diff --git a/setup.py b/setup.py
index c5c78c1f68..d2593888e1 100644
--- a/setup.py
+++ b/setup.py
@@ -41,6 +41,7 @@ entry_points = {
                              'any2lrf   = libprs500.ebooks.lrf.any.convert_from:main',
                              'lrf2lrs   = libprs500.ebooks.lrf.parser:main',
                              'lrs2lrf   = libprs500.ebooks.lrf.lrs.convert_from:main',
+                             'pdfreflow = libprs500.ebooks.lrf.pdf.reflow:main',
                              'isbndb    = libprs500.ebooks.metadata.isbndb:main',
                              'librarything = libprs500.ebooks.metadata.library_thing:main',
                              'lrf2html  = libprs500.ebooks.lrf.html.convert_to:main',                             
diff --git a/src/libprs500/__init__.py b/src/libprs500/__init__.py
index 93f022ebb1..500acb8b3a 100644
--- a/src/libprs500/__init__.py
+++ b/src/libprs500/__init__.py
@@ -13,7 +13,7 @@
 ##    with this program; if not, write to the Free Software Foundation, Inc.,
 ##    51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
 ''' E-book management software'''
-__version__   = "0.4.35"
+__version__   = "0.4.36"
 __docformat__ = "epytext"
 __author__    = "Kovid Goyal <kovid@kovidgoyal.net>"
 __appname__   = 'libprs500'
diff --git a/src/libprs500/ebooks/lrf/pdf/reflow.py b/src/libprs500/ebooks/lrf/pdf/reflow.py
new file mode 100644
index 0000000000..12683c1e90
--- /dev/null
+++ b/src/libprs500/ebooks/lrf/pdf/reflow.py
@@ -0,0 +1,436 @@
+##    Copyright (C) 2008 Kovid Goyal kovid@kovidgoyal.net
+##    This program is free software; you can redistribute it and/or modify
+##    it under the terms of the GNU General Public License as published by
+##    the Free Software Foundation; either version 2 of the License, or
+##    (at your option) any later version.
+##
+##    This program is distributed in the hope that it will be useful,
+##    but WITHOUT ANY WARRANTY; without even the implied warranty of
+##    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+##    GNU General Public License for more details.
+##
+##    You should have received a copy of the GNU General Public License along
+##    with this program; if not, write to the Free Software Foundation, Inc.,
+##    51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+'''
+Convert PDF to a reflowable format using pdftoxml.exe as the PDF parsing backend.
+'''
+
+import sys, os, re, tempfile, subprocess, atexit, shutil, logging, xml.parsers.expat
+from xml.etree.ElementTree import parse
+
+from libprs500 import isosx, OptionParser, setup_cli_handlers
+from libprs500.ebooks import ConversionError
+
+PDFTOXML = 'pdftoxml.exe'
+if isosx and hasattr(sys, 'frameworks_dir'):
+    PDFTOXML = os.path.join(getattr(sys, 'frameworks_dir'), PDFTOXML)
+
+class Page(object):
+    
+    def __init__(self, attrs):
+        for a in ('number', 'width', 'height'):
+            setattr(self, a, float(attrs[a]))
+        self.id     = attrs['id']
+        self.current_line = None
+        self.lines = []
+        
+    def end_line(self):
+        if self.current_line is not None:
+            self.current_line.finalize()
+            self.lines.append(self.current_line)
+            self.current_line = None
+            
+    def finalize(self):
+        self.identify_groups()
+        self.look_for_page_break()
+    
+    def identify_groups(self):
+        groups = []
+        in_group = False
+        for i in range(len(self.lines)):
+            if not in_group:
+                groups.append(i)
+                in_group = True
+            else:
+                pl = self.lines[i-1]
+                cl = self.lines[i]
+                if cl.left != pl.left and cl.width != pl.width:
+                    groups.append(i)
+        self.groups = []
+        for i in range(len(groups)):
+            start = groups[i]
+            if i +1 == len(groups):
+                stop = len(self.lines)
+            else:
+                stop = groups[i+i]
+            self.groups.append(self.lines[start:stop])
+        
+        if len(self.groups) > 1:
+            self.group[0].test_header(self.width, self.height)
+            self.groups[-1].test_footer(self.width, self.height)
+            
+    def look_for_page_break(self):
+        max = 0
+        for g in self.groups:
+            if not g.is_footer and g.bottom > max:
+                max = g.bottom
+        self.page_break_after = max < 0.8*self.height
+        
+
+class Group(object, StyleContainer):
+    
+    def __init__(self, lines):
+        self.lines = lines
+        self.set_style(self.lines)
+        self.width = max([i.width for i in self.lines])
+        self.bottom = max([i.bottom for i in self.lines])
+        tot, ltot = 0, 0
+        for i in range(1, len(self.lines)):
+            bot = self.lines[i-1].bottom
+            top = self.lines[i].top
+            tot += abs(top - bot)
+            ltot += self.lines[i].left
+        self.average_line_spacing = tot/float(len(self.lines)-1)
+        ltot += self.lines[0].left
+        self.average_left_margin = ltot/float(len(self.lines))
+        self.left_margin = min([i.left for i in self.lines])
