KG 0.7.7 update

Changelog.yaml

@@ -4,6 +4,61 @@
 # for important features/bug fixes.
 # Also, each release can have new and improved recipes.
 
+- version: 0.7.7
+  date: 2010-07-02
+
+  new features:
+    - title: "Support for the Nokia E52"
+
+    - title: "Searching on the size column"
+
+    - title: "iTunes driver: Add option to disable cover fetching for speeding up the fetching of large book collections"
+
+  bug fixes:
+    - title: "SONY driver: Only update metadata when books are sent to device."
+
+    - title: "TXT Input: Ensure the generated html is splittable"
+      tickets: [5904]
+
+    - title: "Fix infinite loop in default cover generation."
+      tickets: [6061]
+
+    - title: "HTML Input: Fix a parsing bug that was triggered in rare conditions"
+      tickets: [6064]
+
+    - title: "HTML2Zip plugin: Do not replace ligatures"
+      tickets: [6019]
+
+    - title: "iTunes driver: Fix transmission of non integral series numbers"
+      tickets: [6046]
+
+    - title: "Simplify implementation of cover caching and ensure cover browser is updated when covers are changed"
+
+    - title: "PDF metadata: Fix last character corrupted when setting metadata in encrypted files."
+
+    - title: "PDF metadata: Update the version of PoDoFo used to set metadata to 0.8.1. Hopefully that means more PDF files will work"
+
+    - title: "Device drivers: Speedup for dumping metadata cache to devices on Windows XP"
+
+    - title: "EPUB Output: Ensure that language setting is conformant to the specs"
+
+    - title: "MOBI Output: Fix a memory leak and a crash in the palmdoc compression routine"
+
+    - title: "Metadata download: Fix a regressiont at resulted in a failed download for some books"
+
+  new recipes:
+    - title: "Foreign Policy and Alo!"
+      author: Darko Miletic
+
+    - title: Statesman and ifzm
+      author: rty
+
+  improved recipes:
+    - Akter
+    - The Old New Thing
+
+
+
 - version: 0.7.6
   date: 2010-06-28
 

resources/recipes/china_press.recipe

@@ -7,7 +7,7 @@ class AdvancedUserRecipe1277228948(BasicNewsRecipe):
 
     __author__            = 'rty'
     __version__            = '1.0'
-    language = 'zh_CN'
+    language = 'zh'
     pubisher  = 'www.chinapressusa.com'
     description           = 'Overseas Chinese Network Newspaper in the USA'
     category              = 'News in Chinese, USA'

resources/recipes/foreign_policy.recipe

@@ -0,0 +1,45 @@
+__license__   = 'GPL v3'
+__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>'
+'''
+www.foreignpolicy.com
+'''
+
+from calibre.web.feeds.news import BasicNewsRecipe
+
+class ForeignPolicy(BasicNewsRecipe):
+    title                 = 'Foreign Policy'
+    __author__            = 'Darko Miletic'
+    description           = 'International News'
+    publisher             = 'Washingtonpost.Newsweek Interactive, LLC'
+    category              = 'news, politics, USA'
+    oldest_article        = 31
+    max_articles_per_feed = 200
+    no_stylesheets        = True
+    encoding              = 'utf8'
+    use_embedded_content  = False
+    language              = 'en'
+    remove_empty_feeds    = True
+    extra_css             = ' body{font-family: Georgia,"Times New Roman",Times,serif } img{margin-bottom: 0.4em} h1,h2,h3,h4,h5,h6{font-family: Arial,Helvetica,sans-serif} '
+
+    conversion_options = {
+                          'comment'   : description
+                        , 'tags'      : category
+                        , 'publisher' : publisher
+                        , 'language'  : language
+                        }
+
+    keep_only_tags    = [dict(attrs={'id':['art-mast','art-body','auth-bio']})]
+    remove_tags       = [dict(name='iframe'),dict(attrs={'id':['share-box','base-ad']})]
+    remove_attributes = ['height','width']
+
+
+    feeds = [(u'Articles', u'http://www.foreignpolicy.com/node/feed')]
+
+    def print_version(self, url):
+        return url + '?print=yes&page=full'
+
+    def preprocess_html(self, soup):
+        for item in soup.findAll(style=True):
+            del item['style']
+        return soup
+

