Merge pull request #85 from nexB/83-add-copyright-diff

Add Copyright class and copyright_diff() #82 #83

Author: steven-esser

src/deltacode/__init__.py

@@ -57,6 +57,7 @@ def __init__(self, new_path, old_path, options):
             self.determine_delta()
             self.determine_moved()
             self.license_diff()
+            self.copyright_diff()
             # Sort deltas by score, descending, i.e., high > low.
             self.deltas.sort(key=lambda Delta: Delta.score, reverse=True)
 
@@ -191,28 +192,57 @@ def license_diff(self):
         attribute -- if there has been a license change and depending on the
         nature of that change.
         """
-        for i in self.deltas:
-            if 20 <= i.score < 100:
+        for delta in self.deltas:
+            if 20 <= delta.score < 100:
 
-                new_licenses = i.new_file.licenses or []
-                old_licenses = i.old_file.licenses or []
+                new_licenses = delta.new_file.licenses or []
+                old_licenses = delta.old_file.licenses or []
 
-                if len(i.new_file.licenses) > 0 and i.old_file.licenses == []:
-                    i.factors.append('license info added')
-                    i.score += 20
+                if len(delta.new_file.licenses) > 0 and delta.old_file.licenses == []:
+                    delta.update(20, 'license info added')
                     return
 
-                if i.new_file.licenses == [] and len(i.old_file.licenses) > 0:
-                    i.factors.append('license info removed')
-                    i.score += 15
+                if delta.new_file.licenses == [] and len(delta.old_file.licenses) > 0:
+                    delta.update(15, 'license info removed')
                     return
 
-                new_keys = set(l.key for l in new_licenses)
-                old_keys = set(l.key for l in old_licenses)
+                new_keys = set(license.key for license in new_licenses)
+                old_keys = set(license.key for license in old_licenses)
 
                 if new_keys != old_keys:
-                    i.factors.append('license change')
-                    i.score += 10
+                    delta.update(10, 'license change')
+
+    def copyright_diff(self):
+        """
+        Compare the copyright details for a pair of 'new' and 'old' File objects
+        in a Delta object and change the Delta object's 'score' attribute --
+        and add an appropriate category (e.g., 'copyright info removed', 'copyright
+        info added' or 'copyright change') to the Delta object's 'factors'
+        attribute -- if there has been a copyright change and depending on the
+        nature of that change.
+        """
+        for delta in self.deltas:
+            if 20 <= delta.score < 100:
+
+                new_copyrights = delta.new_file.copyrights or []
+                old_copyrights = delta.old_file.copyrights or []
+
+                if len(delta.new_file.copyrights) > 0 and delta.old_file.copyrights == []:
+                    delta.update(10, 'copyright info added')
+                    return
+                elif delta.new_file.copyrights == [] and len(delta.old_file.copyrights) > 0:
+                    delta.update(10, 'copyright info removed')
+                    return
+
+                new_statements = set(statement for copyright in new_copyrights for statement in copyright.statements)
+                old_statements = set(statement for copyright in old_copyrights for statement in copyright.statements)
+
+                new_holders = set(holder for copyright in new_copyrights for holder in copyright.holders)
+                old_holders = set(holder for copyright in old_copyrights for holder in copyright.holders)
+
+                if ((new_statements != old_statements) or
+                        (new_holders != old_holders)):
+                    delta.update(5, 'copyright change')
 
     def index_deltas(self, index_key='path', delta_list=[]):
         """
@@ -250,6 +280,15 @@ def __init__(self, score=0, new_file=None, old_file=None):
         self.factors = []
         self.score = score
 
