Mercurial > public > sg101

comparison news/management/commands/fix_news_embeds.py @ 1075:5bba39fafad8

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Added mgmt command to fix news story video embeds. These embeds stopped working when we went to https.
author Brian Neal <bgneal@gmail.com>
date Sat, 16 Apr 2016 00:31:06 -0500
parents
children b24708086bfc
comparison
equal deleted inserted replaced
1074:650ab160cbb9 1075:5bba39fafad8
1 """
2 Custom management command to find and fix the old YouTube <object> embeds in the
3 news stories.
4 """
5 from optparse import make_option
6 import urlparse
7 from urllib2 import HTTPError
8
9 from django.core.management.base import NoArgsCommand
10 import lxml.html
11
12 from news.models import Story
13 from oembed.core import get_oembed
14
15
16 class Command(NoArgsCommand):
17 help = "Rewrite news stories that have old flash YouTube embeds"
18 option_list = NoArgsCommand.option_list + (
19 make_option('-p', '--preview',
20 action='store_true',
21 default=False,
22 help="find and print old embeds but don't update"),
23 )
24
25 def handle_noargs(self, **options):
26 self.preview = options['preview']
27
28 qs = Story.objects.all()
29 for story in qs.iterator():
30 self._process_story(story)
31
32 def _process_story(self, story):
33 r1 = self._process_html(story, 'short_text')
34 r2 = self._process_html(story, 'long_text')
35 r3 = self._process_html(story, 'admin_content')
36 if not self.preview and (r1 or r2 or r3):
37 print "Updating", story.title
38 story.save()
39
40 def _process_html(self, story, field):
41 html = getattr(story, field)
42 s = html.strip()
43 if not s:
44 return False
45
46 root = lxml.html.fragment_fromstring(s, create_parent=True)
47 for obj in root.iter('object'):
48 if story.version != 0:
49 print "*" * 5, story.title, "bad version!"
50 continue
51 for param in obj.iter('param'):
52 value = param.get('value')
53 if value and value.startswith('http'):
54 r = urlparse.urlparse(value)
55 if (r.hostname != 'www.youtube.com' and
56 r.hostname != 'www.youtube-nocookie.com'):
57 print "Unknown source hostname:", r.hostname, ";", story.title
58 continue
59 try:
60 new_embed = self._process_path(story, r.path)
61 except HTTPError as ex:
62 print "*" * 5, story.title, ex
63 continue
64 if not new_embed:
65 continue
66 parent = obj.getparent()
67 new_child = lxml.html.fragment_fromstring(new_embed)
68 parent.replace(obj, new_child)
69 new_html = lxml.html.tostring(root)[5:-6]
70 if self.preview:
71 print story.title
72 print new_html, "\n" * 3
73 return False
74
75 setattr(story, field, new_html)
76 return True
77
78
79 def _process_path(self, story, path):
80 if path.startswith('/v/'):
81 video_id = strip_query(path[3:])
82 url = "https://www.youtube.com/watch?v={}".format(video_id)
83 elif path.startswith('/p/'):
84 playlist_id = strip_query(path[3:])
85 url = "https://www.youtube.com/playlist?list=PL{}".format(playlist_id)
86 else:
87 print "Unknown YouTube path:", path, ";", story.title
88 return None
89
90 oembed = get_oembed("http://www.youtube.com/oembed", url, scheme='https')
91 return oembed['html']
92
93
94 def strip_query(path):
95 n = path.find('&')
96 if n != -1:
97 return path[:n]
98 return path