def test_set_language():
sub1 = SubtitleFile('path', language='en')
sub2 = SubtitleFile('path', language=languages.getlang('es'))
assert isinstance(sub1.language,
str), "Subtitles must be converted to Language class"
assert isinstance(sub2.language,
str), "Subtitles can be passed as Langauge models"
assert sub1.language == 'en', "Subtitles must have a language"
assert sub2.language == 'es', "Subtitles must have a language"
pytest.raises(TypeError, SubtitleFile, 'path', language='notalanguage')
def test_convertible_substitles_ar_ttml(youtube_test_file):
"""
Regression test to make sure correct lang_code is detected from .ttml data.
"""
local_path = os.path.join("tests", "testcontent", "downloaded",
"testsubtitles_ar.ttml")
assert os.path.exists(local_path)
subtitle_file = SubtitleFile(local_path, language='ar')
filename = subtitle_file.process_file()
assert filename, 'conferted filename must exit'
assert filename.endswith(
'.vtt'), 'conferted filename must have .vtt extension'
def test_multiple_subs_can_be_added(video_file):
"""
Baseline check to make sure we're not dropping subtitle files on validate.
"""
assert os.path.exists("tests/testcontent/testsubtitles_ar.srt")
video_node = VideoNode('vid-src-id', "Video", licenses.PUBLIC_DOMAIN)
video_node.add_file(video_file)
sub1 = SubtitleFile("tests/testcontent/testsubtitles_ar.srt", language='en')
video_node.add_file(sub1)
sub2 = SubtitleFile("tests/testcontent/testsubtitles_ar.srt", language='ar')
video_node.add_file(sub2)
video_node.validate()
sub_files = [f for f in video_node.files if isinstance(f, SubtitleFile)]
assert len(sub_files) == 2, 'Missing subtitles files!'
def test_duplicate_language_codes_fixed_by_validate(video_file):
"""
Video nodes should have at most one subtitle file for a particular lang code.
"""
assert os.path.exists("tests/testcontent/testsubtitles_ar.srt")
video_node = VideoNode('vid-src-id', "Video", licenses.PUBLIC_DOMAIN)
video_node.add_file(video_file)
sub1 = SubtitleFile("tests/testcontent/testsubtitles_ar.srt", language='ar')
video_node.add_file(sub1)
# now let's add file with a duplicate language code...
sub2 = SubtitleFile("tests/testcontent/testsubtitles_ar.srt", language='ar')
video_node.add_file(sub2)
video_node.validate()
sub_files = [f for f in video_node.files if isinstance(f, SubtitleFile)]
assert len(sub_files) == 1, 'Duplicate subtitles files not removed!'
def construct_channel(self, *args, **kwargs):
channel = self.get_channel(**kwargs)
videos_topic = TopicNode(source_id="/wiki/Category:Articles_containing_video_clips",
title="Articles containing video clips")
channel.add_child(videos_topic)
thumbnail_url = 'https://upload.wikimedia.org/wikipedia/commons/thumb/e/ee/A_Is_for_Atom_1953.webm/220px--A_Is_for_Atom_1953.webm.jpg'
page = download_wikipedia_page('/wiki/Category:Articles_containing_video_clips',
thumbnail_url, 'A Is for Atom')
videos_topic.add_child(page)
video_url = 'https://upload.wikimedia.org/wikipedia/commons/e/ee/A_Is_for_Atom_1953.webm'
video_file = VideoFile(path=video_url)
video_node = VideoNode(title='A Is for Atom 1953', source_id='A_Is_for_Atom_1953.webm',
files=[video_file], license=licenses.PublicDomainLicense())
subtitle_url = 'https://commons.wikimedia.org/w/api.php?action=timedtext&title=File%3AA_Is_for_Atom_1953.webm&lang={}&trackformat=srt'
subtitle_languages = [
'en',
'es',
]
for lang in subtitle_languages:
subtitle_file = SubtitleFile(path=subtitle_url.format(lang), language=lang, subtitlesformat='srt')
video_node.add_file(subtitle_file)
videos_topic.add_child(video_node)
return channel
def test_convertible_substitles_ar_srt():
"""
Basic check that srt --> vtt conversion works.
"""
local_path = os.path.join("tests", "testcontent", "samples",
"testsubtitles_ar.srt")
assert os.path.exists(local_path)
subtitle_file = SubtitleFile(local_path, language='ar')
filename = subtitle_file.process_file()
assert filename, 'converted filename must exist'
assert filename.endswith(
'.vtt'), 'converted filename must have .vtt extension'
storage_path = config.get_storage_path(filename)
with open(storage_path) as converted_vtt:
filecontents = converted_vtt.read()
check_words = 'لناس على'
assert check_words in filecontents, 'missing check word in converted subs'
def subtitle_file():
local_path = os.path.join("tests", "testcontent", "generated",
"testsubtitles.vtt")
if not os.path.exists(local_path):
with open(local_path, 'wb') as subtitlefile:
subtitlefile.write(b'WEBVTT\n')
subtitlefile.write(b'\n')
subtitlefile.write(b'00:01.000 --> 00:04.250\n')
subtitlefile.write(b'Testing subtitles\n')
return SubtitleFile(local_path, language='en')
def test_convertible_substitles_noext_subtitlesformat():
"""
Check that we handle correctly cases when path doesn't contain extenstion.
