cablegates/pandora/archive/models.py

546 lines
18 KiB
Python
Raw Normal View History

# -*- coding: utf-8 -*-
# vi:si:et:sw=4:sts=4:ts=4
2010-09-18 14:44:35 +00:00
from __future__ import division
2011-01-01 11:44:42 +00:00
import os.path
import re
2010-08-10 22:01:41 +00:00
import time
from django.db import models
2011-06-02 08:19:15 +00:00
from django.db.models import Q
from django.contrib.auth.models import User
from django.conf import settings
2011-04-18 18:50:31 +00:00
from django.db.models.signals import pre_delete
from ox.django import fields
import ox
2011-01-01 11:44:42 +00:00
from ox.normalize import canonicalTitle
2010-09-18 14:44:35 +00:00
import chardet
2010-11-08 16:34:25 +00:00
from item import utils
from item.models import Item
from person.models import get_name_sort
2011-06-27 13:39:35 +00:00
import extract
2010-08-24 17:16:33 +00:00
class File(models.Model):
created = models.DateTimeField(auto_now_add=True)
modified = models.DateTimeField(auto_now=True)
2010-08-07 14:31:20 +00:00
verified = models.BooleanField(default=False)
2011-06-27 13:39:35 +00:00
auto = models.BooleanField(default=True)
2010-08-10 22:01:41 +00:00
oshash = models.CharField(max_length=16, unique=True)
2010-09-23 16:01:48 +00:00
item = models.ForeignKey(Item, related_name='files')
name = models.CharField(max_length=2048, default="") # canoncial path/file
2011-02-01 13:19:34 +00:00
folder = models.CharField(max_length=2048, default="") # canoncial path/file
2011-01-01 11:44:42 +00:00
sort_name = models.CharField(max_length=2048, default="") # sort name
2011-02-01 13:19:34 +00:00
type = models.CharField(default="", max_length=255)
part = models.IntegerField(null=True)
version = models.CharField(default="", max_length=255) # sort path/file name
language = models.CharField(default="", max_length=8)
2010-08-10 22:01:41 +00:00
season = models.IntegerField(default=-1)
episode = models.IntegerField(default=-1)
2010-08-10 22:01:41 +00:00
size = models.BigIntegerField(default=0)
duration = models.FloatField(null=True)
info = fields.DictField(default={})
video_codec = models.CharField(max_length=255)
pixel_format = models.CharField(max_length=255)
display_aspect_ratio = models.CharField(max_length=255)
width = models.IntegerField(default = 0)
height = models.IntegerField(default = 0)
framerate = models.CharField(max_length=255)
audio_codec = models.CharField(max_length=255)
2010-08-10 22:01:41 +00:00
channels = models.IntegerField(default=0)
samplerate = models.IntegerField(default=0)
bits_per_pixel = models.FloatField(default=-1)
pixels = models.BigIntegerField(default=0)
2010-08-07 14:31:20 +00:00
#This is true if derivative is available or subtitles where uploaded
available = models.BooleanField(default = False)
2010-08-10 22:01:41 +00:00
is_audio = models.BooleanField(default=False)
is_video = models.BooleanField(default=False)
is_extra = models.BooleanField(default=False)
is_main = models.BooleanField(default=False)
is_subtitle = models.BooleanField(default=False)
is_version = models.BooleanField(default=False)
2010-08-07 14:31:20 +00:00
def __unicode__(self):
return self.name
2011-06-27 13:39:35 +00:00
def set_state(self):
instance = self.get_instance()
if instance:
if instance.name.lower().startswith('extras/'):
self.is_extra = True
self.is_main = False
elif instance.name.lower().startswith('versions/'):
self.is_version = True
self.is_main = False
else:
self.is_extra = False
self.is_main = True
else:
self.is_main = False
self.is_extra = False
self.is_version = False
self.name = self.get_name()
2011-02-02 07:51:59 +00:00
self.folder = self.get_folder()
self.sort_name = utils.sort_string(canonicalTitle(self.name))
if self.info:
for key in ('duration', 'size'):
setattr(self, key, self.info.get(key, 0))
2011-01-28 08:48:38 +00:00
if 'video' in self.info and self.info['video'] and \
