def test_empty_subtitles():
# regression test for issue #11
text = dedent("""
392
00:29:27,46 --> 00:29:29,83
I'm Liza Minnelli..
393
00:00:00,00 --> 00:00:00,00
394
00:00:00,00 --> 00:00:00,00
""")
ref = SSAFile()
ref.append(
SSAEvent(start=make_time(m=29, s=27, ms=460),
end=make_time(m=29, s=29, ms=830),
text="I'm Liza Minnelli.."))
ref.append(SSAEvent(start=0, end=0, text=""))
ref.append(SSAEvent(start=0, end=0, text=""))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def create_ass_file(subtitle_blocks: List[SubtitleBlock], ass_file,
styles: Dict[str, StyleConfig]):
subs = SSAFile()
colors = [Color(255, 255, 255), Color(100, 100, 255), Color(255, 100, 100)]
for k, name in enumerate(subtitle_blocks[0].names):
my_style = subs.styles["Default"].copy()
my_style.primarycolor = colors[k]
my_style.fontsize = styles[name].fontsize
my_style.shadow = 0
subs.styles[name] = my_style
for sb in subtitle_blocks:
start, end = None, None
for name, text in sb.name_texts:
if len(text) > 0:
text = text.replace("_", " ")
if start is None:
start = sb.start
end = sb.end
sub_line = SSAEvent(
start=start,
end=end,
text=text,
)
sub_line.style = name
subs.append(sub_line)
else:
print(f"WARNING: got empty block! {name} ")
subs.save(ass_file)
def test_write_drawing():
# test for 7bde9a6c3a250cf0880a8a9fe31d1b6a69ff21a0
subs = SSAFile()
e1 = SSAEvent()
e1.start = 0
e1.end = 60000
e1.text = r"{\p1}m 0 0 l 100 0 100 100 0 100{\p0}test"
e2 = SSAEvent()
e2.start = 60000
e2.end = 120000
e2.text = "Subtitle number\\Ntwo."
subs.append(e1)
subs.append(e2)
ref = dedent("""\
1
00:01:00,000 --> 00:02:00,000
Subtitle number
two.
""")
text = subs.to_string("srt")
assert text.strip() == ref.strip()
def build(self, **extra):
from pysubs2 import SSAFile, SSAEvent, Color # type: ignore[import]
millis = lambda td: td / timedelta(milliseconds=1)
subs = (
SSAEvent(
start=millis(t - self.start),
end =millis(t - self.start + timedelta(seconds=length)),
text=text.replace('\n', r'\N'), # \N necessary for SSA files
) for t, text, length in self.l
)
sf = SSAFile()
style = sf.styles['Default'].copy()
style.fontsize = 16 # default is 20, bit too much??
style.outlinecolor = Color(0, 0, 0, 50) # semitransparent
style.shadow = 0.0
style.outline = 0.1
style.borderstyle = 3 # no idea why 3, but it makes the background apperar in conjunction with outline
for k, v in extra.items():
setattr(style, k, v)
sf.styles['Default'] = style
for s in subs:
sf.append(s)
return sf.to_string('ass')
def append_subs(combined_subs,
new_subs,
style=None,
formatter=None,
exclude=None):
"""
Append a set of subs to a current set avoiding a clash if needed. Also allows for styling and formatting
"""
if exclude is None:
exclude = []
new_combined_subs = SSAFile()
if combined_subs:
# First add the subs we are keeping
new_combined_subs.extend(combined_subs)
for sub in new_subs:
# Add a style
if style:
sub.style = style
# Perform the formatting
if formatter:
sub.text = formatter(sub.text)
# See if we want to cater for clashes
sub, second_sub = _adjust_for_clashing_subs(combined_subs, sub,
exclude)
# Prepare results
if sub:
new_combined_subs.append(sub)
if second_sub:
new_combined_subs.append(second_sub)
new_combined_subs.sort()
return new_combined_subs
def test_simple_write():
subs = SSAFile()
e1 = SSAEvent()
e1.start = 0
e1.end = 60000
e1.text = "ten--chars"
e2 = SSAEvent()
e2.start = 60000
e2.end = 120000
e2.text = "ten--chars-ten-chars"
e3 = SSAEvent()
e3.start = 60000
e3.end = 120000
e3.text = "Invisible subtitle."
