def test_insert(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(500, 800, "Hey 1.5"),
(1000, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
result = diff(set_1, set_2)
self.assertEqual(result['changed'], True)
# for both time_change and text_changed, we calculate them as follows:
# there are 9 total subs. 8 of those are matches and 1 is new in
# set_2. So the change amount is 1/9
self.assertAlmostEqual(result['time_changed'], 1/9.0)
self.assertAlmostEqual(result['text_changed'], 1/9.0)
self.assertEqual(len(result['subtitle_data']), 5)
# check the lines that haven't changed
self.check_unchanged_subtitle_data(result, set_1, set_2, 0, 2, 3, 4)
# check the line that was inserted
insert_sub_data = result['subtitle_data'][1]
self.assertEquals(insert_sub_data['time_changed'], True)
self.assertEquals(insert_sub_data['text_changed'], True)
self.assertEquals(insert_sub_data['subtitles'][0], self.empty_line())
self.assertEquals(insert_sub_data['subtitles'][1], set_2[1])
def test_simple_replace(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey New 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
result = diff(set_1, set_2)
self.assertEqual(result['changed'], True)
self.assertAlmostEqual(result['time_changed'], 0)
# for text_changed, we calculate as follows: there are 8 total subs.
# 6 of those are matches and 1 is different in both sets. So 2/8.0
# has been changed.
self.assertAlmostEqual(result['text_changed'], 2/8.0)
self.assertEqual(len(result['subtitle_data']), 4)
# check the lines that haven't changed
self.check_unchanged_subtitle_data(result, set_1, set_2, 0, 2, 3)
# check the line that was inserted
insert_sub_data = result['subtitle_data'][1]
self.assertEquals(insert_sub_data['time_changed'], False)
self.assertEquals(insert_sub_data['text_changed'], True)
self.assertEquals(insert_sub_data['subtitles'][0], set_1[1])
self.assertEquals(insert_sub_data['subtitles'][1], set_2[1])
def test_replace_multiple_lines_with_single(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 3000, "Hey 2 and 3"),
(3000, 4000, "Hey 4"),
])
result = diff(set_1, set_2)
self.assertEqual(result['changed'], True)
# for both time_change and text_changed, we calculate them as follows:
# there are 7 total subs. 4 of those are matches and 2 in set_1 were
# replaced with 1 in set_2. So the change amount is 3/7.
self.assertAlmostEqual(result['time_changed'], 3/7.0)
self.assertAlmostEqual(result['text_changed'], 3/7.0)
self.assertEqual(len(result['subtitle_data']), 4)
# check the lines that haven't changed
self.check_unchanged_subtitle_data(result, set_1, set_2, 0, 3)
# check the line that was inserted
line1 = result['subtitle_data'][1]
self.assertEquals(line1['time_changed'], True)
self.assertEquals(line1['text_changed'], True)
self.assertEquals(line1['subtitles'][0], set_1[1])
self.assertEquals(line1['subtitles'][1], set_2[1])
line2 = result['subtitle_data'][2]
self.assertEquals(line2['time_changed'], True)
self.assertEquals(line2['text_changed'], True)
self.assertEquals(line2['subtitles'][0], set_1[2])
self.assertEquals(line2['subtitles'][1], self.empty_line())
def test_one_set_empty(self):
set_1 = SubtitleSet.from_list(
"en", [(0, 1000, "Hey 1"), (1000, 2000, "Hey 2"), (2000, 3000, "Hey 3"), (3000, 4000, "Hey 4")]
)
result = diff(set_1, SubtitleSet("en"))
self.assertEqual(result["changed"], True)
self.assertEqual(result["text_changed"], 1.0)
self.assertEqual(result["time_changed"], 1.0)
