def update():
flickr = FlickrClient(settings.FLICKR_API_KEY)
# Preload the list of licenses
licenses = licenses = flickr.photos.licenses.getInfo()
licenses = dict((l["id"], smart_unicode(l["url"])) for l in licenses["licenses"]["license"])
# Handle update by pages until we see photos we've already handled
last_update_date = Photo.sync.get_last_update()
page = 1
while True:
log.debug("Fetching page %s of photos", page)
resp = flickr.people.getPublicPhotos(user_id=settings.FLICKR_USER_ID, extras="license,date_taken", per_page="500", page=str(page))
photos = resp["photos"]
if page > photos["pages"]:
log.debug("Ran out of photos; stopping.")
break
for photodict in photos["photo"]:
timestamp = utils.parsedate(str(photodict["datetaken"]))
if timestamp < last_update_date:
log.debug("Hit an old photo (taken %s; last update was %s); stopping.", timestamp, last_update_date)
break
photo_id = utils.safeint(photodict["id"])
license = licenses[photodict["license"]]
secret = smart_unicode(photodict["secret"])
server = smart_unicode(photodict["server"])
_handle_photo(flickr, photo_id, secret, license, timestamp)
page += 1
def _handle_photo(flickr, photo_id, secret, license, timestamp):
info = flickr.photos.getInfo(photo_id=photo_id, secret=secret)["photo"]
photo_id = str(photo_id)
taken_by = smart_unicode(info["owner"]["username"])
url = "http://www.flickr.com/photos/%s/%s/" % (taken_by, photo_id)
title = smart_unicode(info["title"]["_content"])
description = smart_unicode(info["description"]["_content"])
date_uploaded = datetime.datetime.fromtimestamp(utils.safeint(info["dates"]["posted"]))
tags = _convert_tags(info["tags"])
log.debug("Handling photo: %r (taken %s)" % (title, timestamp))
try:
photo = Photo.objects.get(url = url)
photo.title = title
photo.description = description
photo.pub_date = date_uploaded
photo.tags = tags
photo.save()
except Photo.DoesNotExist:
photo = Photo.objects.create(
title = title,
url = url,
description = description,
pub_date = date_uploaded,
tags = tags
)
print "Added %s" % photo
def _tags_for_url(self, url):
tags = set()
xml = utils.getxml(url)
for t in xml.getiterator("tag"):
count = utils.safeint(t.find("count").text)
if count >= self.tag_usage_threshold:
tag = slugify(smart_unicode(t.find("name").text))
tags.add(tag[:50])
return tags
tags.update(self._tags_for_url(url))
return tags
def _tags_for_url(self, url):
tags = set()
try:
xml = utils.getxml(url)
except HttpLib2Error, e:
if e.code == 408:
return ""
else:
raise
except SyntaxError:
return ""
for t in xml.getiterator("tag"):
count = utils.safeint(t.find("count").text)
if count >= self.tag_usage_threshold:
tag = slugify(smart_unicode(t.find("name").text))
tags.add(tag[:50])
return tags
# Memoize tags to avoid unnecessary API calls.
_tag_cache = {}
_tags_for_url = memoize(_tags_for_url, _tag_cache, 1)
def _handle_track(self, artist_name, artist_mbid, track_name, track_mbid,
url, timestamp, tags):
t, created = Track.objects.get_or_create(
artist_name = artist_name,
track_name = track_name,
pub_date = timestamp,