OK, so there are going to be a lot of combinations to try here. I would propose we start with a fairly simple tokens file, exhaust all the possibilities, and then expand upwards if we don't get a match. I would suggest your first token file should be as follows:
^r1^| ^r1^%s| ^r1^|%s ^r1^%s|%s ^r1^%s
+ ^r2^%[Cc]h%[1i]na
^r3^| ^r3^%s| ^r3^|%s ^r3^%s|%s ^r3^%s
+ ^r4^%[Rr]ussia
^r5^| ^r5^%s| ^r5^|%s ^r5^%s|%s ^r5^%s
+ ^r6^111 ^r6^222 ^r6^333
+ ^r7^USA ^r7^usa
^r8^| ^r8^%s| ^r8^|%s ^r8^%s|%s ^r8^%s
Let me break this down for you.
The ^r1^ string defines the position of each line in the passphrase.
%s is a space.
The + symbol requires that that line must be included. The absence of the + means that line is optional.
Therefore, line 1 will try the | character on its own, a space before, a space after, a space before and after, a space on its own, or nothing at all.
The %[Cc] means to pick one of the characters between the [ ] symbols. So line two will attempt China, china, Ch1na, and ch1na.
Line 6 will try one of those three numbers.
Line 7 will try either USA or usa.
Once you start adding other symbols and changing the position of the number, things become significantly more complicated and time consuming. I would try this first, and then we can take it from there if there is no match.
You will obviously need to adjust based on your actual words and numbers. If you are struggling and feel comfortable sharing them, I can adjust for you. You should definitely use the --listpass option to check your set up before running it. See this link:
https://btcrecover.readthedocs.io/en/latest/TUTORIAL/#testing-your-config