From 084f8a7495e22c1de9a462889955b6404adf7020 Mon Sep 17 00:00:00 2001
From: Ryan Cross <rcross@amsl.com>
Date: Fri, 10 Mar 2017 23:04:06 +0000
Subject: [PATCH] Refactor session audio file import.  Add informative email. 
 Fixes #2164.  Commit ready for merge  - Legacy-Id: 12990

---
 .../0048_import_shared_audio_files.py         |  43 ++++++
 ietf/meeting/utils.py                         |   2 +
 ietf/secr/proceedings/proc_utils.py           | 133 +++++++++++++-----
 ietf/secr/proceedings/tests.py                | 119 +++++++++++++++-
 .../proceedings/audio_import_warning.txt      |   9 ++
 ietf/settings.py                              |   3 +-
 6 files changed, 274 insertions(+), 35 deletions(-)
 create mode 100644 ietf/meeting/migrations/0048_import_shared_audio_files.py
 create mode 100644 ietf/secr/templates/proceedings/audio_import_warning.txt

diff --git a/ietf/meeting/migrations/0048_import_shared_audio_files.py b/ietf/meeting/migrations/0048_import_shared_audio_files.py
new file mode 100644
index 000000000..3859491b6
--- /dev/null
+++ b/ietf/meeting/migrations/0048_import_shared_audio_files.py
@@ -0,0 +1,43 @@
+# -*- coding: utf-8 -*-
+# Generated by Django 1.10.5 on 2017-03-07 11:59
+from __future__ import unicode_literals
+
+import os
+
+from django.db import migrations
+from ietf.secr.proceedings.proc_utils import import_audio_files
+
+
+def purge_missing_files(apps, meeting):
+    Document = apps.get_model('doc', 'Document')
+    url = 'https://www.ietf.org/audio/ietf{}'.format(meeting.number)
+    documents = Document.objects.filter(external_url__startswith=url)
+    for document in documents:
+        filename = document.external_url.split('/')[-1]
+        if not os.path.exists(os.path.join('/a/www/audio/ietf{}'.format(meeting.number),filename)):
+            print "Removing missing recording: {} ({})".format(filename,document.pk)
+            document.delete()
+
+def forward(apps, schema_editor):
+    Meeting = apps.get_model('meeting', 'Meeting')
+    Document = apps.get_model('doc', 'Document')
+    for meeting in Meeting.objects.filter(number__in=range(94,98)):
+        print '\nMeeting #{}:'.format(meeting.number)
+        purge_missing_files(apps, meeting)
+        before = Document.objects.filter(type='recording').count()
+        import_audio_files(meeting)
+        after = Document.objects.filter(type='recording').count()
+        print '    {} Documents Added'.format(after - before)
+
+def backward(apps, schema_editor):
+        pass
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('meeting', '0047_fix_shared_timeslot_audio'),
+    ]
+
+    operations = [
+        migrations.RunPython(forward, backward)
+    ]
diff --git a/ietf/meeting/utils.py b/ietf/meeting/utils.py
index 632a9338a..fbafc839c 100644
--- a/ietf/meeting/utils.py
+++ b/ietf/meeting/utils.py
@@ -9,6 +9,7 @@ from django.template.loader import render_to_string
 from ietf.dbtemplate.models import DBTemplate
 from ietf.meeting.models import Session
 from ietf.group.utils import can_manage_materials
+from ietf.secr.proceedings.proc_utils import import_audio_files
 
 def group_sessions(sessions):
 
@@ -120,6 +121,7 @@ def finalize(meeting):
                 sp.rev = '00'
             sp.save()
     
+    import_audio_files(meeting)
     create_proceedings_templates(meeting)
     meeting.proceedings_final = True
     meeting.save()
diff --git a/ietf/secr/proceedings/proc_utils.py b/ietf/secr/proceedings/proc_utils.py
index b9d98e4a7..b931fdbbe 100644
--- a/ietf/secr/proceedings/proc_utils.py
+++ b/ietf/secr/proceedings/proc_utils.py
@@ -7,21 +7,23 @@ from urllib2 import urlopen
 import datetime
 import glob
 import os
+import re
 import shutil
 import subprocess
 
 import debug        # pyflakes:ignore
 
 from django.conf import settings
+from django.core.exceptions import ObjectDoesNotExist
 from django.http import HttpRequest
 from django.shortcuts import render_to_response, render
 from django.db.utils import ConnectionDoesNotExist
 
-from ietf.doc.models import Document, RelatedDocument, DocEvent, NewRevisionDocEvent, State
+from ietf.doc.models import Document, DocAlias, RelatedDocument, DocEvent, NewRevisionDocEvent, State
 from ietf.group.models import Group, Role
 from ietf.group.utils import get_charter_text
 from ietf.meeting.helpers import get_schedule
-from ietf.meeting.models import Session, Meeting, SchedTimeSessAssignment, SessionPresentation
+from ietf.meeting.models import Session, Meeting, SchedTimeSessAssignment, SessionPresentation, TimeSlot
 from ietf.person.models import Person
 from ietf.secr.proceedings.models import InterimMeeting    # proxy model
 from ietf.secr.proceedings.models import Registration
@@ -29,36 +31,86 @@ from ietf.secr.utils.document import get_rfc_num
 from ietf.secr.utils.group import groups_by_session
 from ietf.secr.utils.meeting import get_proceedings_path, get_materials, get_session
 from ietf.utils.log import log
+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})')
 
 # -------------------------------------------------
 # Helper Functions
 # -------------------------------------------------
-def check_audio_files(group,meeting):
+def import_audio_files(meeting):
     '''
     Checks for audio files and creates corresponding materials (docs) for the Session
-    Expects audio files in the format ietf[meeting num]-[room]-YYYMMDD-HHMM-*,
-    
-    Example: ietf90-salonb-20140721-1710-pm3.mp3
+    Expects audio files in the format ietf[meeting num]-[room]-YYYMMDD-HHMM.*,
     
+    Example: ietf90-salonb-20140721-1710.mp3
     '''
-    for session in Session.objects.filter(group=group,
-                                          meeting=meeting,
-                                          status=('sched'),
-                                          timeslotassignments__schedule=meeting.agenda):
-        timeslot = session.official_timeslotassignment().timeslot
-        if not (timeslot.location and timeslot.time):
-            continue
-        room = timeslot.location.name.lower()
-        room = room.replace(' ','')
-        room = room.replace('/','_')
-        time = timeslot.time.strftime("%Y%m%d-%H%M")
-        filename = 'ietf{}-{}-{}*'.format(meeting.number,room,time)
-        path = os.path.join(settings.MEETING_RECORDINGS_DIR,'ietf{}'.format(meeting.number),filename)
-        for file in glob.glob(path):
-            url = 'https://www.ietf.org/audio/ietf{}/{}'.format(meeting.number,os.path.basename(file))
-            doc = Document.objects.filter(external_url=url).first()
-            if not doc:
-                create_recording(session,url)
+    unmatched_files = []
+    path = os.path.join(settings.MEETING_RECORDINGS_DIR, meeting.type.slug + meeting.number)
+    if not os.path.exists(path):
+        return None
+    for filename in os.listdir(path):
+        timeslot = get_timeslot_for_filename(filename)
+        if timeslot:
+            sessionassignments = timeslot.sessionassignments.filter(
+                schedule=timeslot.meeting.agenda,
+                session__status='sched',
+                ).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)
+        else:
+            # use for reconciliation email
+            unmatched_files.append(filename)
+    
+    if unmatched_files:
+        send_audio_import_warning(unmatched_files)
+
+
+def get_timeslot_for_filename(filename):
+    '''Returns a timeslot matching the filename given.
+    NOTE: currently only works with ietfNN prefix (regular meetings)
+    '''
+    basename, _ = os.path.splitext(filename)
+    match = AUDIO_FILE_RE.match(basename)
+    if match:
+        try:
+            meeting = Meeting.objects.get(number=match.groupdict()['number'])
+            room_mapping = {normalize_room_name(room.name): room.name for room in meeting.room_set.all()}
+            time = datetime.datetime.strptime(match.groupdict()['time'],'%Y%m%d-%H%M')
+            return TimeSlot.objects.get(
+                meeting=meeting,
+                location__name=room_mapping[match.groupdict()['room']],
+                time=time)
+        except (ObjectDoesNotExist, KeyError):
+            return None
+
+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)
+    try:
+        doc = Document.objects.get(external_url=url)
+        return doc
+    except ObjectDoesNotExist:
+        pass
+    return create_recording(session,url)
 
