+#!/usr/bin/python
#
# ljdump.py - livejournal archiver
# Greg Hewgill <greg@hewgill.com> http://hewgill.com
-# Version 1.0
-#
-# $Id$
-#
-# This program reads the journal entries from a livejournal (or compatible)
-# blog site and archives them in a subdirectory named after the journal name.
-#
-# The configuration is read from "ljdump.config". A sample configuration is
-# provided in "ljdump.config.sample", which should be copied and then edited.
-# The configuration settings are:
-#
-# server - The XMLRPC server URL. This should only need to be changed
-# if you are dumping a journal that is livejournal-compatible
-# but is not livejournal itself.
-#
-# username - The livejournal user name. A subdirectory will be created
-# with this same name to store the journal entries.
-#
-# password - The account password. This password is never sent in the
-# clear; the livejournal "challenge" password mechanism is used.
-#
-# This program may be run as often as needed to bring the backup copy up
-# to date. Only new items are downloaded.
+# Version 1.3.2
#
# LICENSE
#
# misrepresented as being the original software.
# 3. This notice may not be removed or altered from any source distribution.
#
-# Copyright (c) 2005 Greg Hewgill
+# Copyright (c) 2005-2009 Greg Hewgill
-import codecs, md5, os, pprint, sys, xml.dom.minidom, xmlrpclib
+import codecs, md5, os, pickle, pprint, re, shutil, sys, urllib2, xml.dom.minidom, xmlrpclib
from xml.sax import saxutils
-def dochallenge(params, password):
+MimeExtensions = {
+ "image/gif": ".gif",
+ "image/jpeg": ".jpg",
+ "image/png": ".png",
+}
+
+def calcchallenge(challenge, password):
+ return md5.new(challenge+md5.new(password).hexdigest()).hexdigest()
+
+def flatresponse(response):
+ r = {}
+ while True:
+ name = response.readline()
+ if len(name) == 0:
+ break
+ if name[-1] == '\n':
+ name = name[:len(name)-1]
+ value = response.readline()
+ if value[-1] == '\n':
+ value = value[:len(value)-1]
+ r[name] = value
+ return r
+
+def getljsession(server, username, password):
+ r = urllib2.urlopen(server+"/interface/flat", "mode=getchallenge")
+ response = flatresponse(r)
+ r.close()
+ r = urllib2.urlopen(server+"/interface/flat", "mode=sessiongenerate&user=%s&auth_method=challenge&auth_challenge=%s&auth_response=%s" % (username, response['challenge'], calcchallenge(response['challenge'], password)))
+ response = flatresponse(r)
+ r.close()
+ return response['ljsession']
+
+def dochallenge(server, params, password):
challenge = server.LJ.XMLRPC.getchallenge()
params.update({
'auth_method': "challenge",
'auth_challenge': challenge['challenge'],
- 'auth_response': md5.new(challenge['challenge']+md5.new(password).hexdigest()).hexdigest()
+ 'auth_response': calcchallenge(challenge['challenge'], password)
})
return params
if isinstance(e[k], {}.__class__):
dumpelement(f, k, e[k])
else:
- s = unicode(str(e[k]), "UTF-8")
+ try:
+ s = unicode(str(e[k]), "UTF-8")
+ except UnicodeDecodeError:
+ # fall back to Latin-1 for old entries that aren't UTF-8
+ s = unicode(str(e[k]), "cp1252")
f.write("<%s>%s</%s>\n" % (k, saxutils.escape(s), k))
f.write("</%s>\n" % name)
dumpelement(f, "event", event)
f.close()
-config = xml.dom.minidom.parse("ljdump.config")
-Server = config.documentElement.getElementsByTagName("server")[0].childNodes[0].data
-Username = config.documentElement.getElementsByTagName("username")[0].childNodes[0].data
-Password = config.documentElement.getElementsByTagName("password")[0].childNodes[0].data
+def writelast(username, lastsync, lastmaxid):
