When the documents generated by warc-indexer are written to a single file (for debugging or later batch-index into a Solr installation), a flush is missing, triggering an exception and potentially causing a truncated output.
This pull request adds a flush in the proper place and makes some of the JavaDocs for the DocumentConsumer more explicit on what will happen.
When the documents generated by warc-indexer are written to a single file (for debugging or later batch-index into a Solr installation), a flush is missing, triggering an exception and potentially causing a truncated output.
This pull request adds a flush in the proper place and makes some of the JavaDocs for the
DocumentConsumermore explicit on what will happen.