The sample_text and sample_glyphs entries should (in my opinion) migrate to the language textproto files while the languages entries also present on the METADATA.pb file should be enough to cross-reference the lang database:
languages: "koi_Perm" # Komi-Permyak, Old Permic
languages: "kv_Perm" # Komi, Old Permic
While updating the METADATA.pb schema description on FontBakery (https://github.com/googlefonts/fontbakery/issues/3443) and gf-tools (https://github.com/googlefonts/gf-docs/issues/114), I noticed that some files such as https://github.com/google/fonts/blob/main/ofl/notosansoldpermic/METADATA.pb have
sample_text
andsample_glyphs
entries and this seems redundant to the kind of content I see in some of the language textprotos available at https://github.com/google/fonts/blob/main/lang/languages/On this specific case, though, https://github.com/google/fonts/blob/main/lang/languages/koi_Perm.textproto and https://github.com/google/fonts/blob/main/lang/languages/kv_Perm.textproto lack such sample entries.
The
sample_text
andsample_glyphs
entries should (in my opinion) migrate to the language textproto files while thelanguages
entries also present on the METADATA.pb file should be enough to cross-reference the lang database: