I just don't understand the source coding theorem.

Do you mean I can think of it as the average code length L = entropy (H)?

For one source S, I think it is correct to interpret that the "shortest" average code length L equals entropy H(S).

As H(S) LL (where は is an arbitrary positive number) in the source site, the source coding theorem says, "No algorithm can reduce the average code length L to less than entropy H(S)."