+        
+        self.detect_paragraphs()
+        
+        
+        
+    def detect_paragraphs(self):
+        if not self.lines:
+            return
+        indent_buffer = 5
+        self.lines[0].is_para_start = self.lines[0].left > self.average_left_margin+indent_buffer 
+        for i in range(1, len(self.lines)):
+            pl, l = self.lines[i-1:i+1]
+            c1 = pl.bottom - l.top > self.average_line_spacing
+            c2 = l.left > self.average_left_margin+indent_buffer
+            c3 = pl.width < 0.8 * self.width
+            l.is_para_start = c1 or c2 or c3
+            
+    def test_header(self, page_width, page_height):
+        self.is_header = len(self.lines) == 1 and self.lines[0].width < 0.5*page_width 
+        
+    def test_footer(self, page_width, page_height):
+        self.is_footer = len(self.lines) == 1 and self.lines[0].width < 0.5*page_width
+
+class Text(object):
+    
+    def __init__(self, attrs):
+        for a in ('x', 'y', 'width', 'height'):
+            setattr(self, a, float(attrs[a]))
+        self.id = attrs['id']
+        self.objects = []
+        
+    def add_token(self, tok):
+        if not self.objects:
+            self.objects.append(tok)
+        else:
+            ptok = self.objects[-1]
+            if tok == ptok:
+                ptok.text += ' ' + tok.text
+            else:
+                self.objects.append(tok)
+    
+    def add(self, object):
+        if isinstance(object, Token):
+            self.add_token(object)
+        else:
+            print 'WARNING: Unhandled object', object.__class__.__name__
+            
+    def to_xhtml(self):
+        res = []
+        for obj in self.objects:
+            if isinstance(obj, Token):
+                res.append(obj.to_xhtml())
+        return ' '.join(res)
+                
+
+class StyleContainer(object):
+    
+    def set_style(self, iterator):
+        styles = set([])
+        for tok in iterator:
+            if hasattr(tok, 'style'):
+                styles.add(tok.style)
+        counts = [0*i for i in range(len(styles))]
+        for i in range(len(styles)):
+            counts[i] = sum([1 for j in self if j.style == styles[i]])
+        max = max(counts)
+        for i in range(len(counts)):
+            if counts[i] == max:
+                break
+        self.style = styles[i]
+        for obj in iterator:
+            if obj.style == self.style:
+                obj.style = None
+    
+class Line(list, StyleContainer):
+    
+    def calculate_geometry(self):
+        self.left   = self[0].x
+        self.width  = self[-1].x + self[-1].width - self.left
+        self.top    = min(o.y for o in self)
+        self.bottom = max(o.height+o.y for o in self)
+        
+    def finalize(self):
+        self.calculate_geometry()
+        self.set_style(self)
+        
+    def to_xhtml(self, group_id):
+        ans = '<span class="%s" '%group_id
+        if self.style is not None:
+            ans += 'style="%s"'%self.style.to_css(inline=True)
+        ans += '>%s</span>'
+        res = []
+        for object in self:
+            if isinstance(object, Text):
+                res.append(object.to_xhtml())
+                
+        return ans%(' '.join(res))
+                
+        
+class TextStyle(object):
+    
+    def __init__(self, tok):
+        self.bold   = tok.bold
+        self.italic = tok.italic
+        self.font_name = tok.font_name
+        self.font_size = tok.font_size
+        self.color     = tok.font_color
+        
+    def __eq__(self, other):
+        if isinstance(other, self.__class__):
+            for a in ('font_size', 'bold', 'italic', 'font_name', 'color'):
+                if getattr(self, a) != getattr(other, a):
+                    return False
+            return True
+        return False
+    
+    def to_css(self, inline=False):
+        fw  = 'bold' if self.bold else 'normal'
+        fs  = 'italic' if self.italic else 'normal'
+        fsz = '%dpt'%self.font_size
+        props = ['font-weight: %s;'%fw, 'font-style: %s;'%fs, 'font-size: %s;'%fsz,
+                 'color: rgb(%d, %d, %d);'%self.color]
+        joiner = ' '
+        if not inline:
+            joiner = '\n'
+            props = ['{'] + props + ['}']
+        return joiner.join(props) 
+
+class Token(object):
+    
+    def __init__(self, attrs):
+        for a in ('x', 'y', 'width', 'height', 'rotation', 'angle', 'font-size'):
+            setattr(self, a.replace('-', '_'), float(attrs[a]))
+        for a in ('bold', 'italic'):
+            setattr(self, a, attrs[a]=='yes')
+        self.font_name = attrs['font-name']
+        fc = re.compile(r'#([a-f0-9]{2})([a-f0-9]{2})([a-f0-9]{2})', re.IGNORECASE)
+        fc = fc.match(attrs['font-color'])
+        self.font_color = (int(fc.group(1), 16), int(fc.group(2), 16), int(fc.group(3), 16))
+        self.id = attrs['id']
+        self.text = u''
+        self.style = TextStyle(self)
+        
+    def handle_char_data(self, data):
+        self.text += data
+        
+    def __eq__(self, other):
+        if isinstance(other, self.__class__):