resources/recipes/ifzm.recipe

@@ -0,0 +1,50 @@
+from calibre.web.feeds.news import BasicNewsRecipe
+
+class AdvancedUserRecipe1277305250(BasicNewsRecipe):
+    title          = u'infzm - China Southern Weekly'
+    oldest_article = 14
+    max_articles_per_feed = 100
+
+    feeds          = [(u'\u5357\u65b9\u5468\u672b-\u70ed\u70b9\u65b0\u95fb', u'http://www.infzm.com/rss/home/rss2.0.xml'),
+	(u'\u5357\u65b9\u5468\u672b-\u7ecf\u6d4e\u65b0\u95fb', u'http://www.infzm.com/rss/economic.xml'),
+	(u'\u5357\u65b9\u5468\u672b-\u6587\u5316\u65b0\u95fb', u'http://www.infzm.com/rss/culture.xml'),
+	(u'\u5357\u65b9\u5468\u672b-\u751f\u6d3b\u65f6\u5c1a', u'http://www.infzm.com/rss/lifestyle.xml'),
+	(u'\u5357\u65b9\u5468\u672b-\u89c2\u70b9', u'http://www.infzm.com/rss/opinion.xml')
+	]
+    __author__            = 'rty'
+    __version__            = '1.0'
+    language = 'zh'
+    pubisher  = 'http://www.infzm.com'
+    description           = 'Chinese Weekly Tabloid'
+    category              = 'News, China'
+    remove_javascript = True
+    use_embedded_content   = False
+    no_stylesheets = True
+    #encoding               = 'GB2312'
+    encoding               = 'UTF-8'
+    conversion_options = {'linearize_tables':True}
+    masthead_url = 'http://i50.tinypic.com/2qmfb7l.jpg'
+
+    extra_css = '''
+             @font-face { font-family: "DroidFont", serif, sans-serif;  src: url(res:///system/fonts/DroidSansFallback.ttf); }\n
+             body {
+                  margin-right: 8pt;
+                  font-family: 'DroidFont', serif;}
+             .detailContent {font-family: 'DroidFont', serif, sans-serif}
+            '''
+
+    keep_only_tags = [
+                              dict(name='div', attrs={'id':'detailContent'}),
+                               ]
+    remove_tags = [
+                    dict(name='div', attrs={'id':['detailTools', 'detailSideL', 'pageNum']}),
+                         ]
+    remove_tags_after = [
+                  dict(name='div', attrs={'id':'pageNum'}),
+                         ]
+    def preprocess_html(self, soup):
+        for item in soup.findAll(color=True):
+            del item['font']
+        for item in soup.findAll(style=True):
+            del item['style']
+        return soup

resources/recipes/statesman.recipe

@@ -0,0 +1,35 @@
+from calibre.web.feeds.news import BasicNewsRecipe
+
+class AdvancedUserRecipe1278049615(BasicNewsRecipe):
+    title          = u'Statesman'
+    pubisher  = 'http://www.statesman.com/'
+    description           = 'Austin Texas Daily Newspaper'
+    category              = 'News, Austin, Texas'
+    __author__            = 'rty'
+    oldest_article = 3
+
+    max_articles_per_feed = 100
+
+    feeds          = [(u'News', u'http://www.statesman.com/section-rss.do?source=news&includeSubSections=true'),
+	(u'Business', u'http://www.statesman.com/section-rss.do?source=business&includeSubSections=true'),
+	(u'Life', u'http://www.statesman.com/section-rss.do?source=life&includesubsection=true'),
+	(u'Editorial', u'http://www.statesman.com/section-rss.do?source=opinion&includesubsections=true'),
+	(u'Sports', u'http://www.statesman.com/section-rss.do?source=sports&includeSubSections=true')
+	]
+    masthead_url = "http://www.statesman.com/images/cmg-logo.gif"
+    #temp_files = []
+    #articles_are_obfuscated = True
+
+    remove_javascript = True
+    use_embedded_content   = False
+    no_stylesheets = True
+    language = 'en'
+    encoding               = 'utf-8'
+    conversion_options = {'linearize_tables':True}
+    remove_tags = [
+                    dict(name='div', attrs={'id':'cxArticleOptions'}),
+                        ]
+    keep_only_tags = [
+	 dict(name='div', attrs={'class':'cxArticleHeader'}),
+                     dict(name='div', attrs={'id':'cxArticleBodyText'}),
+                               ]