+    def update(self, score=0, factor=''):
+        """
+        Add the score to the Delta object's 'score' attribute and add a string,
+        summarizing the factor associated with the score, to the object's
+        'factors' attribute (a list).
+        """
+        self.factors.append(factor)
+        self.score += score
+
     def to_dict(self):
         """
         Return an OrderedDict comprising the 'factors', 'score' and new and old

src/deltacode/models.py

@@ -161,13 +161,20 @@ def __init__(self, dictionary={}):
         self.sha1 = dictionary.get('sha1', '')
         self.original_path = ''
         self.licenses = self.get_licenses(dictionary) if dictionary.get('licenses') else []
+        self.copyrights = self.get_copyrights(dictionary) if dictionary.get('copyrights') else []
 
     def get_licenses(self, dictionary):
         if dictionary.get('licenses') == []:
             return []
         else:
             return [License(l) for l in dictionary.get('licenses')]
 
+    def get_copyrights(self, dictionary):
+        if dictionary.get('copyrights') == []:
+            return []
+        else:
+            return [Copyright(l) for l in dictionary.get('copyrights')]
+
     def to_dict(self):
         d = OrderedDict([
             ('path', self.path),
@@ -180,6 +187,13 @@ def to_dict(self):
 
         if self.licenses:
             d['licenses'] = [l.to_dict() for l in self.licenses]
+        else:
+            d['licenses'] = []
+
+        if self.copyrights:
+            d['copyrights'] = [l.to_dict() for l in self.copyrights]
+        else:
+            d['copyrights'] = []
 
         return d
 
@@ -199,7 +213,8 @@ def __repr__(self):
 
 class License(object):
     """
-    License object created from the 'license' field in an ABCD formatted 'file' dictionary.
+    License object created from the 'license' field in an ABCD formatted 'file'
+    dictionary.
     """
     def __init__(self, dictionary={}):
         self.key = dictionary.get('key')
@@ -225,7 +240,37 @@ def to_dict(self):
 
     def __repr__(self):
         """
-        Return string containing a printable representation of the License object.
+        Return string containing a printable representation of the License
+        object.
+        """
+        return "%s" % self.__dict__
+
+
+class Copyright(object):
+    """
+    Copyright object created from the 'copyrights' field in an ABCD formatted
+    'file' dictionary.
+    """
+    def __init__(self, dictionary={}):
+        self.statements = dictionary.get('statements')
+        self.holders = dictionary.get('holders')
+
+    def to_dict(self):
+        """
+        Given a Copyright object, return an OrderedDict with the full
+        set of fields from the ScanCode 'copyrights' value.
+        """
+        d = OrderedDict([
+            ('statements', self.statements),
+            ('holders', self.holders)
+        ])
+
+        return d
+
+    def __repr__(self):
+        """
+        Return string containing a printable representation of the Copyright
+        object.
         """
         return "%s" % self.__dict__
 

tests/data/cli/copyright_and_license_info_added.csv

@@ -0,0 +1,2 @@
+Factors,Score,Path,Name,Type,Size,Old Path
+modified license info added copyright info added,50,path.txt,path.txt,file,300,

tests/data/cli/copyright_and_license_info_added_new.json

@@ -0,0 +1,63 @@
+{
+    "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": "new/default.txt",
+            "type": "file",
+            "name": "default.txt",
+            "size": 100,
+            "sha1": "a",
+            "licenses": [],
+            "copyrights": []
+        },
+        {
+            "path": "new/path.txt",
+            "type": "file",
+            "name": "path.txt",
+            "size": 300,
+            "sha1": "b_modified",
+            "licenses": [
+                {
+                    "key": "gpl-1.0-plus",
+                    "score": 20.0,
+                    "short_name": "GPL 1.0 or later",
+                    "category": "Copyleft",
+                    "owner": "Free Software Foundation (FSF)",
+                    "homepage_url": "http://www.gnu.org/licenses/old-licenses/gpl-1.0-standalone.html",
+                    "text_url": "http://www.gnu.org/licenses/old-licenses/gpl-1.0-standalone.html",
+                    "reference_url": "https://enterprise.dejacode.com/urn/urn:dje:license:gpl-1.0-plus",
+                    "spdx_license_key": "GPL-1.0+",
+                    "spdx_url": "https://spdx.org/licenses/GPL-1.0",
+                    "start_line": 353,
+                    "end_line": 353,
+                    "matched_rule": {
+                      "identifier": "gpl_63.RULE",
+                      "license_choice": false,
+                      "licenses": [
+                        "gpl-1.0-plus"
+                      ]
+                    }
+                  }
+              ],
+            "copyrights": [
+                {
+                    "statements": [
+                        "Copyright (c) 2016 Mark Adler"
+                    ],
+                    "holders": [
+                        "Mark Adler"
+                    ],
+                    "authors": [],
+                    "start_line": 1,
+                    "end_line": 3
+                }
+            ]
+        }
+    ]
+}

