params = {'api_key': API_KEY, 'num': count}

        self.note_count = post['note_count']
        try:            
            self.source_title = post['source_title']
            self.source_url = post['source_url']
        except KeyError:
            self.source_title = ''
            self.source_url = ''

    def type_callback(option, opt, value, parser):
        value = value.replace('text', 'regular')
        value = value.replace('chat', 'conversation')
        value = value.replace('photoset', 'photo')
        csv_callback(option, opt, value, parser)

--- tumblr_backup_apiv2.py.orig 2014-10-06 14:07:00.964349600 +0200
+++ tumblr_backup_apiv2.py  2014-10-06 14:06:31.007685400 +0200
@@ -155,7 +155,7 @@


 def apiparse(base, count, start=0):
-    params = {'api_key': API_KEY, 'num': count}
+    params = {'api_key': API_KEY, 'limit': count}
     if start > 0:
         params['offset'] = start
     url = base + '?' + urllib.urlencode(params)
@@ -450,6 +450,13 @@
         self.tm = time.localtime(self.date)
         self.title = ''
         self.tags = post['tags']
+        self.note_count = post['note_count']
+        try:            
+            self.source_title = post['source_title']
+            self.source_url = post['source_url']
+        except KeyError:
+            self.source_title = ''
+            self.source_url = ''
         if options.tags:
             self.tags_lower = set(t.lower() for t in self.tags)
         self.file_name = join(self.ident, dir_index) if options.dirs else self.ident + post_ext
@@ -692,10 +699,9 @@
         csv_callback(option, opt, value.lower(), parser)

     def type_callback(option, opt, value, parser):
-        value = value.replace('text', 'regular')
-        value = value.replace('chat', 'conversation')
-        value = value.replace('photoset', 'photo')
-        csv_callback(option, opt, value, parser)
+        values_list = value.lower().split(',')
+        set(values_list).intersection(['text','quote','link', 'answer','video','audio','photo','chat'])
+        setattr(parser.values, option.dest, values_list)

     parser = optparse.OptionParser("Usage: %prog [options] blog-name ...",
         description="Makes a local backup of Tumblr blogs."
@@ -748,7 +754,9 @@
         " case-insensitive)"
     )
     parser.add_option('-T', '--type', type='string', action='callback',
-        callback=type_callback, help="save only posts of type TYPE (comma-separated values)"
+        callback=type_callback, help="save only posts of type TYPE (comma-separated values)."
+        " Type can be:  text, quote, link, answer, video, audio, photo, chat"
+        
     )
     parser.add_option('-I', '--image-names', type='choice', choices=('o', 'i', 'bi'),
         default='o', metavar='FMT',

--- tumblr_backup_apiv2.py.orig 2014-10-08 12:21:28.322480700 +0200
+++ tumblr_backup_apiv2_inlinesupport.py    2014-10-08 14:01:19.565999100 +0200
@@ -162,7 +162,7 @@


 def apiparse(base, count, start=0):
-    params = {'api_key': API_KEY, 'num': count}
+    params = {'api_key': API_KEY, 'limit': count}
     if start > 0:
         params['offset'] = start
     url = base + '?' + urllib.urlencode(params)
@@ -530,9 +530,11 @@
         def get_try(elt):
             return post.get(elt)

+
         def append_try(elt, fmt=u'%s'):
             elt = get_try(elt)
             if elt:
+                elt = re.sub(r'(<img [^\>]*src\s*=\s*["\'])(.*?)(["\'][^\>]*>)', self.get_inline_url, elt, flags=re.I)
                 append(elt, fmt)

         if self.typ == 'text':
@@ -576,7 +578,7 @@

         elif self.typ == 'answer':
             self.title = post['question']
-            append(post['answer'])
+            append_try('answer')

         elif self.typ == 'chat':
             self.title = get_try('title')
@@ -599,6 +601,44 @@

         self.save_post()

+    def get_inline_url(self, match):
+        """Saves an inline image if not saved yet. Returns the new URL or
+        the original URL in case of download errors."""
+        
+        self.image_dir = join(post_dir, self.ident) if options.dirs else image_dir
+        self.image_folder = path_to(self.image_dir)
+        image_url = match.group(2)
+
+        def _url(fn):
+            return match.group(1) + u'%s%s/%s' % (save_dir, self.image_dir, fn) + match.group(3)
+
+        image_filename = image_url.split('/')[-1]
+        # check if a file with this name already exists
+        known_extension = '.' in image_filename[-5:]
+        image_glob = glob(join(self.image_folder, image_filename +
+            ('' if known_extension else '.*')
+        ))
+        if image_glob:
+            return _url(split(image_glob[0])[1])
+        # download the image data
+        try:
+            image_response = urllib2.urlopen(image_url, timeout=HTTP_TIMEOUT)
+            image_data = image_response.read()
+            image_response.close()
+        except:
+            # return the original URL
+            return match.group(0)
+        # determine the file type if it's unknown
+        if not known_extension:
+            image_type = imghdr.what(None, image_data[:32])
+            if image_type:
+                image_filename += '.' + image_type.replace('jpeg', 'jpg')
+        # save the image
+        with open_image(self.image_dir, image_filename) as image_file:
+            image_file.write(image_data)
+        return _url(image_filename)
+
+
     def get_image_url(self, image_url, offset):
         """Saves an image if not saved yet. Returns the new URL or
         the original URL in case of download errors."""

--- tumblr_backup_apiv2.py.orig 2014-10-08 12:21:28.322480700 +0200
+++ tumblr_backup_apiv2_fixphotosetcaption.py   2014-10-08 14:34:34.414669700 +0200
@@ -554,7 +554,7 @@
                     content[-1] = u'<a href="%s">%s</a>' % (escape(url), content[-1])
                 content[-1] = '<p>' + content[-1] + '</p>'
                 if p['caption']:
-                    append(o['caption'], u'<p>%s</p>')
+                    append(p['caption'], u'<p>%s</p>')
             append_try('caption')

         elif self.typ == 'link':

--- tumblr_backup_apiv2.py.orig 2014-10-07 11:29:56.562755600 +0200
+++ tumblr_backup_mediasupport.py   2014-10-07 17:37:38.295584200 +0200
@@ -22,6 +22,7 @@
 import time
 import urllib
 import urllib2
+import socket
 from xml.sax.saxutils import escape

 # extra optional packages
@@ -29,6 +30,10 @@
     import pyexiv2
 except ImportError:
     pyexiv2 = None
+try:
+    import youtube_dl
+except ImportError:
+    youtube_dl = None

 # default blog name(s)
 DEFAULT_BLOGS = ['bbolli']
@@ -52,12 +57,14 @@
 # variable directory names, will be set in TumblrBackup.backup()
 save_folder = ''
 image_folder = ''
+media_folder = ''

 # constant names
 root_folder = os.getcwdu()
 post_dir = 'posts'
 json_dir = 'json'
 image_dir = 'images'
+media_dir = 'medias'
 archive_dir = 'archive'
 theme_dir = 'theme'
 save_dir = '../'
@@ -130,6 +137,10 @@
 def open_image(*parts):
     return open_file(lambda f: open(f, 'wb'), parts)

+    
+def open_media(*parts):
+    return open_file(lambda f: open(f, 'ab'), parts)
+

 def strftime(format, t=None):
     if t is None:
@@ -162,7 +173,7 @@


 def apiparse(base, count, start=0):
-    params = {'api_key': API_KEY, 'num': count}
+    params = {'api_key': API_KEY, 'limit': count}
     if start > 0:
         params['offset'] = start
     url = base + '?' + urllib.urlencode(params)
@@ -334,7 +345,7 @@
         base = get_api_url(account)

         # make sure there are folders to save in
-        global save_folder, image_folder, post_ext, post_dir, save_dir, have_custom_css
+        global save_folder, image_folder, media_folder, post_ext, post_dir, save_dir, have_custom_css
         if options.blosxom:
             save_folder = root_folder
             post_ext = '.txt'
@@ -343,6 +354,7 @@
         else:
             save_folder = join(root_folder, options.outdir or account)
             image_folder = path_to(image_dir)
+            media_folder = path_to(media_dir)
             if options.dirs:
                 post_ext = ''
                 save_dir = '../../'
@@ -465,6 +477,16 @@
         self.tags = post['tags']
         if options.tags:
             self.tags_lower = set(t.lower() for t in self.tags)
+        try:
+            self.note_count = post['note_count']
+        except KeyError:
+            self.note_count = ''
+        try:
+            self.source_title = post['source_title']
+            self.source_url = post['source_url']
+        except KeyError:
+            self.source_title = ''
+            self.source_url = ''
         self.file_name = join(self.ident, dir_index) if options.dirs else self.ident + post_ext
         self.llink = self.ident if options.dirs else self.file_name

@@ -516,11 +538,47 @@
             append_try('source', u'<p>%s</p>')

         elif self.typ == 'video':
-            append(post['player'][-1]['embed_code'])
+            src = ''
+            if not options.skip_images:
+                self.media_dir = join(post_dir, self.ident) if options.dirs else media_dir
+                self.media_folder = path_to(self.media_dir)
+                if post['video_type'] == 'tumblr':
+                    src = self.get_media_url(post['video_url'], '.mp4')
+                elif youtube_dl:
+                    if post['html5_capable']:
+                        try:
+                            src = self.get_youtube_url(post['permalink_url'])
+                        except:
+                            sys.stdout.write(u'Unknown video type in post #%s%-50s\n' % (self.ident, ' '))
+                    else:
+                        try:
+                            src = self.get_youtube_url(post['source_url'])
+                        except:
+                            sys.stdout.write(u'Unknown video type in post #%s%-50s\n' % (self.ident, ' '))
+            if src:
+                append(u'<p><video controls><source src="%s" type="video/mp4">Your browser does not support the video element.<br /><a href="%s" >Video file</a></video></p>' % (src, src))
+            else:
+                append(post['player'][-1]['embed_code'])
             append_try('caption')

         elif self.typ == 'audio':
-            append(post['player'])
+            src = ''
+            if not options.skip_images:
+                self.media_dir = join(post_dir, self.ident) if options.dirs else media_dir
+                self.media_folder = path_to(self.media_dir)
+                if post['audio_type'] == 'tumblr':
+                    audio_url = post['audio_url']
+                    if audio_url.startswith('http://a.tumblr.com/'):
+                        src = self.get_media_url(audio_url, '.mp3')
+                    elif audio_url.startswith('https://www.tumblr.com/audio_file/'):
+                        audio_url = u'http://a.tumblr.com/%so1.mp3' % audio_url.split('/')[-1]
+                        src = self.get_media_url(audio_url, '.mp3')
+                elif  post['audio_type'] == 'soundcloud':
+                    src = self.get_media_url(post['audio_url'], '.mp3')
+            if src:
+                append(u'<p><audio controls><source src="%s" type="audio/mpeg">Your browser does not support the audio element.<br /><a href="%s" >Audio file</a></audio></p>' % (src, src))
+            else:
+                append(post['player'])
             append_try('caption')

         elif self.typ == 'answer':
@@ -548,6 +606,79 @@

         self.save_post()

+
+    def get_youtube_url(self, youtube_url):
+        # determine the media file name
+        ydl = youtube_dl.YoutubeDL({'outtmpl': join(self.media_folder, u'%(id)s_%(uploader_id)s_%(title)s.%(ext)s'), 'quiet': True, 'restrictfilenames': True, 'noplaylist': True})
+        ydl.add_default_info_extractors()
+        try:
+            result = ydl.extract_info(youtube_url, download=False)
+            media_filename = ydl.prepare_filename(result)
+        except:
+            return ''
+
+        # check if a file with this name already exists
+        media_glob = glob(media_filename)
+        if media_glob:
+            return u'%s%s/%s' % (save_dir, self.media_dir, split(media_glob[0])[1])
+
+        try:
+            result = ydl.extract_info(youtube_url, download=True)
+        except:
+            return ''
+        return u'%s%s/%s' % (save_dir, self.media_dir,os.path.split(media_filename)[1])
+
+
+    def get_media_url(self, media_url, extension=''):
+        def _url(fn):
+            return u'%s%s/%s' % (save_dir, self.media_dir, fn)
+
+        # determine the media file name
+        if options.image_names == 'i':
+            media_filename = self.ident
+        elif options.image_names == 'bi':
+            media_filename = account + '_' + self.ident
+        else:
+            media_filename = media_url.split('/')[-1]
+
+        if '.' in media_filename[-4]:
+            media_filename = media_filename[:-4]
+        media_filename += extension
+
+        # check if a file with this name already exists
+        media_glob = glob(join(self.media_folder, media_filename))
+        if media_glob:
+            return _url(split(media_glob[0])[1])
+
+        # download the media data
+        media_part_glob = glob(join(self.media_folder, media_filename + '.part'))
+        if media_part_glob:
+            try:
+                os.remove(media_part_glob[0])
+            except Exception, e:
+                sys.stderr.write('Error deleting the temporary file: %s' % e)
+                return ''
+
+        try:
+            media_response = urllib2.urlopen(media_url)
+            while True:
+                media_data = media_response.read(1024*1024)
+                if not media_data:
+                    break
+                # save the media
+                with open_media(self.media_dir, media_filename + '.part') as media_file:
+                    media_file.write(media_data)
+            try:
+                os.rename(join(self.media_folder, media_filename + '.part'), join(self.media_folder, media_filename))
+            except Exception, e:
+                sys.stderr.write('Error writing the media file: %s' % e)
+                return ''
+            media_response.close()
+        except (urllib2.URLError, urllib2.HTTPError, socket.error):
+            return ''
+        return _url(media_filename)
+
+
     def get_image_url(self, image_url, offset):
         """Saves an image if not saved yet. Returns the new URL or
         the original URL in case of download errors."""
@@ -567,7 +698,7 @@
             image_filename = account + '_' + self.ident + offset
         else:
             image_filename = image_url.split('/')[-1]
-        glob_filter = '' if '.' in image_filename else '.*'
+        glob_filter = '' if '.' in image_filename[-4] else '.*'
         # check if a file with this name already exists
         image_glob = glob(join(self.image_folder, image_filename + glob_filter))
         if image_glob:
@@ -582,7 +713,7 @@
             # return the original URL
             return image_url
         # determine the file type if it's unknown
-        if '.' not in image_filename:
+        if '.' not in image_filename[-4]:
             image_type = imghdr.what(None, image_data[:32])
             if image_type:
                 image_filename += '.' + image_type.replace('jpeg', 'jpg')