e3.is_comment = True
subs.append(e1)
subs.append(e2)
subs.append(e3)
ref = dedent("""\
00:00:00:ten--chars
00:01:00:ten--chars-ten-chars
""")
text = subs.to_string("tmp")
assert text.strip() == ref.strip()
def test_read_position_styling():
"""position is ignored, italic is converted, color is ignored"""
text = dedent("""\
1
00:00:10,500 --> 00:00:13,000 X1:63 X2:223 Y1:43 Y2:58
<i>Elephant's Dream</i>
2
00:00:15,000 --> 00:00:18,000 X1:53 X2:303 Y1:438 Y2:453
<font color="cyan">At the left we can see...</font>
""")
ref = SSAFile()
ref.append(
SSAEvent(start=make_time(s=10.5),
end=make_time(s=13),
text="{\\i1}Elephant's Dream{\\i0}"))
ref.append(
SSAEvent(start=make_time(s=15),
end=make_time(s=18),
text="At the left we can see..."))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def save_to_srt(results, file_name):
"""Converts the output of the model to the standard subtitle format .srt"""
subs = SSAFile()
for result in results:
event = SSAEvent(start=make_time(s=result['start']),
end=make_time(s=result['end']),
text=result['word'])
subs.append(event)
subs.save(file_name)
def test_writer_strips_tags():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10, text="Let me tell you{a secret}."))
f = dedent("""\
{0}{0}1000
{0}{10}Let me tell you.
""")
assert subs.to_string("microdvd", fps=1000) == f
def test_insertion_of_wrong_type():
subs = SSAFile()
subs.append(SSAEvent())
with assert_raises(TypeError):
subs.append(42)
with assert_raises(TypeError):
subs.insert(42)
with assert_raises(TypeError):
subs[0] = 42
def test_writer_strips_tags():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10, text="Let me tell you{a secret}."))
f = dedent("""\
{0}{0}1000
{0}{10}Let me tell you.
""")
assert subs.to_string("microdvd", fps=1000) == f
def test_insertion_of_wrong_type():
subs = SSAFile()
subs.append(SSAEvent())
with assert_raises(TypeError):
subs.append(42)
with assert_raises(TypeError):
subs.insert(42)
with assert_raises(TypeError):
subs[0] = 42
def test_writer_uses_original_fps():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10, text="Hello!"))
subs.fps = 1000
f = dedent("""\
{0}{0}1000
{0}{10}Hello!
""")
assert subs.to_string("microdvd") == f
def test_writer_handles_whitespace():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10,
text=r"Hello,\hworld!\NSo many\N\nNewlines."))
f = dedent("""\
{0}{0}1000
{0}{10}Hello, world!|So many||Newlines.
""")
assert subs.to_string("microdvd", fps=1000) == f
def test_writer_uses_original_fps():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10, text="Hello!"))
subs.fps = 1000
f = dedent("""\
{0}{0}1000
{0}{10}Hello!
""")
assert subs.to_string("microdvd") == f
def test_writer_skips_comment_lines():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10, text="Hello!"))
subs.append(SSAEvent(start=0, end=10, text="World!"))
subs[0].is_comment = True
f = dedent("""\
{0}{0}1000
{0}{10}World!
""")
assert subs.to_string("microdvd", fps=1000) == f
def test_writer_handles_whitespace():
subs = SSAFile()
subs.append(
SSAEvent(start=0, end=10,
text=r"Hello,\hworld!\NSo many\N\nNewlines."))
f = dedent("""\
{0}{0}1000
{0}{10}Hello, world!|So many||Newlines.
