For an alphabet of N symbols and r distinct symbol weights we describe a code for which the number of operations needed for encoding and decoding is equal to clogr+c/sub 1/ instead of clogN+c/sub 2/ ...

For which languages does it work? The comments in prettify.js are authoritative but the lexer should work on a number of languages including C and friends, Java, Python, Bash, SQL, HTML, XML, CSS, ...