mirror of
https://github.com/lrnselfreliance/wrolpi.git
synced 2026-10-26 19:01:39 -04:00
* Videos are marked as validated so they will not be tried again. All previous bulk methods are combined into one. * Passing json to fixture factory to avoid writing the files in every test. * Adding log to explain if refresh is slow. * Generating video poster if one cannot be found. * Generating video poster if no poster can be found. * Videos without a video file (yet to be downloaded) are not validated. * Video title is extracted from the video file name as a last resort. * Keeping existing view count because it may be from the channel, which will be more recent. * Downloads should only be downloaded after they are due, not the same day. * Setting manager on new downloads. * Checking if downloading is disabled in loop. * Adding log when refresh is complete. * Not generating posters by default. This is to avoid deleting files. * Not converting posters unless channel has that enabled. * Not logging on valid because all posters will be checked. * Validating videos in upsert. * Using video name parser in upsert. * Using validate_video in upsert so that data is only read once from disk. * Attaching manager in download_manager.get_download. * Adding 22 to preferred video format for odd 720p resolution. Signed-off-by: Roland <roland@learningselfreliance.com>
93 lines
4.3 KiB
Python
93 lines
4.3 KiB
Python
from pathlib import Path
|
|
|
|
from modules.videos import captions
|
|
from modules.videos.models import Video
|
|
from wrolpi.db import get_db_curs, get_db_context
|
|
from wrolpi.test.common import wrap_test_db, TestAPI
|
|
from wrolpi.vars import PROJECT_DIR
|
|
|
|
|
|
class TestCaption(TestAPI):
|
|
|
|
def setUp(self) -> None:
|
|
super().setUp()
|
|
self.vtt_path1 = Path(self.tmp_dir.name) / 'example1.en.vtt'
|
|
self.vtt_path1.touch()
|
|
self.vtt_path1.write_text((PROJECT_DIR / 'test/example1.en.vtt').read_text())
|
|
self.vtt_path2 = Path(self.tmp_dir.name) / 'example2.en.vtt'
|
|
self.vtt_path2.touch()
|
|
self.vtt_path2.write_text((PROJECT_DIR / 'test/example2.en.vtt').read_text())
|
|
|
|
def test_example_en_vtt(self):
|
|
expected = ['okay welcome to this session this is',
|
|
'okay welcome to this session this is',
|
|
'okay welcome to this session this is\ncalled the kinetic bunny need to meet',
|
|
'called the kinetic bunny need to meet',
|
|
'called the kinetic bunny need to meet\nthe virtual bonnie emilio stop this is',
|
|
'the virtual bonnie emilio stop this is',
|
|
'the virtual bonnie emilio stop this is\njana car who just to give a little bit',
|
|
'jana car who just to give a little bit',
|
|
'jana car who just to give a little bit\nof background information we come from a',
|
|
'of background information we come from a',
|
|
'of background information we come from a\ncompany called le code from finland and']
|
|
self.assertEqual(list(captions.get_caption_text(self.vtt_path1)), expected)
|
|
|
|
expected = ['okay welcome to this session this is',
|
|
'called the kinetic bunny need to meet',
|
|
'the virtual bonnie emilio stop this is',
|
|
'jana car who just to give a little bit',
|
|
'of background information we come from a',
|
|
'company called le code from finland and']
|
|
self.assertEqual(list(captions.get_unique_caption_lines(self.vtt_path1)), expected)
|
|
|
|
@wrap_test_db
|
|
def test_process_captions(self):
|
|
_, session = get_db_context()
|
|
video1 = Video(title='scream', caption_path=str(self.vtt_path1))
|
|
session.add(video1)
|
|
session.commit()
|
|
video1.caption = captions.get_video_captions(video1)
|
|
video2 = Video(title='bar', caption_path=str(self.vtt_path2))
|
|
session.add(video2)
|
|
session.commit()
|
|
video2.caption = captions.get_video_captions(video2)
|
|
|
|
session.commit()
|
|
session.flush()
|
|
session.refresh(video1)
|
|
session.refresh(video2)
|
|
|
|
# Get the video from the DB
|
|
video1 = session.query(Video).filter_by(id=video1.id).one()
|
|
self.assertIsNotNone(video1.caption)
|
|
video2 = session.query(Video).filter_by(id=video2.id).one()
|
|
self.assertIsNotNone(video2.caption)
|
|
|
|
# Search using the tsvector, "sessions" never actually appears in the text, but "session" does
|
|
with get_db_curs() as curs:
|
|
def select_textsearch(*args):
|
|
curs.execute('SELECT id FROM video WHERE textsearch @@ to_tsquery(%s) ORDER BY id', args)
|
|
|
|
select_textsearch('sessions')
|
|
self.assertEqual(curs.fetchall(), [[1, ]])
|
|
# Matches video1.title and video2.caption
|
|
select_textsearch('scream')
|
|
self.assertEqual(curs.fetchall(), [[1, ], [2, ]])
|
|
# Matches video1.title and video2.caption
|
|
select_textsearch('scream | sessions')
|
|
self.assertEqual(curs.fetchall(), [[1, ], [2, ]])
|
|
# Only matches video1.title
|
|
select_textsearch('scream & sessions')
|
|
self.assertEqual(curs.fetchall(), [[1, ]])
|
|
# Matches neither
|
|
select_textsearch('scream & sess')
|
|
self.assertEqual(curs.fetchall(), [])
|
|
# Matches video2.caption
|
|
select_textsearch('yawn | sess')
|
|
self.assertEqual(curs.fetchall(), [[2, ]])
|
|
# Matches video2.caption
|
|
select_textsearch('yawn')
|
|
self.assertEqual(curs.fetchall(), [[2, ]])
|
|
# Matches video2.title
|
|
select_textsearch('bar')
|
|
self.assertEqual(curs.fetchall(), [[2, ]])
|