fix(jira): include fields=*all in Jira Server issue search query (fixes #9122)

[Thundercloud12]

Summary

When collecting issues from Jira Server (on-premise) via /api/2/search in setupIssueV2Collector, the fields query parameter was omitted.

According to Atlassian's REST API specifications, when fields is omitted, /rest/api/2/search defaults to returning only *navigable fields. Since worklog is not a navigable field, Jira Server omits the embedded fields.worklog array from the search response payload.

This resulted in the following downstream collection failure:

1. extractIssues could not extract the embedded worklogs, which contain the first page of up to 20 worklogs, leaving issue.worklog_total = 0.
2. collectWorklogs only queries issues where worklog_total > 20, assuming issues with 20 or fewer worklogs were already collected during the first step.
3. Consequently, Jira Server issues with 20 or fewer worklogs had 0 worklogs recorded in jira_issue_worklogs.

This PR fixes the issue by:

• Adding query.Set("fields", "*all") to setupIssueV2Collector in issue_collector.go, bringing it in line with setupIssueV3Collector for Jira Cloud.
• Adding the unit test TestExtractIssuesWorklog in issue_extractor_test.go to ensure embedded worklogs are correctly parsed and recorded.

Does this close any open issues?

Closes #9122

Screenshots

Unit tests passing:

=== RUN   TestExtractIssuesWorklog
=== RUN   TestExtractIssuesWorklog/extracts_embedded_worklog_when_fields_contain_worklog
=== RUN   TestExtractIssuesWorklog/no_worklog_extracted_when_worklog_field_is_absent
--- PASS: TestExtractIssuesWorklog (0.00s)
PASS
ok      github.com/apache/devlake/plugins/jira/tasks    0.020s

Other Information

• Official Atlassian documentation for /rest/api/2/search states that the search resource returns only *navigable fields by default.
• Fields such as comment and worklog can contain large amounts of data, so they are excluded from the default *navigable set to keep API responses performant.
• These fields must be explicitly requested, either individually or by requesting *all.
• setupIssueV3Collector for Jira Cloud already uses query.Set("fields", "*all"); this change aligns Jira Server with the same behavior.

[Thundercloud12]

@klesh please have a look, thank you!