Merge pull request #19 from mardikark-gslab/spacy_version_check

Updated the Spacy library version upper limit to 3.5.0
acryldata · Aug 17, 2023 · fc290dc · fc290dc
2 parents 47bfe54 + d9d0d3a
commit fc290dc
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 4 deletions.
diff --git a/datahub-classify/setup.py b/datahub-classify/setup.py
@@ -18,7 +18,7 @@ def get_long_description():
     "schwifty>=2022.9.0",
     "python-stdnum>=1.17",
     "ipaddress>=1.0.23",
-    "spacy>=3.4.1,<=3.4.3",
+    "spacy>=3.4.1,<=3.5.0",
     "phonenumbers>=8.12.56,<=8.13.0",
 }
 

diff --git a/datahub-classify/tests/expected_output/expected_infotypes_UNIT_TESTING.json b/datahub-classify/tests/expected_output/expected_infotypes_UNIT_TESTING.json
@@ -52,7 +52,7 @@
         "General Manager": "Full_Name"
     },
     "2018-seattle-business-districts": {
-        "Address": "Street_Address",
+        "Address": "no_infotype",
         "E-mail": "Email_Address"
     },
     "Customer_Segmentation": {
@@ -129,7 +129,7 @@
         "Full_Name": "Full_Name"
     },
     "1-MB-Test": {
-        "First and Last Name": "no_infotype",
+        "First and Last Name": "Full_Name",
         "SSN" :"US_Social_Security_Number",
         "Credit Card Number" : "Credit_Debit_Card_Number"
     },

diff --git a/datahub-classify/tests/expected_output/expected_infotypes_confidence_slabs.json b/datahub-classify/tests/expected_output/expected_infotypes_confidence_slabs.json
@@ -54,7 +54,6 @@
         "General Manager": 0.6
     },
     "2018-seattle-business-districts": {
-        "Address": 0.6,
         "E-mail": 1.0
     },
     "Customer_Segmentation": {
@@ -129,6 +128,7 @@
         "Full_Name": 0.9
     },
     "1-MB-Test": {
+        "First and Last Name": 0.7,
         "SSN": 1.0,
         "Credit Card Number": 1.0
     },