Thread: pset, NULL, and COPY-reporting
Greetings -- Due to some level of ignorance at install-time, our "psql" clients on two different production computers appear to have different values for /null/: "" and "NULL" respectively. At the same time, one of these computers reports the number of rows read-in with a "\copy TABLE FROM FILE" command as "COPY 123", while the other computer does not. Is there some reference that tells specifically where the "default at login" value of /null/ is set? Is there some reference that tells what controls the "COPY 123" behavior? Are the /null/ and COPY behaviors linked? If so, how? Thank you for your consideration, John Bell ARM Archive Oak Ridge National Lab
On Tuesday, June 2, 2015, John Bell <jzb@ornl.gov> wrote:
Greetings --
Due to some level of ignorance at install-time, our "psql" clients on two
different production computers appear to have different values for /null/: ""
and "NULL" respectively.
At the same time, one of these computers reports the number of rows read-in with
a "\copy TABLE FROM FILE" command as "COPY 123", while the other computer does
not.
Is there some reference that tells specifically where the "default at login"
value of /null/ is set?
Is there some reference that tells what controls the "COPY 123" behavior?
Are the /null/ and COPY behaviors linked? If so, how?
Pretty sure the copy behavior is always on but version dependent.
You can decide what to have null output as using pset variables. See the documentation. The default could have changed between versions, maybe...
David J.
"David G. Johnston" <david.g.johnston@gmail.com> writes: > On Tuesday, June 2, 2015, John Bell <jzb@ornl.gov> wrote: >> Due to some level of ignorance at install-time, our "psql" clients on two >> different production computers appear to have different values for /null/: > You can decide what to have null output as using pset variables. See the > documentation. The default could have changed between versions, maybe... No, I don't think so --- but possibly one machine has a setting in ~/.psqlrc that the other doesn't. >> At the same time, one of these computers reports the number of rows >> read-in with >> a "\copy TABLE FROM FILE" command as "COPY 123", while the other computer >> does not. > Pretty sure the copy behavior is always on but version dependent. Yeah, psql 9.4 started doing that. regards, tom lane