com.atlassian.jira.rest.client.api.domain.Issue.getIssueLinks() returns null unexpectedly

[Tim Vellacott]

the javadoc for com.atlassian.jira.rest.client.api.domain.Issue.getIssueLinks() says it returns issue links for this issue (possibly nothing) or null when issue links are deactivated for this JIRA instance

now, in our instance (on JIRA 7.3.3), links are enabled, and we have issues that contain links.  but getIssueLinks() is returning null.  i'm wondering if there's some other setting (or if i'm doing it wrong).  here's my code snippet (the first branch is always invoked):

Promise<SearchResult> searchJqlPromise = jiraRestClient.getSearchClient().searchJql(this.getQuery(proj), ISSUES_PER_CALL, startAt, fieldSet);
for (Issue issue : searchJqlPromise.claim().getIssues()) {
if (null == issue.getIssueLinks()) {
  System.err.println("for issue " + issue.getKey() + "; getIssueLinks() returned null");
} else {
  for (IssueLink link : issue.getIssueLinks()) {
    String targetKey = link.getTargetIssueKey();
    System.err.println("for issue " + issue.getKey() + " there is a link with key " + targetKey);
  }
}

thanks!

 I was using JRJC 3.0.0 but have now found 4.0.0 exhibits the same behaviour.

[Tim Vellacott]

For the benefit of all, I solved my own problem by trawling through the source code.

In the jiraRestClient.getSearchClient().searchJql() call  to fetch the issues, the last parameter is a set of Strings that specify the fields to return.  This set has to include "issuelinks" for getIssueLinks() to return the list.  ("*all" will also work). 

To reduce data transfer volume (I'm often retrieving 30,000 issues down a WAN) I had cut down the list of fields to the minimum.  Adding in the "issuelinks" explicitly has solved the problem.

JRJC team, you could maybe consider put a small caveat in the javadoc for getIssueLinks() along these lines.