Split each author into its own EXTH Creator field, even with the 'Use author_sort as author' setting.

Also, in a slightly unrelated note, sprinkle azw3 awareness in a few places.

src/calibre/ebooks/__init__.py

@@ -163,7 +163,7 @@ def render_html(path_to_html, width=590, height=750, as_xhtml=True):
 
 def check_ebook_format(stream, current_guess):
     ans = current_guess
-    if current_guess.lower() in ('prc', 'mobi', 'azw', 'azw1'):
+    if current_guess.lower() in ('prc', 'mobi', 'azw', 'azw1', 'azw3'):
         stream.seek(0)
         if stream.read(3) == 'TPZ':
             ans = 'tpz'

src/calibre/ebooks/metadata/archive.py

@@ -70,7 +70,7 @@ class ArchiveExtract(FileTypePlugin):
         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', 'fb2'):
+                'mp3', 'pdb', 'azw', 'azw1', 'azw3', 'fb2'):
             return archive
 
         of = self.temporary_file('_archive_extract.'+ext)

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', 'azw', 'snb'
+                       'rb', 'imp', 'azw', 'azw3', 'azw1' 'snb'
                       ]
 
 # The priorities for loading metadata from different file types
@@ -85,7 +85,7 @@ def _get_metadata(stream, stream_type, use_libprs_metadata,
     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', 'azw'):
+    if stream_type in ('mobi', 'prc', 'azw', 'azw1', 'azw3'):
         stream_type = 'mobi'
     if stream_type in ('odt', 'ods', 'odp', 'odg', 'odf'):
         stream_type = 'odt'

src/calibre/ebooks/metadata/mobi.py

@@ -341,11 +341,14 @@ class MetadataUpdater(object):
             kindle_pdoc = None
             share_not_sync = False
         if mi.author_sort and pas:
-            authors = mi.author_sort
+            # We want an EXTH field per author...
-            update_exth_record((100, normalize(authors).encode(self.codec, 'replace')))
+            authors = mi.author_sort.split(' & ')
+            for author in authors:
+                update_exth_record((100, normalize(author).encode(self.codec, 'replace')))
         elif mi.authors:
-            authors = ';'.join(mi.authors)
+            authors = mi.authors
-            update_exth_record((100, normalize(authors).encode(self.codec, 'replace')))
+            for author in authors:
+                update_exth_record((100, normalize(author).encode(self.codec, 'replace')))
         if mi.publisher:
             update_exth_record((101, normalize(mi.publisher).encode(self.codec, 'replace')))
         if mi.comments:
@@ -360,6 +363,7 @@ class MetadataUpdater(object):
         if mi.isbn:
             update_exth_record((104, mi.isbn.encode(self.codec, 'replace')))
         if mi.tags:
+            # FIXME: Keep a single subject per EXTH field?
             subjects = '; '.join(mi.tags)
             update_exth_record((105, normalize(subjects).encode(self.codec, 'replace')))
 

src/calibre/ebooks/metadata/rar.py

@@ -32,7 +32,7 @@ def get_metadata(stream):
         if stream_type:
             stream_type = stream_type[1:]
             if stream_type in ('lit', 'opf', 'prc', 'mobi', 'fb2', 'epub',
-                               'rb', 'imp', 'pdf', 'lrf', 'azw'):
+                               'rb', 'imp', 'pdf', 'lrf', 'azw', 'azw1', 'azw3'):
                 with TemporaryDirectory() as tdir:
                     with CurrentDir(tdir):
                        stream = extract_member(path, match=None, name=f,

src/calibre/ebooks/metadata/zip.py

@@ -23,7 +23,7 @@ def get_metadata(stream):
         if stream_type:
             stream_type = stream_type[1:]
             if stream_type in ('lit', 'opf', 'prc', 'mobi', 'fb2', 'epub',
-                               'rb', 'imp', 'pdf', 'lrf', 'azw'):
+                               'rb', 'imp', 'pdf', 'lrf', 'azw', 'azw1', 'azw3'):
                 with TemporaryDirectory() as tdir:
                     with CurrentDir(tdir):
                         path = zf.extract(f)

src/calibre/ebooks/mobi/writer8/exth.py

@@ -54,8 +54,16 @@ def build_exth(metadata, prefer_author_sort=False, is_periodical=False,
         items = metadata[term]
         if term == 'creator':
             if prefer_author_sort:
-                creators = [unicode(c.file_as or c) for c in
+                # This is a bit hackish... We only get the first item in the creators list,
+                # because we only care about the file_as property, and it contains *all* the authors in every creator markup,
+                # so we only need one, or we end up with duplicates ;).
+                # We then end up with a single item in our list, that contains every authors, in author sort syntax, separated by an ' & ' character.
+                # That's not good enough, because we want each author in a separate entry in the list, so we just split this on every & ;).
+                # This way, we properly end up with multiple Creator fields in the EXTH header, one for each author, like KindleGen :).
+                all_creators = [unicode(c.file_as or c) for c in
                         items][:1]
+                for creator in all_creators:
+                    creators = creator.split(' & ')
             else:
                 creators = [unicode(c) for c in items]
             items = creators

src/calibre/gui2/actions/add.py

@@ -30,7 +30,7 @@ def get_filters():
             (_('LRF Books'), ['lrf']),
             (_('HTML Books'), ['htm', 'html', 'xhtm', 'xhtml']),
             (_('LIT Books'), ['lit']),
-            (_('MOBI Books'), ['mobi', 'prc', 'azw']),
+            (_('MOBI Books'), ['mobi', 'prc', 'azw', 'azw3']),
             (_('Topaz books'), ['tpz','azw1']),
             (_('Text books'), ['txt', 'text', 'rtf']),
             (_('PDF Books'), ['pdf', 'azw4']),

src/calibre/gui2/wizard/__init__.py

@@ -446,7 +446,7 @@ class KindlePage(QWizardPage, KindleUI):
             if not accounts: accounts = {}
             for y in accounts.values():
                 y[2] = False
-            accounts[x] = ['AZW, MOBI, TPZ, PRC, AZW1', True, True]
+            accounts[x] = ['AZW, MOBI, TPZ, PRC, AZW1, AZW3', True, True]
             conf.set('accounts', accounts)
 
     def nextId(self):