--- tumblr_backup.py.orig   2014-10-05 11:55:41.019549600 +0200
+++ tumblr_backup_pagination.py 2014-10-06 10:29:59.294856000 +0200
@@ -271,6 +271,7 @@
             for year in sorted(self.index.keys(), reverse=options.reverse_index):
                 self.save_year(idx, year)
             idx.write(u'<p>Generated on %s.</p>\n' % strftime('%x %X'))
+            idx.write(self.footer())

     def save_year(self, idx, year):
         idx.write('<h3>%s</h3>\n<ul>\n' % year)
@@ -283,23 +284,66 @@
             ))
         idx.write('</ul>\n\n')

+    def get_page(self, direction, year, month):
+        list_years = sorted(self.index.keys(), reverse=direction)
+        list_years = list_years[list_years.index(year):]
+        for it_year in list_years:
+            if it_year == year:
+                list_months = sorted(self.index[it_year].keys(), reverse=direction) # false for next page
+                list_months = list_months[list_months.index(month)+1:]
+            else:
+                list_months = sorted(self.index[it_year].keys(), reverse=direction)
+            for it_month in list_months:
+                if len(self.index[it_year][it_month]):
+                    return it_year, it_month
+        return False, False
+
     def save_month(self, year, month, tm):
-        file_name = '%d-%02d' % (year, month)
-        if options.dirs:
-            arch = open_text(archive_dir, file_name, dir_index)
-            file_name += '/'
-        else:
-            file_name += '.html'
-            arch = open_text(archive_dir, file_name)
-        with arch:
-            arch.write('\n\n'.join([
-                self.header(strftime('%B %Y', tm), body_class='archive'),
-                '\n'.join(p.get_post() for p in sorted(
-                    self.index[year][month], key=lambda x: x.date, reverse=options.reverse_month
-                )),
-                '<p><a href=%s rel=contents>Index</a></p>\n' % save_dir
-            ]))
-        return file_name
+        postsperpage = options.number_page if options.number_page >= 1 else len(self.index[year][month])
+        file_stream = '\n\n'.join([self.header(strftime('%B %Y', tm), body_class='archive')])
+        for n, p in enumerate(sorted(self.index[year][month], key=lambda x: x.date, reverse=options.reverse_month), start=1):
+            file_stream += '\n'.join([p.get_post()])
+            if ((n % postsperpage == 0) or (n == len(self.index[year][month]))):
+                page = n / postsperpage if n % postsperpage == 0 else n / postsperpage + 1
+                file_name = '%d-%02d-p%s' % (year, month, page)
+                if options.dirs:
+                    arch = open_text(archive_dir, file_name, dir_index)
+                    file_name += '/%s' % dir_index
+
+                    if page > 1:
+                        pp = '%s%s/%d-%02d-p%s/%s' % (save_dir, archive_dir, year, month, page - 1, dir_index)
+                    else:
+                        previous_year, previous_month = self.get_page(False if options.reverse_month else True, year, month)
+                        pp = '%s%s/%d-%02d-p%s/%s' % (save_dir, archive_dir, previous_year, previous_month, len(self.index[previous_year][previous_month]) / postsperpage + 1, dir_index)  if previous_year else ''
+                    if len(self.index[year][month]) - n > 0:
+                        np = '%s%s/%d-%02d-p%s/%s' % (save_dir, archive_dir, year, month, page + 1, dir_index)
+                    else:
+                        next_year, next_month = self.get_page(True if options.reverse_month else False, year, month)
+                        np = '%s%s/%d-%02d-p%s/%s' % (save_dir, archive_dir, next_year, next_month, 1, dir_index) if next_year else ''
+
+                else:
+                    file_name += '.html'
+                    arch = open_text(archive_dir, file_name)
+
+                    if page > 1:
+                        pp = '%d-%02d-p%s.html' % (year, month, page - 1)
+                    else:
+                        previous_year, previous_month = self.get_page(False if options.reverse_month else True, year, month)
+                        pp = '%d-%02d-p%s.html' % (previous_year, previous_month, len(self.index[previous_year][previous_month]) / postsperpage + 1)  if previous_year else ''
+                    if len(self.index[year][month]) - n > 0:
+                        np = '%d-%02d-p%s.html' % (year, month, page + 1)
+                    else:
+                        next_year, next_month = self.get_page(True if options.reverse_month else False, year, month)
+                        np = '%d-%02d-p%s.html' % (next_year, next_month, 1) if next_year else ''
+
+                file_stream += '\n'.join([self.footer(previous_page=pp,next_page=np)])
+                with arch:
+                    arch.write(file_stream)
+                file_stream = '\n\n'.join([self.header(strftime('%B %Y', tm), body_class='archive')])
+                if page == 1:
+                    first_file = file_name
+
+        return first_file

