sorter.py 9.71 KB
Newer Older
Andreas Valder's avatar
Andreas Valder committed
1
from server import *
2
import traceback
Julian Rother's avatar
Julian Rother committed
3
import os.path
Andreas Valder's avatar
Andreas Valder committed
4

5
@app.route('/internal/sort/log')
6
7
8
9
10
11
12
13
14
15
16
17
18
@register_navbar('Sortierlog', icon='sort-by-attributes-alt')
@mod_required
def sort_log():
	return render_template('sortlog.html',sortlog=query('''
			SELECT 
				sortlog.*,
				lectures.id as lecture_id,
				lectures.title as lecture_title,
				lectures.course_id as course_id, 
				courses.title as course_title
			FROM sortlog 
			JOIN lectures ON lectures.id = sortlog.lecture_id
			JOIN courses ON courses.id = lectures.course_id 
19
			ORDER BY sortlog.`when` DESC
Andreas Valder's avatar
Andreas Valder committed
20
			LIMIT 50
21
		'''),sorterrorlog=query('SELECT * FROM sorterrorlog ORDER BY sorterrorlog.`when` DESC'))
22

23
24
25
26
27
def to_ascii(inputstring):
	asciistring = inputstring
	for charset in [('ä', 'ae'), ('ö', 'oe'), ('ü', 'ue'), ('ß', 'ss')]:
		asciistring = asciistring.replace(charset[0],charset[1])
	return asciistring
28

29
30
31
32
33
34
35
36
37
38
39
40
@job_handler('probe', 'remux', 'transcode')
def update_video_metadata(jobid, jobtype, data, state, status):
	if 'video_id' not in data:
		return
	if jobtype not in ['remux', 'transcode']:
		video = query('SELECT * FROM videos WHERE id = ?', data['video_id'])[0]
		if video['hash'] and video['hash'] != status['hash']:
			print('Hash mismatch for video', data['video_id'])
			return
	modify('UPDATE videos_data SET hash = ?, file_size = ? WHERE id = ?',
			status['hash'], status['filesize'], data['video_id'])

41
def insert_video(lectureid, dbfilepath, fileformatid, hash="", filesize=-1):
42
	video_id = modify('''INSERT INTO videos_data 
43
		(lecture_id, visible, path, video_format, title, comment, internal, file_modified, time_created, time_updated, created_by, hash, file_size)
44
		VALUES 
45
46
		(?, 0, ?, ?, "", "", "", ?, ?, ?, ?, ?, ?)''',
		lectureid, dbfilepath, fileformatid, datetime.now(), datetime.now(), datetime.now(), -1, hash, filesize)
47
	query('INSERT INTO sortlog (lecture_id,video_id,path,`when`) VALUES (?,?,?,?)', lectureid, video_id, dbfilepath, datetime.now())
Andreas Valder's avatar
Andreas Valder committed
48
	schedule_thumbnail(lectureid)
Julian Rother's avatar
Julian Rother committed
49
	schedule_job('probe', {'path': dbfilepath, 'lecture_id': lectureid, 'video_id': video_id, 'import-chapters': True})
50

Andreas Valder's avatar
Andreas Valder committed
51
def schedule_thumbnail(lectureid, filePath=None):
Andreas Valder's avatar
Andreas Valder committed
52
53
54
55
56
57
58
59
	videos = query('''
			SELECT videos.path
			FROM videos
			JOIN formats ON (videos.video_format = formats.id)
			WHERE videos.lecture_id = ?
			ORDER BY formats.prio DESC''', lectureid )
	if videos:
		path = videos[0]['path']
Andreas Valder's avatar
Andreas Valder committed
60
61
	elif filePath:
		path = filePath
Andreas Valder's avatar
Andreas Valder committed
62
63
64
65
	else:
		return
	data = '{"lectureid": "'+str(lectureid)+'", "path": "'+path+'"}'
	query('INSERT INTO jobs (type, data, time_created) VALUES ("thumbnail", ?, ?)', data, datetime.now());
66

