1
|
|
|
import logging |
2
|
|
|
import os |
3
|
|
|
|
4
|
|
|
from pathlib import Path |
5
|
|
|
from yurl import URL |
6
|
|
|
from youtube_dl import YoutubeDL |
7
|
|
|
from hashlib import md5 |
8
|
|
|
|
9
|
|
|
from tumdlr.downloader import sanitize_filename, download |
10
|
|
|
from tumdlr.errors import TumdlrDownloadError, TumdlrParserError |
11
|
|
|
|
12
|
|
|
|
13
|
|
|
class TumblrPost: |
14
|
|
|
""" |
15
|
|
|
This is the base container class for all Tumblr post types. It contains data that is always available with any |
16
|
|
|
type of post. |
17
|
|
|
|
18
|
|
|
Additional supported post types may extend this class to provide additional metadata parsing |
19
|
|
|
""" |
20
|
|
|
def __init__(self, post, blog): |
21
|
|
|
""" |
22
|
|
|
Args: |
23
|
|
|
post(dict): API response |
24
|
|
|
blog(tumdlr.api.TumblrBlog): Parent blog |
25
|
|
|
""" |
26
|
|
|
self._post = post |
27
|
|
|
self.blog = blog |
28
|
|
|
self.log = logging.getLogger('tumdlr.containers.post') |
29
|
|
|
|
30
|
|
|
self.id = None # type: int |
31
|
|
|
self.type = None # type: str |
32
|
|
|
self.url = None # type: URL |
33
|
|
|
self.tags = set() |
34
|
|
|
self.post_date = None # type: str |
35
|
|
|
self.note_count = None # type: int |
36
|
|
|
|
37
|
|
|
self.files = [] |
38
|
|
|
|
39
|
|
|
try: |
40
|
|
|
self._parse_post() |
41
|
|
|
except Exception as e: |
42
|
|
|
self.log.warn('Failed to parse post data: %r', self, exc_info=e) |
43
|
|
|
raise TumdlrParserError(post_data=post) |
44
|
|
|
|
45
|
|
|
@property |
46
|
|
|
def is_text(self): |
47
|
|
|
""" |
48
|
|
|
Returns: |
49
|
|
|
bool |
50
|
|
|
""" |
51
|
|
|
return self.type == 'text' |
52
|
|
|
|
53
|
|
|
@property |
54
|
|
|
def is_photo(self): |
55
|
|
|
""" |
56
|
|
|
Returns: |
57
|
|
|
bool |
58
|
|
|
""" |
59
|
|
|
return self.type in ['photo', 'link'] |
60
|
|
|
|
61
|
|
|
@property |
62
|
|
|
def is_video(self): |
63
|
|
|
""" |
64
|
|
|
Returns: |
65
|
|
|
bool |
66
|
|
|
""" |
67
|
|
|
return self.type == 'video' |
68
|
|
|
|
69
|
|
|
def _parse_post(self): |
70
|
|
|
self.id = self._post['id'] |
71
|
|
|
self.type = self._post['type'] |
72
|
|
|
self.url = URL(self._post['post_url']) if 'post_url' in self._post else None |
73
|
|
|
self.tags = set(self._post.get('tags', [])) |
74
|
|
|
self.note_count = self._post.get('note_count') |
75
|
|
|
self.post_date = self._post['date'] |
76
|
|
|
|
77
|
|
|
def __repr__(self): |
78
|
|
|
return "<TumblrPost id='{id}' type='{type}' url='{url}'>"\ |
79
|
|
|
.format(id=self.id, type=self.type, url=self.url) |
80
|
|
|
|
81
|
|
|
def __str__(self): |
82
|
|
|
return self.url.as_string() if self.url else '' |
83
|
|
|
|
84
|
|
|
|
85
|
|
|
class TumblrPhotoSet(TumblrPost): |
86
|
|
|
""" |
87
|
|
|
Container class for Photo and Photo Link post types |
88
|
|
|
""" |
89
|
|
|
def __init__(self, post, blog): |
90
|
|
|
""" |
91
|
|
|
Args: |
92
|
|
|
post(dict): API response |
93
|
|
|
blog(tumdlr.api.blog.TumblrBlog): Parent blog |
94
|
|
|
""" |
95
|
