The current DeduplicationCDXExtractionBatchJob makes invalid CDX'es if a deduplicationmigration record exists in the metadata-file

The waybackindexer uses the dk.netarkivet.wayback.indexer.ArchiveFile.index method to index arc and warc-files.

In the case of metadata-files, it currently uses the DeduplicationCDXExtractionBatchJob batchjob to generate deduplicationCDX'es from the duplicate entries in the crawllog.

This will not work for metadata-files with a deduplicationmigration record.

Instead we should fetch the deduplicationmigration and the crawllog from the metadatafile, and
then do the replacement, as we do in the RawMetadataCache.migrateDuplicates method