parse biltrans too

[unhammer]

biltrans-output has tags on the "wordform" (stuff between ^ and first /)

[unhammer]

and we can have a final "empty" analysis in anaphora resolution style:

^El<det><def><m><pl>/The<det><def><m><pl>/$

[khannatanmai]

Do you want the readings to be differentiated by SL/TL? Also if there's a reading in the ref side after anaphora resolution do you want to differentiate that?