Julian Rother's avatar
Julian Rother committed
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
@job_handler('transcode')
def insert_transcoded_video(jobid, jobtype, data, state, status):
	if 'lecture_id' not in data or 'source_id' not in data or 'format_id' not in data:
		return
	if 'video_id' in data:
		return
	video_id = modify('''INSERT INTO videos_data 
		(lecture_id, visible, path, video_format, title, comment, internal, file_modified, time_created, time_updated, created_by, hash, file_size, source)
		VALUES 
		(?, 0, ?, ?, "", "", "", ?, ?, ?, ?, ?, ?, ?)''',
		data['lecture_id'], data['output']['path'], data['format_id'],
		datetime.now(), datetime.now(), datetime.now(), -1, status['hash'],
		status['filesize'], data['source_id'])
	schedule_thumbnail(data['lecture_id'])

82
83
84
85
86
87
def sort_file(filename, course=None, lectures=None):
	# filenames: <handle>-<sorter>-<format>.mp4
	# "sorter" musst be found with fuzzy matching. "sorter" musst be one or more of the following types: (inside the loop)
	# '_' and ' ' are handled like '-'
	splitfilename = filename.replace('_','-').replace(' ','-').split('-')
	if not course:
88
89
90
		handle = splitfilename[0]
		if splitfilename[0].endswith('ws') or splitfilename[0].endswith('ss'):
			handle = '-'.join(splitfilename[:2])
91
92
		courses = query('SELECT * FROM courses WHERE handle = ?', handle)
		if not courses:
93
			return [], 0
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
		course = courses[0]
	if not lectures:
		lectures = query('SELECT * from lectures where course_id = ?', course['id'])
	# we save all extraced data in a dict
	data = {'keywords': []}
	# parse the file name and save all data in 'data'
	for s in splitfilename:
		s = s.replace('.mp4','')
		#-<YYMMDD> (date)
		#-<HHMM> (time)
		#-<keyword>
		#	Looking for keywords in: title,speaker,comment, comma seperated list in internal
		try:
			if len(s) == 6:
				data['date'] = datetime.strptime(s,'%y%m%d').date()
			elif  len(s) == 4:
				data['time'] = datetime.strptime(s,'%H%M').time()
			else:	
				data['keywords'].append(s)
		except ValueError:
			# if its not a date or time, handle it as keyword
			data['keywords'].append(s)
	# try to match the file on a single lecture
	matches = []
	# first try date and time (if one of them is set)
	if ('date' in data) or ('time' in data):
		for lecture in lectures:
			if not ('time' in lecture) or not lecture['time']:
				continue
			if ('date' in data) and (lecture['time'].date() != data['date']):
					continue
			if ('time' in data) and (lecture['time'].time() != data['time']):
					continue
			matches.append(lecture)
	# if we can't match exactly  based on date and time, we have to match keywords
	if ((len(matches) != 1) and (len(data['keywords']) > 0)):
		#only test lectures with the correct date/time, if we have any. Else test for matches in all lectures of this course
		if len(matches) == 0:
			matches.extend(lectures)
		found = False
		for field in ['title','speaker','comment','internal']:
			for lecture in matches:
				for keyword in data['keywords']:
					# first test for exact match, else make it asci and try substring test
					if (keyword == lecture[field]) or \
						 (str(keyword).lower() in str(to_ascii(lecture[field]).lower())):
						found = True
						matches = [lecture]
					if found:
						break
				if found:
					break
			if found:
				break
	# now we should have found exactly one match
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
	# default format is "unknown", with id 0
	fmt = 0
	formats = query('SELECT * FROM formats ORDER BY prio DESC')
	for videoformat in formats:
		# we match the last part of the file name without the extension
		formatstring = splitfilename[-1].split('.',1)[0].lower()
		if formatstring in videoformat['keywords'].replace(',',' ').split(' '):
			fmt = videoformat['id']
			break
	return matches, fmt

def log_sort_error(course_id, path, matches):
	matches_id = []
	for match in matches:
		matches_id.append(str(match['id']))
	query('INSERT INTO sorterrorlog_data (course_id, path, matches, `when`, time_updated, time_created) VALUES (?, ?, ?, ?, ?, ?)',
			course_id, path, ','.join(matches_id), datetime.now(), datetime.now(), datetime.now())

def sort_api_token_required(func):
	@wraps(func)
	def decorator(*args, **kwargs):
		if 'apikey' in request.values:
			token = request.values['apikey']
		elif request.get_json() and ('apikey' in request.get_json()):
			token = request.get_json()['apikey']
		else:
			token = None
		if not token == config.get('SORTER_API_KEY', [None]):
			return 'Permission denied', 403
		else:
			return func(*args, **kwargs)
	return decorator

