Emmanuel Cecchet wrote:
> Greg Smith wrote:
>> On Fri, 11 Sep 2009, Emmanuel Cecchet wrote:
>>
>>> I guess the problem with extra or missing columns is to make sure
>>> that you know exactly which data belongs to which column so that you
>>> don't put data in the wrong columns which is likely to happen if
>>> this is fully automated.
>>
>> Allowing the extra column case is easy: everwhere in copy.c you find
>> the error message "extra data after last expected column", just
>> ignore the overflow fields rather than rejecting the line just based
>> on that. And the default information I mentioned you might want to
>> substitute for missing columns is already being collected by the code
>> block with the comment "Get default info if needed".
> If I understand it well, you expect the garbage to be after the last
> column. But what if the extra or missing column is somewhere upfront
> or in the middle? Sometimes you might have a type conflict problem
> that will help you detect the problem, sometimes you will just insert
> garbage. This might call for another mechanism that would log the
> lines that are automatically 'adjusted' to be able to rollback any
> mistake that might happen during this automated process.
>
>
Garbage off to the right is exactly the case that we have. Judging from
what I'm hearing a number of other people are too.
Nobody suggests that a facility to ignore extra columns will handle
every case. It will handle what increasingly appears to be a common case.
cheers
andrew
