Merge branch '0.8.x'

* 0.8.x:
  fix: ID3 v2.3 to v2.4 date conversion
  fix: Match mp3 mime-types against all possible mime-types

src/eyed3/core.py

@@ -79,13 +79,16 @@ def load(path, tag_version=None):
     else:
         raise IOError("file not found: %s" % path)
 
-    mtype = guessMimetype(path)
-    log.debug("File mime-type: %s" % mtype)
+    mtypes = guessMimetype(path, all_types=True)
+    log.debug(f"File mime-type: {mtypes}")
 
-    if (mtype in mp3.MIME_TYPES or
-            (mtype in mp3.OTHER_MIME_TYPES and path.suffix.lower() in mp3.EXTENSIONS)):
+    if set(mtypes).intersection(set(mp3.MIME_TYPES)):
         return mp3.Mp3AudioFile(path, tag_version)
-    elif mtype == "application/x-id3":
+    elif (set(mtypes).intersection(set(mp3.OTHER_MIME_TYPES)) and
+            path.suffix.lower() in mp3.EXTENSIONS):
+        # Same as above, but mp3 was not typed detected; making this odd/special
+        return mp3.Mp3AudioFile(path, tag_version)
+    elif "application/x-id3" in mtypes:
         return id3.TagFile(path, tag_version)
     else:
         return None

src/eyed3/id3/tag.py

@@ -1057,6 +1057,7 @@ def _saveV2Tag(self, version, encoding, max_padding):
 
     def _convertFrames(self, std_frames, convert_list, version):
         """Maps frame incompatibilities between ID3 v2.3 and v2.4.
+
         The items in ``std_frames`` need no conversion, but the list/frames
         may be edited if necessary (e.g. a converted frame replaces a frame
         in the list).  The items in ``convert_list`` are the frames to convert
@@ -1078,6 +1079,14 @@ def _convertFrames(self, std_frames, convert_list, version):
                     date_frames[f.id] = f
 
         if date_frames:
+            def fidHandled(fid):
+                # A duplicate text frame (illegal ID3 but oft seen) may exist. The date_frames dict
+                # will have one, but the flist has multiple, hence the loop.
+                for frame in list(flist):
+                    if frame.id == fid:
+                        flist.remove(frame)
+                del date_frames[fid]
+
             if version == ID3_V2_4:
                 if b"TORY" in date_frames or b"XDOR" in date_frames:
                     # XDOR -> TDOR (full date)
@@ -1087,8 +1096,7 @@ def _convertFrames(self, std_frames, convert_list, version):
                         converted_frames.append(DateFrame(b"TDOR", date))
                     for fid in (b"TORY", b"XDOR"):
                         if fid in flist:
-                            flist.remove(date_frames[fid])
-                            del date_frames[fid]
+                            fidHandled(fid)
 
                 # TYER, TDAT, TIME -> TDRC
                 if (b"TYER" in date_frames or b"TDAT" in date_frames or
@@ -1098,16 +1106,14 @@ def _convertFrames(self, std_frames, convert_list, version):
                         converted_frames.append(DateFrame(b"TDRC", date))
                     for fid in [b"TYER", b"TDAT", b"TIME"]:
                         if fid in date_frames:
-                            flist.remove(date_frames[fid])
-                            del date_frames[fid]
+                            fidHandled(fid)
 
             elif version == ID3_V2_3:
                 if b"TDOR" in date_frames:
                     date = date_frames[b"TDOR"].date
                     if date:
                         converted_frames.append(DateFrame(b"TORY", str(date.year)))
-                    flist.remove(date_frames[b"TDOR"])
-                    del date_frames[b"TDOR"]
+                    fidHandled(b"TDOR")
 
                 if b"TDRC" in date_frames:
                     date = date_frames[b"TDRC"].date
@@ -1127,16 +1133,14 @@ def _convertFrames(self, std_frames, convert_list, version):
                             converted_frames.append(TextFrame(b"TIME",
                                                               date_str))
 
-                    flist.remove(date_frames[b"TDRC"])
-                    del date_frames[b"TDRC"]
+                    fidHandled(b"TDRC")
 
                 if b"TDRL" in date_frames:
                     # TDRL -> XDOR
                     date = date_frames[b"TDRL"].date
                     if date:
                         converted_frames.append(DateFrame(b"XDOR", str(date)))
-                    flist.remove(date_frames[b"TDRL"])
-                    del date_frames[b"TDRL"]
+                    fidHandled(b"TDRL")
 
             # All other date frames have no conversion
             for fid in date_frames:
@@ -1171,9 +1175,8 @@ def _convertFrames(self, std_frames, convert_list, version):
         if len(flist) != 0:
             unconverted = ", ".join([f.id.decode("ascii") for f in flist])
             if version[0] != 1:
-                raise TagException("Unable to covert the following frames to "
-                                   "version %s: %s" % (versionToString(version),
-                                                       unconverted))
+                raise TagException("Unable to convert the following frames to "
+                                   f"version {versionToString(version)}: {unconverted}")
 
         # Some frames in converted_frames may replace/edit frames in std_frames.
         for cframe in converted_frames:

src/eyed3/utils/__init__.py

@@ -31,32 +31,40 @@ def os_walk_unpack(w):
 
 class MagicTypes(magic.Magic):
     def __init__(self):
-        magic.Magic.__init__(self, mime=True, mime_encoding=False,
-                             keep_going=False)
+        magic.Magic.__init__(self, mime=True, mime_encoding=False, keep_going=True)
 
-    def guess_type(self, filename):
+    def guess_type(self, filename, all_types=False):
         if os.path.splitext(filename)[1] in ID3_MIME_TYPE_EXTENSIONS:
-            return ID3_MIME_TYPE
+            return ID3_MIME_TYPE if not all_types else [ID3_MIME_TYPE]
         try:
-            return self.from_file(filename)
+            types = self.from_file(filename)
         except UnicodeEncodeError:
             # https://github.com/ahupp/python-magic/pull/144
-            return self.from_file(filename.encode("utf-8", 'surrogateescape'))
+            types = self.from_file(filename.encode("utf-8", 'surrogateescape'))
+
+        delim = r"\012- "
+        if all_types:
+            return types.split(delim)
+        else:
+            return types.split(delim)[0]
+
 
 
 _mime_types = MagicTypes()
 
 
-def guessMimetype(filename, with_encoding=False):
-    """Return the mime-type for ``filename``. If ``with_encoding`` is True
-    the encoding is included and a 2-tuple is returned, (mine, enc)."""
+def guessMimetype(filename, with_encoding=False, all_types=False):
+    """Return the mime-type for ``filename`` (or list of possible types when `all_types` is True).
+
+    If ``with_encoding`` is True the encoding is included and a 2-tuple is returned, (mine, enc).
+    """
 
     filename = str(filename) if isinstance(filename, pathlib.Path) else filename
-    mime = _mime_types.guess_type(filename)
+    mime = _mime_types.guess_type(filename, all_types=all_types)
     if not with_encoding:
         return mime
     else:
-        warnings.warn("File character encoding no lopng return, value is None",
+        warnings.warn("File character encoding no longer returned, value is None",
                       UserWarning, stacklevel=2)
         return mime, None