this is a confusion between entity and its name. the solution can be
judri semantic frame or possibly se judri.

i mi ca tavla fo lo se judri be li jy by obu bei me'o ibu sy obu 69 pi'e 3

(probably not pi'e. Also me'o in judri3 might not be correct.)

A degenerate semantic frame can be an option for this specific case of
ISO-names of languages. E.g.

x1 is a name of language x2 in ISO-system, its part of standard (ISO 639-1,
ISO 639-3 etc.) automatically determined by the number of symbols in x1.

A potential problem can be with the list of pars of the standard.
ISO 639-2 has the same number of symbols as ISO 639-3.
ISO 639-6 was abandoned.

Repurposing existing codes for other languages will break older Lojban
texts so eventually conservative Lojban system will emerge just as what is
happening with ISO-names of countries.