diff --git a/src/calibre/ebooks/metadata/archive.py b/src/calibre/ebooks/metadata/archive.py
index adbea86521..c0734fd78f 100644
--- a/src/calibre/ebooks/metadata/archive.py
+++ b/src/calibre/ebooks/metadata/archive.py
@@ -60,7 +60,19 @@ class ArchiveExtract(FileTypePlugin):
         else:
             fnames = zf.namelist()
 
-        fnames = [x for x in fnames if '.' in x and x.lower().rpartition('/')[-1] != 'thumbs.db']
+        def fname_ok(fname):
+            bn = os.path.basename(fname).lower()
+            if bn == 'thumbs.db':
+                return False
+            if '.' not in bn:
+                return False
+            if bn.rpartition('.')[-1] in {'diz', 'nfo'}:
+                return False
+            if '__MACOSX' in fname.split('/'):
+                return False
+            return True
+
+        fnames = list(filter(fname_ok, fnames))
         if is_comic(fnames):
             ext = '.cbr' if is_rar else '.cbz'
             of = self.temporary_file('_archive_extract'+ext)
@@ -72,8 +84,9 @@ class ArchiveExtract(FileTypePlugin):
             return archive
         fname = fnames[0]
         ext = os.path.splitext(fname)[1][1:]
-        if ext.lower() not in ('lit', 'epub', 'mobi', 'prc', 'rtf', 'pdf',
-                'mp3', 'pdb', 'azw', 'azw1', 'azw3', 'fb2'):
+        if ext.lower() not in {
+                'lit', 'epub', 'mobi', 'prc', 'rtf', 'pdf', 'mp3', 'pdb',
+                'azw', 'azw1', 'azw3', 'fb2', 'docx', 'doc', 'odt'}:
             return archive
 
         of = self.temporary_file('_archive_extract.'+ext)