peertrilcke
commented
2 years ago - Clean up (leave IDs for speech acts); see script for plaintext: https://earlyprint.org/assets/xsl/plaintext_from_epxml.xsl
- generating
(possibly by using their machine generated cast list - check their list first) - Ingo provides mapping for metadata (see https://github.com/dracor-org/engdracor/issues/5)