Slippinj

Hotfix 3.0.1/Fixed Tlacuilo for single .yml file:
* Now Tlacuilo works correctly when a single .yml file is provided to the -w parameter, instead of autodiscovering them when a folder path is provided.

- Removed default_configuration.py to avoid coupling with our templates definition
- Removed swamp & hdfs_deploy_folder params from Tlacuilo to avoid coupling, they should be used as --extra parameters now
- Added -l --local mode to run Hersir from inside EMR (no AWS or ssh calls should be used)
- Added -j --job-file-name on Cooper to be able to avoid interactive mode if one knows which .properties file should be called by Oozie
- Now if "hdfs_deploy_folder" variablre is present on the configuration files, it will be used as the HDFS deploy folder for Anabasii. If not, you can still provide the deploy folder via -d HDFS_DEPLOY_FOLDER.
- Updated requirements.txt to add new cx_Oracle version 6.1 (Dec 2017)

Feature/2.1.0: Tlacuilo support infering EMR config
- Now the Tlacuilo module asks for EMR ID if you specify the --interactive flag, so there's no longer need to pass EMR-related parameters (like private IP) when calling Hersir with --extra parameters.

- Deprecated all Hive & Sqoop parameters from Scribe generation. If still want to be used, you should go back to tag 1.7.0.

- Added new "auto" feature for Tlacuilo/Hersir: if there are any .py files inside the folders where the .yml selected files are located, they'll be "packed" on the generated workflow, under workflow_root_folder/lib/ path. This way one can start using pyspark scripts on new workflows easily.