Sync to trunk

src/calibre/debug.py

@@ -43,7 +43,11 @@ def update_module(mod, path):
         zp = os.path.join(os.path.dirname(sys.executable), 'library.zip')
     elif isosx:
         zp = os.path.join(os.path.dirname(getattr(sys, 'frameworks_dir')),
-                            'Resources', 'lib', 'python2.5', 'site-packages.zip')
+                            'Resources', 'lib', 
+                            'python'+'.'.join(map(str, sys.version_info[:2])), 
+                            'site-packages.zip')
+    else:
+        zp = os.path.join(getattr(sys, 'frozen_path'), 'loader.zip')
     if zp is not None:
         update_zipfile(zp, mod, path)
     else:

src/calibre/ebooks/html.py

@@ -335,7 +335,7 @@ class PreProcessor(object):
     # Fix pdftohtml markup
     PDFTOHTML  = [
                   # Remove <hr> tags
-                  (re.compile(r'<hr.*?>', re.IGNORECASE), lambda match: '<span style="page-break-after:always"> </span>'),
+                  (re.compile(r'<hr.*?>', re.IGNORECASE), lambda match: '<br />'),
                   # Remove page numbers
                   (re.compile(r'\d+<br>', re.IGNORECASE), lambda match: ''),
                   # Remove <br> and replace <br><br> with <p>

src/calibre/ebooks/lit/oeb.py

@@ -267,6 +267,14 @@ class Manifest(object):
                 return result
             return cmp(self.id, other.id)
         
+        def abshref(self, href):
+            if '/' not in self.href:
+                return href
+            dirname = os.path.dirname(self.href)
+            href = os.path.join(dirname, href)
+            href = os.path.normpath(href).replace('\\', '/')
+            return href
+    
     def __init__(self, oeb):
         self.oeb = oeb
         self.items = {}

src/calibre/ebooks/lit/writer.py

@@ -143,17 +143,16 @@ def warn(x):
 class ReBinary(object):
     NSRMAP = {'': None, XML_NS: 'xml'}
     
-    def __init__(self, root, path, oeb, map=HTML_MAP, logger=FauxLogger()):
+    def __init__(self, root, item, oeb, map=HTML_MAP, logger=FauxLogger()):
-        self.path = path
+        self.item = item
         self.logger = logger
-        self.dir = os.path.dirname(path)
         self.manifest = oeb.manifest
         self.tags, self.tattrs = map
         self.buf = StringIO()
         self.anchors = []
         self.page_breaks = []
         self.is_html  = is_html = map is HTML_MAP
-        self.stylizer = Stylizer(root, path, oeb) if is_html else None
+        self.stylizer = Stylizer(root, item.href, oeb) if is_html else None
         self.tree_to_binary(root)
         self.content = self.buf.getvalue()
         self.ahc = self.build_ahc() if is_html else None
@@ -210,6 +209,8 @@ class ReBinary(object):
             if attr in ('href', 'src'):
                 value = urlnormalize(value)
                 path, frag = urldefrag(value)
+                if self.item:
+                    path = self.item.abshref(path)
                 prefix = unichr(3)
                 if path in self.manifest.hrefs:
                     prefix = unichr(2)
@@ -222,7 +223,7 @@ class ReBinary(object):
             elif attr.startswith('ms--'):
                 attr = '%' + attr[4:]
             elif tag == 'link' and attr == 'type' and value in OEB_STYLES:
-                value = OEB_CSS_MIME
+                value = CSS_MIME
             if attr in tattrs:
                 self.write(tattrs[attr])
             else:
@@ -275,7 +276,7 @@ class ReBinary(object):
     def build_ahc(self):
         if len(self.anchors) > 6:
             self.logger.log_warn("More than six anchors in file %r. " \
-                "Some links may not work properly." % self.path)
+                "Some links may not work properly." % self.item.href)
         data = StringIO()
         data.write(unichr(len(self.anchors)).encode('utf-8'))
         for anchor, offset in self.anchors:
@@ -479,7 +480,7 @@ class LitWriter(object):
             secnum = 0
             if not isinstance(data, basestring):
                 self._add_folder(name)
-                rebin = ReBinary(data, item.href, self._oeb, map=HTML_MAP,
+                rebin = ReBinary(data, item, self._oeb, map=HTML_MAP,
                                  logger=self._logger)
                 self._add_file(name + '/ahc', rebin.ahc, 0)
                 self._add_file(name + '/aht', rebin.aht, 0)
@@ -559,7 +560,7 @@ class LitWriter(object):
         meta.attrib['ms--minimum_level'] = '0'
         meta.attrib['ms--attr5'] = '1'
         meta.attrib['ms--guid'] = '{%s}' % str(uuid.uuid4()).upper()
-        rebin = ReBinary(meta, 'content.opf', self._oeb, map=OPF_MAP,
+        rebin = ReBinary(meta, None, self._oeb, map=OPF_MAP,
                          logger=self._logger)
         meta = rebin.content
         self._meta = meta

