[Zubersoft]

The mapping in most cases is very straight forward, i.e. Artist -> Artist, Collection->Collection, Year -> Year, etc. These are not standard fields though - it's more like custom metadata (typically using the xmpDM namespace). For the more standardized fields, Author maps to Composer, Books maps to Album, Subject maps to Genre, Keywords maps to Keywords and Title maps to Title. I don't support Creator or Producer.

This is the full list of fields I look for:

                    "Title", "Artist", "Album", "Keywords", "Composer", "Genre", "Books",
                    "Collection", "Setlist", "Key", "Signature", "Year", "Subject", "Author",
                    "Duration", "CustomGroup", "SourceType", "Custom", "Custom2"

using the PDF library method of getting metadata, and then 

                            "composer", "album", "artist", "genre",
                            "key", "Composer", "Album", "Artist", "Genre", "Key",
                            "Collection", "collection", "Setlist", "setlist", "signature", "Signature",
                            "Year", "year", "Duration", "duration", "CustomGroup", "customGroup",
                            "SourceType", "sourceType", "custom", "Custom", "custom2", "Custom2"

using an xpath query of ".//*/Description/<field>" of the XMP data retrieved from the PDF. I also do a search for .//xmpDM:<field> for composer, album, artist, genre and key just to cover all possibilities.

Mike