Code Review

The PR adds variant search result document indexing logic to VariantSummaryCurationIndexer and then disables the indexer via config. Two issues worth addressing before re-enabling:

Bug: Potential NPE / IndexOutOfBoundsException in converter (VariantSearchResultConverter.java:21)

CuratedVariantGenomicLocationAssociation variantLocation = doc.getVariantList().get(0).getCuratedVariantGenomicLocations().get(0);

This chain will throw NullPointerException if getVariantList() returns null, or IndexOutOfBoundsException if either list is empty. The null check on line 22 only guards the final result, not the intermediate calls. If any VariantSummaryDocument has no variants or no genomic locations, the indexer will crash mid-batch.

Consider guarding against null/empty lists before dereferencing:

if (doc.getVariantList() == null || doc.getVariantList().isEmpty()) continue;
var variant = doc.getVariantList().get(0);
if (variant.getCuratedVariantGenomicLocations() == null || variant.getCuratedVariantGenomicLocations().isEmpty()) continue;
CuratedVariantGenomicLocationAssociation variantLocation = variant.getCuratedVariantGenomicLocations().get(0);

Minor: Commented-out enum entry (IndexerConfig.java:41)

The import for VariantSummaryCurationIndexer was removed but the enum entry was commented out rather than removed. If someone uncomments it later, it won't compile without re-adding the import. Consider either keeping the import or adding a // TODO comment noting the dependency, so it's clear this is intentionally disabled and what's needed to re-enable.

The rest of the converter logic and the dual indexDocuments calls (summary docs to one index, search result docs to another) look correct.