""")
assert subs.to_string("microdvd", fps=1000) == f
def test_simple_read():
text = dedent("""\
00:00:00:ten--chars
00:01:00:ten--chars-ten-chars
""")
#calculate endtime from starttime + 500 miliseconds + 67 miliseconds per each character (15 chars per second)
ref = SSAFile()
ref.append(SSAEvent(start=0, end=make_time(ms=1840), text="ten--chars"))
ref.append(SSAEvent(start=make_time(m=1), end=make_time(ms=62510), text="ten--chars-ten-chars"))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def test_writer_skips_comment_lines():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10, text="Hello!"))
subs.append(SSAEvent(start=0, end=10, text="World!"))
subs[0].is_comment = True
f = dedent("""\
{0}{0}1000
{0}{10}World!
""")
assert subs.to_string("microdvd", fps=1000) == f
def filter_subs(combined_subs):
filtered_subs = SSAFile()
last_sub = None
last_top_sub = None
for sub in combined_subs:
if sub.style == 'bottom':
filtered_subs.append(sub)
if last_top_sub and last_top_sub.end > sub.start:
last_top_sub.end = sub.start
last_sub = sub
elif sub.style == 'top':
if last_sub and last_sub.end > sub.start:
sub.start = last_sub.end
if sub.start < sub.end:
filtered_subs.append(sub)
last_top_sub = sub
last_sub = sub
else:
filtered_subs.append(sub)
filtered_removed_empty_subs = SSAFile()
filtered_removed_empty_subs.styles = combined_subs.styles
for sub in filtered_subs:
if sub.end > sub.start:
filtered_removed_empty_subs.append(sub)
logger.info(
f'Filtered subtitles from {len(combined_subs)} to {len(filtered_removed_empty_subs)} subtitles'
)
return filtered_removed_empty_subs
def test_keep_unknown_html_tags():
# see issue #26
text = dedent("""\
1
00:00:10,500 --> 00:00:13,000
<i>Elephant's <sub>Little</sub> Dream</i>
2
00:00:15,000 --> 00:00:18,000
<font color="cyan">At the left we can see...</font>
""")
ref_default = SSAFile()
ref_default.append(SSAEvent(start=make_time(s=10.5), end=make_time(s=13), text="{\\i1}Elephant's Little Dream{\\i0}"))
ref_default.append(SSAEvent(start=make_time(s=15), end=make_time(s=18), text="At the left we can see..."))
ref_keep = SSAFile()
ref_keep.append(SSAEvent(start=make_time(s=10.5), end=make_time(s=13), text="{\\i1}Elephant's <sub>Little</sub> Dream{\\i0}"))
ref_keep.append(SSAEvent(start=make_time(s=15), end=make_time(s=18), text="<font color=\"cyan\">At the left we can see...</font>"))
subs_default = SSAFile.from_string(text)
subs_keep = SSAFile.from_string(text, keep_unknown_html_tags=True)
assert subs_default.equals(ref_default)
assert subs_keep.equals(ref_keep)
assert subs_keep.to_string("srt") == ref_keep.to_string("srt")
def test_write_drawing():
subs = SSAFile()
subs.append(
SSAEvent(start=0,
end=10,
text=r"{\p1}m 0 0 l 100 0 100 100 0 100{\p0}test"))
subs.append(SSAEvent(start=10, end=20, text="Let me tell you."))
f = dedent("""\
{0}{0}1000
{10}{20}Let me tell you.
""")
assert subs.to_string("microdvd", fps=1000) == f
def test_write_read():
subs = SSAFile()
e1 = SSAEvent(text="Hello, world!")