src/calibre/constants.py

@@ -2,7 +2,7 @@ __license__   = 'GPL v3'
 __copyright__ = '2008, Kovid Goyal kovid@kovidgoyal.net'
 __docformat__ = 'restructuredtext en'
 __appname__   = 'calibre'
-__version__   = '0.7.6'
+__version__   = '0.7.7'
 __author__    = "Kovid Goyal <kovid@kovidgoyal.net>"
 
 import re

src/calibre/customize/builtins.py

@@ -30,6 +30,7 @@ every time you add an HTML file to the library.\
 
         with TemporaryDirectory('_plugin_html2zip') as tdir:
             recs =[('debug_pipeline', tdir, OptionRecommendation.HIGH)]
+            recs.append(['keep_ligatures', True, OptionRecommendation.HIGH])
             if self.site_customization and self.site_customization.strip():
                 recs.append(['input_encoding', self.site_customization.strip(),
                     OptionRecommendation.HIGH])
@@ -81,7 +82,7 @@ class PML2PMLZ(FileTypePlugin):
 
         return of.name
 
-
+# Metadata reader plugins {{{
 class ComicMetadataReader(MetadataReaderPlugin):
 
     name = 'Read comic metadata'
@@ -319,7 +320,9 @@ class ZipMetadataReader(MetadataReaderPlugin):
     def get_metadata(self, stream, ftype):
         from calibre.ebooks.metadata.zip import get_metadata
         return get_metadata(stream)
+# }}}
 
+# Metadata writer plugins {{{
 
 class EPUBMetadataWriter(MetadataWriterPlugin):
 
@@ -395,6 +398,7 @@ class TOPAZMetadataWriter(MetadataWriterPlugin):
         from calibre.ebooks.metadata.topaz import set_metadata
         set_metadata(stream, mi)
 
+# }}}
 
 from calibre.ebooks.comic.input import ComicInput
 from calibre.ebooks.epub.input import EPUBInput
@@ -444,7 +448,7 @@ from calibre.devices.kindle.driver import KINDLE, KINDLE2, KINDLE_DX
 from calibre.devices.nook.driver import NOOK
 from calibre.devices.prs505.driver import PRS505
 from calibre.devices.android.driver import ANDROID, S60
-from calibre.devices.nokia.driver import N770, N810, E71X
+from calibre.devices.nokia.driver import N770, N810, E71X, E52
 from calibre.devices.eslick.driver import ESLICK, EBK52
 from calibre.devices.nuut2.driver import NUUT2
 from calibre.devices.iriver.driver import IRIVER_STORY
@@ -519,6 +523,7 @@ plugins += [
     S60,
     N770,
     E71X,
+    E52,
     N810,
     COOL_ER,
     ESLICK,

src/calibre/devices/apple/driver.py

@@ -20,7 +20,7 @@ from calibre.utils.config import config_dir
 from calibre.utils.date import isoformat, now, parse_date
 from calibre.utils.localization import get_lang
 from calibre.utils.logging import Log
-from calibre.utils.zipfile import ZipFile, safe_replace
+from calibre.utils.zipfile import ZipFile
 
 from PIL import Image as PILImage
 
@@ -1727,7 +1727,6 @@ class ITUNES(DriverBase):
             return thumb_data
 
         thumb_path = book_path.rpartition('.')[0] + '.jpg'
-        format = book_path.rpartition('.')[2].lower()
         if isosx:
             title = book.name()
         elif iswindows:

src/calibre/devices/nokia/driver.py

@@ -67,3 +67,24 @@ class E71X(USBMS):
     VENDOR_NAME      = 'NOKIA'
     WINDOWS_MAIN_MEM = 'S60'
 