 
 def create_recording(session,url):
@@ -94,6 +146,30 @@ def create_recording(session,url):
     pres = SessionPresentation.objects.create(session=session,document=doc,rev=doc.rev)
     session.sessionpresentation_set.add(pres)
 
+    return doc
+
+def get_next_sequence(group,meeting,type):
+    '''
+    Returns the next sequence number to use for a document of type = type.
+    Takes a group=Group object, meeting=Meeting object, type = string
+    '''
+    aliases = DocAlias.objects.filter(name__startswith='{}-{}-{}-'.format(type,meeting.number,group.acronym))
+    if not aliases:
+        return 1
+    aliases = aliases.order_by('name')
+    sequence = int(aliases.last().name.split('-')[-1]) + 1
+    return sequence
+
+def send_audio_import_warning(unmatched_files):
+    '''Send email to interested parties that some audio files weren't matched to timeslots'''
+    send_mail(request = None,
+              to       = settings.AUDIO_IMPORT_EMAIL,
+              frm      = "IETF Secretariat <ietf-secretariat@ietf.org>",
+              subject  = "Audio file import warning",
+              template = "proceedings/audio_import_warning.txt",
+              context  = dict(unmatched_files=unmatched_files),
+              extra    = {})
+
 def mycomp(timeslot):
     '''
     This takes a timeslot object and returns a key to sort by the area acronym or None
@@ -172,13 +248,6 @@ def get_progress_stats(sdate,edate):
 
     return data
 
-def get_next_sequence(group,meeting,type):
-    '''
-    Returns the next sequence number to use for a document of type = type.
-    Takes a group=Group object, meeting=Meeting object, type = string
-    '''
-    return Document.objects.filter(name__startswith='{}-{}-{}-'.format(type,meeting.number,group.acronym)).count() + 1
-
 def write_html(path,content):
     f = open(path,'w')
     f.write(content)
@@ -226,7 +295,7 @@ def create_proceedings(meeting, group, is_final=False):
     if meeting.type_id == 'ietf' and int(meeting.number) < 79:
         return
 
-    check_audio_files(group,meeting)
+    #check_audio_files(group,meeting)
     materials = get_materials(group,meeting)
 
     chairs = group.role_set.filter(name='chair')
diff --git a/ietf/secr/proceedings/tests.py b/ietf/secr/proceedings/tests.py
index 448934fa2..859d3cf93 100644
--- a/ietf/secr/proceedings/tests.py
+++ b/ietf/secr/proceedings/tests.py
@@ -5,16 +5,21 @@ import shutil
 from django.conf import settings
 from django.core.urlresolvers import reverse
 
+from ietf.doc.models import Document
 from ietf.group.models import Group
-from ietf.meeting.models import Session
+from ietf.meeting.models import Session, TimeSlot, SchedTimeSessAssignment
 from ietf.meeting.test_data import make_meeting_test_data
 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
+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)
+
 
 SECR_USER='secretary'
 
