If you include empty files in the input directory, stupid assumptions made by buildNtd in CsvBuilder cause the document IDs to fall out of sync. This wreaks havoc on much of the output, making it quite meaningless! (It appears that the metadata file remains correct, but that's about it, sadly.)
If you include empty files in the input directory, stupid assumptions made by
buildNtd
inCsvBuilder
cause the document IDs to fall out of sync. This wreaks havoc on much of the output, making it quite meaningless! (It appears that the metadata file remains correct, but that's about it, sadly.)