fix(nanno): preserve user-requested field order in AnnotationSourceTSV

[Copilot]

AnnotationSourceTSV.getHeaderNameAndPositions returned a HashMap, causing keySet() iteration in the constructor to produce non-deterministic ordering for fieldNames/fieldPositions arrays — potentially mismatching the user-requested CSV field order and emptyRecordResult.

Changes

• AnnotationSourceTSV.java: Replace HashMap with LinkedHashMap in getHeaderNameAndPositions so insertion order (driven by fieldNames.split(",")) is preserved when populating the precomputed arrays
• AnnotateTest.java: Updated expected value in jsonInputsTSV test to match the now-deterministic user-requested field order (aaref,HGVSc_VEP,HGVSp_VEP)
• Executor.java: Fixed two pre-existing bugs that caused endToEnd and endToEndSnpEff tests to fail:
  • Replaced the literal "java" string with getJavaPath() (derived from java.home) so the correct JVM binary is used regardless of PATH
  • Removed shell-style quotes around the classpath argument — ProcessBuilder does not use a shell, so the literal " characters were being passed to the JVM, preventing class loading
  • Switched Executor(String[], String) and Executor(String jvmArgs, String arguments, String qualifiedMainClassName) constructors to use ProcessBuilder directly with an unquoted classpath

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Existing and updated unit tests in AnnotationSourceTSVTest and AnnotateTest cover the changes.

• AnnotationSourceTSVTest.extractFieldsFromRecord
• AnnotationSourceTSVTest.getHeaderNameAndPositions
• AnnotateTest.jsonInputsTSV
• AnnotateTest.endToEnd
• AnnotateTest.endToEndSnpEff

Are WDL Updates Required?

No.

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[holmeso]

@copilot this causes a failure in a qannotate test

[Copilot]

@copilot this causes a failure in a qannotate test

Fixed in commit d646599. There were two separate issues:

1. AnnotateTest.jsonInputsTSV – the expected value was written for the old HashMap (non-deterministic) ordering. Updated to "aaref=M\tHGVSc_VEP=c.1A>C\tHGVSp_VEP=p.Met1?" which matches the user-requested field order "aaref,HGVSc_VEP,HGVSp_VEP" preserved by the LinkedHashMap.

2. AnnotateTest.endToEnd / endToEndSnpEff – pre-existing bugs in Executor.java: the 2-arg constructor passed a shell-quoted classpath ("\"...\") to ProcessBuilder, which doesn't use a shell, so the literal quote characters prevented the JVM from loading classes. Also, the "java" literal was replaced with getJavaPath() (derived from java.home) for robustness. List.getFirst()/getLast() are standard Java 21 features — no --enable-preview is needed.