meltano_run_op: expected value to be of Python type dict.
Closed this issue · 0 comments
ReneTC commented
Again I am really sorry spamming the repo. But I do think this is a bug.
I have to jobs:
jobs:
- name: fetch_raw_data
tasks:
- utils:raw_data
- name: raw_data_to_duckdb
tasks:
- tap-spreadsheets-anywhere target-duckdb
They work perfectly and can even be executed with this dagster with load_jobs_from_meltano_project()
from this repo.
However, if I make these two as meltano_run_op()
instead:
@job(resource_defs={"meltano": meltano_resource})
def meltano_etl():
raw_data = meltano_run_op("utils:raw_data")()
make_duck_db = meltano_run_op("tap-spreadsheets-anywhere target-duckdb")(raw_data)
@repository
def dev_meltano_etl():
return [
meltano_etl
]
Second op will fail, with dagster returning:
dagster._core.errors.DagsterTypeCheckDidNotPass: Type check failed for step input "env" - expected type "dict?". Description: Value of type <class 'str'> failed type check for Dagster type dict, expected value to be of Python type dict.