Merge pull request #4 from dedupeio/jfc/attempted-match

Define indicator variable for attempted match and configure deployments

Author: jeancochrane

.github/workflows/main.yml

@@ -4,6 +4,8 @@ on:
   push:
     branches:
     - master
+    tags:
+    - v*
   pull_request:
     branches:
     - master
@@ -25,6 +27,31 @@ jobs:
       run: |
         pip install --upgrade pip
         pip install -e .[tests]
-        pip uninstall -y ilcs-parser
-        pip install -i https://test.pypi.org/simple/ ilcs-parser
         pytest
+  deploy:
+    if: github.event_name == 'push' && (startsWith(github.ref, 'refs/tags') || endsWith(github.ref, 'master'))
+    needs: test
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v1
+    - uses: actions/setup-python@v1
+      with:
+        python-version: 3.8
+    - name: Build distribution files
+      run: |
+        pip install --upgrade pip setuptools wheel
+        pip install -e .[tests]
+        python setup.py sdist bdist_wheel
+    - name: Publish to test PyPI
+      if: endsWith(github.ref, 'master')
+      uses: pypa/[email protected]
+      with:
+        user: __token__
+        password: ${{ secrets.test_pypi_password }}
+        repository_url: https://test.pypi.org/legacy/
+    - name: Publish to PyPI
+      if: startsWith(github.ref, 'refs/tags')
+      uses: pypa/[email protected]
+      with:
+        user: __token__
+        password: ${{ secrets.pypi_password }}

MANIFEST.in

@@ -0,0 +1,2 @@
+README.md
+LICENSE

dedupe/variables/ilcs.py

@@ -1,17 +1,12 @@
-import string
-import math
-import functools
-
 import numpy
 import ilcs_parser
-from parseratorvariable import ParseratorType, consolidate
+from parseratorvariable import ParseratorType
 
 CITATION = (
     ('chapter', ('Chapter',)),
     ('act prefix', ('ActPrefix',)),
     ('section', ('Section',)),
     ('subsection', ('SubSection',)),
-    ('attempted', ('Attempted',))
 )
 
 
@@ -25,29 +20,44 @@ def __init__(self, definition):
         self.components = (('Citation', self.compareFields, CITATION),)
         block_parts = ('Citation',)
         super().__init__(definition, ilcs_parser.tag, block_parts)
-        # Add exact match to the distance vector
-        self.expanded_size += 1
+        # Add exact match and attempted indicators to the distance vector
+        self.num_additional_indicators = 2
+        self.expanded_size += self.num_additional_indicators
 
     def fields(self, field):
         """
         Override the parent method to append an exact match field.
         """
         fields = super().fields(field)
-        fields += [('exact match', 'Exact')]
+        fields += [('attempted match', 'Dummy'), ('exact match', 'Exact')]
         return fields
 
     def comparator(self, field_1, field_2):
         """
         Override the parent method to append an exact match field.
         """
-        # Temporarily subtract the exact match indicator from expanded_size,
-        # since the parent method assumes that the last element of the distance
-        # vector is the full-string comparison.
-        self.expanded_size -= 1
+        # Temporarily subtract the exact and attempted match indicators from
+        # expanded_size, since the parent method assumes that the last element
+        # of the distance vector is the full-string comparison
+        self.expanded_size -= self.num_additional_indicators
         distances = super().comparator(field_1, field_2)
-        self.expanded_size += 1
+        self.expanded_size += self.num_additional_indicators
+
+        # Set the attempted match indicator variable
+        try:
+            parsed_variable_1 = self.tagger(field_1)
+            parsed_variable_2 = self.tagger(field_2)
+        except TypeError:
+            attempted = 0
+        else:
+            variable_type_1, variable_type_2 = parsed_variable_1[1], parsed_variable_2[1]
+            if 'Ambiguous' in (variable_type_1, variable_type_2):
+                attempted = 0
+            else:
+                attempted = int(parsed_variable_1.is_attempted == parsed_variable_2.is_attempted)
+        distances = numpy.append(distances, attempted)
 