def test_data_ordering(self):
set_1 = SubtitleSet.from_list("en", [(0, 1000, "Hey 1")])
set_2 = SubtitleSet.from_list("en", [(0, 1000, "Hey 1"), (1200, 2000, "Hey 2"), (2000, 3000, "Hey 3")])
result = diff(set_1, set_2)
subs_result = result["subtitle_data"][2]["subtitles"]
# make sure the 0 index subs is for set_1, test
# we respect the ordering of arguments passed to diff
self.assertEqual(subs_result[0].text, None)
self.assertEqual(subs_result[1].text, "Hey 3")
def test_one_set_empty(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
result = diff(set_1, SubtitleSet('en'))
self.assertEqual(result['changed'], True)
self.assertEqual(result['text_changed'], 1.0)
self.assertEqual(result['time_changed'], 1.0)
def test_unsynced_reflect_time_changes(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(None, None, "Hey 2"),
])
result = diff(set_1, set_2)
self.assertAlmostEqual(result['time_changed'], 1/3.0)
def test_unsynced_reflect_time_changes(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(None, None, "Hey 2"),
])
result = diff(set_1, set_2)
self.assertAlmostEqual(result['time_changed'], 1/3.0)
def test_text_changes(self):
set_1 = SubtitleSet.from_list(
"en", [(0, 1000, "Hey 1"), (1000, 2000, "Hey 2"), (2000, 3000, "Hey 3"), (3000, 4000, "Hey 4")]
)
set_2 = SubtitleSet.from_list(
"en", [(0, 1000, "Hey 1"), (1000, 2000, "Hey 22"), (2000, 3000, "Hey 3"), (3000, 4000, "Hey 4")]
)
result = diff(set_1, set_2)
self.assertEqual(result["changed"], True)
self.assertEqual(result["text_changed"], 1 / 4.0)
self.assertEqual(result["time_changed"], 0)
self.assertEqual(len(result["subtitle_data"]), 4)
# only sub #2 should have text changed
for i, sub_data in enumerate(result["subtitle_data"]):
self.assertEqual(sub_data["text_changed"], i == 1)
def test_data_ordering(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1200, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
])
result = diff(set_1, set_2)
subs_result = result['subtitle_data'][2]['subtitles']
# make sure the 0 index subs is for set_1, test
# we respect the ordering of arguments passed to diff
self.assertEqual(subs_result[0].text , None)
self.assertEqual(subs_result[1].text , "Hey 3")
def test_text_changes(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey 22"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
result = diff(set_1, set_2)
self.assertEqual(result['changed'], True)
self.assertEqual(result['text_changed'], 1/4.0)
self.assertEqual(result['time_changed'], 0)
self.assertEqual(len(result['subtitle_data']), 4)
# only sub #2 should have text changed
for i,sub_data in enumerate(result['subtitle_data']):
self.assertEqual(sub_data['text_changed'], i ==1)
def test_time_changes(self):
set_1 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1000, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
set_2 = SubtitleSet.from_list('en', [
(0, 1000, "Hey 1"),
(1200, 2000, "Hey 2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
])
result = diff(set_1, set_2)
self.assertEqual(result['changed'], True)
self.assertEqual(result['time_changed'], 1/4.0)
self.assertEqual(result['text_changed'], 0)
self.assertEqual(len(result['subtitle_data']), 4)
# only sub #2 should have text changed
for i,sub_data in enumerate(result['subtitle_data']):
self.assertEqual(sub_data['time_changed'], i ==1)
self.assertFalse(sub_data['text_changed'])
def test_delete(self):
set_1 = SubtitleSet.from_list(
