[#15 state:fixed] Improved tie breaker in cases where filenames end with digits inside brackets.

core/scanner.py

@@ -7,7 +7,7 @@
 # http://www.hardcoded.net/licenses/hs_license
 
 import logging
-
+import re
 
 from hscommon import job
 from hsutil import io
@@ -27,6 +27,15 @@ SCAN_TYPE_CONTENT_AUDIO) = range(7)
 
 SCANNABLE_TAGS = ['track', 'artist', 'album', 'title', 'genre', 'year']
 
+RE_DIGIT_ENDING = re.compile(r'\d+|\(\d+\)|\[\d+\]|{\d+}')
+
+def is_same_with_digit(name, refname):
+    # Returns True if name is the same as refname, but with digits (with brackets or not) at the end
+    if not name.startswith(refname):
+        return False
+    end = name[len(refname):].strip()
+    return RE_DIGIT_ENDING.match(end) is not None
+
 class Scanner(object):
     def __init__(self):
         self.ignore_list = IgnoreList()
@@ -71,9 +80,9 @@ class Scanner(object):
             return False
         if 'copy' in refname:
             return True
-        if dupename.startswith(refname) and (dupename[len(refname):].strip().isdigit()):
+        if is_same_with_digit(dupename, refname):
             return False
-        if refname.startswith(dupename) and (refname[len(dupename):].strip().isdigit()):
+        if is_same_with_digit(refname, dupename):
             return True
         return len(dupe.path) > len(ref.path)
     

core/tests/scanner_test.py

@@ -429,11 +429,20 @@ class ScannerTestFakeFiles(TestCase):
         # if ref has the same words as dupe, but has some just one extra word which is a digit, it
         # becomes a dupe
         s = Scanner()
-        o1, o2 = no('foo bar 42'), no('foo bar')
+        o1 = no('foo bar 42')
+        o2 = no('foo bar [42]')
+        o3 = no('foo bar (42)')
+        o4 = no('foo bar {42}')
+        o5 = no('foo bar')
+        # all numbered names have deeper paths, so they'll end up ref if the digits aren't correctly
+        # used as tie breakers
         o1.path = Path('deeper/path')
-        o2.path = Path('foo')
-        [group] = s.GetDupeGroups([o1, o2])
-        assert group.ref is o2
+        o2.path = Path('deeper/path')
+        o3.path = Path('deeper/path')
+        o4.path = Path('deeper/path')
+        o5.path = Path('foo')
+        [group] = s.GetDupeGroups([o1, o2, o3, o4, o5])
+        assert group.ref is o5
     
     def test_partial_group_match(self):
         # Count the number od discarded matches (when a file doesn't match all other dupes of the