sorter.py 8.09 KB
Newer Older
Andreas Valder's avatar
Andreas Valder committed
1
from server import *
2
import traceback
Andreas Valder's avatar
Andreas Valder committed
3

4
@app.route('/internal/sort/log')
5
6
7
8
9
10
11
12
13
14
15
16
17
@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 
18
			ORDER BY sortlog.`when` DESC
Andreas Valder's avatar
Andreas Valder committed
19
			LIMIT 50
20
		'''),sorterrorlog=query('SELECT * FROM sorterrorlog ORDER BY sorterrorlog.`when` DESC'))
21

22
23
24
25
26
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
27

28
def insert_video(lectureid, dbfilepath, fileformatid, hash="", filesize=-1):
29
	video_id = modify('''INSERT INTO videos_data 
30
		(lecture_id, visible, path, video_format, title, comment, internal, file_modified, time_created, time_updated, created_by, hash, file_size)
31
		VALUES 
32
33
		(?, 0, ?, ?, "", "", "", ?, ?, ?, ?, ?, ?)''',
		lectureid, dbfilepath, fileformatid, datetime.now(), datetime.now(), datetime.now(), -1, hash, filesize)
34
	query('INSERT INTO sortlog (lecture_id,video_id,path,`when`) VALUES (?,?,?,?)', lectureid, video_id, dbfilepath, datetime.now())
Andreas Valder's avatar
Andreas Valder committed
35
	schedule_thumbnail(lectureid)
Julian Rother's avatar
Julian Rother committed
36
	schedule_job('probe', {'path': dbfilepath, 'lecture_id': lectureid, 'video_id': video_id, 'import-chapters': True})
37

Andreas Valder's avatar
Andreas Valder committed
38
def schedule_thumbnail(lectureid, filePath=None):
Andreas Valder's avatar
Andreas Valder committed
39
40
41
42
43
44
45
46
	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
47
48
	elif filePath:
		path = filePath
Andreas Valder's avatar
Andreas Valder committed
49
50
51
52
	else:
		return
	data = '{"lectureid": "'+str(lectureid)+'", "path": "'+path+'"}'
	query('INSERT INTO jobs (type, data, time_created) VALUES ("thumbnail", ?, ?)', data, datetime.now());
53

54
55
56
57
58
59
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:
60
61
62
		handle = splitfilename[0]
		if splitfilename[0].endswith('ws') or splitfilename[0].endswith('ss'):
			handle = '-'.join(splitfilename[:2])
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
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
		courses = query('SELECT * FROM courses WHERE handle = ?', handle)
		if not courses:
			return []
		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
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
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
	# 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

@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'])
172

173
@app.route('/internal/sort/now')
174
@mod_required
Andreas Valder's avatar
Andreas Valder committed
175
@sched_func(600)
176
177
def sort_now():
	courses = query('SELECT * FROM courses')
178
	for course in courses:
Andreas Valder's avatar
Andreas Valder committed
179
		modify('BEGIN')
Andreas Valder's avatar
Andreas Valder committed
180
		for mountpoint in config['VIDEOMOUNT']:
181
182
183
184
185
			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']
186
187
188
189
			try:
				files = os.listdir(coursepath)
			except FileNotFoundError:
				files = []
190
			for filename in files:
191
				try:
192
193
194
195
					# 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
196
						# vpnonline/08ws-swt/08ws-swt-081118.mp4
197
198
						if os.path.basename(filename) == os.path.basename(file_to_ignore['path']):
							ignore = True
199
							break
200
					if ignore:
201
						continue
202
					if not os.path.splitext(filename)[1] == '.mp4':
203
						continue
204
					matches, fmt = sort_file(filename, course=course, lectures=lectures)
205
					dbfilepath = mountpoint['prefix']+course['handle']+'/'+filename
206
					if len(matches) == 1:
207
						insert_video(matches[0]['id'], dbfilepath, fmt)
Andreas Valder's avatar
Andreas Valder committed
208
					else:
209
						log_sort_error(course['id'], dbfilepath, matches)
210
211
				except Exception:
					traceback.print_exc()
Andreas Valder's avatar
Andreas Valder committed
212
		modify('COMMIT')
213
214
215
	if 'ref' in request.values:
		return redirect(request.values['ref'])
	else:
216
		return 'OK', 200
Andreas Valder's avatar
Andreas Valder committed
217