src/calibre/ebooks/lrf/html/convert_from.py

@@ -109,6 +109,10 @@ class HTMLConverter(object, LoggingInterface):
                         # Remove self closing script tags as they also mess up BeautifulSoup
                         (re.compile(r'(?i)<script[^<>]+?/>'), lambda match: ''),
                         
+                        # BeautifulSoup treats self closing <div> tags as open <div> tags
+                        (re.compile(r'(?i)<\s*div([^>]*)/\s*>'), 
+                         lambda match: '<div%s></div>'%match.group(1))
+                        
                         ]
     # Fix Baen markup
     BAEN = [ 
@@ -122,7 +126,7 @@ class HTMLConverter(object, LoggingInterface):
     # Fix pdftohtml markup
     PDFTOHTML  = [
                   # Remove <hr> tags
-                  (re.compile(r'<hr.*?>', re.IGNORECASE), lambda match: '<span style="page-break-after:always"> </span>'),
+                  (re.compile(r'<hr.*?>', re.IGNORECASE), lambda match: '<br />'),
                   # Remove page numbers
                   (re.compile(r'\d+<br>', re.IGNORECASE), lambda match: ''),
                   # Remove <br> and replace <br><br> with <p>
@@ -576,20 +580,20 @@ class HTMLConverter(object, LoggingInterface):
         if (css.has_key('display') and css['display'].lower() == 'none') or \
            (css.has_key('visibility') and css['visibility'].lower() == 'hidden'):
             return ''
-        text = u''
+        text, alt_text = u'', u''
         for c in tag.contents:
             if limit != None and len(text) > limit:
                 break
             if isinstance(c, HTMLConverter.IGNORED_TAGS):
-                return u''
+                continue
             if isinstance(c, NavigableString):
                 text += unicode(c)                
             elif isinstance(c, Tag):
                 if c.name.lower() == 'img' and c.has_key('alt'):
-                    text += c['alt']
+                    alt_text += c['alt']
-                    return text
+                    continue
                 text += self.get_text(c)
-        return text
+        return text if text.strip() else alt_text
     
     def process_links(self):
         def add_toc_entry(text, target):

src/calibre/ebooks/lrf/objects.py

@@ -799,18 +799,39 @@ class Text(LRFStream):
         length = len(self.stream)
         style = self.style.as_dict()
         current_style = style.copy()
+        text_tags = set(list(TextAttr.tag_map.keys()) + \
+                        list(Text.text_tags.keys()) + \
+                        list(ruby_tags.keys()))
+        text_tags -= set([0xf500+i for i in range(10)])
+        text_tags.add(0xf5cc)
         
         while stream.tell() < length:
         
-            # Is there some text beofre a tag?
+            # Is there some text before a tag?
-            pos = self.stream.find('\xf5', stream.tell()) - 1
+            def find_first_tag(start):
-            if pos > 0:
+                pos = self.stream.find('\xf5', start)
-                self.add_text(self.stream[stream.tell():pos])
+                if pos == -1:
-                stream.seek(pos)
+                    return -1
-            elif pos == -2: # No tags in this stream
+                try:
+                    stream.seek(pos-1)
+                    _t = Tag(stream)
+                    if _t.id in text_tags:
+                        return pos-1
+                    return find_first_tag(pos+1)
+                    
+                    
+                except:
+                    return find_first_tag(pos+1)
+                    
+            start_pos = stream.tell()        
+            tag_pos = find_first_tag(start_pos)
+            if tag_pos >= start_pos:
+                if tag_pos > start_pos:
+                    self.add_text(self.stream[start_pos:tag_pos])
+                stream.seek(tag_pos)
+            else: # No tags in this stream
                 self.add_text(self.stream)
                 stream.seek(0, 2)
-                print repr(self.stream)
                 break
             
