Refactor tests, Delta.to_dict() #12

src/deltacode/__init__.py

@@ -84,7 +84,11 @@ def determine_delta(self):
                         deltas['unmodified'].append(Delta(new_file, f, 'unmodified'))
                         continue
                     else:
-                        deltas['modified'].append(Delta(new_file, f, 'modified'))
+                        delta = Delta(new_file, f, 'modified')
+                        # Change the Delta object's 'category' attribute to
+                        # 'license change' if a substantial license change has been detected.
+                        delta.license_diff()
+                        deltas['modified'].append(delta)
 
         # now time to find the added.
         for path, old_files in old_index.items():
@@ -124,11 +128,11 @@ def get_stats(self):
 
     def to_dict(self):
         """
-        Given an OrderedDict of Delta objects, return an OrderedDict with a
-        'deltacode_version' field, a 'deltacode_stats' field, a 'deltas_count'
-        field, and a 'deltas' field containing a list of our Delta objects.
+        Given an OrderedDict of Delta objects, return an OrderedDict of Delta
+        objects grouping the objects under the keys 'added', 'removed',
+        'modified' or 'unmodified'.
         """
-        if self.deltas == None:
+        if self.deltas is None:
             return
 
         return OrderedDict([
@@ -151,8 +155,40 @@ def __init__(self, new_file=None, old_file=None, delta_type=None):
         self.old_file = old_file
         self.category = delta_type
 
+    def license_diff(self):
+        """
+        Compare the license details for a pair of 'new' and 'old' File objects
+        in a Delta object and change the Delta object's 'category' attribute to
+        'license change' if those details differ and the cutoff score test is
+        satisfied.
+        """
+        if self.new_file is None or self.old_file is None:
+            return
+
+        cutoff_score = 50
+
+        if self.new_file.licenses:
+            new_keys = [l.key for l in self.new_file.licenses if l.score >= cutoff_score]
+        else:
+            new_keys = []
+
+        if self.old_file.licenses:
+            old_keys = [l.key for l in self.old_file.licenses if l.score >= cutoff_score]
+        else:
+            old_keys = []
+
+        new_keys = list(set(new_keys))
+        old_keys = list(set(old_keys))
+
+        if new_keys != old_keys:
+            self.category = 'license change'
+
     def to_dict(self):
-        if self.new_file == None and self.old_file == None:
+        """
+        Check the 'category' attribute of the Delta object and return an
+        OrderedDict comprising the 'category' and 'path' of the object.
+        """
+        if self.new_file is None and self.old_file is None:
             return
 
         if self.category == 'added':
@@ -168,10 +204,15 @@ def to_dict(self):
         elif self.category == 'modified':
             return OrderedDict([
                 ('category', 'modified'),
-                ('path', self.old_file.path)
+                ('path', self.new_file.path)
+            ])
+        elif self.category == 'license change':
+            return OrderedDict([
+                ('category', 'license change'),
+                ('path', self.new_file.path)
             ])
         else:
             return OrderedDict([
                 ('category', 'unmodified'),
-                ('path', self.old_file.path)
+                ('path', self.new_file.path)
             ])

tests/data/cli/modified_new_license_added.csv

@@ -0,0 +1,4 @@
+Type of delta,Path
+modified,some/path/c/c1.py
+license change,some/path/a/a1.py
+license change,some/path/b/b1.py

tests/data/cli/modified_new_license_added_low_score.csv

@@ -0,0 +1,3 @@
+Type of delta,Path
+modified,some/path/a/a1.py
+modified,some/path/b/b1.py

