In the publicly available NACHA schema, we are only able to successfully parse test data that contains "nolines" in the file name. A quick glance shows that other files have CRLF line endings, which are optional (parsed with WSP*) which daffodil is failing to parse correctly.