e2 = SSAEvent(text="The other subtitle.\\NWith two lines.", style="custom style")
s1 = SSAStyle(italic=True, primarycolor=Color(r=255, g=0, b=0, a=0))
subs.append(e1)
subs.append(e2)
subs.styles["custom style"] = s1
json_text = subs.to_string("json")
subs2 = SSAFile.from_string(json_text, "json")
assert subs2.equals(subs)
def test_read_bad_tags():
"""missing opening/closing tag, bad nesting, extra whitespace"""
text = dedent("""\
1
00:00:10,500 --> 00:00:13,000
< u><i><font color="red" >Elephant's < s>Dream< / i > Is Long</s> And Badly Nested</xyz>
""")
ref = SSAFile()
ref.append(SSAEvent(start=make_time(s=10.5), end=make_time(s=13), text="{\\u1}{\\i1}Elephant's {\\s1}Dream{\\i0} Is Long{\\s0} And Badly Nested"))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def test_read_bad_tags():
"""missing opening/closing tag, bad nesting, extra whitespace"""
text = dedent("""\
1
00:00:10,500 --> 00:00:13,000
< u><i><font color="red" >Elephant's < s>Dream< / i > Is Long</s> And Badly Nested</xyz>
""")
ref = SSAFile()
ref.append(SSAEvent(start=make_time(s=10.5), end=make_time(s=13), text="{\\u1}{\\i1}Elephant's {\\s1}Dream{\\i0} Is Long{\\s0} And Badly Nested"))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def test_write_read():
subs = SSAFile()
e1 = SSAEvent(text="Hello, world!")
e2 = SSAEvent(text="The other subtitle.\\NWith two lines.",
style="custom style")
s1 = SSAStyle(italic=True, primarycolor=Color(r=255, g=0, b=0, a=0))
subs.append(e1)
subs.append(e2)
subs.styles["custom style"] = s1
json_text = subs.to_string("json")
subs2 = SSAFile.from_string(json_text, "json")
assert subs2.equals(subs)
def test_transform_framerate():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10))
subs.append(SSAEvent(start=1000, end=1010))
with assert_raises(ValueError):
subs.transform_framerate(1, 0)
with assert_raises(ValueError):
subs.transform_framerate(1, -1)
with assert_raises(ValueError):
subs.transform_framerate(0, 1)
with assert_raises(ValueError):
subs.transform_framerate(-1, 1)
subs.transform_framerate(10, 20)
assert subs[0] == SSAEvent(start=0, end=5)
assert subs[1] == SSAEvent(start=500, end=505)
def test_transform_framerate():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=10))
subs.append(SSAEvent(start=1000, end=1010))
with assert_raises(ValueError):
subs.transform_framerate(1, 0)
with assert_raises(ValueError):
subs.transform_framerate(1, -1)
with assert_raises(ValueError):
subs.transform_framerate(0, 1)
with assert_raises(ValueError):
subs.transform_framerate(-1, 1)
subs.transform_framerate(10, 20)
assert subs[0] == SSAEvent(start=0, end=5)
assert subs[1] == SSAEvent(start=500, end=505)
def merge_subs(starting_subs, tolerance_millis=1000, style=None):
"""
Take some subs and eliminate any blank spots where they are less than a tolerance (default of 1 second)
"""
merged_subs = SSAFile()
for sub in starting_subs:
if style and sub.style != style:
continue
if merged_subs and merged_subs[-1].end + tolerance_millis >= sub.start:
merged_subs[-1].end = sub.start
merged_subs.append(sub)
if style:
for sub in starting_subs:
if sub.style != style:
merged_subs.append(sub)
merged_subs.sort()
return merged_subs
def test_read_malformed():
"""no line number, no empty line, leading whitespace, bad timestamp format"""
text = dedent("""\
00:00:00.000 ->00:01:00.000
An example subtitle.
0:01:00,00 --> 0:02:00,00
Subtitle number
two.
""")
ref = SSAFile()
ref.append(SSAEvent(start=0, end=make_time(m=1), text="An example subtitle."))
ref.append(SSAEvent(start=make_time(m=1), end=make_time(m=2), text="Subtitle number\\Ntwo."))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def add_effects(subtitles):
effected_subs = SSAFile()
for sub in subtitles:
content = sub.plaintext.strip().replace('\n', ' ')
time_per_syllable = (sub.end - sub.start) / syllables.estimate(content)
current_time = sub.start
current_index = 0
for word in content.split(' '):
sylls = syllables.estimate(word)
sub_end_time = current_time + time_per_syllable * sylls
current_index += len(word) if current_index == 0 else len(word) + 1
text = content[:current_index] + '{\\alpha&HFF}' + content[
current_index:] # adds transparency
effected_subs.append(
SSAEvent(start=current_time, end=sub_end_time, text=text))
current_time = sub_end_time
return effected_subs
def test_read_malformed():
"""no line number, no empty line, leading whitespace, bad timestamp format"""
text = dedent("""\
00:00:00.000 ->00:01:00.000
An example subtitle.
