MOBI Output: Add option to control setting of PDOC flag

resources/recipes/nytimes.recipe

@@ -9,7 +9,8 @@ import re
 import time
 from calibre import entity_to_unicode
 from calibre.web.feeds.recipes import BasicNewsRecipe
-from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, NavigableString, Comment
+from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, NavigableString, \
+Comment, BeautifulStoneSoup
 
 class NYTimes(BasicNewsRecipe):
 
@@ -256,7 +257,6 @@ class NYTimes(BasicNewsRecipe):
         # Fetch the outer table
         table = soup.find('table')
         previousTable = table
-        contentTable = None
 
         # Find the deepest table containing the stories
         while True :

resources/recipes/usatoday.recipe

@@ -128,7 +128,6 @@ class USAToday(BasicNewsRecipe):
     def parse_feeds(self, *args, **kwargs):
         parsed_feeds = BasicNewsRecipe.parse_feeds(self, *args, **kwargs)
         # Count articles for progress dialog
-        content_feeds = []
         article_count = 0
         for feed in parsed_feeds:
             article_count += len(feed)
@@ -297,7 +296,6 @@ class USAToday(BasicNewsRecipe):
 
             if not headline_found:
                 # Monolithic <div> - restructure
-                insert_loc = 0
                 tag = body.find(True)
                 while True:
                     insertLoc += 1

src/calibre/ebooks/metadata/mobi.py

@@ -322,11 +322,13 @@ class MetadataUpdater(object):
 
         recs = []
         try:
-             from calibre.ebooks.conversion.config import load_defaults
+            from calibre.ebooks.conversion.config import load_defaults
-             prefs = load_defaults('mobi_output')
+            prefs = load_defaults('mobi_output')
-             pas = prefs.get('prefer_author_sort', False)
+            pas = prefs.get('prefer_author_sort', False)
+            kindle_pdoc = prefs.get('personal_doc', None)
         except:
             pas = False
+            kindle_pdoc = None
         if mi.author_sort and pas:
             authors = mi.author_sort
             update_exth_record((100, authors.encode(self.codec, 'replace')))
@@ -350,15 +352,8 @@ class MetadataUpdater(object):
             subjects = '; '.join(mi.tags)
             update_exth_record((105, subjects.encode(self.codec, 'replace')))
 
-            # >>> Begin patch for ticket #4652 <<<
+            if kindle_pdoc and kindle_pdoc in mi.tags:
-            kindle_doc_types = set([u'[kindle_ebok]',u'[kindle_pdoc]'])
+                update_exth_record((501, str('PDOC')))
-            doc_type = list(kindle_doc_types.intersection(set(mi.tags)))[0]
-            if doc_type:
-                if doc_type == '[kindle_ebok]':
-                    update_exth_record((501,str('EBOK')))
-                elif doc_type == '[kindle_pdoc]':
-                    update_exth_record((501, str('PDOC')))
-            # >>> End patch
 
         if mi.pubdate:
             update_exth_record((106, str(mi.pubdate).encode(self.codec, 'replace')))

src/calibre/ebooks/metadata/topaz.py

@@ -83,7 +83,7 @@ class MetadataUpdater(object):
 
         sig = self.data[:4]
         if not sig.startswith('TPZ'):
-            raise ValueError('Not a Topaz file')
+            raise ValueError("'%s': Not a Topaz file" % getattr(stream, 'name', 'Unnamed stream'))
         offset = 4
 
         self.header_records, consumed = self.decode_vwi(self.data[offset:offset+4])
@@ -92,13 +92,13 @@ class MetadataUpdater(object):
 
         # First integrity test - metadata header
         if not 'metadata' in self.topaz_headers:
-            raise ValueError('Invalid Topaz format - no metadata record')
+            raise ValueError("'%s': Invalid Topaz format - no metadata record" % getattr(stream, 'name', 'Unnamed stream'))
 
         # Second integrity test - metadata body
         md_offset = self.topaz_headers['metadata']['blocks'][0]['offset']
         md_offset += self.base
         if self.data[md_offset+1:md_offset+9] != 'metadata':
-            raise ValueError('Damaged metadata record')
+            raise ValueError("'%s': Damaged metadata record" % getattr(stream, 'name', 'Unnamed stream'))
 
     def book_length(self):
         ''' convenience method for retrieving book length '''

src/calibre/ebooks/mobi/output.py

@@ -36,6 +36,9 @@ class MOBIOutput(OutputFormatPlugin):
             recommended_value=False, level=OptionRecommendation.LOW,
             help=_('Disable compression of the file contents.')
         ),
+        OptionRecommendation(name='personal_doc', recommended_value='[kindle_pdoc]',
+            help=_('Tag marking book to be filed with Personal Docs')
+        ),
     ])
 
     def check_for_periodical(self):

src/calibre/gui2/convert/mobi_output.py

@@ -24,7 +24,7 @@ class PluginWidget(Widget, Ui_Form):
     def __init__(self, parent, get_option, get_help, db=None, book_id=None):
         Widget.__init__(self, parent, 'mobi_output',
                 ['prefer_author_sort', 'rescale_images', 'toc_title',
-                'dont_compress', 'no_inline_toc', 'masthead_font']
+                'dont_compress', 'no_inline_toc', 'masthead_font','personal_doc']
                 )
         self.db, self.book_id = db, book_id
 

src/calibre/gui2/convert/mobi_output.ui

@@ -6,8 +6,8 @@
    <rect>
     <x>0</x>
     <y>0</y>
-    <width>421</width>
+    <width>521</width>
-    <height>300</height>
+    <height>331</height>
    </rect>
   </property>
   <property name="windowTitle">
@@ -64,13 +64,27 @@
       <item>
        <widget class="QLabel" name="label_2">
         <property name="text">
-         <string>Masthead font:</string>
+         <string>Periodical masthead font:</string>
         </property>
        </widget>
       </item>
       <item>
        <widget class="QComboBox" name="opt_masthead_font"/>
       </item>
+      <item>
+       <layout class="QHBoxLayout" name="horizontalLayout">
+        <item>
+         <widget class="QLabel" name="label_3">
+          <property name="text">
+           <string>Personal Doc tag:</string>
+          </property>
+         </widget>
+        </item>
+        <item>
+         <widget class="QLineEdit" name="opt_personal_doc"/>
+        </item>
+       </layout>
+      </item>
       <item>
        <spacer name="verticalSpacer">
         <property name="orientation">
@@ -79,7 +93,7 @@
         <property name="sizeHint" stdset="0">
          <size>
           <width>20</width>
-          <height>55</height>
+          <height>40</height>
          </size>
         </property>
        </spacer>