+class E52(USBMS):
+
+    name = 'Nokia E52 device interface'
+    gui_name = 'Nokia E52'
+    description = _('Communicate with the Nokia E52')
+    author = 'David Ignjic'
+    supported_platforms = ['windows', 'linux', 'osx']
+
+    VENDOR_ID = [0x421]
+    PRODUCT_ID = [0x1CD]
+    BCD = [0x100]
+
+
+    FORMATS = ['mobi', 'prc']
+
+    EBOOK_DIR_MAIN = 'eBooks'
+    SUPPORTS_SUB_DIRS = True
+
+    VENDOR_NAME = 'NOKIA'
+    WINDOWS_MAIN_MEM = 'S60'
+

src/calibre/devices/prs505/sony_cache.py

@@ -350,9 +350,7 @@ class XMLCache(object):
                 record = lpath_map.get(book.lpath, None)
                 if record is None:
                     record = self.create_text_record(root, i, book.lpath)
-                date = self.check_timestamp(record, book, path)
+                self.update_text_record(record, book, path, i)
-                if date is not None:
-                    self.update_text_record(record, book, date, path, i)
                 # Ensure the collections in the XML database are recorded for
                 # this book
                 if book.device_collections is None:
@@ -390,23 +388,35 @@ class XMLCache(object):
                 debug_print('WARNING: Some elements in the JSON cache were not'
                         ' found in the XML cache')
             records = [x for x in records if x is not None]
-            ids = set()
+            # Ensure each book has an ID.
             for rec in records:
-                id = rec.get('id', None)
+                if rec.get('id', None) is None:
-                if id is None:
                     rec.set('id', str(self.max_id(root)+1))
-                    id = rec.get('id', None)
+            ids = [x.get('id', None) for x in records]
-                ids.add(id)
+            # Given that we set the ids, there shouldn't be any None's. But
-            # ids cannot contain None, so no reason to check
+            # better to be safe...
+            if None in ids:
+                debug_print('WARNING: Some <text> elements do not have ids')
+                ids = [x for x in ids if x is not None]
 
             playlist = self.get_or_create_playlist(bl_index, category)
-            # Reduce ids to books not already in the playlist
+            # Get the books currently in the playlist. We will need them to be
+            # sure to put back any books that were manually added.
+            playlist_ids = []
             for item in playlist:
                 id_ = item.get('id', None)
                 if id_ is not None:
-                    ids.discard(id_)
+                    playlist_ids.append(id_)
-            # Add the books in ids that were not already in the playlist
+            # Empty the playlist. We do this so that the playlist will have the
-            for id_ in ids:
+            # order specified by get_collections
+            for item in list(playlist):
+                playlist.remove(item)
+
+            # Get a list of ids not known by get_collections
+            extra_ids = [x for x in playlist_ids if x not in ids]
+            # Rebuild the collection in the order specified by get_collections. Then
+            # add the ids that get_collections didn't know about.
+            for id_ in ids + extra_ids:
                 item = playlist.makeelement(
                         '{%s}item'%self.namespaces[bl_index],
                         nsmap=playlist.nsmap, attrib={'id':id_})
@@ -443,23 +453,15 @@ class XMLCache(object):
         root.append(ans)
         return ans
 