             tag = Tag(stream)
@@ -1166,7 +1187,8 @@ class TOCObject(LRFStream):
             refpage = struct.unpack("<I", stream.read(4))[0]
             refobj  = struct.unpack("<I", stream.read(4))[0]
             cnt = struct.unpack("<H", stream.read(2))[0]
-            label = unicode(stream.read(cnt), "utf_16")
+            raw = stream.read(cnt)
+            label = raw.decode('utf_16_le')
             self._contents.append(TocLabel(refpage, refobj, label))
             c -= 1
             

src/calibre/ebooks/mobi/reader.py

@@ -33,7 +33,6 @@ class EXTHHeader(object):
         self.length, self.num_items = struct.unpack('>LL', raw[4:12])
         raw = raw[12:]
         pos = 0
-        
         self.mi = MetaInformation('Unknown', ['Unknown'])
         self.has_fake_cover = True
         
@@ -49,9 +48,17 @@ class EXTHHeader(object):
                 self.cover_offset, = struct.unpack('>L', content)
             elif id == 202:
                 self.thumbnail_offset, = struct.unpack('>L', content)
+            #else:
+            #    print 'unknown record', id, repr(content)
         title = re.search(r'\0+([^\0]+)\0+', raw[pos:])
         if title:
-            self.mi.title = title.group(1).decode(codec, 'ignore')
+            title = title.group(1).decode(codec, 'replace')
+            if len(title) > 2:
+                self.mi.title = title
+            else:
+                title = re.search(r'\0+([^\0]+)\0+', ''.join(reversed(raw[pos:])))
+                if title:
+                    self.mi.title = ''.join(reversed(title.group(1).decode(codec, 'replace')))
             
                 
     def process_metadata(self, id, content, codec):
@@ -67,7 +74,8 @@ class EXTHHeader(object):
             if not self.mi.tags:
                 self.mi.tags = []
             self.mi.tags.append(content.decode(codec, 'ignore'))
-         
+        #else:
+        #    print 'unhandled metadata record', id, repr(content), codec 
             
 
 class BookHeader(object):
@@ -466,6 +474,10 @@ def get_metadata(stream):
             cover =  os.path.join(tdir, mi.cover)
             if os.access(cover, os.R_OK):
                 mi.cover_data = ('JPEG', open(os.path.join(tdir, mi.cover), 'rb').read())
+        else:
+            path = os.path.join(tdir, 'images', '00001.jpg')
+            if os.access(path, os.R_OK):
+                mi.cover_data = ('JPEG', open(path, 'rb').read())
     return mi
         
 def option_parser():

src/calibre/gui2/main.py

@@ -1482,7 +1482,8 @@ in which you want to store your books files. Any existing books will be automati
         return True
 
     
-    def shutdown(self):
+    def shutdown(self, write_settings=True):
+        if write_settings:
             self.write_settings()
         self.job_manager.terminate_all_jobs()
         self.device_manager.keep_going = False
@@ -1503,6 +1504,7 @@ in which you want to store your books files. Any existing books will be automati
 
     
     def closeEvent(self, e):
+        self.write_settings()
         if self.system_tray_icon.isVisible():
             if not dynamic['systray_msg'] and not isosx:
                 info_dialog(self, 'calibre', 'calibre '+_('will keep running in the system tray. To close it, choose <b>Quit</b> in the context menu of the system tray.')).exec_()
@@ -1512,7 +1514,7 @@ in which you want to store your books files. Any existing books will be automati
         else:
             if self.confirm_quit():
                 try:
-                    self.shutdown()
+                    self.shutdown(write_settings=False)
                 except:
                     pass
                 e.accept()

src/calibre/utils/zipfile.py

@@ -338,7 +338,7 @@ class ZipInfo (object):
         if isinstance(self.filename, unicode):
             try:
                 return self.filename.encode('ascii'), self.flag_bits
-            except UnicodeEncodeError:
+            except:
                 return self.filename.encode('utf-8'), self.flag_bits | 0x800
         else:
             return self.filename, self.flag_bits