X-Git-Url: https://git.mdrn.pl/redakcja.git/blobdiff_plain/96cb5dd17c7f23a7c63706cee2b7516840d75407..0cd571520ee3f3a1dfaed2f5340961e39a941ad1:/apps/cover/utils.py

diff --git a/apps/cover/utils.py b/apps/cover/utils.py
index e22fa727..51aee190 100755
--- a/apps/cover/utils.py
+++ b/apps/cover/utils.py
@@ -3,7 +3,10 @@
 # This file is part of FNP-Redakcja, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
+import json
+import re
 from urllib import FancyURLopener
+
 from django.contrib.sites.models import Site
 
 
@@ -11,3 +14,60 @@ class URLOpener(FancyURLopener):
     @property
     def version(self):
         return 'FNP Redakcja (http://%s)' % Site.objects.get_current()
+
+
+class FlickrError(Exception):
+    pass
+
+
+def get_flickr_data(url):
+    m = re.match(r'(https?://)?(www\.|secure\.)?flickr\.com/photos/(?P<author>[^/]+)/(?P<img>\d+)/?', url)
+    if not m:
+        raise FlickrError("It doesn't look like Flickr URL.")
+    author_slug, img_id = m.group('author'), m.group('img')
+    base_url = "https://www.flickr.com/photos/%s/%s/" % (author_slug, img_id)
+    try:
+        html = URLOpener().open(url).read().decode('utf-8')
+    except IOError:
+        raise FlickrError('Error reading page')
+    match = re.search(r'<a href="([^"]*)"[^>]* rel="license ', html)
+    if not match:
+        raise FlickrError('License not found.')
+    else:
+        license_url = match.group(1)
+        re_license = re.compile(r'https?://creativecommons.org/licenses/([^/]*)/([^/]*)/.*')
+        m = re_license.match(license_url)
+        if not m:
+            re_pd = re.compile(r'https?://creativecommons.org/publicdomain/([^/]*)/([^/]*)/.*')
+            m = re_pd.match(license_url)
+            if not m:
+                raise FlickrError('License does not look like CC: %s' % license_url)
+            if m.group(1).lower() == 'zero':
+                license_name = 'Public domain (CC0 %s)' % m.group(2)
+            else:
+                license_name = 'Public domain'
+        else:
+            license_name = 'CC %s %s' % (m.group(1).upper(), m.group(2))
+    m = re.search(r'<a[^>]* class="owner-name [^>]*>([^<]*)<', html)
+    if m:
+        author = "%s@Flickr" % m.group(1)
+    else:
+        raise FlickrError('Error reading author name.')
+    m = re.search(r'<h1[^>]*>(.*?)</h1>', html, re.S)
+    if not m:
+        raise FlickrError('Error reading image title.')
+    title = m.group(1).strip()
+    m = re.search(r'modelExport: (\{.*\})', html)
+    try:
+        assert m
+        download_url = 'https:' + json.loads(m.group(1))['main']['photo-models'][0]['sizes']['o']['url']
+    except (AssertionError, ValueError, IndexError, KeyError):
+        raise FlickrError('Error reading image URL.')
+    return {
+        'source_url': base_url,
+        'license_url': license_url,
+        'license_name': license_name,
+        'author': author,
+        'title': title,
+        'download_url': download_url,
+    }