Thread: Why doesn't COPY support the HEADER options for tab-separated output?
Mostly just curious, as this is preventing me from using tab-separated output. I'd like there to be a header in my files. I have to use CSVs instead.
Joe
On Mon, Aug 12, 2013 at 02:15:25PM -0700, Joe Van Dyk wrote: > Mostly just curious, as this is preventing me from using tab-separated output. > I'd like there to be a header in my files. I have to use CSVs instead. [ "this" means the fact that COPY doesn't support the HEADER option. ] We assume CSV is going to be read into a tool that can use the header names; delimited files don't usually have a use for those headers. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + It's impossible for everything to be true. +
On Mon, Aug 12, 2013 at 4:21 PM, Bruce Momjian <bruce@momjian.us> wrote: > On Mon, Aug 12, 2013 at 02:15:25PM -0700, Joe Van Dyk wrote: >> Mostly just curious, as this is preventing me from using tab-separated output. >> I'd like there to be a header in my files. I have to use CSVs instead. > > [ "this" means the fact that COPY doesn't support the HEADER option. ] > > We assume CSV is going to be read into a tool that can use the header > names; delimited files don't usually have a use for those headers. that's a terrible assumption. merlin
On Mon, Aug 12, 2013 at 2:21 PM, Bruce Momjian <bruce@momjian.us> wrote: > On Mon, Aug 12, 2013 at 02:15:25PM -0700, Joe Van Dyk wrote: >> Mostly just curious, as this is preventing me from using tab-separated output. >> I'd like there to be a header in my files. I have to use CSVs instead. > > [ "this" means the fact that COPY doesn't support the HEADER option. ] > > We assume CSV is going to be read into a tool that can use the header > names; delimited files don't usually have a use for those headers. There have been a few cases where I wanted this, too. I can see why implementing it is less important than having it for CSV, but is there a reason to consider this harmful for TEXT format? I.e. if someone submitted a patch, would it be rejected? Cheers, Jeff
On Mon, Aug 12, 2013 at 03:17:00PM -0700, Jeff Janes wrote: > On Mon, Aug 12, 2013 at 2:21 PM, Bruce Momjian <bruce@momjian.us> wrote: > > On Mon, Aug 12, 2013 at 02:15:25PM -0700, Joe Van Dyk wrote: > >> Mostly just curious, as this is preventing me from using tab-separated output. > >> I'd like there to be a header in my files. I have to use CSVs instead. > > > > [ "this" means the fact that COPY doesn't support the HEADER option. ] > > > > We assume CSV is going to be read into a tool that can use the header > > names; delimited files don't usually have a use for those headers. > > There have been a few cases where I wanted this, too. I can see why > implementing it is less important than having it for CSV, but is there > a reason to consider this harmful for TEXT format? I.e. if someone > submitted a patch, would it be rejected? I see no reason _not_ to support it, so sure, either we add this to 9.4, or add it to the TODO list. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + It's impossible for everything to be true. +
On 08/12/2013 02:15 PM, Joe Van Dyk wrote: > Mostly just curious, as this is preventing me from using tab-separated > output. I'd like there to be a header in my files. I have to use CSVs > instead. I am not sure I understand. Do you mean you are converting the file to comma separated or that you are using CSV and having to specify the delimiter as '\t'? > > Joe -- Adrian Klaver adrian.klaver@gmail.com
On 12/08/13 23:18, Bruce Momjian wrote: > On Mon, Aug 12, 2013 at 03:17:00PM -0700, Jeff Janes wrote: >> On Mon, Aug 12, 2013 at 2:21 PM, Bruce Momjian <bruce@momjian.us> wrote: >>> On Mon, Aug 12, 2013 at 02:15:25PM -0700, Joe Van Dyk wrote: >>>> Mostly just curious, as this is preventing me from using tab-separated output. >>>> I'd like there to be a header in my files. I have to use CSVs instead. Late to the discussion, but it does work to set format=csv and delimiter = E'\t' to get tab-separated. Be nice not to have to though. -- Richard Huxton Archonet Ltd
On Thu, Aug 15, 2013 at 09:25:07AM +0100, Richard Huxton wrote:
> On 12/08/13 23:18, Bruce Momjian wrote:
> >On Mon, Aug 12, 2013 at 03:17:00PM -0700, Jeff Janes wrote:
> >>On Mon, Aug 12, 2013 at 2:21 PM, Bruce Momjian <bruce@momjian.us> wrote:
> >>>On Mon, Aug 12, 2013 at 02:15:25PM -0700, Joe Van Dyk wrote:
> >>>>Mostly just curious, as this is preventing me from using tab-separated output.
> >>>>I'd like there to be a header in my files. I have to use CSVs instead.
>
> Late to the discussion, but it does work to set format=csv and
> delimiter = E'\t' to get tab-separated. Be nice not to have to
> though.
I assume that is going to add quotes to a field that contains quotes:
CREATE TABLE test(x TEXT);
INSERT INTO test VALUES ('"a"');
COPY test TO STDOUT WITH (FORMAT CSV, DELIMITER E'\t');
"""a"""
I have added this to the TODO list:
Allow COPY "text" format to output a header
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ It's impossible for everything to be true. +
On Thu, Aug 15, 2013 at 1:25 AM, Richard Huxton <dev@archonet.com> wrote: > On 12/08/13 23:18, Bruce Momjian wrote: >> >> On Mon, Aug 12, 2013 at 03:17:00PM -0700, Jeff Janes wrote: >>> >>> On Mon, Aug 12, 2013 at 2:21 PM, Bruce Momjian <bruce@momjian.us> wrote: >>>> >>>> On Mon, Aug 12, 2013 at 02:15:25PM -0700, Joe Van Dyk wrote: >>>>> >>>>> Mostly just curious, as this is preventing me from using tab-separated >>>>> output. >>>>> I'd like there to be a header in my files. I have to use CSVs instead. > > > Late to the discussion, but it does work to set format=csv and delimiter = > E'\t' to get tab-separated. Be nice not to have to though. Also, the escaping/quoting mechanisms are different. I often find the escapes of the TEXT format more convenient than the CSV with E'\t'. Cheers, Jeff