merge from trunk

recipes/attac_es.recipe

@@ -0,0 +1,21 @@
+from calibre.web.feeds.news import BasicNewsRecipe
+
+class AttacEspanaRecipe (BasicNewsRecipe):
+   __author__ = u'Marc Busqué'
+   __url__ = 'http://www.lamarciana.com'
+   __version__ = '1.0'
+   __license__   = 'GPL v3'
+   __copyright__ = u'2012, Marc Busqué <marc@lamarciana.com>'
+   title = u'attac.es'
+   description = u'La Asociación por la Tasación de las Transacciones Financieras y por la Ayuda a los Ciudadanos (ATTAC) es un movimiento internacional altermundialista que promueve el control democrático de los mercados financieros y las instituciones encargadas de su control mediante la reflexión política y la movilización social.'
+   url = 'http://www.attac.es'
+   language = 'es'
+   tags = 'contrainformación, información alternativa'
+   oldest_article = 7
+   remove_empty_feeds = True
+   no_stylesheets = True
+   cover_url = u'http://www.attac.es/wp-content/themes/attacweb/images/attaces.jpg'
+
+   feeds = [
+         (u'Attac', u'http://www.attac.es/feed'),
+         ]

recipes/bwmagazine.recipe

@@ -15,6 +15,7 @@ class BusinessWeek(BasicNewsRecipe):
     oldest_article        = 7
     max_articles_per_feed = 200
     no_stylesheets        = True
+    auto_cleanup = True
     encoding              = 'utf8'
     use_embedded_content  = False
     language              = 'en'
@@ -36,12 +37,12 @@ class BusinessWeek(BasicNewsRecipe):
                         , 'language'  : language
                         }
 
-    remove_tags       = [
+    #remove_tags       = [
-                           dict(attrs={'class':'inStory'})
+                           #dict(attrs={'class':'inStory'})
-                          ,dict(name=['meta','link','iframe','base','embed','object','table','th','tr','td'])
+                          #,dict(name=['meta','link','iframe','base','embed','object','table','th','tr','td'])
-                          ,dict(attrs={'id':['inset','videoDisplay']})
+                          #,dict(attrs={'id':['inset','videoDisplay']})
-                        ]
+                        #]
-    keep_only_tags    = [dict(name='div', attrs={'id':['story-body','storyBody']})]
+    #keep_only_tags    = [dict(name='div', attrs={'id':['story-body','storyBody']})]
     remove_attributes = ['lang']
     match_regexps     = [r'http://www.businessweek.com/.*_page_[1-9].*']
 
@@ -100,3 +101,4 @@ class BusinessWeek(BasicNewsRecipe):
                tstr = alink.string
                alink.replaceWith(tstr)
         return soup
+ 

recipes/metro_news_nl.recipe

@@ -34,6 +34,8 @@ from BeautifulSoup import BeautifulSoup
     Added new feeds
     Updated css
     Changed order of regex to speedup proces
+ Version 1.9.3 23-05-2012
+    Updated Cover image
 '''
 
 class AdvancedUserRecipe1306097511(BasicNewsRecipe):
@@ -51,7 +53,7 @@ class AdvancedUserRecipe1306097511(BasicNewsRecipe):
     no_stylesheets = True
     remove_javascript = True
     remove_empty_feeds = True
-    cover_url = 'http://www.oldreadmetro.com/img/en/metroholland/last/1/small.jpg'
+    cover_url = 'http://www.readmetro.com/en/holland/metro-holland/image/large/last/'
     publication_type = 'newspaper'
     encoding = 'utf-8'
     remove_attributes = ['style', 'font', 'width', 'height', 'itemtype', 'itemprop', 'itemscope']#, 'href']

session.vim

@@ -6,7 +6,7 @@ fun! CalibreLog()
     " making a release.
     enew
     read ! bzr log -l 500
-    set nomodifiable noswapfile buftype=nofile
+    setl nomodifiable noswapfile buftype=nofile
     edit Changelog.yaml
     edit src/calibre/constants.py
 endfun

src/calibre/ebooks/conversion/preprocess.py

@@ -626,7 +626,10 @@ class HTMLPreProcessor(object):
         if getattr(self.extra_opts, 'smarten_punctuation', False):
             html = self.smarten_punctuation(html)
 
-        unsupported_unicode_chars = self.extra_opts.output_profile.unsupported_unicode_chars
+        try:
+            unsupported_unicode_chars = self.extra_opts.output_profile.unsupported_unicode_chars
+        except AttributeError:
+            unsupported_unicode_chars = u''
         if unsupported_unicode_chars:
             from calibre.utils.localization import get_udc
             unihandecoder = get_udc()

src/calibre/ebooks/htmlz/oeb2html.py

@@ -120,13 +120,12 @@ class OEB2HTML(object):
                 el.attrib['id'] = self.get_link_id(page.href, el.attrib['id'])[1:]
 
     def get_css(self, oeb_book):
-        css = u''
+        css = b''
         for item in oeb_book.manifest:
             if item.media_type == 'text/css':
-                css = item.data.cssText
+                css += item.data.cssText + b'\n\n'
-                break
         return css
-    
+
     def prepare_string_for_html(self, raw):
         raw = prepare_string_for_xml(raw)
         raw = raw.replace(u'\u00ad', '­')