-    def check_timestamp(self, record, book, path):
+    def update_text_record(self, record, book, path, bl_index):
-        '''
-        Checks the timestamp in the Sony DB against the file. If different,
-        return the file timestamp. Otherwise return None.
-        '''
-        timestamp = os.path.getmtime(path)
-        date = strftime(timestamp)
-        if date != record.get('date', None):
-            return date
-        return None
-
-    def update_text_record(self, record, book, date, path, bl_index):
         '''
         Update the Sony database from the book. This is done if the timestamp in
         the db differs from the timestamp on the file.
         '''
-        record.set('date', date)
+        timestamp = os.path.getmtime(path)
+        date = strftime(timestamp)
+        if date != record.get('date', None):
+            record.set('date', date)
         record.set('size', str(os.stat(path).st_size))
         title = book.title if book.title else _('Unknown')
         record.set('title', title)

src/calibre/devices/usbms/books.py

@@ -174,8 +174,8 @@ class CollectionsBookList(BookList):
                     if lpath not in collections_lpaths[category]:
                         collections_lpaths[category].add(lpath)
                         collections[category].append(book)
-                        if attr == 'series':
+                    if attr == 'series':
-                            series_categories.add(category)
+                        series_categories.add(category)
         # Sort collections
         for category, books in collections.items():
             def tgetter(x):

src/calibre/ebooks/metadata/covers.py

@@ -0,0 +1,15 @@
+#!/usr/bin/env python
+# vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
+
+__license__   = 'GPL v3'
+__copyright__ = '2010, Kovid Goyal <kovid@kovidgoyal.net>'
+__docformat__ = 'restructuredtext en'
+
+
+from calibre.customize import Plugin
+
+class CoverDownload(Plugin):
+
+    supported_platforms = ['windows', 'osx', 'linux']
+    author = 'Kovid Goyal'
+    type = _('Cover download')

src/calibre/ebooks/metadata/douban.py

@@ -15,7 +15,6 @@ from calibre.utils.config import OptionParser
 from calibre.ebooks.metadata.fetch import MetadataSource
 from calibre.utils.date import parse_date, utcnow
 
-DOUBAN_API_KEY = None
 NAMESPACES = {
               'openSearch':'http://a9.com/-/spec/opensearchrss/1.0/',
               'atom' : 'http://www.w3.org/2005/Atom',
@@ -35,13 +34,15 @@ date           = XPath("descendant::db:attribute[@name='pubdate']")
 creator        = XPath("descendant::db:attribute[@name='author']")
 tag            = XPath("descendant::db:tag")
 
+CALIBRE_DOUBAN_API_KEY = '0bd1672394eb1ebf2374356abec15c3d'
+
 class DoubanBooks(MetadataSource):
 
     name = 'Douban Books'
     description = _('Downloads metadata from Douban.com')
     supported_platforms = ['windows', 'osx', 'linux'] # Platforms this plugin will run on
     author              = 'Li Fanxi <lifanxi@freemindworld.com>' # The author of this plugin
-    version             = (1, 0, 0)   # The version number of this plugin
+    version             = (1, 0, 1)   # The version number of this plugin
 
     def fetch(self):
         try:
@@ -65,7 +66,7 @@ class Query(object):
     type = "search"
 
     def __init__(self, title=None, author=None, publisher=None, isbn=None,
-                 max_results=20, start_index=1):
+                 max_results=20, start_index=1, api_key=''):
         assert not(title is None and author is None and publisher is None and \
                    isbn is None)
         assert (int(max_results) < 21)
@@ -89,16 +90,16 @@ class Query(object):
 
         if self.type == "isbn":
             self.url = self.ISBN_URL + q
-            if DOUBAN_API_KEY is not None:
+            if api_key != '':
-                self.url = self.url + "?apikey=" + DOUBAN_API_KEY
+                self.url = self.url + "?apikey=" + api_key
         else:
             self.url = self.SEARCH_URL+urlencode({
                     'q':q,
                     'max-results':max_results,
                     'start-index':start_index,
                     })
-            if DOUBAN_API_KEY is not None:
+            if api_key != '':
-                self.url = self.url + "&apikey=" + DOUBAN_API_KEY
+                self.url = self.url + "&apikey=" + api_key
 
     def __call__(self, browser, verbose):
         if verbose:
@@ -177,7 +178,7 @@ class ResultList(list):
             d = None
         return d
 