'width' in self.info['video'][0]:
2011-01-22 10:14:30 +00:00
video = self.info['video'][0]
self.video_codec = video['codec']
self.width = video['width']
self.height = video['height']
self.framerate = video['framerate']
if 'display_aspect_ratio' in video:
self.display_aspect_ratio = video['display_aspect_ratio']
else:
self.display_aspect_ratio = "%s:%s" % (self.width, self.height)
self.is_video = True
self.is_audio = False
2011-01-22 10:14:30 +00:00
if self.name.endswith('.jpg') or \
self.name.endswith('.png') or \
self.duration == 0.04:
2010-12-27 05:07:15 +00:00
self.is_video = False
else:
self.is_video = False
2011-02-23 11:51:32 +00:00
self.display_aspect_ratio = "4:3"
self.width = '320'
self.height = '240'
if 'audio' in self.info and self.info['audio']:
2011-01-22 10:14:30 +00:00
audio = self.info['audio'][0]
self.audio_codec = audio['codec']
self.samplerate = audio.get('samplerate', 0)
self.channels = audio.get('channels', 0)
if not self.is_video:
self.is_audio = True
else:
self.is_audio = False
if self.framerate:
2010-12-22 15:17:38 +00:00
self.pixels = int(self.width * self.height * float(utils.parse_decimal(self.framerate)) * self.duration)
2011-02-01 13:19:34 +00:00
else:
self.is_video = os.path.splitext(self.name)[-1] in ('.avi', '.mkv', '.dv', '.ogv', '.mpeg', '.mov')
self.is_audio = os.path.splitext(self.name)[-1] in ('.mp3', '.wav', '.ogg', '.flac')
2011-06-02 08:19:15 +00:00
self.is_subtitle = os.path.splitext(self.name)[-1] in ('.srt', )
2011-02-01 13:19:34 +00:00
if not self.is_audio and not self.is_video and self.name.endswith('.srt'):
self.is_subtitle = True
2011-06-02 08:19:15 +00:00
else:
self.is_subtitle = False
2011-02-01 13:19:34 +00:00
self.type = self.get_type()
2011-06-01 16:55:30 +00:00
self.language = self.get_language()
self.part = self.get_part()
2011-02-02 07:51:59 +00:00
2011-02-01 13:19:34 +00:00
if self.type not in ('audio', 'video'):
self.duration = None
2011-06-27 13:39:35 +00:00
def save(self, *args, **kwargs):
if self.auto:
self.set_state()
super(File, self).save(*args, **kwargs)
2010-09-03 13:28:44 +00:00
#upload and data handling
2011-01-22 10:14:30 +00:00
data = models.FileField(null=True, blank=True,
upload_to=lambda f, x: f.path('data.bin'))
2010-12-04 01:09:10 +00:00
def path(self, name):
h = self.oshash
return os.path.join('files', h[:2], h[2:4], h[4:6], h[6:], name)
2010-09-03 13:28:44 +00:00
2010-08-24 17:16:33 +00:00
def contents(self):
2010-09-17 21:06:01 +00:00
if self.data != None:
self.data.seek(0)
2010-09-03 13:28:44 +00:00
return self.data.read()
2010-08-24 17:16:33 +00:00
return None
def srt(self, offset=0):
2011-01-01 11:44:42 +00:00
2010-09-18 14:44:35 +00:00
def _detectEncoding(fp):
2011-01-01 11:44:42 +00:00
bomDict={ # bytepattern : name
(0x00, 0x00, 0xFE, 0xFF): "utf_32_be",
(0xFF, 0xFE, 0x00, 0x00): "utf_32_le",
(0xFE, 0xFF, None, None): "utf_16_be",
(0xFF, 0xFE, None, None): "utf_16_le",
(0xEF, 0xBB, 0xBF, None): "utf_8",
2010-09-18 14:44:35 +00:00
}
# go to beginning of file and get the first 4 bytes
oldFP = fp.tell()
fp.seek(0)
(byte1, byte2, byte3, byte4) = tuple(map(ord, fp.read(4)))
# try bom detection using 4 bytes, 3 bytes, or 2 bytes
bomDetection = bomDict.get((byte1, byte2, byte3, byte4))
2011-01-01 11:44:42 +00:00
if not bomDetection:
2010-09-18 14:44:35 +00:00
bomDetection = bomDict.get((byte1, byte2, byte3, None))
2011-01-01 11:44:42 +00:00
if not bomDetection:
2010-09-18 14:44:35 +00:00
bomDetection = bomDict.get((byte1, byte2, None, None))
## if BOM detected, we're done :-)
fp.seek(oldFP)
2011-01-01 11:44:42 +00:00
if bomDetection:
return bomDetection
2010-09-18 14:44:35 +00:00
encoding = 'latin-1'
#more character detecting magick using http://chardet.feedparser.org/
fp.seek(0)
rawdata = fp.read()
encoding = chardet.detect(rawdata)['encoding']
