Handle more exceptions in title tag normalization code - 62

desbma · Apr 9, 2018 · e39345e · e39345e
1 parent dd8d3de
commit e39345e
Show file tree

Hide file tree

Showing 4 changed files with 20 additions and 5 deletions.
diff --git a/amg/sanitize.py b/amg/sanitize.py
@@ -26,11 +26,12 @@ def normalize_tag_case(s):
   new_words = []
   prev_word = None
   roman_letters = frozenset("IVXLCDM")
+  punct_followed_all_uppercase = set(".-")
   punct_followed_uppercase = set(string.punctuation)
   punct_followed_uppercase.remove("'")
   for i, old_word in enumerate(old_words):
     if (((prev_word is not None) and
-            ((prev_word[-1] in punct_followed_uppercase) and old_word[0].isupper())) or
+            ((prev_word[-1] in punct_followed_all_uppercase) and old_word[0].isupper())) or
             ("." in old_word)):
       new_word = old_word
     elif old_word[0] in "(-":
@@ -40,7 +41,7 @@ def normalize_tag_case(s):
         new_word = "'".join((old_word[0].lower(), old_word[2:].capitalize()))
       else:
         new_word = old_word
-    elif (i != 0) and (old_word.lower() in TAG_LOWERCASE_WORDS):
+    elif (i != 0) and (old_word.lower() in TAG_LOWERCASE_WORDS) and (prev_word[-1] not in punct_followed_uppercase):
       new_word = old_word.lower()
     elif all(map(roman_letters.__contains__,
                  old_word.strip(string.punctuation))):

diff --git a/amg/tag.py b/amg/tag.py
@@ -92,7 +92,8 @@ def __init__(self, artist, album):
     expressions.extend(("pre-orders available", "preorders available", "hd",
                         "official", "pre-listening", "prelistening", "trollzorn",
                         "uncensored", "s/t", "sw exclusive",
-                        "transcending obscurity india"))  # crappy label suffixes TODO how to handle that?
+                        "transcending obscurity india",  # crappy label suffixes TODO how to handle that?
+                        "trailer for the upcoming album"))
     year = datetime.datetime.today().year
     for y in range(year - 5, year + 1):
       expressions.append(str(y))
@@ -123,6 +124,9 @@ def __init__(self, artist, album):
     # normalize case
     self.registerCleaner(FunctionCleaner(sanitize.normalize_tag_case, execute_once=True))
 
+    # post normalize case fix
+    self.registerCleaner(FunctionCleaner(lambda x: x.replace("PT.", "pt."), execute_once=True))
+
   def registerCleaner(self, cleaner, args=()):
     assert(isinstance(cleaner, TitleCleanerBase))
     self.cleaners.append((cleaner, args))

diff --git a/tests/test_sanitize.py b/tests/test_sanitize.py
@@ -16,6 +16,8 @@ def test_normalize_tag_case(self):
                   "VI VI VI": "VI VI VI",
                   "Episode VI: name": "Episode VI: Name",
                   "Matsya - The Fish": "Matsya - The Fish",
-                  "I'M ALIVE!": "I'm Alive!"}
+                  "I'M ALIVE!": "I'm Alive!",
+                  "MARK OF THE BEAST PT. 2: SCION OF DARKNESS": "Mark of the Beast PT. 2: Scion of Darkness",
+                  "BZZ: THE": "Bzz: The"}
     for before, after in references.items():
       self.assertEqual(sanitize.normalize_tag_case(before), after)
diff --git a/tests/test_tag.py b/tests/test_tag.py
@@ -465,7 +465,15 @@ def test_normalize_title_tag(self):
                   ("Altars of Grief - Desolation [From album; Iris; 2018]",
                    "Altars of Grief",
                    "Iris",
-                   "Desolation"))
+                   "Desolation"),
+                  ("MONOTHEIST - MARK OF THE BEAST PT. 2: SCION OF DARKNESS",
+                   "Monotheist",
+                   "Scourge",
+                   "Mark of the Beast pt. 2: Scion of Darkness"),
+                  ("SHADOWKEEP - Trailer for the upcoming album \"ShadowKeep\" (PURE STEEL RECORDS)",
+                   "Shadowkeep",
+                   "Shadowkeep",
+                   "Shadowkeep"))
 
     for source, artist, album, expected_result in references:
       with self.subTest(source=source, expected_result=expected_result, artist=artist, album=album):