Add the possibility to outputs sqllite or DuckDB file database

[Ben8t]

Feature description

Using in-memory database system such SQLlite or DuckDB, user might want to pass the file database artifact from task to task. Currently it's possible to input a database:

id: ingest_database
namespace: kestra.qa

tasks:

    - id: create_db
      type: io.kestra.plugin.scripts.shell.Commands
      commands:
        - touch file.db
      outputFiles:
        - "file.db"

    - id: update
      type: io.kestra.plugin.jdbc.sqlite.Queries
      url: jdbc:sqlite:file.db
      sqliteFile: '{{ outputs.create_db.outputFiles["file.db"]}}'
      sql: |
        CREATE TABLE IF NOT EXISTS {{flow.id}}_{{ execution.id }} (
          "time" VARCHAR,
          "temperature_2m" VARCHAR,
          "latitude" VARCHAR,
          "longitude" VARCHAR,
          "city" VARCHAR
        );

        CREATE TABLE IF NOT EXISTS temperature (
          "time" VARCHAR,
          "temperature_2m" VARCHAR,
          "latitude" VARCHAR,
          "longitude" VARCHAR,
          "city" VARCHAR,
          primary key(time, city)
        );
      fetchType: STORE

But it's not possible to output the file.db and use it in downstream tasks

[Ben8t]

Even using a simple

    - id: update
      type: io.kestra.plugin.jdbc.sqlite.Queries
      url: jdbc:sqlite:file.db
      sql: |
        CREATE TABLE IF NOT EXISTS {{flow.id}}_{{ execution.id }} (
          "time" VARCHAR,
          "temperature_2m" VARCHAR,
          "latitude" VARCHAR,
          "longitude" VARCHAR,
          "city" VARCHAR
        );

        CREATE TABLE IF NOT EXISTS temperature (
          "time" VARCHAR,
          "temperature_2m" VARCHAR,
          "latitude" VARCHAR,
          "longitude" VARCHAR,
          "city" VARCHAR,
          primary key(time, city)
        );
      fetchType: STORE

It's not possible to access the file.db