseThrow();
new WebLinkProcessor.Builder().build().process(linkHeader);

[sven1103]

@devluiscavalcante is right.

Hint: The HTTP Semantics specifies for fields, that , "Field sections are composed of any number of "field lines [...]" (see https://datatracker.ietf.org/doc/html/rfc9110#section-5.2).

So you might also consider to fetch all "link" fields and combine them by concatenating them with ",", which is mentioned here: https://datatracker.ietf.org/doc/html/rfc9110#name-field-order.

Linksmith can then parse all web links.