Add import of YouTube session videos using YouTube Data API. Fixes ietf-tools#2249. Commit ready for merge.

 - Legacy-Id: 13485

Author: rpcross

ietf/api/tests.py

@@ -2,6 +2,7 @@
 import sys
 import json
 from importlib import import_module
+from mock import patch
 
 from django.apps import apps
 from django.test import Client
@@ -23,7 +24,8 @@
 )
 
 class CustomApiTestCase(TestCase):
-    def test_notify_meeting_import_audio_files(self):
+    @patch('ietf.secr.proceedings.proc_utils.import_youtube_video_urls')
+    def test_notify_meeting_import_audio_files(self, mock_import):
         meeting = make_meeting_test_data()
         client = Client(Accept='application/json')
         # try invalid method GET

ietf/meeting/forms.py

@@ -80,13 +80,14 @@ def clean(self):
                 if date:
                     dates.append(date)
             if len(dates) < 2:
-                return
+                return self.cleaned_data
             dates.sort()
             last_date = dates[0]
             for date in dates[1:]:
-                if last_date.day + 1 != date.day:
+                if last_date + datetime.timedelta(days=1) != date:
                     raise forms.ValidationError('For Multi-Day meetings, days must be consecutive')
                 last_date = date
+        return self.cleaned_data
 
 class InterimMeetingModelForm(forms.ModelForm):
     group = GroupModelChoiceField(queryset=Group.objects.filter(type__in=('wg', 'rg'), state__in=('active', 'proposed', 'bof')).order_by('acronym'), required=False)

ietf/meeting/tests_views.py

@@ -970,7 +970,6 @@ def test_interim_request_multi_day(self):
                 'session_set-INITIAL_FORMS':0}
 
         r = self.client.post(urlreverse("ietf.meeting.views.interim_request"),data)
-        
         self.assertRedirects(r,urlreverse('ietf.meeting.views.upcoming'))
         meeting = Meeting.objects.order_by('id').last()
         self.assertEqual(meeting.type_id,'interim')

ietf/meeting/views.py

@@ -58,7 +58,8 @@
 from ietf.meeting.helpers import send_interim_announcement_request
 from ietf.meeting.utils import finalize
 from ietf.secr.proceedings.utils import handle_upload_file
-from ietf.secr.proceedings.proc_utils import get_progress_stats, post_process, import_audio_files
+from ietf.secr.proceedings.proc_utils import (get_progress_stats, post_process, import_audio_files,
+    import_youtube_video_urls)
 from ietf.utils import log
 from ietf.utils.mail import send_mail_message
 from ietf.utils.pipe import pipe
@@ -2173,6 +2174,7 @@ def api_import_recordings(request, number):
     if request.method == 'POST':
         meeting = get_meeting(number)
         import_audio_files(meeting)
+        import_youtube_video_urls(meeting)
         return HttpResponse(status=201)
     else:
         return HttpResponse(status=405)

ietf/secr/proceedings/proc_utils.py

@@ -3,16 +3,19 @@
 
 This module contains all the functions for generating static proceedings pages
 '''
-from urllib2 import urlopen
 import datetime
 import glob
+import httplib2
 import os
 import re
 import shutil
 import subprocess
+import urllib2
+from urllib import urlencode
 
 import debug        # pyflakes:ignore
 
+from apiclient.discovery import build
 from django.conf import settings
 from django.core.exceptions import ObjectDoesNotExist
 from django.http import HttpRequest
@@ -34,10 +37,84 @@
 from ietf.utils.mail import send_mail
 
 AUDIO_FILE_RE = re.compile(r'ietf(?P<number>[\d]+)-(?P<room>.*)-(?P<time>[\d]{8}-[\d]{4})')
+VIDEO_TITLE_RE = re.compile(r'IETF(?P<number>\d{2})-(?P<name>.*)-(?P<date>\d{8})-(?P<time>\d{4})')
 
 # -------------------------------------------------
-# Helper Functions
+# Recording Functions
 # -------------------------------------------------
+
+def import_youtube_video_urls(meeting, http=httplib2.Http()):
+    '''Create Document and set external_url for session videos'''
+    youtube = build(settings.YOUTUBE_API_SERVICE_NAME, settings.YOUTUBE_API_VERSION,
+        developerKey=settings.YOUTUBE_API_KEY, http=http)
+    playlistid = get_youtube_playlistid(youtube, 'IETF' + meeting.number)
+    if playlistid is None:
+        return None
+    for video in get_youtube_videos(youtube, playlistid):
+        match = VIDEO_TITLE_RE.match(video['title'])
+        if match:
+            session = _get_session(**match.groupdict())
+            if session:
+                url = video['url']
+                get_or_create_recording_document(url,session)
+
+def get_youtube_playlistid(youtube, title, http=httplib2.Http()):
+    '''Returns the youtube playlistId matching title string, a string'''
+    request = youtube.search().list(
+        q=title,
+        part='id,snippet',
+        channelId=settings.YOUTUBE_IETF_CHANNEL_ID,
+        type='playlist',
+        maxResults=1
+    )
+    search_response = request.execute(http=http)
+
+    try:
+        playlistid = search_response['items'][0]['id']['playlistId']
+    except (KeyError, IndexError):
+        return None
+    return playlistid
+
+def get_youtube_videos(youtube, playlistid, http=httplib2.Http()):
+    '''Returns list of dictionaries with title, urls keys'''
+    videos = []
+    kwargs = dict(part="snippet",playlistId=playlistid,maxResults=50)
+    playlistitems = youtube.playlistItems()
+    request = playlistitems.list(**kwargs)
+    # handle pagination
+    while request is not None:
+        playlistitems_doc = request.execute(http=http)
+        videos.extend(_get_urls_from_json(playlistitems_doc))
+        request = playlistitems.list_next(request, playlistitems_doc)
+    return videos
+
+def _get_session(number,name,date,time):
+    '''Lookup session using data from video title'''
+    meeting = Meeting.objects.get(number=number)
+    schedule = meeting.agenda
+    timeslot_time = datetime.datetime.strptime(date + time,'%Y%m%d%H%M')
+    try:
+        assignment = SchedTimeSessAssignment.objects.get(
+            schedule = schedule,
+            session__group__acronym = name.lower(),
+            timeslot__time = timeslot_time,
+        )
+    except (SchedTimeSessAssignment.DoesNotExist, SchedTimeSessAssignment.MultipleObjectsReturned):
+        return None
+
+    return assignment.session
+
+def _get_urls_from_json(doc):
+    '''Returns list of dictonary titel,url from search results'''
+    urls = []
+    for item in doc['items']:
+        title = item['snippet']['title']
+        #params = dict(v=item['snippet']['resourceId']['videoId'], list=item['snippet']['playlistId'])
+        params = [('v',item['snippet']['resourceId']['videoId']), ('list',item['snippet']['playlistId'])]
+        url = settings.YOUTUBE_BASE_URL + '?' + urlencode(params)
+        urls.append(dict(title=title, url=url))
+    return urls
+
 def import_audio_files(meeting):
     '''
     Checks for audio files and creates corresponding materials (docs) for the Session
@@ -58,20 +135,9 @@ def import_audio_files(meeting):
                 ).exclude(session__agenda_note__icontains='canceled').order_by('timeslot__time')
             if not sessionassignments:
                 continue
-            doc = get_or_create_recording_document(filename,sessionassignments[0].session)
-            for sessionassignment in sessionassignments:
-                session = sessionassignment.session
-                if doc not in session.materials.all():
-                    # add document to session
-                    presentation = SessionPresentation.objects.create(
-                        session=session,
-                        document=doc,
-                        rev=doc.rev)
-                    session.sessionpresentation_set.add(presentation)
-                    if not doc.docalias_set.filter(name__startswith='recording-{}-{}'.format(meeting.number,session.group.acronym)):
-                        sequence = get_next_sequence(session.group,session.meeting,'recording')
-                        name = 'recording-{}-{}-{}'.format(session.meeting.number,session.group.acronym,sequence)
-                        doc.docalias_set.create(name=name)
+            url = settings.IETF_AUDIO_URL + 'ietf{}/{}'.format(meeting.number, filename)
+            doc = get_or_create_recording_document(url,sessionassignments[0].session)
+            attach_recording(doc, [ x.session for x in sessionassignments ])
         else:
             # use for reconciliation email
             unmatched_files.append(filename)
@@ -98,20 +164,30 @@ def get_timeslot_for_filename(filename):
         except (ObjectDoesNotExist, KeyError):
             return None
 