src/calibre/ebooks/mobi/debug/headers.py

@@ -377,6 +377,7 @@ class MOBIHeader(object): # {{{
 
         a('Compression: %s'%self.compression)
         a('Unused: %r'%self.unused)
+        a('Text length: %d'%self.text_length)
         a('Number of text records: %d'%self.number_of_text_records)
         a('Text record size: %d'%self.text_record_size)
         a('Encryption: %s'%self.encryption_type)

src/calibre/ebooks/mobi/writer2/main.py

@@ -411,6 +411,7 @@ class MobiWriter(object):
         header_fields['fdst_count'] = 1 # Why not 0? Kindlegen uses 1
         header_fields['flis_record'] = flis_number
         header_fields['fcis_record'] = fcis_number
+        header_fields['text_length'] = self.text_length
         extra_data_flags = 0b1 # Has multibyte overlap bytes
         if self.primary_index_record_idx is not None:
             extra_data_flags |= 0b10

src/calibre/gui2/preferences/tweaks.py

@@ -42,7 +42,9 @@ class Tweak(object): # {{{
     def __init__(self, name, doc, var_names, defaults, custom):
         translate = _
         self.name = translate(name)
-        self.doc = translate(doc.strip())
+        self.doc = doc.strip()
+        if self.doc:
+            self.doc = translate(self.doc)
         self.var_names = var_names
         self.default_values = {}
         for x in var_names:

src/calibre/utils/smartypants.py

@@ -597,15 +597,17 @@ def educateQuotes(str):
 	str = re.sub(r"""(?<=\W)'(?=\w)""", r"""&#8216;""", str)
 	str = re.sub(r"""(?<=\w)"(?=\W)""", r"""&#8221;""", str)
 	str = re.sub(r"""(?<=\w)'(?=\W)""", r"""&#8217;""", str)
-	
+
+    # The following are commented out as smartypants tokenizes text by
+    # stripping out html tags. Therefore, there is no guarantee that the
+    # start-of-line and end-ol-line regex operators will match anything
+    # meaningful
+
 	# Special case for Quotes at end of line with a preceeding space (may change just to end of line)
-	str = re.sub(r"""(?<=\s)"$""", r"""&#8221;""", str)
+	#str = re.sub(r"""(?<=\s)"$""", r"""&#8221;""", str)
-	str = re.sub(r"""(?<=\s)'$""", r"""&#8217;""", str)
+	#str = re.sub(r"""(?<=\s)'$""", r"""&#8217;""", str)
-	
+
 	# Special case for Quotes at beginning of line with a space - multiparagraph quoted text:
-	# This case commented out as it works under the assumption that the regex pattern will always
-	# evaluate a complete sentence - the tokenize function called earlier in smarty will break up
-	# the text based on tags, so sentence fragments can be passed to the patterns as well.
 	#str = re.sub(r"""^"(?=\s)""", r"""&#8220;""", str)
 	#str = re.sub(r"""^'(?=\s)""", r"""&#8216;""", str)
 

src/odf/odf2xhtml.py

@@ -854,9 +854,24 @@ ol, ul { padding-left: 2em; }
             else:
                 css_styles[css2] = [name]
 
+        def filter_margins(css2):
+            names = { k for k, v in css2 }
+            ignore = set()
+            if {'margin-left', 'margin-right', 'margin-top',
+                    'margin-bottom'}.issubset(names):
+                # These come from XML and we cannot preserve XML attribute
+                # order so we assume that margin is to be overridden See
+                # https://bugs.launchpad.net/calibre/+bug/941134 and
+                # https://bugs.launchpad.net/calibre/+bug/1002702
+                ignore.add('margin')
+            css2 = sorted(css2, key=lambda x:{'margin':0}.get(x[0], 1))
+            for k, v in css2:
+                if k not in ignore:
+                    yield k, v
+
         for css2, names in css_styles.iteritems():
             self.writeout("%s {\n" % ', '.join(names))
-            for style, val in css2:
+            for style, val in filter_margins(css2):
                 self.writeout("\t%s: %s;\n" % (style, val) )
             self.writeout("}\n")
 
@@ -941,20 +956,8 @@ ol, ul { padding-left: 2em; }
         if self.currentstyle is None: # Added by Kovid
             return
 
-        # Added by Kovid
-        names = {x[1]:x for x in attrs.iterkeys()}
-        ignore_keys = set()
-        if ('margin' in names and 'margin-top' in names and 'margin-left' in
-                names and 'margin-right' in names and 'margin-bottom' in
-                names):
-            # These come from XML and we cannot preserve XML attribute order so
-            # we assume that margin is to be overridden
-            # See https://bugs.launchpad.net/calibre/+bug/941134
-            ignore_keys.add(names['margin'])
-
         for key,attr in attrs.items():
-            if key not in ignore_keys:
+            self.styledict[self.currentstyle][key] = attr
-                self.styledict[self.currentstyle][key] = attr
 
 
     familymap = {'frame':'frame', 'paragraph':'p', 'presentation':'presentation',