+            for a in ('rotation', 'angle', 'font_size', 'bold', 'italic', 'font_name', 'font_color'):
+                if getattr(self, a) != getattr(other, a):
+                    return False
+            return True
+        return False
+    
+    def to_xhtml(self):
+        if self.style is not None:
+            ans = u'<span style="%s">%s</span>'%(self.style.to_css(inline=True), self.text)
+        else:
+            ans = self.text
+        return ans
+
+class PDFDocument(object):
+    
+    SKIPPED_TAGS = ('DOCUMENT', 'METADATA', 'PDFFILENAME', 'PROCESS', 'VERSION',
+                    'COMMENT', 'CREATIONDATE')
+    
+    def __init__(self, filename):
+        parser = xml.parsers.expat.ParserCreate('UTF-8')
+        parser.buffer_text          = True
+        parser.returns_unicode      = True
+        parser.StartElementHandler  = self.start_element
+        parser.EndElementHandler    = self.end_element
+        
+        self.pages = []
+        self.current_page = None
+        self.current_token = None
+        
+        src = open(filename, 'rb').read()
+        self.parser = parser
+        parser.Parse(src)
+        
+        
+    def start_element(self, name, attrs):
+        if name == 'TOKEN':
+            self.current_token = Token(attrs)
+            self.parser.CharacterDataHandler = self.current_token.handle_char_data
+        elif name == 'TEXT':
+            text = Text(attrs)
+            if self.current_page.current_line is None:
+                self.current_page.current_line = Line()
+                self.current_page.current_line.append(text)
+            else:
+                y, height = self.current_page.current_line[0].y, self.current_page.current_line[0].height
+                if y == text.y or y+height == text.y + text.height:
+                    self.current_page.current_line.append(text)
+                else:
+                    self.current_page.end_line()
+                    self.current_page.current_line = Line()
+                    self.current_page.current_line.append(text)
+        elif name == 'PAGE':
+            self.current_page = Page(attrs)
+        elif name.lower() == 'xi:include':
+            print 'WARNING: Skipping vector image'
+        elif name in self.SKIPPED_TAGS:
+            pass
+        else:
+            print 'WARNING: Unhandled element', name
+        
+    def end_element(self, name):
+        if name == 'TOKEN':
+            if self.current_token.angle == 0 and self.current_token.rotation == 0:
+                self.current_page.current_line[-1].add(self.current_token)
+            self.current_token = None
+            self.parser.CharacterDataHandler = None
+        elif name == 'PAGE':
+            self.current_page.finalize()
+            self.pages.append(self.current_page)
+            self.current_page = None
+    
+    
+    def to_xhtml(self):
+        header = u'''\
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
+    "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml"
+     xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+     xsi:schemaLocation="http://www.w3.org/MarkUp/SCHEMA/xhtml11.xsd" >
+<head>
+<style type="text/css">
+%(style)s
+</style>
+</head>
+<body>
+%(body)s
+</body>
+</html>
+'''
+        res = []
+        para = []
+        styles = []
+        for page in self.pages:
+            res.append(u'<a name="%s" />'%page.id)
+            for group in page.groups:
+                if group.is_header or group.is_footer:
+                    continue
+                if group.style is not None:
+                    styles.append(u'.%s %s\n'%(group.id, group.style.to_css()))
+                for line in group.lines:
+                    if line.is_para_start:
+                        indent = group.left_margin - line.left
+                        if para:
+                            res.append(u'<p style="text-indent: %dpt">%s</p>'%(indent, ''.join(para)))
+                            para = []
+                    para.append(line.to_xhtml(group.id))
+            if page.page_break_after:
+                res.append(u'<br style="page-break-after:always" />')
+                if para:
+                    res.append(u'<p>%s</p>'%(''.join(para)))
+                    para = []
+                    
+        return (header%dict(style='\n'.join(styles), body='\n'.join(res))).encode('utf-8')
+
+class PDFConverter(object):
+
+    @classmethod
+    def generate_xml(cls, pathtopdf, logger):
+        pathtopdf = os.path.abspath(pathtopdf)
+        tdir = tempfile.mkdtemp('pdf2xml', 'libprs500')
+        atexit.register(shutil.rmtree, tdir)
+        xmlfile = os.path.basename(pathtopdf)+'.xml'
+        os.chdir(tdir)
+        cmd = PDFTOXML + ' -outline "%s" "%s"'%(pathtopdf, xmlfile)
+        p = subprocess.Popen(cmd, shell=True, stderr=subprocess.STDOUT, 
+                             stdout=subprocess.PIPE)
+        log = p.stdout.read()
+        ret = p.wait()
+        if ret != 0:
+            raise ConversionError, log
+        xmlfile = os.path.join(tdir, xmlfile)
+        if os.stat(xmlfile).st_size < 20:
+            raise ConversionError(os.path.basename(pathtopdf) + ' does not allow copying of text.')
