As Adrian points out, it is a bit of a strange use case, since the
jsonb itself could be an array. If you really want to do this,
however, have a look at pgcopy: https://pgcopy.readthedocs.io/
pgcopy makes COPY with psycopg2 easy, and supports jsonb[]
(Note: I am the author)
If you want to roll your own, feel free to have a look under the hood
On Mon, Feb 20, 2023 at 11:34:19PM -0500, Samuel Marks wrote:
> How do I insert into a table with a `json[]` column using the `COPY FROM`
> syntax?
>
> Attempt:
> https://gist.github.com/SamuelMarks/fec744a620e2abd0257671aa6f2a96b4
>
> Error:
> psycopg2.errors.InvalidTextRepresentation: malformed array literal:
> "{"jj":null,"text":"bop"}"
>
> Or when I try to json.dump the dict twice I get:
> psycopg2.errors.InvalidTextRepresentation: malformed array literal:
> ""{"jj":null,"text":"bop"}"" DETAIL: Array value must start with "{" or
> dimension information. CONTEXT: COPY my_table, line 1, column
> json_arr_col: ""{"jj":null,"text":"bop"}""
>
> I'm using your `copy_expert` function. PS: Also asked on
> https://stackoverflow.com/q/75511919
>
> Am I meant to represent the lists with braces rather than square brackets?
> - Or should I be using some internal psycopg function as opposed to my
> hacked together `parse_col` function?
>
> Thanks,
>
> Samuel Marks
> Charity <https://sydneyscientific.org> | consultancy <https://offscale.io>
> | open-source <https://github.com/offscale> | LinkedIn
> <https://linkedin.com/in/samuelmarks>