tests/data/cli/copyright_and_license_info_added_old.json

@@ -0,0 +1,29 @@
+{
+    "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": "old/default.txt",
+            "type": "file",
+            "name": "default.txt",
+            "size": 100,
+            "sha1": "a",
+            "licenses": [],
+            "copyrights": []
+        },
+        {
+            "path": "old/path.txt",
+            "type": "file",
+            "name": "path.txt",
+            "size": 300,
+            "sha1": "b",
+            "licenses": [],
+            "copyrights": []
+        }
+    ]
+}

tests/data/cli/copyright_and_license_info_removed.csv

@@ -0,0 +1,2 @@
+Factors,Score,Path,Name,Type,Size,Old Path
+modified license info removed copyright info removed,45,path.txt,path.txt,file,300,

tests/data/cli/copyright_and_license_info_removed_new.json

@@ -0,0 +1,29 @@
+{
+    "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": "new/default.txt",
+            "type": "file",
+            "name": "default.txt",
+            "size": 100,
+            "sha1": "a",
+            "licenses": [],
+            "copyrights": []
+        },
+        {
+            "path": "new/path.txt",
+            "type": "file",
+            "name": "path.txt",
+            "size": 300,
+            "sha1": "b_modified",
+            "licenses": [],
+            "copyrights": []
+        }
+    ]
+}

tests/data/cli/copyright_and_license_info_removed_old.json

@@ -0,0 +1,63 @@
+{
+    "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": "old/default.txt",
+            "type": "file",
+            "name": "default.txt",
+            "size": 100,
+            "sha1": "a",
+            "licenses": [],
+            "copyrights": []
+        },
+        {
+            "path": "old/path.txt",
+            "type": "file",
+            "name": "path.txt",
+            "size": 300,
+            "sha1": "b",
+            "licenses": [
+                {
+                    "key": "gpl-1.0-plus",
+                    "score": 20.0,
+                    "short_name": "GPL 1.0 or later",
+                    "category": "Copyleft",
+                    "owner": "Free Software Foundation (FSF)",
+                    "homepage_url": "http://www.gnu.org/licenses/old-licenses/gpl-1.0-standalone.html",
+                    "text_url": "http://www.gnu.org/licenses/old-licenses/gpl-1.0-standalone.html",
+                    "reference_url": "https://enterprise.dejacode.com/urn/urn:dje:license:gpl-1.0-plus",
+                    "spdx_license_key": "GPL-1.0+",
+                    "spdx_url": "https://spdx.org/licenses/GPL-1.0",
+                    "start_line": 353,
+                    "end_line": 353,
+                    "matched_rule": {
+                      "identifier": "gpl_63.RULE",
+                      "license_choice": false,
+                      "licenses": [
+                        "gpl-1.0-plus"
+                      ]
+                    }
+                  }
+              ],
+            "copyrights": [
+                {
+                    "statements": [
+                        "Copyright (c) 2016 Mark Adler"
+                    ],
+                    "holders": [
+                        "Mark Adler"
+                    ],
+                    "authors": [],
+                    "start_line": 1,
+                    "end_line": 3
+                }
+            ]
+        }
+    ]
+}

tests/data/cli/copyright_change_no_license_change.csv

@@ -0,0 +1,2 @@
+Factors,Score,Path,Name,Type,Size,Old Path
+modified copyright change,25,path.txt,path.txt,file,300,