+ f = open("%s/.last" % username, "w")
+ f.write("%s\n" % lastsync)
+ f.write("%s\n" % lastmaxid)
+ f.close()
+
+def createxml(doc, name, map):
+ e = doc.createElement(name)
+ for k in map.keys():
+ me = doc.createElement(k)
+ me.appendChild(doc.createTextNode(map[k]))
+ e.appendChild(me)
+ return e
+
+def gettext(e):
+ if len(e) == 0:
+ return ""
+ return e[0].firstChild.nodeValue
-print "Fetching journal entries for: %s" % Username
-try:
- os.mkdir(Username)
- print "Created subdirectory: %s" % Username
-except:
- pass
+def ljdump(Server, Username, Password):
+ m = re.search("(.*)/interface/xmlrpc", Server)
+ if m:
+ Server = m.group(1)
-server = xmlrpclib.ServerProxy(Server)
+ print "Fetching journal entries for: %s" % Username
+ try:
+ os.mkdir(Username)
+ print "Created subdirectory: %s" % Username
+ except:
+ pass
-total = 0
-fetched = 0
-errors = 0
+ ljsession = getljsession(Server, Username, Password)
+
+ server = xmlrpclib.ServerProxy(Server+"/interface/xmlrpc")
+
+ newentries = 0
+ newcomments = 0
+ errors = 0
+
+ lastsync = ""
+ lastmaxid = 0
+ try:
+ f = open("%s/.last" % Username, "r")
+ lastsync = f.readline()
+ if lastsync[-1] == '\n':
+ lastsync = lastsync[:len(lastsync)-1]
+ lastmaxid = f.readline()
+ if len(lastmaxid) > 0 and lastmaxid[-1] == '\n':
+ lastmaxid = lastmaxid[:len(lastmaxid)-1]
+ if lastmaxid == "":
+ lastmaxid = 0
+ else:
+ lastmaxid = int(lastmaxid)
+ f.close()
+ except:
+ pass
+ origlastsync = lastsync
-last = ""
-while True:
- r = server.LJ.XMLRPC.syncitems(dochallenge({
+ r = server.LJ.XMLRPC.login(dochallenge(server, {
'username': Username,
'ver': 1,
- 'lastsync': last,
+ 'getpickws': 1,
+ 'getpickwurls': 1,
}, Password))
- #pprint.pprint(r)
- if len(r['syncitems']) == 0:
- break
- for item in r['syncitems']:
- if item['item'][0] == 'L':
- fn = "%s/%s" % (Username, item['item'])
- if not os.access(fn, os.F_OK):
- print "Fetching journal entry %s" % item['item']
+ userpics = dict(zip(map(str, r['pickws']), r['pickwurls']))
+ userpics['*'] = r['defaultpicurl']
+
+ while True:
+ r = server.LJ.XMLRPC.syncitems(dochallenge(server, {
+ 'username': Username,
+ 'ver': 1,
+ 'lastsync': lastsync,
+ }, Password))
+ #pprint.pprint(r)
+ if len(r['syncitems']) == 0:
+ break
+ for item in r['syncitems']:
+ if item['item'][0] == 'L':
+ print "Fetching journal entry %s (%s)" % (item['item'], item['action'])
try:
- e = server.LJ.XMLRPC.getevents(dochallenge({
+ e = server.LJ.XMLRPC.getevents(dochallenge(server, {
'username': Username,
'ver': 1,
'selecttype': "one",
'itemid': item['item'][2:],
}, Password))
- writedump(fn, e['events'][0])
- fetched += 1
+ if e['events']:
+ writedump("%s/%s" % (Username, item['item']), e['events'][0])
+ newentries += 1
+ else:
+ print "Unexpected empty item: %s" % item['item']
+ errors += 1
except xmlrpclib.Fault, x:
print "Error getting item: %s" % item['item']
pprint.pprint(x)
errors += 1
- last = item['time']
- total += 1
-print "%d total entries" % total
-print "%d fetched entries" % fetched
-if errors > 0:
- print "%d errors" % errors
+ lastsync = item['time']
+ writelast(Username, lastsync, lastmaxid)
+
+ # The following code doesn't work because the server rejects our repeated calls.
+ # http://www.livejournal.com/doc/server/ljp.csp.xml-rpc.getevents.html
+ # contains the statement "You should use the syncitems selecttype in
+ # conjuntions [sic] with the syncitems protocol mode", but provides
+ # no other explanation about how these two function calls should
+ # interact. Therefore we just do the above slow one-at-a-time method.
