Re: Perform COPY FROM encoding conversions in larger chunks

On 02/02/2021 23:42, John Naylor wrote:
> Although a new patch is likely forthcoming, I did take a brief look and
> found the following:
>
>
> In copyfromparse.c, this is now out of date:
>
>  * Read the next input line and stash it in line_buf, with conversion to
>  * server encoding.
>
>
> One of your FIXME comments seems to allude to this, but if we really
> need a difference here, maybe it should be explained:
>
> +#define INPUT_BUF_SIZE 65536 /* we palloc INPUT_BUF_SIZE+1 bytes */
>
> +#define RAW_BUF_SIZE 65536 /* allocated size of the buffer */

We do in fact still need the +1 for the NUL terminator. It was missing
from the last patch version, but that was wrong; my fuzz testing
actually uncovered a bug caused by that. Fixed.

Attached are new patch versions. The first patch is same as before, but
rebased, pgindented, and with a couple of tiny fixes where conversion
functions were still missing the "if (noError) break;" checks.

I've hacked on the second patch more, doing more refactoring and
commenting for readability. I think it's in pretty good shape now.

- Heikki