-        # Set the exact match indicator variable.
+        # Set the exact match indicator variable
         exact_match = 1 if field_1 == field_2 else 0
         distances = numpy.append(distances, exact_match)
 

setup.py

@@ -7,15 +7,24 @@
         "for installing setuptools"
     )
 
+
+def readme():
+    with open('README.md') as f:
+        return f.read()
+
+
 setup(
     version='0.0.0',
+    author='DataMade',
     url='https://github.com/dedupeio/dedupe-variable-ilcs',
     description='Dedupe variable for Illinois Compiled Statute (ILCS) codes',
+    long_description=readme(),
+    long_description_content_type='text/markdown',
     name='dedupe-variable-ilcs',
     packages=['dedupe.variables'],
     license='The MIT License: http://www.opensource.org/licenses/mit-license.php',
     install_requires=[
-        'ilcs-parser @ https://github.com/datamade/ilcs-parser/archive/master.zip#egg=ilcs-parser-0.0.0',
+        'ilcs-parser',
         'parseratorvariable'
     ],
     extras_require={'tests': ['pytest', 'parserator']},

tests/test_ilcs.py

@@ -3,23 +3,30 @@
 
 def test_all_distances(ilcs):
     numpy.testing.assert_almost_equal(
-        ilcs.comparator('125 55/21-a (att)', '126 55/21-b (atttt)'),
+        ilcs.comparator('125 55/21-a (att)', '126 55/21-b (att)'),
         numpy.array([
-            1, 0, 1,
-            ilcs.compareString('126', '125'),
-            ilcs.compareString('55', '55'),
-            ilcs.compareString('21', '21'),
-            ilcs.compareString('a', 'b'),
-            ilcs.compareString('att', 'atttt'),
-            1, 1, 1, 1, 1, 0, 0
+            1,  # citation: Not Missing
+            0,  # ambiguous: Dummy
+            1,  # same name type?: Dummy
+            ilcs.compareString('126', '125'),  # chapter: Derived
+            ilcs.compareString('55', '55'),  # act prefix: Derived
+            ilcs.compareString('21', '21'),  # section: Derived
+            ilcs.compareString('a', 'b'),  # subsection: Derived
+            1,  # chapter: Not Missing
+            1,  # act prefix: Not Missing
+            1,  # section: Not Missing
+            1,  # subsection: Not Missing
+            0,  # full string: String
+            1,  # attempted match: Dummy
+            0   # exact match: Exact
         ])
     )
 
 
 def test_exact_match(ilcs):
     numpy.testing.assert_almost_equal(
         ilcs.comparator('125 55/21 (att)', '125 55/21 (att)'),
-        numpy.array([1, 0, 1, 0.5, 0.5, 0.5, 0, 0.5, 1, 1, 1, 0, 1, 0, 1])
+        numpy.array([1, 0, 1, 0.5, 0.5, 0.5, 0, 1, 1, 1, 0, 0, 1, 1])
     )
 
 
@@ -29,8 +36,21 @@ def test_mismatched_elements(ilcs):
         numpy.array([
             1, 0, 1,
             ilcs.compareString('125', '125'),
-            ilcs.compareString('56', '55'),
+            ilcs.compareString('55', '56'),
+            ilcs.compareString('21', '21'),
+            0, 1, 1, 1, 0, 0, 0, 0
+        ])
+    )
+
+
+def test_attempted_match(ilcs):
+    numpy.testing.assert_almost_equal(
+        ilcs.comparator('725 5/21-a (att)', '720-5/8-4 725 6/21'),
+        numpy.array([
+            1, 0, 1,
+            ilcs.compareString('125', '125'),
+            ilcs.compareString('5', '6'),
             ilcs.compareString('21', '21'),
-            0, 0, 1, 1, 1, 0, 0, 0, 0
+            0, 1, 1, 1, 0, 0, 1, 0
         ])
     )