tests/data/cli/scan_modified_new_license_added.json

@@ -0,0 +1,92 @@
+{
+    "scancode_notice": "Generated with ScanCode and provided on an \"AS IS\" BASIS, WITHOUT WARRANTIES\nOR CONDITIONS OF ANY KIND, either express or implied. No content created from\nScanCode should be considered or used as legal advice. Consult an Attorney\nfor any legal advice.\nScanCode is a free software code scanning tool from nexB Inc. and others.\nVisit https://github.com/nexB/scancode-toolkit/ for support and download.",
+    "scancode_version": "2.1.0",
+    "scancode_options": {
+        "--license": true,
+        "--info": true
+    },
+    "files_count": 3,
+    "files": [
+      {
+        "path": "some/path/a/a1.py",
+        "type": "file",
+        "name": "a1.py",
+        "size": 300,
+        "sha1": "000647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "apache-2.0",
+            "score": 80.0,
+            "short_name": "Apache 2.0",
+            "category": "Permissive"
+          },
+          {
+            "key": "agpl-2.0",
+            "score": 70.0,
+            "short_name": "AGPL 2.0",
+            "category": "Copyleft"
+          },
+          {
+            "key": "bsd-simplified",
+            "score": 100.0,
+            "short_name": "BSD-Simplified",
+            "category": "Permissive"
+          },
+          {
+            "key": "mit",
+            "score": 100.0,
+            "short_name": "MIT License",
+            "category": "Permissive"
+          }
+        ]
+      },
+      {
+        "path": "some/path/b/b1.py",
+        "type": "file",
+        "name": "b1.py",
+        "size": 300,
+        "sha1": "111647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "apache-2.0",
+            "score": 80.0,
+            "short_name": "Apache 2.0",
+            "category": "Permissive"
+          },
+          {
+            "key": "gpl-2.0",
+            "score": 60.0,
+            "short_name": "GPL 2.0",
+            "category": "Copyleft"
+          },
+          {
+            "key": "bsd-simplified",
+            "score": 100.0,
+            "short_name": "BSD-Simplified",
+            "category": "Permissive"
+          },
+          {
+            "key": "mit",
+            "score": 100.0,
+            "short_name": "MIT License",
+            "category": "Permissive"
+          }
+        ]
+      },
+      {
+        "path": "some/path/c/c1.py",
+        "type": "file",
+        "name": "c1.py",
+        "size": 300,
+        "sha1": "333647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "mit",
+            "score": 100.0,
+            "short_name": "MIT License",
+            "category": "Permissive"
+          }
+        ]
+      }
+    ]
+  }

tests/data/cli/scan_modified_new_license_added_low_score.json

@@ -0,0 +1,59 @@
+{
+    "scancode_notice": "Generated with ScanCode and provided on an \"AS IS\" BASIS, WITHOUT WARRANTIES\nOR CONDITIONS OF ANY KIND, either express or implied. No content created from\nScanCode should be considered or used as legal advice. Consult an Attorney\nfor any legal advice.\nScanCode is a free software code scanning tool from nexB Inc. and others.\nVisit https://github.com/nexB/scancode-toolkit/ for support and download.",
+    "scancode_version": "2.1.0",
+    "scancode_options": {
+        "--license": true,
+        "--info": true
+    },
+    "files_count": 2,
+    "files": [
+      {
+        "path": "some/path/a/a1.py",
+        "type": "file",
+        "name": "a1.py",
+        "size": 350,
+        "sha1": "222647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "apache-2.0",
+            "score": 80.0,
+            "short_name": "Apache 2.0",
+            "category": "Permissive"
+          },
+          {
+            "key": "agpl-2.0",
+            "score": 70.0,
+            "short_name": "AGPL 2.0",
+            "category": "Copyleft"
+          },
+          {
+            "key": "bsd-simplified",
+            "score": 40.0,
+            "short_name": "BSD-Simplified",
+            "category": "Permissive"
+          }
+        ]
+      },
+      {
+        "path": "some/path/b/b1.py",
+        "type": "file",
+        "name": "b1.py",
+        "size": 290,
+        "sha1": "333647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "mit",
+            "score": 100.0,
+            "short_name": "MIT License",
+            "category": "Permissive"
+          },
+          {
+            "key": "bsd-simplified",
+            "score": 30.0,
+            "short_name": "BSD-Simplified",
+            "category": "Permissive"
+          }
+        ]
+      }
+    ]
+  }