     def header(self, title='', body_class='', subtitle='', avatar=''):
         root_rel = '' if body_class == 'index' else save_dir
@@ -323,6 +367,19 @@
             h += u'<p class=subtitle>%s</p>\n' % subtitle
         return h

+    def footer(self, previous_page='', next_page=''):
+        f = u'<footer id="footer">'
+        f += u'<div id="pagination">\n'
+        f += u'<a href="%s%s" class="index">Index</a>\n' % (save_dir, dir_index)
+        if ((previous_page) or (next_page)):
+            if (previous_page):
+                f += u'<a href="%s" class="previous">Previous</a>\n' % previous_page
+            if (next_page):
+                f += u'<a href="%s" class="next">Next</a>\n'% next_page
+        f += u'</div>\n'
+        f += u'</footer>\n</body>\n</html>'
+        return f
+
     def backup(self, account):
         """makes single files and an index for every post on a public Tumblr blog account"""

@@ -378,6 +435,7 @@

         # use the meta information to create a HTML header
         TumblrPost.post_header = self.header(body_class='post')
+        TumblrPost.post_footer = self.footer()

         # find the post number limit to back up
         last_post = options.count + options.skip if options.count else int(soup.posts('total'))
@@ -442,6 +500,7 @@
 class TumblrPost:

     post_header = ''    # set by TumblrBackup.backup()
+    post_footer = ''

     def __init__(self, post):
         self.content = ''
@@ -604,13 +663,14 @@
         post = self.post_header + u'<article class=%s id=p-%s>\n' % (self.typ, self.ident)
         post += u'<p class=meta><span class=date>%s</span>\n' % strftime('%x %X', self.tm)
         post += u'<a class=llink href=%s%s/%s>¶</a>\n' % (save_dir, post_dir, self.llink)
-        post += u'<a href=%s rel=canonical>●</a></p>\n' % self.url
+        post += u'<a href="%s">●</a></p>\n' % self.url
         if self.title:
             post += u'<h2>%s</h2>\n' % self.title
         post += self.content
         if self.tags:
             post += u'\n<p class=tags>%s</p>' % u''.join(self.tag_link(t) for t in self.tags)
         post += '\n</article>\n'
+        post += self.post_footer
         return post

     @staticmethod
@@ -760,6 +820,9 @@
     parser.add_option('-p', '--period', help="limit the backup to PERIOD"
         " ('y', 'm', 'd' or YYYY[MM[DD]])"
     )
+    parser.add_option('-N', '--number-page', type='int', default=50,
+        help="set NUMBER_PAGE posts per month page"
+    )
     parser.add_option('-P', '--private', help="password for a private tumblr",
         metavar='PASSWORD'
     )

--- tumblr_backup.py.orig   2014-10-01 09:57:32.142436800 +0200
+++ tumblr_backup_threading.py  2014-10-01 16:51:47.713677400 +0200
@@ -5,6 +5,8 @@
 from __future__ import with_statement
 import os
 from os.path import join, split, splitext
+import threading
+from threading import Thread, Event
 import sys
 import urllib
 import urllib2
@@ -68,6 +70,7 @@
 have_custom_css = False

 MAX_POSTS = 50
+imagePool = {}

 # ensure the right date/time format
 try:
@@ -199,6 +202,57 @@
         sys.stderr.write('Writing metadata failed for tags: %s in: %s\n' % (tags, image_name))


