Merge pull request #1133 from dedupeio/fix_levenshtein_search_dep

Fix levenshtein search dep

.github/workflows/pythonpackage.yml

@@ -49,12 +49,11 @@ jobs:
       run: pip install -r requirements.txt
     - name: pytest
       run: pytest
-    - name: Submit to coveralls
-      run: coveralls --service=github
-      env:
-        COVERALLS_PARALLEL: true
-        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        COVERALLS_FLAG_NAME: job-${{ matrix.os}}-${{ matrix.python-version }}
+    - name: Code Coverage
+      uses: codecov/codecov-action@v3
+      with:
+        flags: job-${{ matrix.os}}-${{ matrix.python-version }}
+        verbose: true # optional (default = false)      
     - name: Integration tests
       # Do everything twice: The first time is training and generates settings,
       # the second time it tests using a static settings file.
@@ -66,18 +65,6 @@ jobs:
         python benchmarks/benchmarks/canonical_matching.py
         python benchmarks/benchmarks/canonical_gazetteer.py
         python benchmarks/benchmarks/canonical_gazetteer.py
-  coveralls_finish:
-    name: Indicate completion to coveralls.io
-    needs: test
-    runs-on: ubuntu-latest
-    container: python:3-slim
-    steps:
-    - name: Finished
-      run: |
-        pip3 install --upgrade coveralls
-        coveralls --service=github --finish
-      env:
-        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
   settings_file_persists:
     runs-on: ubuntu-latest
     steps:

README.md

@@ -1,6 +1,6 @@
 # Dedupe Python Library
 
-[![Tests Passing](https://github.com/dedupeio/dedupe/workflows/tests/badge.svg)](https://github.com/dedupeio/dedupe/actions?query=workflow%3Atests)[![Coverage Status](https://coveralls.io/repos/github/dedupeio/dedupe/badge.svg)](https://coveralls.io/github/dedupeio/dedupe)
+[![Tests Passing](https://github.com/dedupeio/dedupe/workflows/tests/badge.svg)](https://github.com/dedupeio/dedupe/actions?query=workflow%3Atests)[![codecov](https://codecov.io/gh/dedupeio/dedupe/branch/master/graph/badge.svg?token=aauKUrTEgh)](https://codecov.io/gh/dedupeio/dedupe)
 
 _dedupe is a python library that uses machine learning to perform fuzzy matching, deduplication and entity resolution quickly on structured data._
 

dedupe/core.py

@@ -94,21 +94,21 @@ def fieldDistance(self, record_pairs: RecordPairs) -> None:
         if not mask.any():
             return
         scores = scores[mask]
-        record_ids = numpy.array(record_ids)[mask]
+        record_id_array = numpy.array(record_ids)[mask]
 
         with self.offset.get_lock():
             fp: Scores
             fp = numpy.memmap(
                 self.score_file_path,
                 dtype=self.dtype,
                 offset=self.offset.value,
-                shape=(len(record_ids),),
+                shape=(len(record_id_array),),
             )
-            fp["pairs"] = record_ids
+            fp["pairs"] = record_id_array
             fp["score"] = scores
             fp.flush()
 
-            self.offset.value += len(record_ids) * self.dtype.itemsize
+            self.offset.value += len(record_id_array) * self.dtype.itemsize
 
 
 def scoreDuplicates(

pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "dedupe"
 description = "A python library for accurate and scaleable data deduplication and entity-resolution"
-version = "2.0.20"
+version = "2.0.21"
 readme = "README.md"
 requires-python = ">=3.7"
 license = {file = "LICENSE"}
@@ -36,7 +36,7 @@ dependencies = [
     "haversine>=0.4.1",
     "BTrees>=4.1.4",
     "zope.index",
-    "Levenshtein_search==1.4.5",
+    "dedupe_Levenshtein_search",
     "typing_extensions",
 ]
 
@@ -68,7 +68,7 @@ check_untyped_defs = true
 
 [tool.pytest.ini_options]
 minversion = "7.1"
-addopts = "--cov dedupe --cov-report html"
+addopts = "--cov dedupe --cov-report xml"
 testpaths = ["tests", "dedupe"]
 
 [tool.isort]