tlcconfig
#
Configure the tlc
package and Object Service
When importing the tlc package from a Jupyter notebook (e.g. import tlc
), the default configuration will be picked
from the config file and environment variables (if set). If the users wants to override these values, one can import the
tlcconfig
module and set the configuration options before importing the tlc
package.
- Example Usage:
import tlcconfig.options
import tlcconfig.option_loader
# Create a new OptionLoader instance. Pass false to constructor to prevent
# options being loaded from default configuration file and environment variables.
myOptionLoader = tlcconfig.option_loader.OptionLoader(initialize=False)
# Set an option:
scan_urls = ["s3://bucket/path/to/tables", "file://home/data/tables"]
myOptionLoader.set(tlcconfig.options.TABLE_SCAN_URLS, scan_urls)
# Set the global OptionLoader instance to the one we just created.
OptionLoader.setInstance(myOptionLoader)
import tlc # This will use the options set above
# Use tlc package with your custom options
Modules and Packages#
Module |
Description |
---|---|
Configuration options for 3LC |
|
This module contains the OptionLoader class, which is used to load and store the 3LC options. |
|
Logging configuration for the 3LC module, supporting default and customizable configurations. |