Bug fixes & improve duplicate checks

[normalizedwater546]

• When using --meta --cbz with a duplicated cbz file will result in this error.
  • This was caused by the downloader being skipped, but not the rest of the process.

File "\?\A:\nhentai.venv\Scripts\nhentai-script.py", line 33, in <module>
    sys.exit(load_entry_point('nhentai==0.5.7', 'console_scripts', 'nhentai')())
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "A:\nhentai.venv\Lib\site-packages\nhentai-0.5.7-py3.12.egg\nhentai\command.py", line 94, in main
    generate_metadata_file(options.output_dir, table, doujinshi)
  File "A:\nhentai.venv\Lib\site-packages\nhentai-0.5.7-py3.12.egg\nhentai\utils.py", line 313, in generate_metadata_file
    f = open(os.path.join(doujinshi_dir, 'info.txt'), 'w', encoding='utf-8')

• Fixes pdf generation reading non-image files (info.txt, ComicInfo.xml)
  • This assumes the only image types are png, jpg, jpeg, and gif from nhentai.

Traceback (most recent call last):
  File "A:\nhentai.venv\Lib\site-packages\img2pdf.py", line 1817, in read_images
    imgdata = Image.open(im)
              ^^^^^^^^^^^^^^
  File "A:\nhentai.venv\Lib\site-packages\PIL\Image.py", line 3498, in open
    raise UnidentifiedImageError(msg)
PIL.UnidentifiedImageError: cannot identify image file <_io.BytesIO object at 0x0000026F22120720>


Traceback (most recent call last):
  File "\\?\A:\nhentai.venv\Scripts\nhentai-script.py", line 33, in <module>
    sys.exit(load_entry_point('nhentai==0.5.7', 'console_scripts', 'nhentai')())
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "A:\nhentai.venv\Lib\site-packages\nhentai-0.5.7-py3.12.egg\nhentai\command.py", line 114, in main
    generate_pdf(options.output_dir, doujinshi, options.rm_origin_dir, options.move_to_folder)
  File "A:\nhentai.venv\Lib\site-packages\nhentai-0.5.7-py3.12.egg\nhentai\utils.py", line 243, in generate_pdf
    pdf_f.write(img2pdf.convert(full_path_list, rotation=img2pdf.Rotation.ifvalid))
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "A:\nhentai.venv\Lib\site-packages\img2pdf.py", line 2733, in convert
    ) in read_images(
         ^^^^^^^^^^^^
  File "A:\nhentai.venv\Lib\site-packages\img2pdf.py", line 1829, in read_images
    raise ImageOpenError(
img2pdf.ImageOpenError: cannot read input image (not jpeg2000). PIL: error reading image: cannot identify image file <_io.BytesIO object at 0x0000026F22120720>

• When a .cbz file is already generated, duplicate downloads will be ignored even when the flags are different (i.e. --pdf instead of --cbz).
  • This was caused by a hard-coded .cbz file extension check.
• De-duped file exists checks into helper method for more consistent behavior.
  • Logger message is done in parent function rather than helper to maintain logger context.
• Removed warning when folder already exists.
  • Nothing is wrong with this. Proceed silently.
• Optimized path call in generating filename path at 1 and 2
  • os.path.join(doujinshi_dir, '..') was essentially equivalent to output_dir

Not sure what the expected behavior should be here. I assumed this should continue with the rest of the process.

[RicterZ]

In some cases we don't need to download files, use export DEBUG=NODOWNLOAD to skip rest process. In other cases we can use export DEBUG=1 to enable debug mode and continue download process

[normalizedwater546]

This should be fine then. Returning True will proceed with everything else in main, as it did previously.

[RicterZ]

Some problems still exists:

1. If --pdf --cbz option specified, only generate cbz file.
2. If a cbz file exists, will ignore --pdf option.
3. start_download method is coupled with too many options such as regenerate_cbz, file_type, maybe optimization is needed