Hi, I got

Getting subtitles for each frame
Traceback (most recent call last):
  File "src/main.py", line 80, in <module>
    runner.run(sys.argv[1:])
  File "src/main.py", line 46, in run
    self.__run__(
  File "src/main.py", line 61, in __run__
    segment_finder = SubtitleSegmentFinder(subtitle_parser.get_subtitle_parts())
  File "C:\Users\Tilman\AppData\Local\Programs\Python\Python38\src\subtitle_segment_finder.py", line 52, in get_subtitle_parts
    for caption in webvtt.read(self.input_file):
  File "C:\Users\Tilman\AppData\Local\Programs\Python\Python38\lib\site-packages\webvtt\webvtt.py", line 60, in read
    parser = WebVTTParser().read(file)
  File "C:\Users\Tilman\AppData\Local\Programs\Python\Python38\lib\site-packages\webvtt\parsers.py", line 25, in read
    self._validate(content)
  File "C:\Users\Tilman\AppData\Local\Programs\Python\Python38\lib\site-packages\webvtt\parsers.py", line 258, in _validate
    raise MalformedFileError('The file does not have a valid format')
webvtt.errors.MalformedFileError: The file does not have a valid format

when trying to use a .srt subtitle file. Is it possible to add support for that?

I'm getting this error - In my subtitle file I dont have any dots (.) I think that is the problem,

This is the output

Number of frames: 103
Getting subtitles for each frame
Traceback (most recent call last):
  File "src/main.py", line 87, in <module>
    runner.run(sys.argv[1:])
  File "src/main.py", line 51, in run
    self.__run__(
  File "src/main.py", line 70, in __run__
    segments = segment_finder.get_subtitle_segments(subtitle_breaks)
  File "D:\Study\Sixth-Sem\Lecture-Video-to-PDF\src\subtitle_segment_finder.py", line 42, in get_subtitle_segments
    pos = self.__get_part_position_of_time_break__(time_break)
  File "D:\Study\Sixth-Sem\Lecture-Video-to-PDF\src\subtitle_segment_finder.py", line 122, in __get_part_position_of_time_break__
    if self.parts[right_part_index].text[right_part_char_index] == ".":
IndexError: string index out of range```

Hi! I like the idea of your project but it's not what I actually want.

I have a few lectures videos without subtitle files and I'm not interested in having subtitles in the PDF document either. I just want to to obtain a document with the slides in landscape orientation.

Please consider adding the option to achieve the mentioned result.

Side note: I noticed that your program depends on CPU power. Have you ever thought about utilizing the GPU instead?

python.exe src/main.py "D:\Downloads\vid1.m4v" -s "D:\Downloads\subt1.vtt" -o output.pdf

C:\Users\User\AppData\Local\Programs\Python\Python38>python.exe src/main.py "D:\Downloads\vid1.m4v" -s "D:\Downloads\subt1.vtt" -o output.pdf
Getting selected frames
Getting subtitles for each frame
Merging frames and subtitles
C:\Users\User\AppData\Local\Programs\Python\Python38\lib\site-packages\fpdf\fpdf.py:710: UserWarning: Substitutting Arial by core font Helvetica
  warnings.warn("Substitutting Arial by core font Helvetica")

resulted in cropped slides in the output.pdf like this:

Also, the subtitles are more or less split into two major parts. They were converted from .srt to .vtt via https://www.happyscribe.com and have the following exemplary format:

1090 00:36:26.960 --> 00:36:29.660 inzwischen wird das war schon ein

Hi Thank you for sharing the code. I tried executing the script as per the instructions in readme file. It works fine if run with the example subtitle files (subtitles_1, subtitles_2, etc) are used. But when using my subtitle file with video, the program throws the following error (link to subtitle file that was used: https://github.com/docstar1/Lecture-Video-to-PDF/blob/2daed0bc3426db515ad8da3ef3611834789c23db/tests/subtitles/subtitles_3.vtt): Traceback (most recent call last): File "src/main.py", line 77, in runner.run(sys.argv[1:]) File "src/main.py", line 47, in run video_segment_finder, video_filepath, subtitle_parser, output_filepath File "src/main.py", line 61, in run segments = segment_finder.get_subtitle_segments(subtitle_breaks) File "..\PycharmProjects\Lecture-Video-to-PDF\src\subtitle_segment_finder.py", line 96, in get_subtitle_segments pos = self.get_part_position_of_time_break(time_break) File "..\PycharmProjects\Lecture-Video-to-PDF\src\subtitle_segment_finder.py", line 152, in get_part_position_of_time_break part = self.parts[part_index] TypeError: list indices must be integers or slices, not NoneType

