Implement the --linearize-tables transform.

src/calibre/customize/profiles.py

@@ -143,7 +143,7 @@ class OutputProfile(Plugin):
 
     # ADE dies an agonizing, long drawn out death if HTML files have more
     # bytes than this.
-    flow_size                 = sys.maxint
+    flow_size                 = -1
     # ADE runs screaming when it sees these characters
     remove_special_chars      = re.compile(u'[\u200b\u00ad]')
     # ADE falls to the ground in a dead faint when it sees an <object>

src/calibre/ebooks/conversion/plumber.py

@@ -94,7 +94,8 @@ OptionRecommendation(name='font_size_mapping',
 OptionRecommendation(name='line_height',
             recommended_value=None, level=OptionRecommendation.LOW,
             help=_('The line height in pts. Controls spacing between consecutive '
-                   'lines of text. By default ??'
+                   'lines of text. By default no line height manipulation is '
+                   'performed.'
                    )
         ),
 
@@ -102,12 +103,25 @@ OptionRecommendation(name='linearize_tables',
             recommended_value=False, level=OptionRecommendation.LOW,
             help=_('Some badly designed documents use tables to control the '
                 'layout of text on the page. When converted these documents '
-                'often have text that runs of the page and other artifacts. '
+                'often have text that runs off the page and other artifacts. '
                 'This option will extract the content from the tables and '
                 'present it in a linear fashion.'
                 )
         ),
 
+OptionRecommendation(name='dont_split_on_page_breaks',
+            recommended_value=False, level=OptionRecommendation.LOW,
+            help=_('Turn off splitting at page breaks. Normally, input '
+                    'files are automatically split at every page break into '
+                    'two files. This gives an output ebook that can be '
+                    'parsed faster and with less resources. However, '
+                    'splitting is slow and if your source file contains a '
+                    'very large number of page breaks, you should turn off '
+                    'splitting on page breaks.'
+                )
+        ),
+
+
 OptionRecommendation(name='read_metadata_from_opf',
             recommended_value=None, level=OptionRecommendation.LOW,
             short_switch='m',
@@ -330,6 +344,17 @@ OptionRecommendation(name='language',
                 untable=self.opts.linearize_tables)
         flattener(self.oeb, self.opts)
 
+        if self.opts.linearize_tables:
+            from calibre.ebooks.oeb.transforms.linearize_tables import LinearizeTables
+            LinearizeTables()(self.oeb, self.opts)
+
+        from calibre.ebooks.oeb.transforms.split import Split
+        pbx = accelerators.get('pagebreaks', None)
+        split = Split(not self.opts.dont_split_on_page_breaks,
+                max_flow_size=self.opts.output_profile.flow_size,
+                page_breaks_xpath=pbx)
+        split(self.oeb, self.opts)
+
         from calibre.ebooks.oeb.transforms.trimmanifest import ManifestTrimmer
 
         self.log.info('Cleaning up manifest...')

src/calibre/ebooks/mobi/input.py

@@ -29,5 +29,5 @@ class MOBIInput(InputFormatPlugin):
             with open(f, 'wb') as q:
                 q.write(html.tostring(root, encoding='utf-8', method='xml',
                     include_meta_content_type=False))
-            accelerators['pagebreaks'] = {f: '//*[@class="mbp_pagebreak"]'}
+                accelerators['pagebreaks'] = '//h:div[@class="mbp_pagebreak"]'
         return mr.created_opf_path

src/calibre/ebooks/oeb/transforms/linearize_tables.py

@@ -0,0 +1,21 @@
+#!/usr/bin/env python
+# vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
+from __future__ import with_statement
+
+__license__   = 'GPL v3'
+__copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
+__docformat__ = 'restructuredtext en'
+
+from calibre.ebooks.oeb.base import OEB_DOCS, XPNSMAP
+
+class LinearizeTables(object):
+
+    def linearize(self, root):
+        for x in root.xpath('//h:table|//h:td|//h:tr|//h:th',
+                namespaces=XPNSMAP):
+            x.tag = 'div'
+
+    def __call__(self, oeb, context):
+        for x in oeb.manifest.items:
+            if x.media_type in OEB_DOCS:
+                self.linearize(x.data)