fixed toml for pyjedai to work with colab GENSIM PROBLEM

Author: Teris45

pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "pyjedai"
-version = "0.3.2"
+version = "0.3.3"
 description = "An open-source library that builds powerful end-to-end Entity Resolution workflows."
 readme = "README.md"
 authors = [
@@ -40,13 +40,12 @@ classifiers = [
 keywords = ["deduplication", "entity-resolution", "link-discovery"]
 requires-python = ">=3.9"
 dependencies = [
-    "gensim",
     "matplotlib",
     "networkx",
     "nltk",
-    "numpy >= 1.7.0,<2.0",
+    "numpy<3.0",
     "pandas",
-    "scipy==1.12",
+    "scipy",
     "seaborn",
     "tqdm",
     "transformers",
@@ -59,8 +58,13 @@ dependencies = [
     "ollama"
 ]
 
+
 [project.optional-dependencies]
 dev = ["pip-tools", "pytest"]
+with-gensim = [
+    "numpy>=1.7, < 2.0", 
+    "gensim",
+    "scipy==1.12"]
 
 [project.urls]
 "Homepage" = "http://pyjedai.rtfd.io"

src/pyjedai/vector_based_blocking.py

@@ -15,7 +15,13 @@
 import platform
 RUNNING_OS = platform.system()
 
-import gensim.downloader as api
+try: 
+    import gensim.downloader as api
+    HAS_GENSIM = True
+except ImportError:
+    HAS_GENSIM = False
+    
+
 import networkx as nx
 import numpy as np
 import torch
@@ -60,11 +66,14 @@ class EmbeddingsNNBlockBuilding(PYJEDAIFeature):
     _method_info = "Creates a set of candidate pais for every entity id " + \
         "based on Embeddings creariot and Similarity search among the vectors."
 
+    
     _gensim_mapping_download = {
         'fasttext' : 'fasttext-wiki-news-subwords-300',
         'glove' : 'glove-wiki-gigaword-300',
         'word2vec' : 'word2vec-google-news-300'
-    }
+        
+    } if HAS_GENSIM else {}
+    
     _sentence_transformer_mapping = {
         'smpnet' : 'all-mpnet-base-v2',
         'st5' : 'gtr-t5-large',
@@ -303,9 +312,11 @@ def _create_gensim_embeddings(self) -> Tuple[np.array, np.array]:
         Returns:
             Tuple[np.array, np.array]: Embeddings from D1 and D2
         """
-        vectors_1 = []
+        if not HAS_GENSIM : 
+            raise ImportError("Reinstall pyjedai with pip install pyjedai[with-gensim] or use other vectorizer")
+            
         vocabulary = api.load(self._gensim_mapping_download[self.vectorizer])
-        
+    
         if not self._d1_loaded:
             for e1 in self._entities_d1:
                 vectors_1.append(self._create_vector(e1, vocabulary))