[itsme]

In case the PDF has bookmarks, PDFsam can use the bookmarks for splitting and MSP can use the bookmarks instead of a CSV for import (just song title and pages, no additional meta data)
If I need a CSV I try at first to find one. See the thread in this forum, ask me or Robipad.
If the PDF has some  kind of table of content copy/paste often works.
If it is a scanned PDF, the TOC can be created via OCR. That usually needs proof reading.
If that fails I search the web. Many websites that sell fakebooks provide a list of contained titles.
There are websites out there that offer fake book indexes
Having created a list of song titles it needs proof reading and least adding or correcting the pages. I usually also add the keys of the songs.

If you have invested all that effort, please share your results