questions
От | Lendvary Gyorgy |
---|---|
Тема | questions |
Дата | |
Msg-id | 357D2423.52EEFBA0@prolan.hu обсуждение исходный текст |
Ответы |
Re: [SQL] questions
|
Список | pgsql-sql |
I have a little problem with arrays. I,ve defined a type dbref. It has two fields: char type[200] and char record[200]. My prolem is that I can't insert a row into a table which has an attribute with type dbref[2]. For instance: Create table boci (hapci dbref[2]); How can I insert a row into this table? Please help me! My next problem is SPI interface. I have used LIBPQ but I need SPI. I have read a description in PostgrSQL Programmer's Guide, but I understand nearly nothing. Please write me where I can find a better description about SPI! I have to treat very big tables from my C program. I update every attribute with update command. When I update an attribute, I must go through every tuple of the table. It takes a long time. I write you a little piece of my program: sprintf(buff, "CREATE TABLE boci (tup_num int, hapci int)"); PQexec(conn, buff); for (i=0; i<10000; i++) { sprintf(buff, "INSERT INTO boci VALUES(i)"); PQexec(conn, buff); } for (i=0; i<10000; i++) { x = GetValue(); /* GetValue isn't an interesting function */ sprintf(buff, "UPDATE boci SET hapci = %d WHERE tup_num = %d", x, i); PQexec(conn, buff); } I don't want to update every tuple indvidually but I want to prepare a 'block write'. I hope you understand what I'd like to. Can you give me a good method for saving a long time? I need a program that is about 10 times faster than mine. Thanks for your attention! gyurika@prolan.hu
В списке pgsql-sql по дате отправления: