SPARK_HOME Issue + DeIdentification Fix + Normalizer Improvement

[danilojsl]

1. There is an issue whit pip installation when SPARK_HOME variable is setup.
2. There is a fix required in DeIdentification unit tests.
3. There is a pending improvement in Normalizer.

Description

Installation of spark-nklp with Python package is working in all environments, but when SPARK_HOME is setup, it does not find sparknlp.jar

Expected Behavior

1. When using spark-nlp under pyhton, it should locate sparknlp.jar automatically.
2. Unit tests should not use pretrained models.
3. Normalizer should work as always after the improvement.

Current Behavior

When SPARK_HOME is setup, spark-nlp is not able to find sparknlp.jar, thus a python user requires to set spark session manually.

Possible Solution

Steps to Reproduce

1. Set SPARK_HOME
2. Install spark-nlp through python package (pip)
3. In a python terminal write: from sparknlp.base import * spark_nlp_session = SparkNLP().spark_session
4. spark-nlp cannot find the jar

Context

Your Environment

• Version used:
• Browser Name and version:
• Operating System and version (desktop or mobile):
• Link to your project:

[maziyarpanahi]

Is this issue resolved? @danilojsl

[github-actions[bot]]

This issue is stale because it has been open 120 days with no activity. Remove stale label or comment or this will be closed in 5 days