comparators/deb: Looser matching for .deb archive members. (Closes: #881937)

diffoscope/comparators/deb.py

@@ -17,6 +17,7 @@
 # You should have received a copy of the GNU General Public License
 # along with diffoscope.  If not, see <https://www.gnu.org/licenses/>.
 
+import os
 import re
 import logging
 
@@ -88,6 +89,16 @@ class DebContainer(LibarchiveContainer):
 
             return specialize(member.as_container.get_member('content'))
 
+    def perform_fuzzy_matching(self, my_members, other_members):
+        matched = set()
+        for name1 in my_members.keys():
+            main, ext = os.path.splitext(name1)
+            candidates = [name2 for name2 in other_members.keys() - matched
+                          if os.path.splitext(name2)[0] == main]
+            if len(candidates) == 1:
+                yield name1, candidates[0], 0
+                matched.add(candidates[0])
+
 
 class DebFile(File):
     CONTAINER_CLASS = DebContainer

diffoscope/comparators/utils/container.py

@@ -79,6 +79,9 @@ class Container(object, metaclass=abc.ABCMeta):
         for name in filter_excludes(self.get_member_names()):
             yield name, self.get_member(name)
 
+    def perform_fuzzy_matching(self, my_members, other_members):
+        return perform_fuzzy_matching(my_members, other_members)
+
     def get_adjusted_members(self):
         """
         Returns an iterable of pairs. The key is what is used to match when
@@ -150,7 +153,7 @@ class Container(object, metaclass=abc.ABCMeta):
             for name in both_names:
                 yield prep_yield(name, name)
 
-            for my_name, other_name, score in perform_fuzzy_matching(my_members, other_members):
+            for my_name, other_name, score in self.perform_fuzzy_matching(my_members, other_members):
                 comment = "Files similar despite different names" \
                     " (difference score: {})".format(score)
                 yield prep_yield(my_name, other_name, comment)

tests/comparators/test_deb.py

@@ -124,3 +124,15 @@ def test_compare_non_existing(monkeypatch, deb1):
     difference = deb1.compare(MissingFile('/nonexisting', deb1))
     assert difference.source2 == '/nonexisting'
     assert difference.details[-1].source2 == '/dev/null'
+
+
+bug881937_deb1 = load_fixture('bug881937_1.deb')
+bug881937_deb2 = load_fixture('bug881937_2.deb')
+
+def test_compare_different_compression(bug881937_deb1, bug881937_deb2):
+    difference = bug881937_deb1.compare(bug881937_deb2)
+    assert difference.details[1].source1 == 'control.tar.gz'
+    assert difference.details[1].source2 == 'control.tar.xz'
+    expected_diff = get_data('bug881937_control_expected_diff')
+    assert difference.details[1].details[2].details[1].unified_diff == expected_diff
+

tests/data/bug881937_control_expected_diff

+@@ -1,16 +1,16 @@
+ Package: aptitude-robot
+-Version: 1.5.2-1
++Version: 1.5.3-1
+ Architecture: all
+ Maintainer: Elmar S. Heeb <elmar@heebs.ch>
+ Installed-Size: 85
+ Depends: aptitude, libmoo-perl, librun-parts-perl, lsb-base (>= 3.0-6), perl, psmisc
+ Recommends: perl-doc
+-Suggests: needrestart, needrestart-session, apt-listbugs, heirloom-mailx | bsd-mailx | mailx, xymon-client | hobbit-client
++Suggests: needrestart, needrestart-session, apt-listbugs, bsd-mailx | mailx, xymon-client | hobbit-client
+ Enhances: aptitude
+ Section: admin
+-Priority: extra
++Priority: optional
+ Homepage: https://github.com/elmar/aptitude-robot
+ Description: Automate package choice management
+  Framework to use aptitude for automated package management including
+  unattended upgrade, installation, removal, hold, etc.  Allows you to automate
+  what you would manually do with aptitude.