+        return xmlfile
+
+    
+    def __init__(self, pathtopdf, logger, opts):
+        self.cwd    = os.getcwdu()
+        self.logger = logger
+        self.opts   = opts
+        try:
+            self.logger.info('Converting PDF to XML')
+            self.xmlfile   = self.generate_xml(pathtopdf, self.logger)
+            self.tdir      = os.path.dirname(self.xmlfile)
+            self.data_dir  = self.xmlfile + '_data'
+            outline_file = self.xmlfile.rpartition('.')[0]+'_outline.xml'
+            self.logger.info('Parsing XML')
+            self.document = PDFDocument(self.xmlfile)
+            self.outline  = parse(outline_file)
+        finally:
+            os.chdir(self.cwd)
+            
+    def convert(self, output_dir):
+        doc = self.document.to_xhtml()
+        open(os.path.join(output_dir, 'document.html'), 'wb').write(doc)
+        
+            
+            
+def option_parser():
+    parser = OptionParser(usage=\
+'''
+%prog [options] myfile.pdf
+
+Convert a PDF file to a HTML file.
+''')
+    parser.add_option('-o', '--output-dir', default='.', 
+                      help=_('Path to output directory in which to create the HTML file. Defaults to current directory.'))
+    parser.add_option('--verbose', default=False, action='store_true',
+                      help=_('Be more verbose.'))
+    return parser    
+
+def main(args=sys.argv, logger=None):
+    parser = option_parser()
+    options, args = parser.parse_args()
+    if logger is None:
+        level = logging.DEBUG if options.verbose else logging.INFO
+        logger = logging.getLogger('pdf2html')
+        setup_cli_handlers(logger, level)
+    if len(args) != 1:
+        parser.print_help()
+        print _('You must specify a single PDF file.')
+        return 1
+    options.output_dir = os.path.abspath(options.output_dir)
+    converter = PDFConverter(os.path.abspath(args[0]), logger, options)
+    converter.convert(options.output_dir)
+    
+    return 0
+
+if __name__ == '__main__':
+    sys.exit(main())
\ No newline at end of file
diff --git a/src/libprs500/linux.py b/src/libprs500/linux.py
index 174e5868ec..0b38873bb5 100644
--- a/src/libprs500/linux.py
+++ b/src/libprs500/linux.py
@@ -97,6 +97,8 @@ def setup_completion():
         from libprs500.ebooks.lrf.meta import option_parser as metaop
         from libprs500.ebooks.lrf.parser import option_parser as lrf2lrsop
         from libprs500.gui2.lrf_renderer.main import option_parser as lrfviewerop
+        from libprs500.ebooks.lrf.pdf.pdftohtml import option_parser as pdfhtmlop
+        
         f = open_file('/etc/bash_completion.d/libprs500')
         
         f.write('# libprs500 Bash Shell Completion\n')
@@ -117,6 +119,7 @@ def setup_completion():
         f.write(opts_and_exts('opf-meta', metaop, ['opf']))
         f.write(opts_and_exts('epub-meta', metaop, ['epub']))
         f.write(opts_and_exts('lrfviewer', lrfviewerop, ['lrf']))
+        f.write(opts_and_exts('pdfrelow', pdfhtmlop, ['pdf']))
         f.write('''
 _prs500_ls()
 {