Re: Finding error in long input file

On 9 July 2024 23:14:33 Rich Shepard <rshepard@appl-ecosys.com> wrote:

> I've a file with 488 rows to be input into a table. I run the script using
> psql with the `\i' option.
>
> After psql found a few missing commas I thought the script was clean. But,
> psql returned this error:
> bustrac=# \i scripts/insert-addrs.sql
> psql:scripts/insert-addrs.sql:488: ERROR:  syntax error at or near ";"
> LINE 488: ...2832,1,default,'85250 Red House Rd','Paisley','OR','97636');
>                                                                         ^ Am
> I correct that some rows were inserted into the table before this error was
> generated somewhere prior to the end of the file? If so, how do I determine
> which rows were inserted and what is the recommended procedure to locate
> just where that error is?


Did you run the entire thing inside a transaction? If so, then it will have 
been rolled back after the error, and no rows will have been inserted.

Ray.




>
> I've learned from experience that using the old coding approach of dividing
> the file in half, then half again and again until the row with the error is
> found produces multiple rows with different PKs. Not what I want to clean
> up.
>
> A clue stick will be much appreciated.
>
> TIA,
>
> Rich