@@ -33,6 +38,17 @@ class ProceedingsTestCase(TestCase):
         self.assertEqual(response.status_code, 200)
 
 class RecordingTestCase(TestCase):
+    def setUp(self):
+        self.meeting_recordings_dir = os.path.abspath("tmp-meeting-recordings-dir")
+        self.saved_meeting_recordings_dir = settings.MEETING_RECORDINGS_DIR
+        settings.MEETING_RECORDINGS_DIR = self.meeting_recordings_dir
+        if not os.path.exists(self.meeting_recordings_dir):
+            os.makedirs(self.meeting_recordings_dir)
+
+    def tearDown(self):
+        shutil.rmtree(self.meeting_recordings_dir)
+        settings.MEETING_RECORDINGS_DIR = self.saved_meeting_recordings_dir
+
     def test_page(self):
         meeting = make_meeting_test_data()
         url = reverse('proceedings_recording', kwargs={'meeting_num':meeting.number})
@@ -62,6 +78,105 @@ class RecordingTestCase(TestCase):
         response = self.client.post(url,dict(external_url=external_url),follow=True)
         self.assertEqual(response.status_code, 200)
         self.failUnless(external_url in response.content)
+            
+    def test_import_audio_files(self):
+        meeting = make_meeting_test_data()
+        group = Group.objects.get(acronym='mars')
+        session = Session.objects.filter(meeting=meeting,group=group).first()
+        status = SessionStatusName.objects.get(slug='sched')
+        session.status = status
+        session.save()
+        timeslot = session.official_timeslotassignment().timeslot
+        self.create_audio_file_for_timeslot(timeslot)
+        import_audio_files(meeting)
+        self.assertEqual(session.materials.filter(type='recording').count(),1)
+
+    def create_audio_file_for_timeslot(self, timeslot):
+        filename = self.get_filename_for_timeslot(timeslot)
+        path = os.path.join(settings.MEETING_RECORDINGS_DIR,'ietf' + timeslot.meeting.number,filename)
+        if not os.path.exists(os.path.dirname(path)):
+            os.makedirs(os.path.dirname(path))
+        with open(path, "w") as f:
+            f.write('dummy')
+
+    def get_filename_for_timeslot(self, timeslot):
+        '''Returns the filename of a session recording given timeslot'''
+        return "{prefix}-{room}-{date}.mp3".format(
+            prefix=timeslot.meeting.type.slug + timeslot.meeting.number,
+            room=normalize_room_name(timeslot.location.name),
+            date=timeslot.time.strftime('%Y%m%d-%H%M'))
+
+    def test_import_audio_files_shared_timeslot(self):
+        meeting = make_meeting_test_data()
+        mars_session = Session.objects.filter(meeting=meeting,group__acronym='mars').first()
+        ames_session = Session.objects.filter(meeting=meeting,group__acronym='ames').first()
+        scheduled = SessionStatusName.objects.get(slug='sched')
+        mars_session.status = scheduled
+        mars_session.save()
+        ames_session.status = scheduled
+        ames_session.save()
+        timeslot = mars_session.official_timeslotassignment().timeslot
+        SchedTimeSessAssignment.objects.create(timeslot=timeslot,session=ames_session,schedule=meeting.agenda)
+        self.create_audio_file_for_timeslot(timeslot)
+        import_audio_files(meeting)
+        doc = mars_session.materials.filter(type='recording').first()
+        self.assertTrue(doc in ames_session.materials.all())
+        self.assertTrue(doc.docalias_set.filter(name='recording-42-mars-1'))
+        self.assertTrue(doc.docalias_set.filter(name='recording-42-ames-1'))
+
+    def test_normalize_room_name(self):
+        self.assertEqual(normalize_room_name('Test Room'),'testroom')
+        self.assertEqual(normalize_room_name('Rome/Venice'), 'rome_venice')
+
+    def test_get_timeslot_for_filename(self):
+        meeting = make_meeting_test_data()
+        timeslot = TimeSlot.objects.filter(meeting=meeting,type='session').first()
+        name = self.get_filename_for_timeslot(timeslot)
+        self.assertEqual(get_timeslot_for_filename(name),timeslot)
+
+    def test_get_or_create_recording_document(self):
+        meeting = make_meeting_test_data()
+        group = Group.objects.get(acronym='mars')
+        session = Session.objects.filter(meeting=meeting,group=group).first()
+        
+        # test create
+        filename = 'ietf42-testroom-20000101-0800.mp3'
+        docs_before = Document.objects.filter(type='recording').count()
+        doc = get_or_create_recording_document(filename,session)
+        docs_after = Document.objects.filter(type='recording').count()
+        self.assertEqual(docs_after,docs_before + 1)
+        self.assertTrue(doc.external_url.endswith(filename))
+
+        # test get
+        docs_before = docs_after
+        doc2 = get_or_create_recording_document(filename,session)
+        docs_after = Document.objects.filter(type='recording').count()
+        self.assertEqual(docs_after,docs_before)
+        self.assertEqual(doc,doc2)
+
+    def test_create_recording(self):
+        meeting = make_meeting_test_data()
+        group = Group.objects.get(acronym='mars')
+        session = Session.objects.filter(meeting=meeting,group=group).first()
+        filename = 'ietf42-testroomt-20000101-0800.mp3'
+        url = settings.IETF_AUDIO_URL + 'ietf{}/{}'.format(meeting.number, filename)
+        doc = create_recording(session, url)
+        self.assertEqual(doc.name,'recording-42-mars-1')
+        self.assertEqual(doc.group,group)
+        self.assertEqual(doc.external_url,url)
+        self.assertTrue(doc in session.materials.all())
+
+    def test_get_next_sequence(self):
+        meeting = make_meeting_test_data()
+        group = Group.objects.get(acronym='mars')
+        sequence = get_next_sequence(group,meeting,'recording')
+        self.assertEqual(sequence,1)
+
+    def test_send_audio_import_warning(self):
+        length_before = len(outbox)
+        send_audio_import_warning(['recording-43-badroom-20000101-0800.mp3'])
+        self.assertEqual(len(outbox), length_before + 1)
+        self.assertTrue('Audio file import' in outbox[-1]['Subject'])
 