fp.seek(oldFP)
return encoding
def parseTime(t):
return offset + ox.time2ms(t.replace(',', '.')) / 1000
2010-09-18 14:44:35 +00:00
srt = []
f = open(self.data.path)
encoding = _detectEncoding(f)
data = f.read()
f.close()
data = data.replace('\r\n', '\n')
2010-09-18 14:44:35 +00:00
try:
data = unicode(data, encoding)
except:
try:
data = unicode(data, 'latin-1')
except:
print "failed to detect encoding, giving up"
return srt
srts = re.compile('(\d\d:\d\d:\d\d[,.]\d\d\d)\s*-->\s*(\d\d:\d\d:\d\d[,.]\d\d\d)\s*(.+?)\n\n', re.DOTALL)
2011-02-09 14:28:52 +00:00
i = 0
2010-09-18 14:44:35 +00:00
for s in srts.findall(data):
2011-02-09 14:28:52 +00:00
_s = {'id': str(i),
'in': parseTime(s[0]), 'out': parseTime(s[1]), 'value': s[2].strip()}
2011-06-15 10:08:15 +00:00
if srt and srt[-1]['out'] > _s['in']:
srt[-1]['out'] = _s['in']
2010-09-18 14:44:35 +00:00
srt.append(_s)
2011-02-09 14:28:52 +00:00
i += 1
2010-09-18 14:44:35 +00:00
return srt
2010-09-03 13:28:44 +00:00
def editable(self, user):
#FIXME: check that user has instance of this file
return True
def save_chunk(self, chunk, chunk_id=-1, done=False):
2010-09-08 17:14:01 +00:00
if not self.available:
stream, created = Stream.objects.get_or_create(
file=self,
resolution=settings.VIDEO_RESOLUTIONS[0],
format=settings.VIDEO_FORMATS[0])
if created:
stream.video.save(stream.name(), chunk)
2010-09-03 13:28:44 +00:00
else:
f = open(stream.video.path, 'a')
2010-09-03 13:28:44 +00:00
#FIXME: should check that chunk_id/offset is right
f.write(chunk.read())
f.close()
if done:
stream.available = True
stream.save()
2010-09-03 13:28:44 +00:00
return True
return False
2011-01-19 12:06:03 +00:00
def json(self, keys=None, user=None):
2011-06-15 12:20:45 +00:00
if keys and not 'instances' in keys:
keys.append('instances')
2011-02-01 13:19:34 +00:00
resolution = (self.width, self.height)
if resolution == (0, 0):
resolution = None
duration = self.duration
if self.get_type() != 'video':
duration = None
2011-01-19 12:06:03 +00:00
data = {
'available': self.available,
2011-02-01 13:19:34 +00:00
'duration': duration,
2011-01-19 12:06:03 +00:00
'framerate': self.framerate,
'height': self.height,
'width': self.width,
2011-02-01 13:19:34 +00:00
'resolution': resolution,
2011-06-27 13:39:35 +00:00
'id': self.oshash,
2011-01-19 12:06:03 +00:00
'samplerate': self.samplerate,
'video_codec': self.video_codec,
'audio_codec': self.audio_codec,
'name': self.name,
'size': self.size,
2011-01-21 09:31:49 +00:00
'info': self.info,
2011-02-01 13:19:34 +00:00
'users': list(set([i.volume.user.username for i in self.instances.all()])),
'instances': [i.json() for i in self.instances.all()],
'folder': self.get_folder(),
'type': self.get_type(),
'is_main': self.is_main,
'part': self.get_part()
2011-01-19 12:06:03 +00:00
}
if keys:
for k in data.keys():
if k not in keys:
del data[k]
return data
2011-01-01 11:44:42 +00:00
2011-02-01 13:19:34 +00:00
def get_part(self):
2011-06-02 08:19:15 +00:00
#FIXME: this breaks for sub/idx/srt
if os.path.splitext(self.name)[-1] in ('.sub', '.idx', '.srt'):
name = os.path.splitext(self.name)[0]
if self.language:
name = name[-(len(self.language)+1)]
qs = self.item.files.filter(Q(is_video=True)|Q(is_audio=True),
is_main=True, name__startswith=name)
if qs.count()>0:
return qs[0].part
if not self.is_extra:
2011-06-01 16:55:30 +00:00
files = list(self.item.files.filter(type=self.type, language=self.language,
is_main=self.is_main).order_by('sort_name'))
if self in files:
return files.index(self) + 1
return None
2011-02-01 13:19:34 +00:00
def get_type(self):
if self.is_video:
return 'video'
if self.is_audio:
return 'audio'
if self.is_subtitle or os.path.splitext(self.name)[-1] in ('.sub', '.idx'):
return 'subtitle'
return 'unknown'
def get_instance(self):
2011-04-08 08:20:59 +00:00
#FIXME: what about other instances?
