Re: why can't see the updated value after SPI_execute("update ....", false, 1); - Mailing list pgsql-general

From sunpeng
Subject Re: why can't see the updated value after SPI_execute("update ....", false, 1);
Date
Msg-id AANLkTinmjCnwMHSn82ibYiASdqD9FUG+-DaF1qxNKWkn@mail.gmail.com
Whole thread Raw
In response to Re: why can't see the updated value after SPI_execute("update ....", false, 1);  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-general
I then added , yet it still doesn't work.



2010/9/12 Tom Lane <tgl@sss.pgh.pa.us>
sunpeng <bluevaley@gmail.com> writes:
> First I use SPI_execute("update ....
>   bool succ;
>    SPI_connect();
>     int ret = SPI_execute("update ....where uid = 1", false, 1);//later will
> add error processing
>     if (ret == SPI_OK_UPDATE && SPI_processed == 1) {
>         succ = true;
>    }
>    SPI_finish();

> Then I use SPI_execute("select .....where uid = 1", ....

>     SPI_connect();
>     int ret = SPI_execute("select .....where uid = 1", true, 1);//later will
> add error processing
>     if (ret == SPI_OK_SELECT && SPI_processed == 1) {
>            HeapTuple tuple;
>            tuple = SPI_tuptable->vals[0];
>               ...
>            datum = heap_getattr(tuple,attrno,SPI_tuptable->tupdesc,&isnull);
>            ....
>     }
> the datum doesn't change, why ?

You're missing a CommandCounterIncrement() call between the two actions.
If you hadn't passed read_only = true to the second SPI_execute, it
would have done one for you.

                       regards, tom lane

pgsql-general by date:

Previous
From: Casey Jones
Date:
Subject: Re: Huge amount of memory errors with libpq
Next
From: Tom Lane
Date:
Subject: Re: Huge amount of memory errors with libpq