0:01:00,00 --> 0:02:00,00
Subtitle number
two.
""")
ref = SSAFile()
ref.append(SSAEvent(start=0, end=make_time(m=1), text="An example subtitle."))
ref.append(SSAEvent(start=make_time(m=1), end=make_time(m=2), text="Subtitle number\\Ntwo."))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def save_to_subtitles(results, formatter):
"""
Save to subtitle file
:param results: Dictionary containing info and start/end times
:param formatter: Apply text formating to the subtitle
:return: New subtitle file
"""
subs = SSAFile()
for result in results:
event = SSAEvent(start=make_time(s=result['start']),
end=make_time(s=result['end']),
text=formatter(result))
if 'highlight' in result and result['highlight']:
event.style = 'red'
subs.append(event)
logger.info(f'Processed {len(results)} results to subtitle events')
return subs
def test_simple_read():
text = dedent("""\
1
00:00:00,000 --> 00:01:00,000
An example subtitle.
2
00:01:00,000 --> 00:02:00,000
Subtitle number
two.
""")
ref = SSAFile()
ref.append(SSAEvent(start=0, end=make_time(m=1), text="An example subtitle."))
ref.append(SSAEvent(start=make_time(m=1), end=make_time(m=2), text="Subtitle number\\Ntwo."))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def test_simple_read():
text = dedent("""\
1
00:00:00,000 --> 00:01:00,000
An example subtitle.
2
00:01:00,000 --> 00:02:00,000
Subtitle number
two.
""")
ref = SSAFile()
ref.append(SSAEvent(start=0, end=make_time(m=1), text="An example subtitle."))
ref.append(SSAEvent(start=make_time(m=1), end=make_time(m=2), text="Subtitle number\\Ntwo."))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def compress_subs(subs,
max_chars=30,
max_stretch_millis=3000,
max_oldest_millis=10000,
style=None):
"""
Mostly for the use of speech subtitles this will take individual words and create a running subtitle
"""
# Phase 1 based on character count so that we dont overflow the screen
# Phase 2 is to make sure that the oldest word on the screen has not been there for too long
# First remove gaps where they exist
merged_subs = merge_subs(subs, max_stretch_millis, style)
char_count = 0
oldest_start_time = 0
compressed_subs = SSAFile()
for sub in merged_subs:
if style and sub.style is not style:
continue
char_count += len(sub.text)
# Check the character count and reset if needed
if char_count > max_chars:
char_count = len(sub.text)
oldest_start_time = sub.start
# Check if subtitle has been on screen for too long then reset
elif sub.start - oldest_start_time > max_oldest_millis:
char_count = len(sub.text)
oldest_start_time = sub.start
# If there is a gap in time between subtitles then reset
elif len(compressed_subs) > 0 and sub.start != compressed_subs[-1].end:
char_count = len(sub.text)
oldest_start_time = sub.start
# Add this sub
elif len(compressed_subs) > 0:
sub.text = compressed_subs[-1].text + ' ' + sub.text
char_count += 1
compressed_subs.append(sub)
# Append all the other subs
if style:
for sub in merged_subs:
if sub.style is not style:
compressed_subs.append(sub)
compressed_subs.sort()
return compressed_subs
def test_read_position_styling():
"""position is ignored, italic is converted, color is ignored"""
text = dedent("""\
1
00:00:10,500 --> 00:00:13,000 X1:63 X2:223 Y1:43 Y2:58