if self.instances.all().count() > 0:
return self.instances.all()[0]
return None
2011-02-01 13:19:34 +00:00
def get_folder(self):
2011-04-12 19:22:03 +00:00
instance = self.get_instance()
if instance:
return instance.folder
name = os.path.splitext(self.get_name())[0]
2011-04-12 19:22:03 +00:00
name = name.replace('. ', '||').split('.')[0].replace('||', '. ')
if self.item:
if settings.USE_IMDB:
director = self.item.get('director', ['Unknown Director'])
director = map(get_name_sort, director)
2011-04-05 10:50:14 +00:00
director = u'; '.join(director)
director = re.sub(r'[:\\/]', '_', director)
name = os.path.join(director, name)
2011-04-22 23:34:01 +00:00
year = self.item.get('year')
if year:
name += u' (%s)' % year
name = os.path.join(name[0].upper(), name)
return name
2011-04-05 10:50:14 +00:00
return u''
2011-01-21 09:31:49 +00:00
2011-02-02 07:51:59 +00:00
def get_name(self):
2011-04-08 08:20:59 +00:00
instance = self.get_instance()
if instance:
2011-04-12 19:22:03 +00:00
return instance.name
if self.item:
name = self.item.get('title', 'Untitled')
name = re.sub(r'[:\\/]', '_', name)
if not name:
name = 'Untitled'
2011-04-12 19:22:03 +00:00
ext = '.unknown'
return name + ext
2011-02-02 07:51:59 +00:00
2011-06-01 16:55:30 +00:00
def get_language(self):
language = self.name.split('.')
if len(language) >= 3 and len(language[-2]) == 2:
return language[-2]
return ''
2011-04-18 18:50:31 +00:00
def delete_file(sender, **kwargs):
f = kwargs['instance']
#FIXME: delete streams here
2011-04-18 18:50:31 +00:00
if f.data:
f.data.delete()
pre_delete.connect(delete_file, sender=File)
2010-08-10 22:01:41 +00:00
class Volume(models.Model):
2011-01-01 11:44:42 +00:00
2010-08-10 22:01:41 +00:00
class Meta:
unique_together = ("user", "name")
created = models.DateTimeField(auto_now_add=True)
modified = models.DateTimeField(auto_now=True)
user = models.ForeignKey(User, related_name='volumes')
name = models.CharField(max_length=1024)
def __unicode__(self):
return u"%s's %s"% (self.user, self.name)
2010-11-08 17:43:59 +00:00
class Instance(models.Model):
2011-01-01 11:44:42 +00:00
2010-08-10 22:01:41 +00:00
class Meta:
unique_together = ("name", "folder", "volume")
created = models.DateTimeField(auto_now_add=True)
modified = models.DateTimeField(auto_now=True)
2010-08-07 14:31:20 +00:00
2010-08-10 22:01:41 +00:00
atime = models.IntegerField(default=lambda: int(time.time()), editable=False)
ctime = models.IntegerField(default=lambda: int(time.time()), editable=False)
mtime = models.IntegerField(default=lambda: int(time.time()), editable=False)
2010-08-10 22:01:41 +00:00
name = models.CharField(max_length=2048)
2011-03-04 16:10:19 +00:00
folder = models.CharField(max_length=2048)
file = models.ForeignKey(File, related_name='instances')
2010-08-10 22:01:41 +00:00
volume = models.ForeignKey(Volume, related_name='files')
def __unicode__(self):
2010-08-10 22:01:41 +00:00
return u"%s's %s <%s>"% (self.volume.user, self.name, self.file.oshash)
@property
2010-09-23 16:01:48 +00:00
def itemId(self):
return File.objects.get(oshash=self.oshash).itemId
2011-02-01 13:19:34 +00:00
def json(self):
return {
'user': self.volume.user.username,
'volume': self.volume.name,
'folder': self.folder,
'name': self.name
}
2011-01-01 11:44:42 +00:00
2010-09-08 17:14:01 +00:00
def frame_path(frame, name):
2010-08-24 17:16:33 +00:00
ext = os.path.splitext(name)[-1]
2010-09-08 17:14:01 +00:00
name = "%s%s" % (frame.position, ext)
2010-12-04 01:09:10 +00:00
return frame.file.path(name)
