Here is what our customer did:
data = pd.DataFrame([('1.0.0',)])
data.to_sql('version', schema='testschema', index=True, index_label='col1',
con=connection, if_exists='replace')
The problem is that the default value for index is True in pandas. It
proceeds to create the default Table object form
On Mon, Feb 11, 2019 at 4:33 PM mark.keller via sqlalchemy
wrote:
>
> Hi,
>
> I'm working on a bug that emerges when a user tries to use Pandas and
> SQLAlchemy with a Snowflake database and tries to push a data frame to the
> database.
> I'm trying to disallow users from submitting a "CREATE
I just found that I could add a before_creation listener to the default
Table object.
However; I would like to still discuss possibly overwriting the methods in
the default SchemaGenerator.
On Monday, February 11, 2019 at 1:33:39 PM UTC-8, mark@snowflake.com
wrote:
>
> Hi,
>
> I'm working
On Mon, Feb 11, 2019 at 3:37 PM Martin Stein wrote:
>
> So I have replaced the regex-based rewrite with the PGValuesParam approach
> that we talked about. One issue with the regex-based rewrite was that in its
> current form it breaks for parameter-lists > 10 elements, because the inner
>
Hi,
I'm working on a bug that emerges when a user tries to use Pandas and
SQLAlchemy with a Snowflake database and tries to push a data frame to the
database.
I'm trying to disallow users from submitting a "CREATE INDEX ..." SQL
commands to snowflake, which pandas tries to do by default, but
So I have replaced the regex-based rewrite with the PGValuesParam approach
that we talked about. One issue with the regex-based rewrite was that in
its current form it breaks for parameter-lists > 10 elements, because the
inner sorted(..) call does string-based sorting:
'IN