@app.route('/internal/sort/encoded/<filename>')
@sort_api_token_required
def sort_encoded(filename):
	matches, fmt = sort_file(filename)
	if len(matches) != 1:
		log_sort_error(-1, 'kodiert/'+filename, matches)
		return "Could not match filename", 400
	lecture = matches[0]
	course = query('SELECT * FROM courses WHERE id = ?', lecture['course_id'])[0]
	if course['autopublish']:
		schedule_job('publish_video', {'source': filename, 'path': 'pub/'+course['handle']+'/'+filename, 'lecture_id': lecture['id'], 'format_id': fmt})
	return 'OK', 200

Julian Rother's avatar
Julian Rother committed
195
196
197
198
199
200
201
202
203
204
205
206
207
@app.route('/internal/sort/autoencode')
@sort_api_token_required
def sort_autoencode():
	filename = request.values['path']
	path = 'autoencode/'+filename
	matches, fmt = sort_file(filename)
	if len(matches) != 1:
		log_sort_error(-1, 'raw/'+path, matches)
		return "Could not match filename", 400
	lecture = matches[0]
	schedule_job('probe-raw', {'path': path, 'lecture_id': lecture['id'], 'import-chapters': True})
	return 'OK', 200

208
209
210
211
212
@job_handler('publish_video')
def handle_published_video(jobid, jobtype, data, state, status):
	if 'lecture_id' not in data or 'format_id' not in data:
		return
	insert_video(data['lecture_id'], data['path'], data['format_id'], hash=status['hash'], filesize=status['filesize'])
213

214
@app.route('/internal/sort/now')
215
@mod_required
Andreas Valder's avatar
Andreas Valder committed
216
@sched_func(600)
217
218
def sort_now():
	courses = query('SELECT * FROM courses')
219
	for course in courses:
Andreas Valder's avatar
Andreas Valder committed
220
		modify('BEGIN')
Andreas Valder's avatar
Andreas Valder committed
221
		for mountpoint in config['VIDEOMOUNT']:
222
223
224
225
226
			existingvideos = query('SELECT videos.path FROM videos JOIN lectures ON (videos.lecture_id = lectures.id) WHERE lectures.course_id = ?',course['id'])
			knownerrors = query('SELECT sorterrorlog.path FROM sorterrorlog WHERE sorterrorlog.course_id = ?',course['id'])
			ignorefiles = existingvideos + knownerrors
			lectures = query('SELECT * from lectures where course_id = ?',course['id'])
			coursepath = mountpoint['mountpoint']+course['handle']
227
228
229
230
			try:
				files = os.listdir(coursepath)
			except FileNotFoundError:
				files = []
231
			for filename in files:
232
				try:
233
234
235
236
					# if the video is in the table "videos" already (with the correct course), skip it
					ignore = False
					for file_to_ignore in ignorefiles:
						# path is something like
237
						# vpnonline/08ws-swt/08ws-swt-081118.mp4
238
239
						if os.path.basename(filename) == os.path.basename(file_to_ignore['path']):
							ignore = True
240
							break
241
					if ignore:
242
						continue
243
					if not os.path.splitext(filename)[1] == '.mp4':
244
						continue
245
					matches, fmt = sort_file(filename, course=course, lectures=lectures)
246
					dbfilepath = mountpoint['prefix']+course['handle']+'/'+filename
247
					if len(matches) == 1:
248
						insert_video(matches[0]['id'], dbfilepath, fmt)
Andreas Valder's avatar
Andreas Valder committed
249
					else:
250
						log_sort_error(course['id'], dbfilepath, matches)
251
252
				except Exception:
					traceback.print_exc()
Andreas Valder's avatar
Andreas Valder committed
253
		modify('COMMIT')
254
255
256
	if 'ref' in request.values:
		return redirect(request.values['ref'])
	else:
257
		return 'OK', 200
Andreas Valder's avatar
Andreas Valder committed
258