Fix bug #4411: Include chapter headins when generating toc from pml files.

src/calibre/ebooks/pdb/ereader/writer.py

@@ -42,7 +42,9 @@ class Writer(FormatWriter):
         pml = unicode(pmlmlizer.extract_content(oeb_book, self.opts)).encode('cp1252', 'replace')
 
         text, text_sizes = self._text(pml)
-        chapter_index = self._index_item(r'(?s)\\C(?P<val>\d)="(?P<text>.+?)"', pml)
+        chapter_index = self._index_item(r'(?s)\\C(?P<val>[0-4)="(?P<text>.+?)"', pml)
+        chapter_index += self.index_item(r'(?s)\\X(?P<val>[0-4])(?P<text>.+?)\\X[0-4]', pml)
+        chapter_index += self.index_item(r'(?s)\\x(?P<text>.+?)\\x', pml)
         link_index = self._index_item(r'(?s)\\Q="(?P<text>.+?)"', pml)
         images = self._images(oeb_book.manifest, pmlmlizer.image_hrefs)
         metadata = [self._metadata(metadata)]

src/calibre/ebooks/pml/pmlconverter.py

@@ -171,6 +171,9 @@ class PML_HTMLizer(object):
         # &. It will display as &
         pml = pml.replace('&', '&')
 
+        pml = re.sub(r'(?<=\\x)(?P<text>.*?)(?=\\x)', lambda match: '="%s"%s' % (self.strip_pml(match.group('text')), match.group('text')), pml)
+        pml = re.sub(r'(?<=\\X[0-4])(?P<text>.*?)(?=\\X[0-4])', lambda match: '="%s"%s' % (self.strip_pml(match.group('text')), match.group('text')), pml)
+
         pml = re.sub(r'\\a(?P<num>\d{3})', lambda match: '&#%s;' % match.group('num'), pml)
         pml = re.sub(r'\\U(?P<num>[0-9a-f]{4})', lambda match: '%s' % my_unichr(int(match.group('num'), 16)), pml)
 
@@ -178,6 +181,19 @@ class PML_HTMLizer(object):
 
         return pml
 
+    def strip_pml(self, pml):
+        pml = re.sub(r'\\.\d=""', '', pml)
+        pml = re.sub(r'\\.=""', '', pml)
+        pml = re.sub(r'\\.\d', '', pml)
+        pml = re.sub(r'\\.', '', pml)
+        pml = re.sub(r'\\a\d\d\d', '', pml)
+        pml = re.sub(r'\\U\d\d\d\d', '', pml)
+        pml.replace('\r\n', ' ')
+        pml.replace('\n', ' ')
+        pml.replace('\r', ' ')
+
+        return pml
+
     def cleanup_html(self, html):
         old = html
         html = self.cleanup_html_remove_redundant(html)
@@ -503,9 +519,9 @@ class PML_HTMLizer(object):
                 if c == '\\':
                     c = line.read(1)
 
-                    if c in 'xqcrtTiIuobBlk':
+                    if c in 'qcrtTiIuobBlk':
                         text = self.process_code(c, line)
-                    elif c in 'FSX':
+                    elif c in 'FS':
                         l = line.read(1)
                         if '%s%s' % (c, l) == 'Fn':
                             text = self.process_code('Fn', line, 'fn')
@@ -515,8 +531,24 @@ class PML_HTMLizer(object):
                             text = self.process_code('SB', line)
                         elif '%s%s' % (c, l) == 'Sd':
                             text = self.process_code('Sd', line, 'sb')
+                    elif c in 'xXC':
+                        # The PML was modified eariler so x and X put the text
+                        # inside of ="" so we don't have do special processing
+                        # for C.
+                        t = ''
+                        if c in 'XC':
+                            level = line.read(1)
+                        id = 'pml_toc-%s' % len(self.toc)
+                        value = self.code_value(line)
+                        if c == 'x':
+                            t = self.process_code(c, line)
+                        elif c == 'X':
+                            t = self.process_code('%s%s' % (c, level), line)
+                        if not value or value == '':
+                            text = t
                         else:
-                            text = self.process_code('%s%s' % (c, l), line)
+                            self.toc.add_item(os.path.basename(self.file_name), id, value)
+                            text = '<span id="%s"></span>%s' % (id, t)
                     elif c == 'm':
                         empty = False
                         src = self.code_value(line)
@@ -528,11 +560,6 @@ class PML_HTMLizer(object):
                     elif c == 'p':
                         empty = False
                         text = '<br /><br style="page-break-after: always;" />'
-                    elif c == 'C':
-                        line.read(1)
-                        id = 'pml_toc-%s' % len(self.toc)
-                        self.toc.add_item(os.path.basename(self.file_name), id, self.code_value(line))
-                        text = '<span id="%s"></span>' % id
                     elif c == 'n':
                         pass
                     elif c == 'w':

src/calibre/ebooks/pml/pmlml.py

@@ -233,7 +233,7 @@ class PMLMLizer(object):
                 w += '="50%"'
             text.append(w)
         toc_id = elem.attrib.get('id', None)
-        if toc_id:
+        if toc_id and tag  not in ('h1', 'h2','h3','h4','h5','h6',):
             if self.toc.get(page.href, None):
                 toc_title = self.toc[page.href].get(toc_id, None)
                 if toc_title: