Is it possible to preserve the order of attachment in the DDS-converted SU-CDTB corpus?

[logan-siyao-peng]

Thank you so much for converting and releasing the SU-CDTB corpus in the DDS format.
I am wondering since SU-CDTB is converted from a constituency-styled structure, is it possible to preserve the order of span attachments in your released DDS data?
This would be the "head-ordered dependency tree" mentioned in Morey et al. 2018.
Maybe you can add an "order" key to the json format or use the 2nd column in a 10-column rsd file?
Thank you again.