Fix #2058 (Calibre reads Kindle metadata wrong)

src/calibre/__init__.py

@@ -26,12 +26,16 @@ mimetypes.add_type('text/x-sony-bbeb+xml',                '.lrs')
 mimetypes.add_type('application/xhtml+xml',               '.xhtml')
 mimetypes.add_type('image/svg+xml',                       '.svg')
 mimetypes.add_type('application/x-sony-bbeb',             '.lrf')
+mimetypes.add_type('application/x-sony-bbeb',             '.lrx')
 mimetypes.add_type('application/x-dtbncx+xml',            '.ncx')
 mimetypes.add_type('application/adobe-page-template+xml', '.xpgt')
 mimetypes.add_type('application/x-font-opentype',         '.otf')
 mimetypes.add_type('application/x-font-truetype',         '.ttf')
 mimetypes.add_type('application/oebps-package+xml',       '.opf')
 mimetypes.add_type('application/ereader',                 '.pdb')
+mimetypes.add_type('application/mobi',                    '.mobi')
+mimetypes.add_type('application/mobi',                    '.prc')
+mimetypes.add_type('application/mobi',                    '.azw')
 guess_type = mimetypes.guess_type
 import cssutils
 cssutils.log.setLevel(logging.WARN)

src/calibre/devices/kindle/driver.py

@@ -4,9 +4,9 @@ __copyright__ = '2009, John Schember <john at nachtimwald.com>'
 Device driver for Amazon's Kindle
 '''
 
-import os
+import os, re
 
-from calibre.devices.usbms.driver import USBMS
+from calibre.devices.usbms.driver import USBMS, metadata_from_formats
 
 class KINDLE(USBMS):
     # Ordered list of supported formats
@@ -30,6 +30,9 @@ class KINDLE(USBMS):
     EBOOK_DIR_CARD = "documents"
     SUPPORTS_SUB_DIRS = True
     
+    WIRELESS_FILE_NAME_PATTERN = re.compile(
+    r'(?P<title>[^-]+)-asin_(?P<asin>[a-zA-Z\d]{10,})-type_(?P<type>\w{4})-v_(?P<index>\d+).*')
+
     def delete_books(self, paths, end_session=True):
         for path in paths:
             if os.path.exists(path):
@@ -41,6 +44,16 @@ class KINDLE(USBMS):
                 if os.path.exists(filepath + '.mbp'):
                     os.unlink(filepath + '.mbp')
                     
+    @classmethod
+    def metadata_from_path(cls, path):
+        mi = metadata_from_formats([path])
+        if mi.title == _('Unknown') or ('-asin' in mi.title and '-type' in mi.title):
+            match = cls.WIRELESS_FILE_NAME_PATTERN.match(os.path.basename(path))
+            if match is not None:
+                mi.title = match.group('title')
+        return mi
+
+
 class KINDLE2(KINDLE):
     
     PRODUCT_ID = [0x0002]

src/calibre/devices/mime.py

@@ -1,19 +1,20 @@
-__license__   = 'GPL v3'
+from __future__ import with_statement
-__copyright__ = '2009, John Schember <john at nachtimwald.com>'
+__license__ = 'GPL 3'
-'''
+__copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
-Global Mime mapping of ebook types.
+__docformat__ = 'restructuredtext en'
-'''
 
-MIME_MAP   = {
+from calibre import guess_type
-                'azw'  : 'application/azw',
-                'epub' : 'application/epub+zip',
-                'html' : 'text/html',
-                'lrf'  : 'application/x-sony-bbeb',
-                'lrx'  : 'application/x-sony-bbeb',
-                'mobi' : 'application/mobi',
-                'pdf'  : 'application/pdf',
-                'prc'  : 'application/prc',
-                'rtf'  : 'application/rtf',
-                'txt'  : 'text/plain',
-              }
 
+def _mt(path):
+    mt = guess_type(path)[0]
+    if not mt:
+        mt = 'application/octet-stream'
+    return mt
+
+def mime_type_ext(ext):
+    if not ext.startswith('.'):
+        ext = '.'+ext
+    return _mt('a'+ext)
+
+def mime_type_path(path):
+    return _mt(path)

src/calibre/devices/usbms/driver.py

@@ -15,7 +15,7 @@ from calibre.ebooks.metadata import authors_to_string
 from calibre.devices.usbms.device import Device
 from calibre.devices.usbms.books import BookList, Book
 from calibre.devices.errors import FreeSpaceError, PathError
-from calibre.devices.mime import MIME_MAP
+from calibre.devices.mime import mime_type_ext
 
 class File(object):
     def __init__(self, path):
@@ -216,14 +216,17 @@ class USBMS(Device):
         if not os.path.isdir(path):
             os.utime(path, None)
 
+    @classmethod
+    def metadata_from_path(cls, path):
+        return metadata_from_formats([path])
+    
     @classmethod
     def book_from_path(cls, path):
         fileext = path_to_ext(path)
-    
+        mi = cls.metadata_from_path(path)
-        mi = metadata_from_formats([path])
+        mime = mime_type_ext(fileext)
-        mime = MIME_MAP[fileext] if fileext in MIME_MAP.keys() else 'Unknown'
-        
         authors = authors_to_string(mi.authors)
         
-        return Book(path, mi.title, authors, mime)
+        book = Book(path, mi.title, authors, mime)
+        return book
 

src/calibre/ebooks/metadata/meta.py

@@ -15,7 +15,7 @@ _METADATA_PRIORITIES = [
                        'html', 'htm', 'xhtml', 'xhtm',
                        'rtf', 'fb2', 'pdf', 'prc', 'odt',
                        'epub', 'lit', 'lrx', 'lrf', 'mobi',
-                       'rb', 'imp'
+                       'rb', 'imp', 'azw'
                       ]
 
 # The priorities for loading metadata from different file types
@@ -41,7 +41,9 @@ def metadata_from_formats(formats):
     for path, ext in zip(formats, extensions):
         with open(path, 'rb') as stream:
             try:
-                mi.smart_update(get_metadata(stream, stream_type=ext, use_libprs_metadata=True))
+                newmi = get_metadata(stream, stream_type=ext, 
+                                     use_libprs_metadata=True)
+                mi.smart_update(newmi)
             except:
                 continue
             if getattr(mi, 'application_id', None) is not None:
@@ -58,7 +60,7 @@ def get_metadata(stream, stream_type='lrf', use_libprs_metadata=False):
     if stream_type: stream_type = stream_type.lower()
     if stream_type in ('html', 'html', 'xhtml', 'xhtm', 'xml'):
         stream_type = 'html'
-    if stream_type in ('mobi', 'prc'):
+    if stream_type in ('mobi', 'prc', 'azw'):
         stream_type = 'mobi'
     if stream_type in ('odt', 'ods', 'odp', 'odg', 'odf'):
         stream_type = 'odt'