-    def populate(self, entries, browser, verbose=False):
+    def populate(self, entries, browser, verbose=False, api_key=''):
         for x in entries:
             try:
                 id_url = entry_id(x)[0].text
@@ -186,8 +187,8 @@ class ResultList(list):
                 report(verbose)
             mi = MetaInformation(title, self.get_authors(x))
             try:
-                if DOUBAN_API_KEY is not None:
+                if api_key != '':
-                    id_url = id_url + "?apikey=" + DOUBAN_API_KEY
+                    id_url = id_url + "?apikey=" + api_key
                 raw = browser.open(id_url).read()
                 feed = etree.fromstring(raw)
                 x = entry(feed)[0]
@@ -203,12 +204,16 @@ class ResultList(list):
             self.append(mi)
 
 def search(title=None, author=None, publisher=None, isbn=None,
-           verbose=False, max_results=40):
+           verbose=False, max_results=40, api_key=None):
     br   = browser()
     start, entries = 1, []
+
+    if api_key is None:
+        api_key = CALIBRE_DOUBAN_API_KEY
+
     while start > 0 and len(entries) <= max_results:
-        new, start = Query(title=title, author=author, publisher=publisher,
+        new, start = Query(title=title, author=author, publisher=publisher, 
-                       isbn=isbn, max_results=max_results, start_index=start)(br, verbose)
+                       isbn=isbn, max_results=max_results, start_index=start, api_key=api_key)(br, verbose)
         if not new:
             break
         entries.extend(new)
@@ -216,7 +221,7 @@ def search(title=None, author=None, publisher=None, isbn=None,
     entries = entries[:max_results]
 
     ans = ResultList()
-    ans.populate(entries, br, verbose)
+    ans.populate(entries, br, verbose, api_key)
     return ans
 
 def option_parser():

src/calibre/ebooks/oeb/base.py

@@ -808,7 +808,8 @@ class Manifest(object):
                         pat = re.compile(r'&(%s);'%('|'.join(user_entities.keys())))
                         data = pat.sub(lambda m:user_entities[m.group(1)], data)
 
-            parser = etree.XMLParser(no_network=True, huge_tree=True)
+            # Setting huge_tree=True causes crashes in windows with large files
+            parser = etree.XMLParser(no_network=True)
             # Try with more & more drastic measures to parse
             def first_pass(data):
                 try:
@@ -844,7 +845,7 @@ class Manifest(object):
                 nroot = etree.fromstring('<html></html>')
                 has_body = False
                 for child in list(data):
-                    if barename(child.tag) == 'body':
+                    if isinstance(child.tag, (unicode, str)) and barename(child.tag) == 'body':
                         has_body = True
                         break
                 parent = nroot

src/calibre/ebooks/oeb/stylizer.py

@@ -25,10 +25,17 @@ from calibre.ebooks.oeb.base import XHTML, XHTML_NS, CSS_MIME, OEB_STYLES
 from calibre.ebooks.oeb.base import XPNSMAP, xpath, urlnormalize
 from calibre.ebooks.oeb.profile import PROFILES
 
-html_css = open(P('templates/html.css'), 'rb').read()
+_html_css_stylesheet = None
+
+def html_css_stylesheet():
+    global _html_css_stylesheet
+    if _html_css_stylesheet is None:
+        html_css = open(P('templates/html.css'), 'rb').read()
+        _html_css_stylesheet = cssutils.parseString(html_css)
+        _html_css_stylesheet.namespaces['h'] = XHTML_NS
+    return _html_css_stylesheet
+
 XHTML_CSS_NAMESPACE = '@namespace "%s";\n' % XHTML_NS
-HTML_CSS_STYLESHEET = cssutils.parseString(html_css)
-HTML_CSS_STYLESHEET.namespaces['h'] = XHTML_NS
 