 class OldProceedingsTestCase(TestCase):
     ''' Ensure coverage of fragments of old proceedings generation until those are removed ''' 
diff --git a/ietf/secr/templates/proceedings/audio_import_warning.txt b/ietf/secr/templates/proceedings/audio_import_warning.txt
new file mode 100644
index 000000000..322a43f5f
--- /dev/null
+++ b/ietf/secr/templates/proceedings/audio_import_warning.txt
@@ -0,0 +1,9 @@
+
+WARNING:
+
+After the last meeting session audio file import there are {{ unmatched_files|length }}
+file(s) that were not matched to a timeslot.
+
+{% for file in unmatched_files %}{{ file }}
+{% endfor %}
+
diff --git a/ietf/settings.py b/ietf/settings.py
index ce526cc01..46e0d6c29 100644
--- a/ietf/settings.py
+++ b/ietf/settings.py
@@ -132,6 +132,7 @@ OLD_PHOTO_DIRS = [
 IETF_HOST_URL = 'https://www.ietf.org/'
 IETF_ID_URL = IETF_HOST_URL + 'id/'
 IETF_ID_ARCHIVE_URL = IETF_HOST_URL + 'archive/id/'
+IETF_AUDIO_URL = IETF_HOST_URL + 'audio/'
 
 
 # Absolute path to the directory static files should be collected to.
@@ -523,7 +524,7 @@ CACHES = {
 }
 
 IPR_EMAIL_FROM = 'ietf-ipr@ietf.org'
-
+AUDIO_IMPORT_EMAIL = ['agenda@ietf.org']
 IANA_EVAL_EMAIL = "drafts-eval@icann.org"
 
 # Put real password in settings_local.py