Add a pg_encoding_mbcliplen() function that is just like pg_mbcliplen()
except the caller can specify the encoding to work in; this will be needed for pg_stat_statements. In passing, do some marginal efficiency hacking and clean up some comments. Also, prevent the single-byte-encoding code path from fetching one byte past the stated length of the string (this last is a bug that might need to be back-patched at some point).
Showing
Please register or sign in to comment