 INHERITED = set(['azimuth', 'border-collapse', 'border-spacing',
                  'caption-side', 'color', 'cursor', 'direction', 'elevation',
@@ -120,7 +127,7 @@ class Stylizer(object):
         item = oeb.manifest.hrefs[path]
         basename = os.path.basename(path)
         cssname = os.path.splitext(basename)[0] + '.css'
-        stylesheets = [HTML_CSS_STYLESHEET]
+        stylesheets = [html_css_stylesheet()]
         head = xpath(tree, '/h:html/h:head')
         if head:
             head = head[0]

src/calibre/ebooks/txt/input.py

@@ -63,7 +63,8 @@ class TXTInput(InputFormatPlugin):
                 raise ValueError('This txt file has malformed markup, it cannot be'
                     ' converted by calibre. See http://daringfireball.net/projects/markdown/syntax')
         else:
-            html = convert_basic(txt)
+            flow_size = getattr(options, 'flow_size', 0)
+            html = convert_basic(txt, epub_split_size_kb=flow_size)
 
         from calibre.customize.ui import plugin_for_input_format
         html_input = plugin_for_input_format('html')

src/calibre/ebooks/txt/processor.py

@@ -17,14 +17,11 @@ __docformat__ = 'restructuredtext en'
 
 HTML_TEMPLATE = u'<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"/><title>%s</title></head><body>\n%s\n</body></html>'
 
-def convert_basic(txt, title=''):
+def convert_basic(txt, title='', epub_split_size_kb=0):
-    lines = []
     # Strip whitespace from the beginning and end of the line. Also replace
     # all line breaks with \n.
-    for line in txt.splitlines():
+    txt = '\n'.join([line.strip() for line in txt.splitlines()])
-        lines.append(line.strip())
+
-    txt = '\n'.join(lines)
-    
     # Condense redundant spaces
     txt = re.sub('[ ]{2,}', ' ', txt)
 
@@ -34,6 +31,15 @@ def convert_basic(txt, title=''):
     # Remove excessive line breaks.
     txt = re.sub('\n{3,}', '\n\n', txt)
 
+    #Takes care if there is no point to split
+    if epub_split_size_kb > 0:
+        length_byte = len(txt.encode('utf-8'))
+        #Calculating the average chunk value for easy splitting as EPUB (+2 as a safe margin)
+        chunk_size = long(length_byte / (int(length_byte / (epub_split_size_kb * 1024) ) + 2 ))
+        #if there are chunks with a superior size then go and break
+        if (len(filter(lambda x: len(x.encode('utf-8')) > chunk_size, txt.split('\n\n')))) :
+            txt = u'\n\n'.join([split_string_separator(line, chunk_size) for line in txt.split('\n\n')])
+
     lines = []
     # Split into paragraphs based on having a blank line between text.
     for line in txt.split('\n\n'):
@@ -71,3 +77,10 @@ def opf_writer(path, opf_name, manifest, spine, mi):
     with open(os.path.join(path, opf_name), 'wb') as opffile:
         opf.render(opffile)
 
+def split_string_separator(txt, size) :
+    if len(txt.encode('utf-8')) > size:
+        txt = u''.join([re.sub(u'\.(?P<ends>[^.]*)$', u'.\n\n\g<ends>',
+            txt[i:i+size], 1) for i in
+            xrange(0, len(txt.encode('utf-8')), size)])
+    return txt
+

src/calibre/gui2/pictureflow/pictureflow.cpp

@@ -84,6 +84,7 @@ typedef unsigned short QRgb565;
 #define REFLECTION_FACTOR 1.5
 
 #define MAX(x, y) ((x > y) ? x : y)
+#define MIN(x, y) ((x < y) ? x : y)
 