+def attach_recording(doc, sessions):
+    '''Associate recording document with sessions'''
+    for session in sessions:
+        if doc not in session.materials.all():
+            # add document to session
+            presentation = SessionPresentation.objects.create(
+                session=session,
+                document=doc,
+                rev=doc.rev)
+            session.sessionpresentation_set.add(presentation)
+            if not doc.docalias_set.filter(name__startswith='recording-{}-{}'.format(session.meeting.number,session.group.acronym)):
+                sequence = get_next_sequence(session.group,session.meeting,'recording')
+                name = 'recording-{}-{}-{}'.format(session.meeting.number,session.group.acronym,sequence)
+                doc.docalias_set.create(name=name)
+
 def normalize_room_name(name):
     '''Returns room name converted to be used as portion of filename'''
     return name.lower().replace(' ','').replace('/','_')
 
-def get_or_create_recording_document(filename,session):
-    meeting = session.meeting
-    url = settings.IETF_AUDIO_URL + 'ietf{}/{}'.format(meeting.number, filename)
+def get_or_create_recording_document(url,session):
     try:
-        doc = Document.objects.get(external_url=url)
-        return doc
+        return Document.objects.get(external_url=url)
     except ObjectDoesNotExist:
-        pass
-    return create_recording(session,url)
-
+        return create_recording(session,url)
 
 def create_recording(session,url):
     '''
@@ -182,6 +258,10 @@ def mycomp(timeslot):
         key = None
     return key
 
+# -------------------------------------------------
+# End Recording Functions
+# -------------------------------------------------
+
 def get_progress_stats(sdate,edate):
     '''
     This function takes a date range and produces a dictionary of statistics / objects for
@@ -489,7 +569,7 @@ def gen_agenda(context):
 
     # get the text agenda from datatracker
     url = 'https://datatracker.ietf.org/meeting/%s/agenda.txt' % meeting.number
-    text = urlopen(url).read()
+    text = urllib2.urlopen(url).read()
     path = os.path.join(settings.SECR_PROCEEDINGS_DIR,meeting.number,'agenda.txt')
     write_html(path,text)
 

ietf/secr/proceedings/tests.py

@@ -1,24 +1,28 @@
 import debug                            # pyflakes:ignore
+import json
 import os
 import shutil
+from apiclient.discovery import build
+from apiclient.http import HttpMock
+from mock import patch
 
 from django.conf import settings
 from django.urls import reverse
 
 from ietf.doc.models import Document
 from ietf.group.models import Group
+from ietf.meeting.factories import SessionFactory
 from ietf.meeting.models import Session, TimeSlot, SchedTimeSessAssignment
 from ietf.meeting.test_data import make_meeting_test_data
+from ietf.name.models import SessionStatusName
 from ietf.utils.test_data import make_test_data
 from ietf.utils.test_utils import TestCase
 from ietf.utils.mail import outbox
-
-from ietf.name.models import SessionStatusName
-from ietf.meeting.factories import SessionFactory
-
 from ietf.secr.proceedings.proc_utils import (create_proceedings, import_audio_files,
     get_timeslot_for_filename, normalize_room_name, send_audio_import_warning,
-    get_or_create_recording_document, create_recording, get_next_sequence)
+    get_or_create_recording_document, create_recording, get_next_sequence,
+    get_youtube_playlistid, get_youtube_videos, import_youtube_video_urls,
+    _get_session, _get_urls_from_json)
 
 
 SECR_USER='secretary'
@@ -37,6 +41,68 @@ def test_main(self):
         response = self.client.get(url)
         self.assertEqual(response.status_code, 200)
 
+
+class VideoRecordingTestCase(TestCase):
+    @patch('ietf.secr.proceedings.proc_utils.get_youtube_videos')
+    @patch('ietf.secr.proceedings.proc_utils.get_youtube_playlistid')
+    def test_import_youtube_video_urls(self, mock_playlistid, mock_videos):
+        meeting = make_meeting_test_data()
+        session = Session.objects.filter(meeting=meeting, group__acronym='mars').first()
+        title = self._get_video_title_for_session(session)
+        url = 'https://youtube.com?v=test'
+        mock_playlistid.return_value = 'PLC86T-6ZTP5g87jdxNqdWV5475U-yEE8M'
+        mock_videos.return_value = [{'title':title,'url':url}]
+        discovery = os.path.join(settings.BASE_DIR, "../test/data/youtube-discovery.json")
+        http = HttpMock(discovery, {'status': '200'})
+        import_youtube_video_urls(meeting=meeting, http=http)
+        doc = Document.objects.get(external_url=url)
+        self.assertTrue(doc in session.materials.all())
+
+    def _get_video_title_for_session(self, session):
+        '''Returns the youtube video title of a session recording given session'''
+        timeslot = session.official_timeslotassignment().timeslot
+        return "{prefix}-{group}-{date}".format(
+            prefix=session.meeting.type.slug + session.meeting.number,
+            group=session.group.acronym,
+            date=timeslot.time.strftime('%Y%m%d-%H%M')).upper()
+
+    def test_get_youtube_playlistid(self):
+        discovery = os.path.join(settings.BASE_DIR, "../test/data/youtube-discovery.json")
+        http = HttpMock(discovery, {'status': '200'})
+        youtube = build(settings.YOUTUBE_API_SERVICE_NAME, settings.YOUTUBE_API_VERSION,
+            developerKey='',http=http)
+        path = os.path.join(settings.BASE_DIR, "../test/data/youtube-playlistid.json")
+        http = HttpMock(path, {'status': '200'})
+        self.assertEqual(get_youtube_playlistid(youtube, 'IETF98', http=http),'PLC86T-test')
+  
+    def test_get_youtube_videos(self):
+        discovery = os.path.join(settings.BASE_DIR, "../test/data/youtube-discovery.json")
+        http = HttpMock(discovery, {'status': '200'})
+        youtube = build(settings.YOUTUBE_API_SERVICE_NAME, settings.YOUTUBE_API_VERSION,
+            developerKey='',http=http)
+        path = os.path.join(settings.BASE_DIR, "../test/data/youtube-playlistitems.json")
+        http = HttpMock(path, {'status': '200'})
+        videos = get_youtube_videos(youtube, 'PLC86T', http=http)
+        self.assertEqual(len(videos),2)
+
+    def test_get_session(self):
+        meeting = make_meeting_test_data()
+        session = Session.objects.filter(meeting=meeting, group__acronym='mars').first()
+        number = meeting.number
+        name = session.group.acronym
+        date = session.official_timeslotassignment().timeslot.time.strftime('%Y%m%d')
+        time = session.official_timeslotassignment().timeslot.time.strftime('%H%M')
+        self.assertEqual(_get_session(number,name,date,time),session)
+
+    def test_get_urls_from_json(self):
+        path = os.path.join(settings.BASE_DIR, "../test/data/youtube-playlistitems.json")
+        with open(path) as f:
+            doc = json.load(f)
+        urls = _get_urls_from_json(doc)
+        self.assertEqual(len(urls),2)
+        self.assertEqual(urls[0]['title'],'IETF98 Wrap Up')
+        self.assertEqual(urls[0]['url'],'https://www.youtube.com/watch?v=lhYWB5FFkg4&list=PLC86T-6ZTP5jo6kIuqdyeYYhsKv9sUwG1')
+        
 class RecordingTestCase(TestCase):
     def setUp(self):
         self.meeting_recordings_dir = self.tempdir('meeting-recordings')

ietf/secr/proceedings/views.py

@@ -289,7 +289,8 @@ def recording(request, meeting_num):
     session.
     '''
     meeting = get_object_or_404(Meeting, number=meeting_num)
-    sessions = meeting.session_set.filter(type__in=('session','plenary','other'),status='sched').order_by('group__acronym')
+    assignments = meeting.agenda.assignments.exclude(session__type__in=('reg','break')).order_by('session__group__acronym')
+    sessions = [ x.session for x in assignments ]
 
     if request.method == 'POST':
         form = RecordingForm(request.POST,meeting=meeting)

ietf/settings.py

@@ -732,6 +732,11 @@ def skip_unreadable_post(record):
 REGISTRATION_ATTENDEES_BASE_URL = 'https://ietf.org/registration/attendees/'
 NEW_PROCEEDINGS_START = 95
 USE_ETAGS=True
+YOUTUBE_API_KEY = ''
+YOUTUBE_API_SERVICE_NAME = 'youtube'
+YOUTUBE_API_VERSION = 'v3'
+YOUTUBE_BASE_URL = 'https://www.youtube.com/watch'
+YOUTUBE_IETF_CHANNEL_ID = 'UC8dtK9njBLdFnBahHFp0eZQ'
 
 PRODUCTION_TIMEZONE = "America/Los_Angeles"
 

requirements.txt

@@ -20,6 +20,7 @@ django-tastypie>=0.13.2
 django-widget-tweaks>=1.3
 docutils>=0.12                  
 factory-boy>=2.8.1
+google-api-python-client
 # Faker				# from factory-boy
 hashids>=1.1.0
 html5lib>=0.90,<0.99999999	# ietf.utils.html needs a rewrite for html5lib 1.x -- major code changes in sanitizer