Description:

This PR is about setting up github actions so that tests are automatically run each time a PR is made Also, a new integ test was made which is to generate a sample pdf file through the cli

Description:

This PR contains multiple changes:

1. Adding no subtitles to the generated PDF
2. Creating integ tests for the main.py file
3. Refactoring the test code

Usage of (1)

To add no subtitles to the generated PDF, one would need to supply the -S or --skip-subtitles flag to the cli tool, like:

python3 -m src.main tests/videos/input_1.mp4 -S -o output.pdf   

Running the command above will output a pdf with no subtitles, like:

output.pdf

Description:

Problem:

When we have a video like https://www.youtube.com/watch?v=KN6OSdUfgyA with subtitles like subtitles_8-old.txt, we get an exception thrown:

Number of frames: 103
Getting subtitles for each frame
Traceback (most recent call last):
  File "src/main.py", line 87, in <module>
    runner.run(sys.argv[1:])
  File "src/main.py", line 51, in run
    self.__run__(
  File "src/main.py", line 70, in __run__
    segments = segment_finder.get_subtitle_segments(subtitle_breaks)
  File "D:\Study\Sixth-Sem\Lecture-Video-to-PDF\src\subtitle_segment_finder.py", line 42, in get_subtitle_segments
    pos = self.__get_part_position_of_time_break__(time_break)
  File "D:\Study\Sixth-Sem\Lecture-Video-to-PDF\src\subtitle_segment_finder.py", line 122, in __get_part_position_of_time_break__
    if self.parts[right_part_index].text[right_part_char_index] == ".":
IndexError: string index out of range```

Cause:

This is because the subtitle file contains subtitle parts that has no content in it. For instance,

This causes the check of self.parts[right_part_index].text[right_part_char_index] to fail since right_part_index is a valid index in self.parts but since the text is empty and we first set right_part_char_index to 0, it will cause self.parts[right_part_index].text[right_part_char_index] to fail.

Solution:

Each subtitle part needs to have some text. A subtitle part with no text is redundant. Hence, when we parse the subtitles, we need to remove the subtitle parts that have no text.

Manual Test:

1. Ran python3 src/main.py tests/videos/input_8.mp4 -s tests/subtitles/subtitles_8.1.vrt -o output.pdf. It did not crash, and it returned this PDF:

output.pdf

2. Unfortunately, the subtitle file is not the best, and the subtitle at tests/subtitles/subtitles_8.srt is a better replacement for that video's subtitle. Running python3 src/main.py tests/videos/input_8.mp4 -s tests/subtitles/subtitles_8.srt -o output.pdf` produced this subtitle:

output.pdf

Description:

Problem:

This PR is about fixing two bugs:

1. When the subtitle has no periods (.), it will go into an infinite loop or it will assign subtitles to each frame unbalanced.
2. When generating a PDF with non-English subtitles, it will throw this error:

Solution:

For (1), the issue is in the code responsible for getting the best break points of a timestamp in the subtitle. In the current codebase, it will find the index in the subtitle based on its timestamp, and then trying to find a . iteratively left and right from the original index. If the index is not found, it will return the entire subtitle file.

The fix for (1) is to return the index if a . is not found left or right of the original index.

For (2), the fix is to add unicode fonts when writing content to the PDF. From this link, this can be done by downloading a unicode font, adding the font to PyFPDF, and using that font.

Manual Test:

1. Ran python3 src/main.py tests/videos/input_7.mp4 -s tests/subtitles/subtitles_7.srt -o output.pdf
2. The output file looks like this:

output.pdf

Problem:

When the resolution of lecture videos are too large, it embeds images into PDF that are larger than the PDF itself. For instance,

Solution:

Set the width of images when embedding them in PDFs

Description:

This PR is about adding .srt support for subtitles. This PR should not affect existing behaviour in parsing .webvtt subtitles.

If the subtitle filename ends in .srt, it will use the SRT subtitle parser. Else, it will use the WebVTT subtitle parser.

Manual Test:

1. Ran pip3 install -r requirements.txt
2. Ran python3 src/main.py tests/videos/input_1.mp4 -s tests/subtitles/subtitles_1.srt -o output.pdf. It outputted the subtitles correctly:

output.pdf

3. Ran python3 src/main.py tests/videos/input_2.mp4 -s tests/subtitles/subtitles_2.srt -o output.pdf. It outputted the subtitles correctly too:

output.pdf

4. Ran cd test && python3 -m unittest * and they all passed.