flickrtouchr.py

   1 #!/usr/bin/env python
   2
   3 #
   4 # FlickrTouchr - a simple python script to grab all your photos from flickr,
   5 #                dump into a directory - organised into folders by set -
   6 #                along with any favourites you have saved.
   7 #
   8 #                You can then sync the photos to an iPod touch.
   9 #
  10 # Version:       1.2
  11 #
  12 # Original Author:      colm - AT - allcosts.net  - Colm MacCarthaigh - 2008-01-21
  13 #
  14 # Modified by:                  Dan Benjamin - http://hivelogic.com
  15 #
  16 # License:                      Apache 2.0 - http://www.apache.org/licenses/LICENSE-2.0.html
  17 #
  18
  19 import xml.dom.minidom
  20 import webbrowser
  21 import urlparse
  22 import urllib2
  23 import cPickle
  24 import md5
  25 import sys
  26 import os
  27
  28 API_KEY       = "e224418b91b4af4e8cdb0564716fa9bd"
  29 SHARED_SECRET = "7cddb9c9716501a0"
  30
  31 #
  32 # Utility functions for dealing with flickr authentication
  33 #
  34 def getText(nodelist):
  35     rc = ""
  36     for node in nodelist:
  37         if node.nodeType == node.TEXT_NODE:
  38             rc = rc + node.data
  39     return rc.encode("utf-8")
  40
  41 #
  42 # Get the frob based on our API_KEY and shared secret
  43 #
  44 def getfrob():
  45     # Create our signing string
  46     string = SHARED_SECRET + "api_key" + API_KEY + "methodflickr.auth.getFrob"
  47     hash   = md5.new(string).digest().encode("hex")
  48
  49     # Formulate the request
  50     url    = "http://api.flickr.com/services/rest/?method=flickr.auth.getFrob"
  51     url   += "&api_key=" + API_KEY + "&api_sig=" + hash
  52
  53     try:
  54         # Make the request and extract the frob
  55         response = urllib2.urlopen(url)
  56
  57         # Parse the XML
  58         dom = xml.dom.minidom.parse(response)
  59
  60         # get the frob
  61         frob = getText(dom.getElementsByTagName("frob")[0].childNodes)
  62
  63         # Free the DOM
  64         dom.unlink()
  65
  66         # Return the frob
  67         return frob
  68
  69     except:
  70         raise "Could not retrieve frob"
  71
  72 #
  73 # Login and get a token
  74 #
  75 def froblogin(frob, perms):
  76     string = SHARED_SECRET + "api_key" + API_KEY + "frob" + frob + "perms" + perms
  77     hash   = md5.new(string).digest().encode("hex")
  78
  79     # Formulate the request
  80     url    = "http://api.flickr.com/services/auth/?"
  81     url   += "api_key=" + API_KEY + "&perms=" + perms
  82     url   += "&frob=" + frob + "&api_sig=" + hash
  83
  84     # Tell the user what's happening
  85     print "In order to allow FlickrTouchr to read your photos and favourites"
  86     print "you need to allow the application. Please press return when you've"
  87     print "granted access at the following url (which should have opened"
  88     print "automatically)."
  89     print
  90     print url
  91     print
  92     print "Waiting for you to press return"
  93
  94     # We now have a login url, open it in a web-browser
  95     webbrowser.open_new(url)
  96
  97     # Wait for input
  98     sys.stdin.readline()
  99
 100     # Now, try and retrieve a token
 101     string = SHARED_SECRET + "api_key" + API_KEY + "frob" + frob + "methodflickr.auth.getToken"
 102     hash   = md5.new(string).digest().encode("hex")
 103
 104     # Formulate the request
 105     url    = "http://api.flickr.com/services/rest/?method=flickr.auth.getToken"
 106     url   += "&api_key=" + API_KEY + "&frob=" + frob
 107     url   += "&api_sig=" + hash
 108
 109     # See if we get a token
 110     try:
 111         # Make the request and extract the frob
 112         response = urllib2.urlopen(url)
 113
 114         # Parse the XML
 115         dom = xml.dom.minidom.parse(response)
 116
 117         # get the token and user-id
 118         token = getText(dom.getElementsByTagName("token")[0].childNodes)
 119         nsid  = dom.getElementsByTagName("user")[0].getAttribute("nsid")
 120
 121         # Free the DOM
 122         dom.unlink()
 123
 124         # Return the token and userid
 125         return (nsid, token)
 126     except:
 127         raise "Login failed"
 128
 129 #
 130 # Sign an arbitrary flickr request with a token
 131 #
 132 def flickrsign(url, token):
 133     query  = urlparse.urlparse(url).query
 134     query += "&api_key=" + API_KEY + "&auth_token=" + token
 135     params = query.split('&')
 136
 137     # Create the string to hash
 138     string = SHARED_SECRET
 139
 140     # Sort the arguments alphabettically
 141     params.sort()
 142     for param in params:
 143         string += param.replace('=', '')
 144     hash   = md5.new(string).digest().encode("hex")
 145
 146     # Now, append the api_key, and the api_sig args
 147     url += "&api_key=" + API_KEY + "&auth_token=" + token + "&api_sig=" + hash
 148
 149     # Return the signed url
 150     return url
 151
 152 #
 153 # Grab the photo from the server
 154 #
 155 def getphoto(id, token, filename):
 156     try:
 157         # Contruct a request to find the sizes
 158         url  = "http://api.flickr.com/services/rest/?method=flickr.photos.getSizes"
 159         url += "&photo_id=" + id
 160
 161         # Sign the request
 162         url = flickrsign(url, token)
 163
 164         # Make the request
 165         response = urllib2.urlopen(url)
 166
 167         # Parse the XML
 168         dom = xml.dom.minidom.parse(response)
 169
 170         # Get the list of sizes
 171         sizes =  dom.getElementsByTagName("size")
 172
 173         # Grab the original if it exists
 174         if (sizes[-2].getAttribute("label") == "Original"):
 175             imgurl = sizes[-2].getAttribute("source")
 176         else:
 177             imgurl = sizes[-1].getAttribute("source")
 178
 179         # Free the DOM memory
 180         dom.unlink()
 181
 182         # Grab the image file
 183         response = urllib2.urlopen(imgurl)
 184         data = response.read()
 185
 186         # Save the file!