 #define RGB565_RED_MASK 0xF800
 #define RGB565_GREEN_MASK 0x07E0
@@ -800,7 +801,7 @@ QRect PictureFlowPrivate::renderCenterSlide(const SlideInfo &slide) {
   QRect rect(buffer.width()/2 - sw/2, 0, sw, h-1);
   int left = rect.left();
 
-  for(int x = 0; x < sh-1; x++)
+  for(int x = 0; x < MIN(h-1, sh-1); x++)
     for(int y = 0; y < sw; y++)
       buffer.setPixel(left + y, 1+x, src->pixel(x, y));
 

src/calibre/manual/faq.rst

@@ -162,6 +162,9 @@ turned into a collection on the reader. Note that the PRS-500 does not support c
 How do I use |app| with my iPad/iPhone/iTouch?
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
+Over the air
+^^^^^^^^^^^^^^
+
 The easiest way to browse your |app| collection on your Apple device (iPad/iPhone/iPod) is by using the *free* Stanza app, available from the Apple app store. You need at least Stanza version 3.0. Stanza allows you to access your |app| collection wirelessly, over the air.
 
 First perform the following steps in |app|
@@ -181,13 +184,13 @@ Replace ``192.168.1.2`` with the local IP address of the computer running |app|.
 
 If you get timeout errors while browsing the calibre catalog in Stanza, try increasing the connection timeout value in the stanza settings. Go to Info->Settings and increase the value of Download Timeout.
 
-Alternative for the iPad
+With the USB cable
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
-As of |app| version 0.7.0, you can plugin your iPad into the computer using its charging cable, and |app| will detect it and show you a list of books on the iPad. You can then use the Send to device button to send books directly to iBooks on the iPad.
+As of |app| version 0.7.0, you can plug your iDevice into the computer using its charging cable, and |app| will detect it and show you a list of books on the device. You can then use the *Send to device button* to send books directly to iBooks on the device. Note that you must have at least iOS 4 installed on your iPhone/iTouch for this to work.
 
-This method only works on Windows XP and higher and OS X 10.5 and higher. Linux is not supported (iTunes is not available in linux) and OS X 10.4 is not supported. For more details, see
+This method only works on Windows XP and higher and OS X 10.5 and higher. Linux is not supported (iTunes is not available in linux) and OS X 10.4 is not supported.
-`this forum post <http://www.mobileread.com/forums/showpost.php?p=944079&postcount=1>`_.
+For more details on how this works, see `this forum post <http://www.mobileread.com/forums/showpost.php?p=944079&postcount=1>`_.
 
 How do I use |app| with my Android phone?
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

src/calibre/utils/magick_draw.py

@@ -65,6 +65,9 @@ class TextLine(object):
         self.bottom_margin = bottom_margin
         self.font_path = font_path
 
+    def __repr__(self):
+        return u'TextLine:%r:%f'%(self.text, self.font_size)
+
 def alloc_wand(name):
     ans = getattr(p, name)()
     if ans < 0:
@@ -120,6 +123,10 @@ def draw_centered_text(img, dw, text, top, margin=10):
     tokens = text.split(' ')
     while tokens:
         line, tokens = _get_line(img, dw, tokens, img_width-2*margin)
+        if not line:
+            # Could not fit the first token on the line
+            line = tokens[:1]
+            tokens = tokens[1:]
         bottom = draw_centered_line(img, dw, ' '.join(line), top)
         top = bottom
     return top

src/calibre/web/feeds/templates.py

@@ -8,7 +8,7 @@ import copy
 
 from lxml import html, etree
 from lxml.html.builder import HTML, HEAD, TITLE, STYLE, DIV, BODY, \
-        STRONG, EM, BR, SPAN, A, HR, UL, LI, H2, H3, IMG, P as PT, \
+        STRONG, BR, SPAN, A, HR, UL, LI, H2, H3, IMG, P as PT, \
         TABLE, TD, TR
 
 from calibre import preferred_encoding, strftime, isbytestring