|
|
self.log = logging.getLogger('tumdlr.containers.post') |
96
|
|
|
super().__init__(post, blog) |
97
|
|
|
|
98
|
|
|
def _parse_post(self): |
99
|
|
|
""" |
100
|
|
|
Parse all available photos using the best image sizes available |
101
|
|
|
""" |
102
|
|
|
super()._parse_post() |
103
|
|
|
self.title = self._post.get('caption', self._post.get('title')) |
104
|
|
|
|
105
|
|
|
photos = self._post.get('photos', []) |
106
|
|
|
is_photoset = (len(photos) > 1) |
107
|
|
|
|
108
|
|
|
for page_no, photo in enumerate(photos, 1): |
109
|
|
|
best_size = photo.get('original_size') or max(photo['alt_sizes'], key='width') |
110
|
|
|
best_size['page_no'] = page_no if is_photoset else False |
111
|
|
|
self.files.append(TumblrPhoto(best_size, self)) |
112
|
|
|
|
113
|
|
|
def __repr__(self): |
114
|
|
|
return "<TumblrPhotoSet title='{title}' id='{id}' photos='{count}'>"\ |
115
|
|
|
.format(title=self.title.split("\n")[0].strip(), id=self.id, count=len(self.files)) |
116
|
|
|
|
117
|
|
|
|
118
|
|
|
class TumblrVideoPost(TumblrPost): |
119
|
|
|
""" |
120
|
|
|
Container class for Video post types |
121
|
|
|
""" |
122
|
|
|
def __init__(self, post, blog): |
123
|
|
|
""" |
124
|
|
|
Args: |
125
|
|
|
post(dict): API response |
126
|
|
|
blog(tumdlr.api.blog.TumblrBlog): Parent blog |
127
|
|
|
""" |
128
|
|
|
self.log = logging.getLogger('tumdlr.containers.post') |
129
|
|
|
|
130
|
|
|
self.title = None |
131
|
|
|
self.description = None |
132
|
|
|
self.duration = None |
133
|
|
|
self.format = None |
134
|
|
|
|
135
|
|
|
super().__init__(post, blog) |
136
|
|
|
|
137
|
|
|
def _parse_post(self): |
138
|
|
|
""" |
139
|
|
|
Parse all available photos using the best image sizes available |
140
|
|
|
""" |
141
|
|
|
super()._parse_post() |
142
|
|
|
|
143
|
|
|
video_info = YoutubeDL().extract_info(self.url.as_string(), False) |
144
|
|
|
|
145
|
|
|
self.title = video_info.get('title') |
146
|
|
|
|
147
|
|
|
self.description = video_info.get('description') |
148
|
|
|
self.duration = int(video_info.get('duration', 0)) |
149
|
|
|
self.format = video_info.get('format', 'Unknown') |
150
|
|
|
|
151
|
|
|
self.files.append(TumblrVideo(video_info, self)) |
152
|
|
|
|
153
|
|
|
def __repr__(self): |
154
|
|
|
return "<TumblrVideoPost id='{id}'>".format(id=self.id) |
155
|
|
|
|
156
|
|
|
|
157
|
|
|
class TumblrFile: |
158
|
|
|
""" |
159
|
|
|
This is the base container class for all downloadable resources associated with Tumblr posts. |
160
|
|
|
""" |
161
|
|
|
|
162
|
|
|
CATEGORY = 'misc' |
163
|
|
|
|
164
|
|
|
def __init__(self, data, container): |
165
|
|
|
""" |
166
|
|
|
Args: |
167
|
|
|
data(dict): API response data |
168
|
|
|
container(TumblrPost): Parent container |
169
|
|
|
""" |
170
|
|
|
self.log = logging.getLogger('tumdlr.containers.file') |
171
|
|
|
|
172
|
|
|
self._data = data |
173
|
|
|
self.container = container |
174
|
|
|
self.url = URL(self._data.get('url', self._data.get('post_url'))) |
175
|
|
|
|
176
|
|
|
def download(self, context, **kwargs): |
177
|
|
|
""" |
178
|
|
|
Args: |
179
|
|
|
context(tumdlr.main.Context): CLI request context |
180
|
|
|
kwargs(dict): Additional arguments to send with the download request |
181
|
|
|
|
182
|
|
|
Returns: |
183
|
|
|
str: Path to the saved file |
184
|
|
|
""" |
185
|
|
|
try: |
186
|
|
|
download(self.url.as_string(), str(self.filepath(context, kwargs)), **kwargs) |
187
|
|
|
except Exception as e: |
188
|
|
|
self.log.warn('Post download failed: %r', self, exc_info=e) |
189
|
|
|
raise TumdlrDownloadError(error_message=str(e), download_url=self.url.as_string()) |
190
|
|
|
|
191
|
|
|
def filepath(self, context, request_data): |
192
|
|
|
""" |
193
|
|
|
Args: |
194
|
|
|
context(tumdlr.main.Context): CLI request context |
195
|
|
|
request_data(Optional[dict]): Additional arguments to send with the download request |
196
|
|
|
|
197
|
|
|
Returns: |
198
|
|
|
Path |
199
|
|
|
""" |
200
|
|
|
# Construct the save basedir |
201
|
|
|
basedir = Path(context.config['Tumdlr']['SavePath']) |
202
|
|
|
|
203
|
|
|
# Are we categorizing by user? |
204
|
|
|
if context.config['Categorization']['User']: |
205
|
|
|
self.log.debug('Categorizing by user: %s', self.container.blog.name) |
206
|
|
|
basedir = basedir.joinpath(sanitize_filename(self.container.blog.name)) |
207
|
|
|
|
208
|
|
|
# Are we categorizing by post type? |
209
|
|
|
if context.config['Categorization']['PostType']: |
210
|
|
|
self.log.debug('Categorizing by type: %s', self.CATEGORY) |
211
|
|
|
basedir = basedir.joinpath(self.CATEGORY) |
212
|
|
|
|
213
|
|
|
self.log.debug('Basedir constructed: %s', basedir) |
214
|
|
|
|
215
|
|
|
return basedir |
216
|
|
|
|
217
|
|
|
|
218
|
|
|
class TumblrPhoto(TumblrFile): |
219
|
|
|
|
220
|
|
|
CATEGORY = 'photos' |
221
|
|
|
|
222
|
|
|
def __init__(self, photo, photoset): |
223
|
|
|
""" |
224
|
|
|
Args: |
225
|
|
|
photo(dict): Photo API data |
226
|
|
|
photoset(TumblrPhotoSet): Parent container |
227
|
|
|
""" |
228
|
|
|
super().__init__(photo, photoset) |
229
|
|
|
|
230
|
|
|
self.width = self._data.get('width') |
231
|
|
|
self.height = self._data.get('height') |
232
|
|
|
self.page_no = self._data.get('page_no', False) |
233
|
|
|
|
234
|
|
|
def filepath(self, context, request_data): |
235
|
|
|
""" |
236
|
|
|
Get the full file path to save the downloaded file to |
237
|
|
|
|
238
|
|
|
Args: |
239
|
|
|
context(tumdlr.main.Context): CLI request context |
240
|
|
|
request_data(Optional[dict]): Additional arguments to send with the download request |
241
|
|
|
|
242
|
|
|
Returns: |
243
|
|
|
Path |
244
|
|
|
""" |
245
|
|
|
assert isinstance(self.container, TumblrPhotoSet) |
246
|
|
|
filepath = super().filepath(context, request_data) |
247
|
|
|
|
248
|
|
|
request_data['progress_data']['Caption'] = self.container.title |
249
|
|
|
|
250
|
|
|
# Are we categorizing by photosets? |
251
|
|
|
if self.page_no and context.config['Categorization']['Photosets']: |
252
|
|
|
