The Donne project has a special markup language for their transcriptions. I’ve created a Radiant filter that can transform it into HTML for viewing. This afternoon, I was able to change the parser in the filter from producing HTML to producing a data structure that I can manipulate in the Fabulator engine.
Combining the Donne document function with the Radiant page access from this morning, I can run the following expression and get a comma separated list of line numbers in the transcription (assuming the transcription is in the ‘body’ part of the /manuscripts/h06/1 page):
f:string-join(donne:document(radiant::/manuscripts/h06/1/@body)/*/@line, ‘, ‘)