Document BibTeX name parsing rules

[omasanori]

Indeed structured name field sucks and what we assume about names are false but we have to deal with major bibliography software assuming (a subset of) European name structure.

Leave complex cases described in the TeX FAQ, a thread at comp.text.tex, tips by Norman Walsh, questions on Stack Exchange and such for now.

[lassik]

How about:

• (author (bibtex "First" "von" "Last" "Jr."))
• (editor (bibtex "First" "von" "Last" "Jr."))

If each name component is its own string, it's easier than parsing them all from one string.

[lassik]

We could also have eg. (author "feeley") and a separate table of people:

(person
  (id "feeley")
  (name (bibtex "Marc" "" "Feeley" ""))
  ... homepage and other optional info here...)

[omasanori]

Thank you for your suggestions, @lassik !

A database of Schemers is a cool idea. We can reserve symbols for such references.

We may leave strings as-is. By doing so, we can make existing entries still valid with the new spec. Parsing rules can be unspecified in the spec.

Next proposal in short: (name-bibtex <first> <particle> <last> <suffix>) or a string. Other values are reserved for future use.

What do you think?

[lassik]

Fine by me.

[omasanori]

Thank you for your review! I have addressed to all comments.

[lassik]

Thanks! LGTM