 187         fh = open(filename, "w")
 188         fh.write(data)
 189         fh.close()
 190
 191         return filename
 192     except:
 193         print "Failed to retrieve photo id " + id
 194
 195 ######## Main Application ##########
 196 if __name__ == '__main__':
 197
 198     # The first, and only argument needs to be a directory
 199     try:
 200         os.chdir(sys.argv[1])
 201     except:
 202         print "usage: %s directory" % sys.argv[0]
 203         sys.exit(1)
 204
 205     # First things first, see if we have a cached user and auth-token
 206     try:
 207         cache = open("touchr.frob.cache", "r")
 208         config = cPickle.load(cache)
 209         cache.close()
 210
 211     # We don't - get a new one
 212     except:
 213         (user, token) = froblogin(getfrob(), "read")
 214         config = { "version":1 , "user":user, "token":token }
 215
 216         # Save it for future use
 217         cache = open("touchr.frob.cache", "w")
 218         cPickle.dump(config, cache)
 219         cache.close()
 220
 221     # Now, construct a query for the list of photo sets
 222     url  = "http://api.flickr.com/services/rest/?method=flickr.photosets.getList"
 223     url += "&user_id=" + config["user"]
 224     url  = flickrsign(url, config["token"])
 225
 226     # get the result
 227     response = urllib2.urlopen(url)
 228
 229     # Parse the XML
 230     dom = xml.dom.minidom.parse(response)
 231
 232     # Get the list of Sets
 233     sets =  dom.getElementsByTagName("photoset")
 234
 235     # For each set - create a url
 236     urls = []
 237     for set in sets:
 238         pid = set.getAttribute("id")
 239         dir = getText(set.getElementsByTagName("title")[0].childNodes)
 240
 241         # Build the list of photos
 242         url   = "http://api.flickr.com/services/rest/?method=flickr.photosets.getPhotos"
 243         url  += "&photoset_id=" + pid
 244
 245         # Append to our list of urls
 246         urls.append( (url , dir) )
 247
 248     # Free the DOM memory
 249     dom.unlink()
 250
 251     # Add the photos which are not in any set
 252     url   = "http://api.flickr.com/services/rest/?method=flickr.photos.getNotInSet"
 253     urls.append( (url, "No Set") )
 254
 255     # Add the user's Favourites
 256     url   = "http://api.flickr.com/services/rest/?method=flickr.favorites.getList"
 257     urls.append( (url, "Favourites") )
 258
 259     # Time to get the photos
 260     inodes = {}
 261     for (url , dir) in urls:
 262         # Create the directory
 263         try:
 264             os.makedirs(dir)
 265         except:
 266             pass
 267
 268         # Get 500 results per page
 269         url += "&per_page=500"
 270         pages = page = 1
 271
 272         while page <= pages:
 273             request = url + "&page=" + str(page)
 274
 275             # Sign the url
 276             request = flickrsign(request, config["token"])
 277
 278             # Make the request
 279             response = urllib2.urlopen(request)
 280
 281             # Parse the XML
 282             dom = xml.dom.minidom.parse(response)
 283
 284             # Get the total
 285             pages = int(dom.getElementsByTagName("photo")[0].parentNode.getAttribute("pages"))
 286
 287             # Grab the photos
 288             for photo in dom.getElementsByTagName("photo"):
 289                 # Tell the user we're grabbing the file
 290                 print photo.getAttribute("title").encode("utf8") + " ... in set ... " + dir
 291
 292                 # Grab the id
 293                 photoid = photo.getAttribute("id")
 294
 295                 # The target
 296                 target = dir + "/" + photoid + ".jpg"
 297
 298                 # Skip files that exist
 299                 if os.access(target, os.R_OK):
 300                     continue
 301
 302                 # Look it up in our dictionary of inodes first
 303                 if inodes.has_key(photoid) and os.access(inodes[photoid], os.R_OK):
 304                     # woo, we have it already, use a hard-link
 305                     os.link(inodes[photoid], target)
 306                 else:
 307                     inodes[photoid] = getphoto(photo.getAttribute("id"), config["token"], target)
 308
 309             # Move on the next page
 310             page = page + 1