Skip unique key detection for contacts

[sbfnk]

Summary

Fixes #278

• Skip unique key detection for contacts entirely in load_survey()
• If merging creates duplicate contact rows, the merge is rejected (conservative approach)
• We don't have methodology for longitudinal contacts, so this is the appropriate behaviour

This fix prevents load_survey() from failing when contact files lack a cont_id column.

Summary by CodeRabbit

• Bug Fixes

  • Fixed load_survey() failing when merging contact files without a cont_id column
  • Enhanced handling of duplicate entries in longitudinal survey data with stricter validation
  • Improved participant key detection with more robust fallback mechanisms for invalid keys

• Documentation

  • Added new package author to acknowledgments

[coderabbitai]

📝 Walkthrough

Walkthrough

This PR adds Nicholas Tierney as a package author to the DESCRIPTION file and fixes a bug in load_survey() where contact files lacking a cont_id column caused failures by improving duplicate handling and key detection logic in the file merging process.

Changes

Cohort / File(s)	Summary
Metadata & Documentation DESCRIPTION, NEWS.md	Added Nicholas Tierney as author with email and ORCID identifier; documented bug fix for missing cont_id column handling.
Survey Loading Logic R/load-survey-utils.R	Tightened duplicate handling in try_merge_additional_files to reject contact file merges with duplicates; adjusted base key determination to consistently use "part_id"; enhanced participant_key validation to check column existence and duplicates; generalized final_detected_key assignment beyond participant-only types.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

Possibly related PRs

• Prepare for 0.5.0 release #276: Adds Nicholas Tierney to Authors@R in DESCRIPTION, directly parallel to author metadata changes in this PR.
• Refactor load surveys i203 #208: Modifies try_merge_additional_files and key-detection logic within the same utility file affected by this PR's fixes.
• Fix loading of longitudinal surveys with repeated observations #259: Adjusts longitudinal merge logic and duplicate/key detection handling in R/load-survey-utils.R, sharing the same target functions as this PR.

Poem

🐰 A contact without ID? No fret!
We'll detect the key, the safest bet.
With Nicholas aboard to guide the way,
Survey loading brightens this fine day! ✨

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'Skip unique key detection for contacts' accurately reflects the main code change in R/load-survey-utils.R where contacts are now excluded from unique key detection.
Linked Issues check	✅ Passed	The PR directly addresses issue #278 by preventing load_survey() from failing when contact files lack cont_id, implementing the solution to skip unique key detection for contacts.
Out of Scope Changes check	✅ Passed	All changes are within scope: DESCRIPTION adds a new author, NEWS.md documents the fix, and R/load-survey-utils.R implements the core fix for issue #278.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch fix-contact-merge-duplicates

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}