#469 Postgres: Save Dataset Adds Quotation Marks

[John L. Poole]

When I perform a File -Save Dataset As => Save to PostGIS Table, select the radio "Cread new or replace..." the table that is creates has column names wrapped with double quotation marks.

This was very confusing as when I tried to perform some queries, I kept getting column not found and then I realized I had to include quotation marks around the column name.

So this works:

select distinct("MAINTENANC"), count(*)
from kimselect 
group by "MAINTENANC"
order by "MAINTENANC"

whereas this fails:

select distinct(MAINTENANC), count(*)
from kimselect 
group by MAINTENANC
order by MAINTENANC

I should like to be able to create a table that matches the attributes/features name without having to include quotation marks.

[michael michaud]

Hi John,

I know this is an annoying problem. The recommended way to write to PostGIS is now to use the Postgis driver (new) which is a driver able to read/write from and to postgis, and which has some more options like "normalize table name" and "normalize column name" to solve the case sensitivity problem you are mentionning.
I just noticed several problems with this driver in the current source code (see new bug reports). I think the version included in the last release (1.12) is usable though.

[ede]

hearing "quotes" i remember this change
https://sourceforge.net/p/jump-pilot/code/5489/

might have nothing to do with it though.. ede

On 11.04.2018 06:09, John L. Poole wrote:

---

[bugs:#469] https://sourceforge.net/p/jump-pilot/bugs/469/ Postgres: Save Dataset Adds Quotation Marks

Status: open
Labels: postgis dataset
Created: Wed Apr 11, 2018 04:09 AM UTC by John L. Poole
Last Updated: Wed Apr 11, 2018 04:09 AM UTC
Owner: nobody

When I perform a File -Save Dataset As => Save to PostGIS Table, select the radio "Cread new or replace..." the table that is creates has column names wrapped with double quotation marks.

This was very confusing as when I tried to perform some queries, I kept getting column not found and then I realized I had to include quotation marks around the column name.

So this works:

select distinct("MAINTENANC"), count(*)
from kimselect
group by "MAINTENANC"
order by "MAINTENANC"

whereas this fails:

select distinct(MAINTENANC), count(*)
from kimselect
group by MAINTENANC
order by MAINTENANC

I should like to be able to create a table that matches the attributes/features name without having to include quotation marks.

---

Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/jump-pilot/bugs/469/

To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/

[michael michaud]

I can port the "normalize table name", "normalize column name" to the old postgis writer. Can be done before or after migration. Not a blocking feature

[ede]

let's tag it as future then