+def get_image_url(self, image_url, offset):
+   """Saves an image if not saved yet. Returns the new URL or
+   the original URL in case of download errors."""
+
+   def _url(fn):
+       return u'%s%s/%s' % (save_dir, image_dir, fn)
+
+   def _addexif(fn):
+       if options.exif and fn.endswith('.jpg'):
+           add_exif(fn, set(self.tags))
+
+   # determine the image file name
+   offset = '_' + offset if offset else ''
+   if options.image_names == 'i':
+       image_filename = self.ident + offset
+   elif options.image_names == 'bi':
+       image_filename = account + '_' + self.ident + offset
+   else:
+       image_filename = image_url.split('/')[-1]
+   glob_filter = '' if '.' in image_filename else '.*'
+   # check if a file with this name already exists
+   image_glob = glob(join(image_folder, image_filename + glob_filter))
+   if image_glob:
+       _addexif(image_glob[0])
+   # download the image data
+   try:
+       image_response = urllib2.urlopen(image_url)
+   except urllib2.HTTPError:
+       # return the original URL
+       return image_url
+   try:
+       image_data = image_response.read()
+   except urllib2.HTTPError:
+       # return the original URL
+       return image_url
+   image_response.close()
+   # determine the file type if it's unknown
+   if '.' not in image_filename:
+       image_type = imghdr.what(None, image_data[:32])
+       if image_type:
+           image_filename += '.' + image_type.replace('jpeg', 'jpg')
+   # save the image
+   with open_image(image_dir, image_filename) as image_file:
+       image_file.write(image_data)
+   _addexif(join(image_folder, image_filename))
+
+
+def add_to_pool(self, image_url, offset):
+    imagePool[image_url] = [self, offset];
+
+
 def save_style():
     with open_text(backup_css) as css:
         css.write('''\
@@ -396,7 +450,16 @@
                 self.post_count += 1
             return True

-        # Get the XML entries from the API, which we can only do for max 50 posts at once.
+        poolThreads = []
+        quitEvent = Event()
+        for t in range(0,7):
+            ts = SavePool(quitEvent)
+            ts.daemon = True
+            poolThreads.append(ts)
+        for j in poolThreads:
+            j.start()
+
+       # Get the XML entries from the API, which we can only do for max 50 posts at once.
         # Posts "arrive" in reverse chronological order. Post #0 is the most recent one.
         i = options.skip
         while i < last_post:
@@ -426,6 +489,9 @@

         log(account, "%d posts backed up\n" % self.post_count)
         self.total_count += self.post_count
+        quitEvent.set()
+        while (threading.activeCount() > 1):
+            time.sleep(1)


 class TumblrPost:
@@ -481,7 +547,8 @@
             url = escape(get_try('photo-link-url'))
             for p in post.photoset['photo':] if hasattr(post, 'photoset') else [post]:
                 src = unicode(p['photo-url'])
-                append(escape(self.get_image_url(src, p().get('offset'))), u'<img alt="" src="%s">')
+                add_to_pool(self, src, p().get('offset'))
+                append(escape(self.get_image_filename(src, p().get('offset'))), u'<img alt="" src="%s">')
                 if url:
                     content[-1] = u'<a href="%s">%s</a>' % (url, content[-1])
                 content[-1] = '<p>' + content[-1] + '</p>'
@@ -541,18 +608,7 @@
         for p in ('<p>(<(%s)>)', '(</(%s)>)</p>'):
             self.content = re.sub(p % 'p|ol|iframe[^>]*', r'\1', self.content)

-    def get_image_url(self, image_url, offset):
-        """Saves an image if not saved yet. Returns the new URL or
-        the original URL in case of download errors."""
-
-        def _url(fn):
-            return u'%s%s/%s' % (save_dir, image_dir, fn)
-
-        def _addexif(fn):
-            if options.exif and fn.endswith('.jpg'):
-                add_exif(fn, set(self.tags))
-
-        # determine the image file name
+    def get_image_filename(self, image_url, offset):
         offset = '_' + offset if offset else ''
         if options.image_names == 'i':
             image_filename = self.ident + offset
@@ -560,30 +616,8 @@
             image_filename = account + '_' + self.ident + offset
         else:
             image_filename = image_url.split('/')[-1]
-        glob_filter = '' if '.' in image_filename else '.*'
-        # check if a file with this name already exists
-        image_glob = glob(join(image_folder, image_filename + glob_filter))
-        if image_glob:
-            _addexif(image_glob[0])
-            return _url(split(image_glob[0])[1])
-        # download the image data
-        try:
-            image_response = urllib2.urlopen(image_url)
-        except urllib2.HTTPError:
-            # return the original URL
-            return image_url
-        image_data = image_response.read()
-        image_response.close()
-        # determine the file type if it's unknown
-        if '.' not in image_filename:
-            image_type = imghdr.what(None, image_data[:32])
-            if image_type:
-                image_filename += '.' + image_type.replace('jpeg', 'jpg')
-        # save the image
-        with open_image(image_dir, image_filename) as image_file:
-            image_file.write(image_data)
-        _addexif(join(image_folder, image_filename))
-        return _url(image_filename)
+        return u'%s%s/%s' % (save_dir, image_dir, image_filename + image_url[-4:])
+

     def get_post(self):
         """returns this post in HTML"""
@@ -621,6 +655,21 @@
                 f.write(self.xml_content)


