600 Release (#1798)

* Bump versions and update docs

* deprecate pkg-resources (#1782)

* More robust shared pyspark ml param import (#1796)

* deprecate pkg-resources

* more robust import of _shared_pyspark_ml_param

* enterprise NLP 600, import BertForAssertionClassification and FhirDeIdentification (#1797)

Author: C-K-Loan

docs/en/jsl/jsl_release_notes.md

@@ -16,6 +16,24 @@ sidebar:
 See [Github Releases](https://github.com/JohnSnowLabs/johnsnowlabs/releases) for detailed information on Release History and Features
 
 
+## 6.0.0
+Release date: 5-4-2025
+
+The John Snow Labs 6.0.0 Library released with the following pre-installed and recommended dependencies
+
+{:.table-model-big}
+| Library                                                                                 | Version    |
+|-----------------------------------------------------------------------------------------|------------|
+| [Visual NLP](https://nlp.johnsnowlabs.com/docs/en/spark_ocr_versions/ocr_release_notes) | `6.0.0`    |
+| [Enterprise NLP](https://nlp.johnsnowlabs.com/docs/en/licensed_annotators)              | `6.0.0`    |
+| [Finance NLP](https://nlp.johnsnowlabs.com/docs/en/financial_release_notes)             | `1.X.X`    |
+| [Legal NLP](https://nlp.johnsnowlabs.com/docs/en/legal_release_notes)                   | `1.X.X`    |
+| [NLU](https://github.com/JohnSnowLabs/nlu/releases)                                     | `5.4.1` |
+| [Spark-NLP-Display](https://sparknlp.org/docs/en/display)                               | `5.0`      |
+| [Spark-NLP](https://github.com/JohnSnowLabs/spark-nlp/releases/)                        | `6.0.0`    |
+| [Pyspark](https://spark.apache.org/docs/latest/api/python/)                             | `3.4.0`    |
+
+
 
 ## 5.5.5
 Release date: 3-14-2025

johnsnowlabs/abstract_base/software_product.py

@@ -3,7 +3,7 @@
 from abc import ABC
 from typing import List, Set, Optional
 
-import pkg_resources
+from importlib.metadata import version
 
 from johnsnowlabs import settings
 from johnsnowlabs.abstract_base.lib_resolver import Py4JJslLibDependencyResolverABC
@@ -93,17 +93,10 @@ def check_installed_correct_version(
         ):
             return False
         try:
-            if (
-                pkg_resources.get_distribution(cls.pypi_name).version
-                == cls.latest_version.as_str()
-            ):
+            if version(cls.pypi_name) == cls.latest_version.as_str():
                 # print(f'👌 Installed version for {cls.logo + cls.name} is correct, no changes made.')
                 return True
             else:
-                # print(f'🤓 Installed version for {cls.logo + cls.name} is incorrect, '
-                #       f'should be {cls.latest_version.as_str()} but is {pkg_resources.get_distribution(cls.pypi_name).version} '
-                #       f'upgrading the package')
-
                 return False
         except Exception as err:
             v = get_pip_lib_version(lib=cls.pypi_name, py_exec=python_exec_path)
@@ -118,11 +111,11 @@ def get_installed_version(
         download_folder: str = None,
         prefer_pip=False,
         fallback_import=False,
-    ) -> bool:
+    ) -> str:
         # Only supported for current Py Exec Path, return True otherwise
         if not prefer_pip:
             try:
-                return pkg_resources.get_distribution(cls.pypi_name).version
+                return version(cls.pypi_name)
             except:
                 pass
         v = get_pip_lib_version(lib=cls.pypi_name, py_exec=python_exec_path)

johnsnowlabs/auto_install/health_checks/ocr_test.py

@@ -53,11 +53,8 @@ def run_test():
         ]
     )
 
-    import pkg_resources
+    pdf_example = str(files("sparkocr").joinpath("resources/ocr/pdfs/tabular-pdf/data.pdf"))
 
-    pdf_example = pkg_resources.resource_filename(
-        "sparkocr", "resources/ocr/pdfs/tabular-pdf/data.pdf"
-    )
     pdf_example_df = spark.read.format("binaryFile").load(pdf_example).cache()
     pipeline.transform(pdf_example_df).show()
 

johnsnowlabs/finance.py

@@ -128,6 +128,8 @@
             ContextualEntityRuler,
             PretrainedZeroShotNER,
             StructuredJsonConverter,
+            BertForAssertionClassification,
+            FhirDeIdentification
         )
 
         from sparknlp_jsl.modelTracer import ModelTracer

johnsnowlabs/legal.py

@@ -126,6 +126,8 @@
             ContextualEntityRuler,
             PretrainedZeroShotNER,
             StructuredJsonConverter,
+            BertForAssertionClassification,
+            FhirDeIdentification
         )
         from sparknlp_jsl.modelTracer import ModelTracer
         from sparknlp_jsl.pipeline_tracer import PipelineTracer

johnsnowlabs/medical.py

@@ -103,6 +103,8 @@
             ContextualEntityRuler,
             PretrainedZeroShotNER,
             StructuredJsonConverter,
+            BertForAssertionClassification,
+            FhirDeIdentification
         )
         from sparknlp_jsl.structured_deidentification import StructuredDeidentification
         from sparknlp_jsl.text_to_documents_columns import TextToDocumentsColumns

johnsnowlabs/nlp.py

@@ -41,7 +41,14 @@
 
 
 if try_import_lib("pyspark"):
-    import pyspark.ml.param.shared as _shared_pyspark_ml_param
+    try:
+        import pyspark.ml.param.shared as _shared_pyspark_ml_param
+    except Exception as e:
+        try:
+            from pyspark.ml.param import shared as _shared_pyspark_ml_param
+        except Exception as e:
+            _shared_pyspark_ml_param = None
+
     import pyspark.sql as SQL
     import pyspark.sql.functions as F
     import pyspark.sql.types as T

johnsnowlabs/settings.py

@@ -10,21 +10,21 @@
 
 # These versions are used for auto-installs and version  checks
 
-raw_version_jsl_lib = "5.5.5"
+raw_version_jsl_lib = "6.0.0"
 
-raw_version_nlp = "5.5.3"
+raw_version_nlp = "6.0.0"
 
 raw_version_nlu = "5.4.1"
 
 
 raw_version_pyspark = "3.4.0"
 raw_version_nlp_display = "5.0"
 
-raw_version_medical = "5.5.3"
-raw_version_secret_medical = "5.5.3"
+raw_version_medical = "6.0.0"
+raw_version_secret_medical = "6.0.0"
 
-raw_version_secret_ocr = "5.5.0"
-raw_version_ocr = "5.5.0"
+raw_version_secret_ocr = "6.0.0"
+raw_version_ocr = "6.0.0"
 
 raw_version_pydantic = "2"
 

johnsnowlabs/utils/pip_utils.py

@@ -17,13 +17,13 @@
 import json
 import sys
 from urllib import request
-from pkg_resources import parse_version
+from packaging.version import parse
 
 
 def get_all_lib_version_on_pypi(pkg_name):
     url = f"https://pypi.python.org/pypi/{pkg_name}/json"
     releases = json.loads(request.urlopen(url).read())["releases"]
-    return sorted(releases, key=parse_version, reverse=True)
+    return sorted(releases, key=parse, reverse=True)
 
 
 def get_latest_lib_version_on_pypi(pkg_name):

johnsnowlabs/visual.py

@@ -9,7 +9,6 @@
 
         from sparkocr.transformers import *
         from sparkocr.enums import *
-        import pkg_resources
         import sparkocr
         from sparkocr.utils import *
         from sparkocr.schemas import *