+
+ #while True:
+ # r = server.LJ.XMLRPC.getevents(dochallenge(server, {
+ # 'username': Username,
+ # 'ver': 1,
+ # 'selecttype': "syncitems",
+ # 'lastsync': lastsync,
+ # }, Password))
+ # pprint.pprint(r)
+ # if len(r['events']) == 0:
+ # break
+ # for item in r['events']:
+ # writedump("%s/L-%d" % (Username, item['itemid']), item)
+ # newentries += 1
+ # lastsync = item['eventtime']
+
+ print "Fetching journal comments for: %s" % Username
+
+ try:
+ f = open("%s/comment.meta" % Username)
+ metacache = pickle.load(f)
+ f.close()
+ except:
+ metacache = {}
+
+ try:
+ f = open("%s/user.map" % Username)
+ usermap = pickle.load(f)
+ f.close()
+ except:
+ usermap = {}
+
+ maxid = lastmaxid
+ while True:
+ r = urllib2.urlopen(urllib2.Request(Server+"/export_comments.bml?get=comment_meta&startid=%d" % (maxid+1), headers = {'Cookie': "ljsession="+ljsession}))
+ meta = xml.dom.minidom.parse(r)
+ r.close()
+ for c in meta.getElementsByTagName("comment"):
+ id = int(c.getAttribute("id"))
+ metacache[id] = {
+ 'posterid': c.getAttribute("posterid"),
+ 'state': c.getAttribute("state"),
+ }
+ if id > maxid:
+ maxid = id
+ for u in meta.getElementsByTagName("usermap"):
+ usermap[u.getAttribute("id")] = u.getAttribute("user")
+ if maxid >= int(meta.getElementsByTagName("maxid")[0].firstChild.nodeValue):
+ break
+
+ f = open("%s/comment.meta" % Username, "w")
+ pickle.dump(metacache, f)
+ f.close()
+
+ f = open("%s/user.map" % Username, "w")
+ pickle.dump(usermap, f)
+ f.close()
+
+ print "Fetching userpics for: %s" % Username
+ f = open("%s/userpics.xml" % Username, "w")
+ print >>f, """<?xml version="1.0"?>"""
+ print >>f, "<userpics>"
+ for p in userpics:
+ print >>f, """<userpic keyword="%s" url="%s" />""" % (p, userpics[p])
+ pic = urllib2.urlopen(userpics[p])
+ ext = MimeExtensions.get(pic.info()["Content-Type"], "")
+ picfn = re.sub(r"[\/]", "_", p)
+ try:
+ picfn = codecs.utf_8_decode(picfn)[0]
+ picf = open("%s/%s%s" % (Username, picfn, ext), "wb")
+ except:
+ # for installations where the above utf_8_decode doesn't work
+ picfn = "".join([ord(x) < 128 and x or "?" for x in picfn])
+ picf = open("%s/%s%s" % (Username, picfn, ext), "wb")
+ shutil.copyfileobj(pic, picf)
+ pic.close()
+ picf.close()
+ print >>f, "</userpics>"
+ f.close()
+
+ newmaxid = maxid
+ maxid = lastmaxid
+ while True:
+ r = urllib2.urlopen(urllib2.Request(Server+"/export_comments.bml?get=comment_body&startid=%d" % (maxid+1), headers = {'Cookie': "ljsession="+ljsession}))
+ meta = xml.dom.minidom.parse(r)
+ r.close()
+ for c in meta.getElementsByTagName("comment"):
+ id = int(c.getAttribute("id"))
+ jitemid = c.getAttribute("jitemid")
+ comment = {
+ 'id': str(id),
+ 'parentid': c.getAttribute("parentid"),
+ 'subject': gettext(c.getElementsByTagName("subject")),
+ 'date': gettext(c.getElementsByTagName("date")),
+ 'body': gettext(c.getElementsByTagName("body")),
+ 'state': metacache[id]['state'],
+ }
+ if usermap.has_key(c.getAttribute("posterid")):
+ comment["user"] = usermap[c.getAttribute("posterid")]
+ try:
+ entry = xml.dom.minidom.parse("%s/C-%s" % (Username, jitemid))
+ except:
+ entry = xml.dom.minidom.getDOMImplementation().createDocument(None, "comments", None)
+ found = False
+ for d in entry.getElementsByTagName("comment"):
+ if int(d.getElementsByTagName("id")[0].firstChild.nodeValue) == id:
+ found = True
+ break
+ if found:
+ print "Warning: downloaded duplicate comment id %d in jitemid %s" % (id, jitemid)
+ else:
+ entry.documentElement.appendChild(createxml(entry, "comment", comment))
+ f = codecs.open("%s/C-%s" % (Username, jitemid), "w", "UTF-8")
+ entry.writexml(f)
+ f.close()
+ newcomments += 1
+ if id > maxid:
+ maxid = id
+ if maxid >= newmaxid:
+ break
+
+ lastmaxid = maxid
+
+ writelast(Username, lastsync, lastmaxid)
+
+ if origlastsync:
+ print "%d new entries, %d new comments (since %s)" % (newentries, newcomments, origlastsync)
+ else:
+ print "%d new entries, %d new comments" % (newentries, newcomments)
+ if errors > 0:
+ print "%d errors" % errors
+
+if __name__ == "__main__":
+ if os.access("ljdump.config", os.F_OK):
+ config = xml.dom.minidom.parse("ljdump.config")
+ server = config.documentElement.getElementsByTagName("server")[0].childNodes[0].data
+ username = config.documentElement.getElementsByTagName("username")[0].childNodes[0].data
+ password = config.documentElement.getElementsByTagName("password")[0].childNodes[0].data
+ else:
+ from getpass import getpass
+ print "ljdump - livejournal archiver"
+ print
+ print "Enter your Livejournal username and password."
+ print
+ server = "http://livejournal.com"
+ username = raw_input("Username: ")
+ password = getpass("Password: ")
+ print
+ ljdump(server, username, password)