<i>Elephant's Dream</i>
2
00:00:15,000 --> 00:00:18,000 X1:53 X2:303 Y1:438 Y2:453
<font color="cyan">At the left we can see...</font>
""")
ref = SSAFile()
ref.append(SSAEvent(start=make_time(s=10.5), end=make_time(s=13), text="{\\i1}Elephant's Dream{\\i0}"))
ref.append(SSAEvent(start=make_time(s=15), end=make_time(s=18), text="At the left we can see..."))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def search_subtitles(self, subtitles, search_str):
ix = create_in(tmp_dir, self.schema)
with ix.writer() as ix_writer:
for i, subtitle in enumerate(subtitles):
ix_writer.add_document(index=i, content=subtitle.text)
with ix.searcher() as ix_searcher:
query = QueryParser('content', ix.schema).parse(search_str)
results = ix_searcher.search(query)
sorted_results = sorted(results, key=itemgetter('index'))
ix.close()
max_idx = len(subtitles) - 1
extracted = []
i = 0
while i < len(sorted_results):
score = 0
subs = SSAFile()
while i+len(subs) < len(sorted_results) \
and sorted_results[i+len(subs)]['index'] == sorted_results[i]['index']+len(subs):
score = max(score, sorted_results[i + len(subs)].score)
subs.append(subtitles[sorted_results[i + len(subs)]['index']])
lower_index = sorted_results[i]['index']
upper_index = sorted_results[i + len(subs) - 1]['index']
extracted.append(
ExtractedSubtitles(
subs,
score,
None if lower_index - 1 < 0 else subtitles[lower_index -
1].end,
None if upper_index + 1 > max_idx else
subtitles[upper_index + 1].start,
))
i += len(subs)
return extracted
def test_read_complex():
# regression test for #30
text = dedent("""\
WEBVTT
X-TIMESTAMP-MAP=LOCAL:00:00:00.000,MPEGTS:0
00:50.099 --> 00:53.299 line:85% align:middle
Cuidem do seu grupo.
Cuidem de suas fileiras.
01:54.255 --> 01:55.455 line:85% align:middle
Parem!
01:58.155 --> 01:59.555 line:85% align:middle
E, parem!
""")
ref = SSAFile()
ref.append(
SSAEvent(start=make_time(s=50, ms=99),
end=make_time(s=53, ms=299),
text=r"Cuidem do seu grupo.\NCuidem de suas fileiras."))
ref.append(
SSAEvent(start=make_time(m=1, s=54, ms=255),
end=make_time(m=1, s=55, ms=455),
text="Parem!"))
ref.append(
SSAEvent(start=make_time(m=1, s=58, ms=155),
end=make_time(m=1, s=59, ms=555),
text="E, parem!"))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def test_empty_subtitles():
# regression test for issue #11
text = dedent("""
392
00:29:27,46 --> 00:29:29,83
I'm Liza Minnelli..
393
00:00:00,00 --> 00:00:00,00
394
00:00:00,00 --> 00:00:00,00
""")
ref = SSAFile()
ref.append(SSAEvent(start=make_time(m=29, s=27, ms=460), end=make_time(m=29, s=29, ms=830), text="I'm Liza Minnelli.."))
ref.append(SSAEvent(start=0, end=0, text=""))
ref.append(SSAEvent(start=0, end=0, text=""))
subs = SSAFile.from_string(text)
assert subs.equals(ref)
def test_simple_writing():
subs = SSAFile()
subs.append(SSAEvent(start=0, end=1000, text="Hello!"))
subs.append(SSAEvent(start=1000, end=2000, text="Hello World!\\NTwo-line subtitle!"))
subs.append(SSAEvent(start=2000, end=3000, text="{\i1}This is an italic line {with some hidden content}"))
subs.append(SSAEvent(start=3000, end=4000, text="This is a comment line"))
subs[-1].is_comment = True
reference_output = dedent("""\
[0][10] Hello!
[10][20] Hello World!|Two-line subtitle!
