1 # This file is part of FNP-Redakcja, licensed under GNU Affero GPLv3 or later.
2 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
5 from io import StringIO
8 from urllib.request import FancyURLopener
9 from django.conf import settings
11 from wikidata.client import Client
12 from catalogue.constants import WIKIDATA
15 class URLOpener(FancyURLopener):
21 class FlickrError(Exception):
25 def get_flickr_data(url):
26 m = re.match(r'(https?://)?(www\.|secure\.)?flickr\.com/photos/(?P<author>[^/]+)/(?P<img>\d+)/?', url)
28 raise FlickrError("It doesn't look like Flickr URL.")
29 author_slug, img_id = m.group('author'), m.group('img')
30 base_url = "https://www.flickr.com/photos/%s/%s/" % (author_slug, img_id)
32 html = URLOpener().open(url).read().decode('utf-8')
34 raise FlickrError('Error reading page')
35 match = re.search(r'<a href="([^"]*)"[^>]* rel="license ', html)
37 raise FlickrError('License not found.')
39 license_url = match.group(1)
40 re_license = re.compile(r'https?://creativecommons.org/licenses/([^/]*)/([^/]*)/.*')
41 m = re_license.match(license_url)
43 re_pd = re.compile(r'https?://creativecommons.org/publicdomain/([^/]*)/([^/]*)/.*')
44 m = re_pd.match(license_url)
46 raise FlickrError('License does not look like CC: %s' % license_url)
47 if m.group(1).lower() == 'zero':
48 license_name = 'Public domain (CC0 %s)' % m.group(2)
50 license_name = 'Public domain'
52 license_name = 'CC %s %s' % (m.group(1).upper(), m.group(2))
53 m = re.search(r'<a[^>]* class="owner-name [^>]*>([^<]*)<', html)
55 author = "%s@Flickr" % m.group(1)
57 raise FlickrError('Error reading author name.')
58 m = re.search(r'<h1[^>]*>(.*?)</h1>', html, re.S)
60 raise FlickrError('Error reading image title.')
61 title = m.group(1).strip()
62 m = re.search(r'modelExport: (\{.*\})', html)
65 download_url = 'https:' + json.loads(m.group(1))['main']['photo-models'][0]['sizes']['o']['url']
66 except (AssertionError, ValueError, IndexError, KeyError):
67 raise FlickrError('Error reading image URL.')
69 'source_url': base_url,
70 'license_url': license_url,
71 'license_name': license_name,
74 'download_url': download_url,
78 def get_wikimedia_data(url):
79 file_name = url.rsplit('/', 1)[-1].rsplit(':', 1)[-1]
80 d = json.loads(URLOpener().open('https://commons.wikimedia.org/w/api.php?action=query&titles=File:{}&prop=imageinfo&iiprop=url|user|extmetadata&iimetadataversion=latest&format=json'.format(file_name)).read().decode('utf-8'))
82 d = list(d['query']['pages'].values())[0]['imageinfo'][0]
83 ext = d['extmetadata']
86 'title': ext['ObjectName']['value'],
88 'source_url': d['descriptionurl'],
89 'download_url': d['url'],
90 'license_url': ext.get('LicenseUrl', {}).get('value', ''),
91 'license_name': ext['LicenseShortName']['value'],
94 # There are Wikidata links in ObjectName sametimes. Let's use it.
95 wikidata_match = re.search(r'wikidata\.org/wiki/(Q\d+)', meta['title'])
96 if wikidata_match is not None:
97 qitem = wikidata_match.group(1)
99 entity = client.get(qitem)
100 meta['title'] = entity.label.get('pl', str(entity.label))
101 author = entity.get(client.get(WIKIDATA.CREATOR))
102 if author is not None:
103 meta['author'] = author.label.get('pl', str(author.label))
107 if meta['license_name'] == 'Public domain':
108 meta['license_name'] = 'domena publiczna'
109 meta['license_url'] = 'https://pl.wikipedia.org/wiki/Domena_publiczna'
115 def get_mnw_data(url):
116 nr = url.rsplit('/', 1)[-1]
121 'https://cyfrowe-api.mnw.art.pl/api/object/{}/csv'.format(nr)
122 ).read().decode('utf-8')
129 while f'twórca/wytwórnia {i}' in d:
130 authors.append(d[f'twórca/wytwórnia {i}'])
134 license_name = d['klasyfikacja praw autorskich 1']
135 if license_name == 'DOMENA PUBLICZNA':
136 license_name = 'domena publiczna'
137 license_url = 'https://pl.wikipedia.org/wiki/Domena_publiczna'
140 'title': d['nazwa/tytuł'],
141 'author': ', '.join(authors),
143 'download_url': 'https://cyfrowe-cdn.mnw.art.pl/upload/multimedia/{}.{}'.format(
144 d['ścieżka wizerunku'],
145 d['rozszerzenie pliku wizerunku'],
147 'license_url': license_url,
148 'license_name': license_name,
151 def get_rawpixel_data(url):
152 photo_id = re.search(r'/(\d+)/', url).group(1)
154 s = requests.Session()
155 cookies = settings.RAWPIXEL_SESSION
158 'https://www.rawpixel.com/api/v1/user/session',
162 h = {'X-CSRF-Token': token, 'Accept': 'application/json'}
165 f'https://www.rawpixel.com/api/v1/image/data/{photo_id}',
166 headers=h, cookies=cookies).json()
167 download_url = s.post(
168 f'https://www.rawpixel.com/api/v1/image/download/{photo_id}/original',
169 headers=h, cookies=cookies
170 ).json()['downloadUrl']
172 title = data['metadata']['title'].rsplit('|', 1)[0].strip()
176 'author': ', '.join(data['metadata']['artist_names']),
177 'source_url': data['url'],
178 'download_url': download_url,
179 'license_url': data['metadata']['licenseUrl'],
180 'license_name': data['metadata']['license'],
184 def get_import_data(url):
185 if re.match(r'(https?://)?(www\.|secure\.)?flickr\.com/', url):
186 return get_flickr_data(url)
187 if re.match(r'(https?://)?(commons|upload)\.wikimedia\.org/', url):
188 return get_wikimedia_data(url)
189 if re.match(r'(https?://)?cyfrowe\.mnw\.art\.pl/', url):
190 return get_mnw_data(url)
191 if re.match(r'(https?://)?www\.rawpixel\.com/', url):
192 return get_rawpixel_data(url)