Extract anything between first underscore and last underscore with exceptions

I have a list of strings:

str1 <- "core21_ap_202003.xlsx"
str2 <- "core21_ap_thailand_202004.xlsx"
str3 <- "core17_eay_201008_b.xlsx"

strings <- list(str1, str2, str3)

I want to extract "ap", "ap_thailand", and "eay". I have tried:

gsub("_[^_]*$|^[^_]*_","", strings, perl=T)

Output:

[1] "ap"  "ap_thailand"  "eay_201008" 

Works for the first two occurences, but not the last one. I need "eay", not "eay_201008".

In other words, that a country name (here, Thailand) is extracted only if it exists, and then date should never be extracted.

Desired output:

[1] "ap"  "ap_thailand"  "eay" 

>Solution :

Find everything to the first underscore followed by everything until an underscore and 6 digits followed by anything. Keep the part between the undersores.

strings |>
  unlist() |>
  sub(".*?_(.*)_\\d{6}.*", "\\1", x = _) 
## [1] "ap"          "ap_thailand" "eay"