2010-08-07 14:31:20 +00:00
2011-01-01 11:44:42 +00:00
class Frame(models.Model):
2011-01-01 11:44:42 +00:00
2010-08-10 22:01:41 +00:00
class Meta:
unique_together = ("file", "position")
created = models.DateTimeField(auto_now_add=True)
modified = models.DateTimeField(auto_now=True)
file = models.ForeignKey(File, related_name="frames")
position = models.FloatField()
2010-08-07 14:31:20 +00:00
frame = models.ImageField(default=None, null=True, upload_to=frame_path)
2010-11-06 16:14:40 +00:00
'''
def save(self, *args, **kwargs):
name = "%d.jpg" % self.position
if file.name != name:
#FIXME: frame path should be renamed on save to match current position
super(Frame, self).save(*args, **kwargs)
'''
def __unicode__(self):
2011-01-22 10:14:30 +00:00
return u'%s/%s' % (self.file, self.position)
2011-04-18 18:50:31 +00:00
def delete_frame(sender, **kwargs):
f = kwargs['instance']
if f.frame:
f.frame.delete()
pre_delete.connect(delete_frame, sender=Frame)
class Stream(models.Model):
class Meta:
unique_together = ("file", "resolution", "format")
file = models.ForeignKey(File, related_name='streams')
resolution = models.IntegerField(default=96)
format = models.CharField(max_length=255, default='webm')
2011-08-18 18:57:56 +00:00
video = models.FileField(default=None, blank=True, upload_to=lambda f, x: f.path(x))
source = models.ForeignKey('Stream', related_name='derivatives', default=None, null=True)
available = models.BooleanField(default=False)
info = fields.DictField(default={})
2011-08-18 19:37:12 +00:00
duration = models.FloatField(default=0)
aspect_ratio = models.FloatField(default=0)
@property
def timeline_prefix(self):
return os.path.join(settings.MEDIA_ROOT, self.path(), 'timeline')
def name(self):
return u"%sp.%s" % (self.resolution, self.format)
def __unicode__(self):
return u"%s/%s" % (self.file, self.name())
def path(self, name=''):
return self.file.path(name)
def extract_derivatives(self):
self.make_timeline()
for resolution in settings.VIDEO_RESOLUTIONS:
for f in settings.VIDEO_FORMATS:
derivative, created = Stream.objects.get_or_create(file=self.file,
resolution=resolution, format=f)
if created:
derivative.source = self
name = derivative.name()
derivative.video.name = os.path.join(os.path.dirname(self.video.name), name)
derivative.encode()
derivative.save()
return True
def encode(self):
if self.source:
video = self.source.video.path
target = self.video.path
info = ox.avinfo(video)
if extract.stream(video, target, self.name(), info):
self.available = True
else:
self.available = False
self.save()
def make_timeline(self):
if self.available and not self.source:
extract.timeline(self.video.path, self.timeline_prefix)
def save(self, *args, **kwargs):
if self.video and not self.info:
self.info = ox.avinfo(self.video.path)
2011-08-18 19:37:12 +00:00
self.duration = self.info.get('duration', 0)
if 'video' in self.info and self.info['video']:
self.aspect_ratio = self.info['video'][0]['width'] / self.info['video'][0]['height']
else:
self.aspect_ratio = 128/80
super(Stream, self).save(*args, **kwargs)
2011-08-18 19:37:12 +00:00
def json(self):
if settings.XSENDFILE or settings.XACCELREDIRECT:
2011-08-18 19:40:15 +00:00
base_url = '/%s' % self.file.itemId
2011-08-18 19:37:12 +00:00
else:
base_url = os.path.dirname(self.video.url)
return {
'duration': self.duration,
'aspectRatio': self.aspect_ratio,
'baseUrl': base_url
}
def delete_stream(sender, **kwargs):
f = kwargs['instance']
if f.video:
f.video.delete()
pre_delete.connect(delete_stream, sender=Stream)