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[-1].getAttribute("label") == "Original"):
 175                                                 imgurl = sizes[-1].getAttribute("source")
 176                                 else:
 177                                                 print "Failed to get original for photo id " + id
 178
 179
 180         # Free the DOM memory
 181         dom.unlink()
 182
 183         # Grab the image file
 184         response = urllib2.urlopen(imgurl)
 185         data = response.read()
 186
 187         # Save the file!
 188         fh = open(filename, "w")
 189         fh.write(data)
 190         fh.close()
 191
 192         return filename
 193     except:
 194         print "Failed to retrieve photo id " + id
 195
 196 ######## Main Application ##########
 197 if __name__ == '__main__':
 198
 199     # The first, and only argument needs to be a directory
 200     try:
 201         os.chdir(sys.argv[1])
 202     except:
 203         print "usage: %s directory" % sys.argv[0]
 204         sys.exit(1)
 205
 206     # First things first, see if we have a cached user and auth-token
 207     try:
 208         cache = open("touchr.frob.cache", "r")
 209         config = cPickle.load(cache)
 210         cache.close()
 211
 212     # We don't - get a new one
 213     except:
 214         (user, token) = froblogin(getfrob(), "read")
 215         config = { "version":1 , "user":user, "token":token }
 216
 217         # Save it for future use
 218         cache = open("touchr.frob.cache", "w")
 219         cPickle.dump(config, cache)
 220         cache.close()
 221
 222     # Now, construct a query for the list of photo sets
 223     url  = "http://api.flickr.com/services/rest/?method=flickr.photosets.getList"
 224     url += "&user_id=" + config["user"]
 225     url  = flickrsign(url, config["token"])
 226
 227     # get the result
 228     response = urllib2.urlopen(url)
 229
 230     # Parse the XML
 231     dom = xml.dom.minidom.parse(response)
 232
 233     # Get the list of Sets
 234     sets =  dom.getElementsByTagName("photoset")
 235
 236     # For each set - create a url
 237     urls = []
 238     for set in sets:
 239         pid = set.getAttribute("id")
 240         dir = getText(set.getElementsByTagName("title")[0].childNodes)
 241
 242         # Build the list of photos
 243         url   = "http://api.flickr.com/services/rest/?method=flickr.photosets.getPhotos"
 244         url  += "&photoset_id=" + pid
 245
 246         # Append to our list of urls
 247         urls.append( (url , dir) )
 248
 249     # Free the DOM memory
 250     dom.unlink()
 251
 252     # Add the photos which are not in any set
 253     url   = "http://api.flickr.com/services/rest/?method=flickr.photos.getNotInSet"
 254     urls.append( (url, "No Set") )
 255
 256     # Add the user's Favourites
 257     url   = "http://api.flickr.com/services/rest/?method=flickr.favorites.getList"
 258     urls.append( (url, "Favourites") )
 259
 260     # Time to get the photos
 261     inodes = {}
 262     for (url , dir) in urls:
 263         # Create the directory
 264         try:
 265             os.makedirs(dir)
 266         except:
 267             pass
 268
 269         # Get 500 results per page
 270         url += "&per_page=500"
 271         pages = page = 1
 272
 273         while page <= pages:
 274             request = url + "&page=" + str(page)
 275
 276             # Sign the url
 277             request = flickrsign(request, config["token"])
 278
 279             # Make the request
 280             response = urllib2.urlopen(request)
 281
 282             # Parse the XML
 283             dom = xml.dom.minidom.parse(response)
 284
 285             # Get the total
 286             pages = int(dom.getElementsByTagName("photo")[0].parentNode.getAttribute("pages"))
 287
 288             # Grab the photos
 289             for photo in dom.getElementsByTagName("photo"):
 290                 # Tell the user we're grabbing the file
 291                 print photo.getAttribute("title").encode("utf8") + " ... in set ... " + dir
 292
 293                 # Grab the id
 294                 photoid = photo.getAttribute("id")
 295
 296                 # The target
 297                 target = dir + "/" + photoid + ".jpg"
 298
 299                 # Skip files that exist
 300                 if os.access(target, os.R_OK):
 301                     continue
 302
 303                 # Look it up in our dictionary of inodes first
 304                 if inodes.has_key(photoid) and os.access(inodes[photoid], os.R_OK):
 305                     # woo, we have it already, use a hard-link
 306                     os.link(inodes[photoid], target)
 307                 else:
 308                     inodes[photoid] = getphoto(photo.getAttribute("id"), config["token"], target)
 309
 310             # Move on the next page
 311             page = page + 1