self.log.debug('Categorizing by photoset: %s', self.container.id) |
253
|
|
|
filepath = filepath.joinpath(sanitize_filename(str(self.container.id))) |
254
|
|
|
|
255
|
|
|
# Prepend the page number for photosets |
256
|
|
|
if self.page_no: |
257
|
|
|
filepath = filepath.joinpath(sanitize_filename('p{pn}_{pt}'.format(pn=self.page_no, |
258
|
|
|
pt=self.container.title))) |
259
|
|
|
request_data['progress_data']['Photoset Page'] = '{cur} / {tot}'\ |
260
|
|
|
.format(cur=self.page_no, tot=len(self.container.files)) |
261
|
|
|
else: |
262
|
|
|
filepath = filepath.joinpath(sanitize_filename(self.container.title)) |
263
|
|
|
|
264
|
|
|
# Work out the file extension and return |
265
|
|
|
return str(filepath) + os.path.splitext(self.url.as_string())[1] |
266
|
|
|
|
267
|
|
|
def __repr__(self): |
268
|
|
|
return "<TumblrPhoto url='{url}' width='{w}' height='{h}'>".format(url=self.url, w=self.width, h=self.height) |
269
|
|
|
|
270
|
|
|
def __str__(self): |
271
|
|
|
return self.url.as_string() |
272
|
|
|
|
273
|
|
|
|
274
|
|
|
class TumblrVideo(TumblrFile): |
275
|
|
|
|
276
|
|
|
CATEGORY = 'videos' |
277
|
|
|
|
278
|
|
|
def __init__(self, video, vpost): |
279
|
|
|
""" |
280
|
|
|
Args: |
281
|
|
|
video(dict): Video API data |
282
|
|
|
vpost(TumblrVideoPost): Parent container |
283
|
|
|
""" |
284
|
|
|
super().__init__(video, vpost) |
285
|
|
|
|
286
|
|
|
def filepath(self, context, request_data): |
287
|
|
|
""" |
288
|
|
|
Get the full file path to save the video to |
289
|
|
|
|
290
|
|
|
Args: |
291
|
|
|
context(tumdlr.main.Context): CLI request context |
292
|
|
|
request_data(Optional[dict]): Additional arguments to send with the download request |
293
|
|
|
|
294
|
|
|
Returns: |
295
|
|
|
Path |
296
|
|
|
""" |
297
|
|
|
assert isinstance(self.container, TumblrVideoPost) |
298
|
|
|
filepath = super().filepath(context, request_data) |
299
|
|
|
|
300
|
|
|
minutes = int(self.container.duration / 60) |
301
|
|
|
seconds = self.container.duration % 60 |
302
|
|
|
duration = '{} minutes {} seconds'.format(minutes, seconds) if minutes else '{} seconds'.format(seconds) |
303
|
|
|
|
304
|
|
|
if self.container.title: |
305
|
|
|
request_data['progress_data']['Title'] = self.container.title |
306
|
|
|
|
307
|
|
|
request_data['progress_data']['Description'] = self.container.description |
308
|
|
|
request_data['progress_data']['Duration'] = duration |
309
|
|
|
request_data['progress_data']['Format'] = self.container.format |
310
|
|
|
|
311
|
|
|
filepath = filepath.joinpath(sanitize_filename( |
312
|
|
|
self.container.description or |
313
|
|
|
md5(self.url.as_string().encode('utf-8')).hexdigest()) |
314
|
|
|
) |
315
|
|
|
|
316
|
|
|
# Work out the file extension and return |
317
|
|
|
return '{}.{}'.format(str(filepath), self._data.get('ext', 'mp4')) |
318
|
|
|
|
319
|
|
|
def __repr__(self): |
320
|
|
|
return "<TumblrVideo id='{i}'>".format(i=self.container.id) |
321
|
|
|
|
322
|
|
|
def __str__(self): |
323
|
|
|
return self.url.as_string() |
324
|
|
|
|