Beena Emerson
memis****@gmail*****
2015年 3月 17日 (火) 22:05:33 JST
Hello, On Tue, Mar 17, 2015 at 2:23 PM, Sawada Masahiko <sawad****@gmail*****> wrote: > > Thank you for your patch! > > These are good patch for making relax the pg_bigm restriction. > Also, 02_update_datatype.patch reduces GIN index size. > > For more increasing indexed column size, when allocation memory size > exceeds limit we can have following two ways, > > 1. Separate the input string into several strings, and concatenate > bigms after generating each bigm. > 2. Recalculate multi-byte word length, and re-try to allocate memory. > (Currently pg_bigm always use byte length, but multi-byte word > could be more than 1 bytes.) > > #1 eliminates the restrictions, #2 would more increase indexed column size. > But pg_bigm allocates memory using by traditional way at first time > for performance reason. > > Thought? > I agree. #1 would remove this limitation entirely. Suggestion #2 do you mean use the character count rather than the string length? Regards, Beena Emerson -------------- next part -------------- HTML$B$NE:IU%U%!%$%k$rJ]4I$7$^$7$?(B... 下载