Andrew Dunstan <andrew(at)dunslane(dot)net> writes:
> On 2026-03-28 Sa 1:48 PM, Tom Lane wrote:
>> but I don't like the other bit
>>
>> + /* If we've hit the end of the compressed stream, stop. */
>> + if (res == Z_STREAM_END)
>> + break;
>>
>> Paying attention to the amount of data consumed seems just as
>> good and takes less code.
> So IIUC, you agree with the patch except for this last piece, on which I
> agree with your reasoning.
Right, I would leave out the above-quoted hunk but the rest of it
looks good.
regards, tom lane