+class SavePool(threading.Thread):
+    def __init__(self, quit):
+        threading.Thread.__init__(self)
+        self.quit = quit
+    def run(self):
+        imagecounter = 0
+        while not self.quit.isSet() or imagePool:
+            if imagePool:
+                key, value = imagePool.popitem()
+                get_image_url(value[0],key,value[1])
+                log(account, "%d images remaining to save\r" % (len(imagePool)))
+                imagecounter += 1 
+        log(account, "%d images backed up\n" % imagecounter)
+
+
 class BlosxomPost(TumblrPost):

     def get_image_url(self, image_url, offset):

--- tumblr_backup.py.orig   2014-10-07 08:59:34.413053500 +0200
+++ tumblr_backup_fixextension.py   2014-10-07 10:03:36.039336300 +0200
@@ -587,7 +587,7 @@
             image_filename = account + '_' + self.ident + offset
         else:
             image_filename = image_url.split('/')[-1]
-        glob_filter = '' if '.' in image_filename else '.*'
+        glob_filter = '' if '.' in image_filename[-4] else '.*'
         # check if a file with this name already exists
         image_glob = glob(join(self.image_folder, image_filename + glob_filter))
         if image_glob:
@@ -602,7 +602,7 @@
             # return the original URL
             return image_url
         # determine the file type if it's unknown
-        if '.' not in image_filename:
+        if '.' not in image_filename[-4]:
             image_type = imghdr.what(None, image_data[:32])
             if image_type:
                 image_filename += '.' + image_type.replace('jpeg', 'jpg')

            post += u'<h2>%s</h2>\n' % self.title

--- tumblr_backup.py.orig   2014-10-05 11:55:41.019549600 +0200
+++ tumblr_backup_poollimit.py  2014-10-06 10:37:31.151061500 +0200
@@ -676,7 +676,7 @@
 class ThreadPool:

     def __init__(self, count=20):
-        self.queue = Queue.Queue()
+        self.queue = Queue.Queue(1000)
         self.quit = threading.Event()
         self.threads = [threading.Thread(target=self.handler) for _ in range(count)]
         for t in self.threads:
@@ -698,6 +698,8 @@
                 if self.quit.is_set():
                     break
             else:
+                if (self.quit.is_set()):
+                    log(account, "%d remaining posts to save\r" % self.queue.qsize())
                 work()
                 self.queue.task_done()

--- tumblr_backup_apiv2.py.orig 2014-10-08 16:05:11.940811700 +0200
+++ tumblr_backup_apiv2_request.py  2014-10-08 19:06:12.277448300 +0200
@@ -74,6 +74,7 @@
     'text', 'quote', 'link', 'answer', 'video', 'audio', 'photo', 'chat'
 )
 POST_TYPES_SET = frozenset(POST_TYPES)
+POST_ANY_TYPES_SET = frozenset(POST_TYPES +('any',))

 MAX_POSTS = 50

@@ -454,6 +455,20 @@
                         continue
                     if post.date < options.p_start:
                         return False
+                if options.request:
+                    if ((post.typ in options.request) or ('any' in options.request)):
+                        if post.typ in options.request:
+                            if ((len(options.request[post.typ])) and (not set(options.request[post.typ]) & post.tags_lower)):
+                                if 'any' in options.request:
+                                    if ((len(options.request['any'])) and (not set(options.request['any']) & post.tags_lower)):
+                                        continue
+                                else:
+                                    continue
+                        else:
+                            if ((len(options.request['any'])) and (not set(options.request['any']) & post.tags_lower)):
+                                continue
+                    else:
+                        continue
                 if options.tags and not options.tags & post.tags_lower:
                     continue
                 if options.type and post.typ not in options.type:
@@ -514,7 +529,7 @@
         self.tm = time.localtime(self.date)
         self.title = ''
         self.tags = post['tags']
-        if options.tags:
+        if options.tags or options.request:
             self.tags_lower = set(t.lower() for t in self.tags)
         self.file_name = join(self.ident, dir_index) if options.dirs else self.ident + post_ext
         self.llink = self.ident if options.dirs else self.file_name
@@ -776,7 +791,17 @@
         if not types <= POST_TYPES_SET:
             parser.error("--type: invalid post types")
         setattr(parser.values, option.dest, types)
-
+    def request_callback(option, opt, value, parser):
+        raw_request = value.lower().split(';')
+        request = {}
+        for elt in raw_request:
+            if ':' in elt:
+                request.setdefault(elt.split(':')[0], elt.split(':')[1].split(','))
+            else:
+                request.setdefault(elt, '')
+        if not set(request.keys()) <= POST_ANY_TYPES_SET:
+            parser.error("--request: invalid post types")
+        setattr(parser.values, option.dest, request)
     parser = optparse.OptionParser("Usage: %prog [options] blog-name ...",
         description="Makes a local backup of Tumblr blogs."
     )
@@ -825,6 +850,10 @@
     )
     parser.add_option('-P', '--private', help="password for a private tumblr",
         metavar='PASSWORD'
+    )    
+    parser.add_option('-Q', '--request', type='string', action='callback',
+        callback=request_callback, help="Complex backup request. TYPE:TAGS;TYPE2:TAG,TAG;TYPE"
+        " case-insensitive)"
     )
     parser.add_option('-t', '--tags', type='string', action='callback',
         callback=tags_callback, help="save only posts tagged TAGS (comma-separated values;"