"en", [(0, 1000, "Hey 1"), (1000, 2000, "Hey 2"), (2000, 3000, "Hey 3"), (3000, 4000, "Hey 4")]
)
set_2 = SubtitleSet.from_list("en", [(0, 1000, "Hey 1"), (2000, 3000, "Hey 3"), (3000, 4000, "Hey 4")])
result = diff(set_1, set_2)
self.assertEqual(result["changed"], True)
# for both time_change and text_changed, we calculate them as follows:
# there are 7 total subs. 6 of those are matches and 1 is new in
# set_2. So the change amount is 1/9
self.assertAlmostEqual(result["time_changed"], 1 / 7.0)
self.assertAlmostEqual(result["text_changed"], 1 / 7.0)
self.assertEqual(len(result["subtitle_data"]), 4)
# check the lines that haven't changed
self.check_unchanged_subtitle_data(result, set_1, set_2, 0, 2, 3)
# check the line that was deleted
delete_sub_data = result["subtitle_data"][1]
self.assertEquals(delete_sub_data["time_changed"], True)
self.assertEquals(delete_sub_data["text_changed"], True)
self.assertEquals(delete_sub_data["subtitles"][1], self.empty_line())
self.assertEquals(delete_sub_data["subtitles"][0], set_1[1])
def test_replace_single_line_with_multiple(self):
set_1 = SubtitleSet.from_list(
"en", [(0, 1000, "Hey 1"), (1000, 2000, "Hey 2"), (2000, 3000, "Hey 3"), (3000, 4000, "Hey 4")]
)
set_2 = SubtitleSet.from_list(
"en",
[
(0, 1000, "Hey 1"),
(1000, 1500, "Hey 2.1"),
(1500, 2000, "Hey 2.2"),
(2000, 3000, "Hey 3"),
(3000, 4000, "Hey 4"),
],
)
result = diff(set_1, set_2)
self.assertEqual(result["changed"], True)
# for both time_change and text_changed, we calculate them as follows:
# there are 9 total subs. 6 of those are matches and 1 in set 1 was
# changed to 2 in set 2. So the change amount is 3/9.
self.assertAlmostEqual(result["time_changed"], 3 / 9.0)
self.assertAlmostEqual(result["text_changed"], 3 / 9.0)
self.assertEqual(len(result["subtitle_data"]), 5)
# check the lines that haven't changed
self.check_unchanged_subtitle_data(result, set_1, set_2, 0, 3, 4)
# line 1 in set_1 was replaced my lines 2 and 3 in set_2
line1 = result["subtitle_data"][1]
self.assertEquals(line1["time_changed"], True)
self.assertEquals(line1["text_changed"], True)
self.assertEquals(line1["subtitles"][0], set_1[1])
self.assertEquals(line1["subtitles"][1], set_2[1])
line2 = result["subtitle_data"][2]
self.assertEquals(line2["time_changed"], True)
self.assertEquals(line2["text_changed"], True)
self.assertEquals(line2["subtitles"][0], self.empty_line())
self.assertEquals(line2["subtitles"][1], set_2[2])
def test_empty_subs(self):
result = diff(SubtitleSet('en'), SubtitleSet('en'))
self.assertEqual(result['changed'], False)
self.assertEqual(result['text_changed'], 0)
self.assertEqual(result['time_changed'], 0)
self.assertEqual(len(result['subtitle_data']), 0)
def test_empty_subs(self):
result = diff(SubtitleSet('en'), SubtitleSet('en'))
self.assertEqual(result['changed'], False)
self.assertEqual(result['text_changed'], 0)
self.assertEqual(result['time_changed'], 0)
self.assertEqual(len(result['subtitle_data']), 0)
def test_empty_subs(self):
result = diff(SubtitleSet("en"), SubtitleSet("en"))
self.assertEqual(result["changed"], False)
self.assertEqual(result["text_changed"], 0)
self.assertEqual(result["time_changed"], 0)
self.assertEqual(len(result["subtitle_data"]), 0)
def _get_new_version_for_save(self,
subtitles,
language,
session,
user,
new_title,
new_description,
new_metadata,
save_for_later=None):
"""Return a new subtitle version for this save, or None if not needed."""