"""
local_path = os.path.join("tests", "testcontent", "downloaded",
"testsubtitles_ar.ttml")
assert os.path.exists(local_path)
local_path_no_ext = local_path.replace('.ttml', '')
copyfile(local_path, local_path_no_ext)
assert os.path.exists(local_path_no_ext)
subtitle_file = SubtitleFile(
local_path_no_ext,
language='ar',
subtitlesformat='ttml' # settting subtitlesformat becaue no ext
)
filename = subtitle_file.process_file()
assert filename, 'conferted filename must exit'
assert filename.endswith(
'.vtt'), 'conferted filename must have .vtt extension'
def test_convertible_substitles_from_pressurcooker(pressurcooker_test_files):
"""
Try to load all the test files used in pressurecooker as riceccooker `SubtitleFile`s.
All subs have the appropriate extension so no need to specify `subtitlesformat`.
"""
for fixture in pressurcooker_test_files:
localpath = fixture['localpath']
assert os.path.exists(
localpath), 'Error mising local test file ' + localpath
subtitle_file = SubtitleFile(localpath, language=fixture['language'])
filename = subtitle_file.process_file()
assert filename, 'conferted filename must exit'
assert filename.endswith(
'.vtt'), 'conferted filename must have .vtt extension'
storage_path = config.get_storage_path(filename)
with open(storage_path) as converted_vtt:
filecontents = converted_vtt.read()
assert fixture[
'check_words'] in filecontents, 'missing check_words in converted subs'
def test_convertible_substitles_weirdext_subtitlesformat():
"""
Check that we handle cases when ext cannot be guessed from URL or localpath.
Passing `subtitlesformat` allows chef authors to manually specify subs format.
"""
subs_url = 'https://commons.wikimedia.org/w/api.php?' \
+ 'action=timedtext&title=File%3AA_Is_for_Atom_1953.webm&lang=es&trackformat=srt'
subtitle_file = SubtitleFile(
subs_url,
language='es',
subtitlesformat=
'srt' # set subtitlesformat when can't inferr ext form url
)
filename = subtitle_file.process_file()
assert filename, 'conferted filename must exit'
assert filename.endswith(
'.vtt'), 'conferted filename must have .vtt extension'
storage_path = config.get_storage_path(filename)
with open(storage_path) as converted_vtt:
filecontents = converted_vtt.read()
assert 'El total de los protones y neutrones de un átomo' in filecontents, \
'missing check words in converted subs'
def test_bad_subtitles_raises(bad_subtitles_file):
subs_file = SubtitleFile(bad_subtitles_file.name, language='en')
pytest.raises(ValueError, subs_file.process_file)
def create_node(node, assessment_dict, subtitle_path, vtt_videos, base_path,
lite_version, lang_code):
kind = node.get('kind')
# Exercise node creation
if kind == 'Exercise':
child_node = ExerciseNode(
source_id=node['id'],
title=node['title'],
exercise_data={
'mastery_model': node.get('suggested_completion_criteria')
},
description='' if node.get("description") is None else node.get(
"description", '')[:400],
license=licenses.ALL_RIGHTS_RESERVED,
thumbnail=node.get('image_url_256'),
)
# build exercise urls for previews
full_path = base_path + node.get('path').strip('khan')
slug = full_path.split('/')[-2]
full_path = full_path.replace(slug, 'e') + slug
# attach Perseus questions to Exercises
for item in node['all_assessment_items']:
# we replace all references to assessment images with the local file path to the image
for match in re.finditer(FILE_URL_REGEX,
assessment_dict[item['id']]["item_data"]):
file_path = str(match.group(0)).replace('\\', '')
file_path = file_path.replace(REPLACE_STRING,
IMAGE_DL_LOCATION)
assessment_dict[item['id']]["item_data"] = re.sub(
FILE_URL_REGEX, file_path,
assessment_dict[item['id']]["item_data"], 1)
question = PerseusQuestion(
id=item['id'],
raw_data=assessment_dict[item['id']]['item_data'],
source_url=full_path if not lite_version else None,
)
child_node.add_question(question)
# Topic node creation
elif kind == 'Topic':
child_node = TopicNode(
source_id=node["id"],
title=node["title"],
description='' if node.get("description") is None else node.get(
"description", '')[:400])
# Video node creation
elif kind == 'Video':
# standard download url for KA videos
download_url = "https://cdn.kastatic.org/KA-youtube-converted/{0}.mp4/{1}.mp4".format(
node['youtube_id'], node['youtube_id'])
files = [VideoFile(download_url)]
if node['youtube_id'] in vtt_videos:
files.append(
SubtitleFile(subtitle_path +
'/{}.vtt'.format(node['youtube_id']),
language=getlang(lang_code)))
child_node = VideoNode(
source_id=node["id"],
title=node["title"],
description='' if node.get("description") is None else node.get(
"description", '')[:400],
files=files,
thumbnail=node.get('image_url'),
license=licenses.CC_BY_NC_SA)
else: # unknown content file format
return None
return child_node
def make_subtitles_file(subtitles_file_file, language='en'):
"""
Creates a VideoFile object with path taken from `video_file_file.name`.
"""
return SubtitleFile(subtitles_file_file.name, language=language)