        # find the post number limit to back up
        last_post = options.count + options.skip if options.count else blog['posts']

        # find the post number limit to back up
        last_post = options.count + options.skip if (options.count and options.count  + options.skip < blog['posts']) else blog['posts']

--- tumblr_backup.py.orig   2014-10-01 09:57:32.142436800 +0200
+++ tumblr_backup_skip.py   2014-10-01 10:24:32.788979900 +0200
@@ -481,7 +481,10 @@
             url = escape(get_try('photo-link-url'))
             for p in post.photoset['photo':] if hasattr(post, 'photoset') else [post]:
                 src = unicode(p['photo-url'])
-                append(escape(self.get_image_url(src, p().get('offset'))), u'<img alt="" src="%s">')
+                if options.skip_images:
+                    append(u'<img alt="" src="%s">' % (escape(src)))
+                else:
+                    append(escape(self.get_image_url(src, p().get('offset'))), u'<img alt="" src="%s">')
                 if url:
                     content[-1] = u'<a href="%s">%s</a>' % (url, content[-1])
                 content[-1] = '<p>' + content[-1] + '</p>'
@@ -689,6 +692,9 @@
     parser.add_option('-i', '--incremental', action='store_true',
         help="incremental backup mode"
     )
+    parser.add_option('-k', '--skip-images', action='store_true',
+        help="Do not save images"
+    )
     parser.add_option('-x', '--xml', action='store_true',
         help="save the original XML source"
     )

--- tumblr_backup.py.orig   2014-10-05 11:55:41.019549600 +0200
+++ tumblr_backup_errorfix.py   2014-10-06 10:52:58.267506700 +0200
@@ -18,6 +18,7 @@
 import time
 import urllib
 import urllib2
+import socket
 from xml.sax import SAXException
 from xml.sax.saxutils import escape

@@ -165,15 +166,19 @@
     url = base + '?' + urllib.urlencode(params)
     for _ in range(10):
         try:
-            resp = urllib2.urlopen(url)
-        except (urllib2.URLError, urllib2.HTTPError) as e:
+            resp = urllib2.urlopen(url, None, 30)
+        except (urllib2.URLError, urllib2.HTTPError, socket.error) as e:
             sys.stderr.write('%s getting %s\n' % (e, url))
             continue
         if resp.info().gettype() == 'text/xml':
             break
     else:
         return None
-    xml = resp.read()
+    try:
+        xml = resp.read()
+    except (urllib2.URLError, urllib2.HTTPError, socket.error) as e:
+        sys.stderr.write('%s getting %s\n' % (e, url))
+        return None
     try:
         doc = xmltramp.parse(xml)
     except SAXException as e:
@@ -409,11 +414,15 @@
         while i < last_post:
             # find the upper bound
             j = min(i + MAX_POSTS, last_post)
-            log(account, "Getting posts %d to %d of %d\r" % (i, j - 1, last_post))
-
-            soup = xmlparse(base, j - i, i)
+            for e in range(3):
+                log(account, "Getting posts %d to %d of %d%s\r" % (i, j - 1, last_post, '' if not e else ', retry ' + str(e)))
+                soup = xmlparse(base, j - i, i)
+                if soup is None:
+                    continue
+                else:
+                    break
             if soup is None:
-                i += 50         # try the next batch
+                i += MAX_POSTS         # try the next batch
                 continue

             posts = soup.posts['post':]
@@ -582,10 +591,10 @@
             return _url(split(image_glob[0])[1])
         # download the image data
         try:
-            image_response = urllib2.urlopen(image_url)
+            image_response = urllib2.urlopen(image_url, None, 30)
             image_data = image_response.read()
             image_response.close()
-        except urllib2.HTTPError:
+        except (urllib2.URLError, urllib2.HTTPError, socket.error):
             # return the original URL
             return image_url
         # determine the file type if it's unknown