new_version = None
previous_version = language.get_tip(public=False)
if previous_version:
title_changed = (new_title is not None
and new_title != previous_version.title)
desc_changed = (new_description is not None and
new_description != previous_version.description)
metadata_changed = (
new_metadata is not None
and new_metadata != previous_version.get_metadata())
else:
title_changed = new_title is not None
desc_changed = new_description is not None
metadata_changed = new_metadata is not None
subtitle_set = None
subs_length = 0
if isinstance(subtitles, basestring):
subtitle_set = SubtitleSet(language.language_code, subtitles)
elif isinstance(subtitles, SubtitleSet):
subtitle_set = subtitles
if subtitle_set:
subs_length = len(subtitle_set)
# subtitles have changed if only one of the version is empty
# or if the versions themselves differ
if not previous_version and not subtitle_set:
subtitles_changed = False
elif not previous_version or not subtitle_set:
subtitles_changed = True
else:
subtitles_changed = diff(previous_version.get_subtitles(),
subtitle_set)['changed']
should_create_new_version = (subtitles_changed or title_changed
or desc_changed or metadata_changed)
if should_create_new_version:
new_version, should_create_task = self._create_version(
session.language,
user,
new_title=new_title,
new_description=new_description,
new_metadata=new_metadata,
subtitles=subtitles,
session=session)
incomplete = not new_version.is_synced() or save_for_later
# Record the origin of this set of subtitles.
#
# We need to record it *before* creating review/approve tasks (if
# any) because that means these subs were from a post-publish edit
# or something similar. If we record the origin after creating the
# review task it'll be marked as originating from review, which
# isn't right because these subs had to come from something else.
#
# :(
record_workflow_origin(new_version,
new_version.video.get_team_video())
if (not incomplete) and should_create_task:
self._create_review_or_approve_task(new_version)
return new_version
def test_diffing(self):
create_langs_and_versions(self.video, ['en'])
eng = self.video.newsubtitlelanguage_set.get(language_code='en')
subtitles = SubtitleSet.from_list('en', [
(10000, 20000, "1 - :D"),
(20000, 30000, "2 - :D"),
(30000, 40000, "3 - :D"),
(40000, 50000, "4 - :D"),
(50000, 60000, "5 - :D"),
])
subtitles2 = SubtitleSet.from_list(
'en',
[
(10000, 20000, "1 - :D"),
(20000, 25000, "2 - :D"), # time change,
(30000, 40000, "Three - :D"), # text change,
# multiple lines replaced by a single line
(40000, 60000, "45 - :D"),
])
first_version = eng.add_version(subtitles=subtitles)
second_version = eng.add_version(subtitles=subtitles2)
# Note on the argument order to diff: we always diff the more recent
# version against the less recent
diff_result = diff(subtitles2, subtitles)
response = self._simple_test('videos:diffing',
[first_version.id, second_version.id])
self.assertEquals(diff_result, response.context['diff_data'])
diff_sub_data = diff_result['subtitle_data']
html = BeautifulSoup(response.content)
diff_list = html.find('ol', {"class": 'subtitles-diff'})
diff_items = diff_list.findAll('li')
# check number of lines
self.assertEquals(len(diff_items), len(diff_sub_data))
def check_column_data(column, sub_data):
"""Check the data in the HTML for a column against the data in
from diff()
"""
# special check for empty lines
if sub_data.text is None:
self.assertEquals(column.string.strip(), "")
return
time_span, text_span = column.findAll('span', recursive=False)
self.assertEquals(text_span.string.strip(), sub_data.text)
time_child_spans = time_span.findAll('span',
{'class': 'stamp_text'})
self.assertEquals(time_child_spans[0].string.strip(),
format_sub_time(sub_data.start_time))
self.assertEquals(time_child_spans[1].string.strip(),
format_sub_time(sub_data.end_time))
for li, diff_sub_data_item in zip(diff_items, diff_sub_data):
# Intuitively, left_column should be compared against
# ['subtitles'][0], but we do the opposite. This is because of
# the way things are ordered:
# - diff() was passed (older_version, newer_version)
# - The rendered HTML has the newer version on the left and the
# older version on the right
check_column_data(li.find('div', {'class': 'left_column'}),
diff_sub_data_item['subtitles'][1])
check_column_data(li.find('div', {'class': 'right_column'}),
diff_sub_data_item['subtitles'][0])
# we use the time_change class for either text or time changes.
time_changes = li.findAll('span', {'class': 'time_change'})
if (diff_sub_data_item['time_changed']
or diff_sub_data_item['text_changed']):
self.assertNotEqual(len(time_changes), 0)
else:
self.assertEquals(len(time_changes), 0)