What does b stand for, and what is the syntax behind b+sr+Latn?

Question

By default, Android has always displayed Serbian in Cyrillic, but since 7.0, users can choose between Cyrillic and Latin script. Apps (mostly from Google) ususally puts the Cyrillic strings in sr and the Latin strings in b+sr+Latn. However, b+sr+Latn is very different from the usual convention of using dashes (e.g. sr-rRS-port).

What does b stand for? What is the syntax, and what more can I specify with this?

Answer 1

It's a BCP 47 language tag. It's starts with b and followed by a language, script, region and variant.
BCP 47 documentation

b+sr+Latn would be Serbian with the Latin script.
b+sr+Latn+RS would be Serbian in Serbia, with the Latin script.

The same can be applied to Azerbaijani:
b+az+Latn – Latin script
b+az+Cyrl – Cyrillic script
b+az+Arab – Arabic script

But you can virtually do any combination possible:
b+ja+Hrkt – Japanese with only Hiragana and Katakana
b+ru+Latn – Russian with the Latin script
b+hu+Hung – Hungarian with the Hungarian Runic script
b+en+Cyrl+US – American English with the Cyrillic script
b+hy+Piqd+DE – German Armenian with the Klingon script
(but support will be very limited)

See also:
ISO 639-3 language list
ISO 15924 script list
ISO 3166-1 alpha-2 country list