Multiple WRAP characters

[cnieman]

In RFC 40, Daniel made a comment that multiple wrap characters would be useful, and I agree. I'm working on a dataset for France, and only having one wrap character for a label is inconvenient. As Daniel mentioned in RFC 40, the logic for some wrap characters would need to be different (e.g. spaces at the end of a line can be dropped while a hyphen should be kept). Examples of text I came up with to illustrate the need for this are:

• "Les Quatre Vents" wrapped on spaces if MAXLENGTH is satisfied
• "Fontenay-sur-Bois" wrapped on hyphens if MAXLENGTH is satisfied, hyphens remain at ends of lines
• "Domaine de Lacroix-Laval" wrapped on spaces and and hyphens if MAXLENGTH is satisfied, hyphens remain at ends of lines

Ideally, all of these cases would be addressed by the same label block (specifically because of the third example).