tests/data/cli/scan_modified_old_license_added.json

@@ -0,0 +1,80 @@
+{
+    "scancode_notice": "Generated with ScanCode and provided on an \"AS IS\" BASIS, WITHOUT WARRANTIES\nOR CONDITIONS OF ANY KIND, either express or implied. No content created from\nScanCode should be considered or used as legal advice. Consult an Attorney\nfor any legal advice.\nScanCode is a free software code scanning tool from nexB Inc. and others.\nVisit https://github.com/nexB/scancode-toolkit/ for support and download.",
+    "scancode_version": "2.1.0",
+    "scancode_options": {
+      "--license": true,
+      "--info": true
+    },
+    "files_count": 3,
+    "files": [
+      {
+        "path": "some/path/a/a1.py",
+        "type": "file",
+        "name": "a1.py",
+        "size": 200,
+        "sha1": "84b647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "apache-2.0",
+            "score": 80.0,
+            "short_name": "Apache 2.0",
+            "category": "Permissive"
+          },
+          {
+            "key": "public-domain",
+            "score": 10.0,
+            "short_name": "Public Domain",
+            "category": "Public Domain"
+          },
+          {
+            "key": "bsd-simplified",
+            "score": 100.0,
+            "short_name": "BSD-Simplified",
+            "category": "Permissive"
+          }
+        ]
+      },
+      {
+        "path": "some/path/b/b1.py",
+        "type": "file",
+        "name": "b2.py",
+        "size": 200,
+        "sha1": "333647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "mpl-2.0",
+            "score": 90.0,
+            "short_name": "MPL 2.0",
+            "category": "Copyleft Limited"
+          },
+          {
+            "key": "public-domain",
+            "score": 10.0,
+            "short_name": "Public Domain",
+            "category": "Public Domain"
+          },
+          {
+            "key": "bsd-simplified",
+            "score": 100.0,
+            "short_name": "BSD-Simplified",
+            "category": "Permissive"
+          }
+        ]
+      },
+      {
+        "path": "some/path/c/c1.py",
+        "type": "file",
+        "name": "c1.py",
+        "size": 300,
+        "sha1": "222647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "mit",
+            "score": 100.0,
+            "short_name": "MIT License",
+            "category": "Permissive"
+          }
+        ]
+      }
+    ]
+  }

tests/data/cli/scan_modified_old_license_added_low_score.json

@@ -0,0 +1,47 @@
+{
+    "scancode_notice": "Generated with ScanCode and provided on an \"AS IS\" BASIS, WITHOUT WARRANTIES\nOR CONDITIONS OF ANY KIND, either express or implied. No content created from\nScanCode should be considered or used as legal advice. Consult an Attorney\nfor any legal advice.\nScanCode is a free software code scanning tool from nexB Inc. and others.\nVisit https://github.com/nexB/scancode-toolkit/ for support and download.",
+    "scancode_version": "2.1.0",
+    "scancode_options": {
+        "--license": true,
+        "--info": true
+    },
+    "files_count": 2,
+    "files": [
+      {
+        "path": "some/path/a/a1.py",
+        "type": "file",
+        "name": "a1.py",
+        "size": 300,
+        "sha1": "000647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "apache-2.0",
+            "score": 80.0,
+            "short_name": "Apache 2.0",
+            "category": "Permissive"
+          },
+          {
+            "key": "agpl-2.0",
+            "score": 70.0,
+            "short_name": "AGPL 2.0",
+            "category": "Copyleft"
+          }
+        ]
+      },
+      {
+        "path": "some/path/b/b1.py",
+        "type": "file",
+        "name": "b1.py",
+        "size": 300,
+        "sha1": "111647771481d39dd3a53f6dc210c26abac37748",
+        "licenses": [
+          {
+            "key": "mit",
+            "score": 100.0,
+            "short_name": "MIT License",
+            "category": "Permissive"
+          }
+        ]
+      }
+    ]
+  }