[20][30] This is an italic line
""")
assert subs.to_string("mpl2").strip() == reference_output.strip()
def test_simple_write():
subs = SSAFile()
e1 = SSAEvent()
e1.start = 0
e1.end = 60000
e1.text = "An example subtitle."
e2 = SSAEvent()
e2.start = 60000
e2.end = 120000
e2.text = "Subtitle number\\Ntwo."
e3 = SSAEvent()
e3.start = 60000
e3.end = 120000
e3.text = "Invisible subtitle."
e3.is_comment = True
subs.append(e1)
subs.append(e2)
subs.append(e3)
ref = dedent("""\
WEBVTT
1
00:00:00,000 --> 00:01:00,000
An example subtitle.
2
00:01:00,000 --> 00:02:00,000
Subtitle number
two.
""")
text = subs.to_string("vtt")
assert text.strip() == ref.strip()
def compress(subs, max_chars=30, max_stretch_time=3, max_oldest_time=10):
new_subs = SSAFile()
# Phase 1 based on character count so that we dont overflow the screen
# Phase 2 if the end of the last subtitle is close to the start of the next we want to stretch out the end
# Phase 3 is to make sure that the oldest word on the screen has not been there for too long
char_count = 0
current_text = ''
current_event = None
oldest_start_time = 0
for sub in subs:
last_event = current_event
current_event = SSAEvent()
current_event.start = sub.start
current_event.end = sub.end
char_count += len(sub.text)
# Check the character count and reset if needed
if char_count > max_chars:
current_text = sub.text
char_count = len(sub.text)
else:
current_text = current_text + ' ' + sub.text
# Check the stretch of subtitles make last one longer
if last_event and current_event.start - last_event.end < max_stretch_time * 1000:
last_event.end = current_event.start
else:
current_text = sub.text
char_count = len(sub.text)
# Make sure that the oldest subtitle on the screen is not too old
if current_event.start - oldest_start_time > max_oldest_time * 1000:
current_text = sub.text
char_count = len(sub.text)
oldest_start_time = sub.start
current_event.text = current_text
new_subs.append(current_event)
logger.info(
f'Compressed {len(subs)} subtitles into {len(new_subs)} subtitles')
return new_subs
def test_repr_simple():
subs = SSAFile()
subs.append(SSAEvent(start=make_time(m=5), end=make_time(m=6)))
subs.append(SSAEvent(start=make_time(m=125), end=make_time(m=126)))
subs.append(SSAEvent(start=make_time(m=15), end=make_time(m=16)))
subs.styles["style1"] = SSAStyle()
subs.styles["style2"] = SSAStyle()
ref = "<SSAFile with 3 events and 3 styles, last timestamp 2:06:00>"
assert repr(subs) == ref
def build_ref():
subs = SSAFile()
subs.info["My Custom Info"] = "Some: Test, String."
subs.styles["left"] = SSAStyle(alignment=7, bold=True)
subs.styles["topleft"] = SSAStyle(alignment=4)
subs.append(SSAEvent(start=0, end=make_time(m=1), text="An, example, subtitle."))
subs.append(SSAEvent(start=0, end=make_time(m=1), type="Comment", text="You can't see this one."))
subs.append(SSAEvent(start=make_time(m=1), end=make_time(m=2), text="Subtitle number\\Ntwo."))
return subs
def test_simple_write():
subs = SSAFile()
e1 = SSAEvent()
e1.start = 0
e1.end = 60000
e1.text = "An example subtitle."
e2 = SSAEvent()
e2.start = 60000
e2.end = 120000
e2.text = "Subtitle number\\Ntwo."
e3 = SSAEvent()
e3.start = 60000
e3.end = 120000
e3.text = "Invisible subtitle."
e3.is_comment = True
subs.append(e1)
subs.append(e2)
subs.append(e3)
ref = dedent("""\
1
00:00:00,000 --> 00:01:00,000
An example subtitle.
2
00:01:00,000 --> 00:02:00,000
Subtitle number
two.
""")
text = subs.to_string("srt")
assert text.strip() == ref.strip()
