Migrate to Python 3 #110
Migrate to Python 3 #110
Conversation
- Convert to Python 3 exclusively - Restructure code into classes (Soundtrack, Song, AudioFile) - Add type hints - Use pathlib for path operations - Implement context managers for resource handling - Introduce custom exception classes for better error reporting - Use persistent HTTP sessions for improved performance - Spoof Chrome user agent and set Referer to avoid 403 errors (compliant with robots.txt) - Stream downloads with chunked transfers - Refactor CLI with argparse - Add verbose mode for detailed progress - Sanitize filenames to avoid filesystem issues
…umentation - Reformat to comply with Google Python style guide: - Consistent docstrings, type hints, constants, and naming conventions - Restructure core classes (AudioFile, Song, Soundtrack) for clarity and reuse - Introduce `--images` / `download_images` option to fetch album artwork - Enhance CLI parsing and error handling (invalid formats, network retries) - Update README.md with usage examples and API docs - Add requirements.txt pinning `requests` and `beautifulsoup4`
Deleted false statements copied from template.
Revised Library and API sections of the README to reflect changes in the Python 3 migration.
Cleaned up superfluity in the README.
|
Lgtm, works well |
This should lower the barrier for complete beginners and ensure a smooth first-time experience.
…, update 2nd docs example to minecraft, since kh3 was broken.
|
Oh, beautiful. Most everything anyone’s ever wanted implemented, isn’t it. Before you merge this, would you like to implement the search functionality of the original ( |
|
I'll add search back, I didn't realize I deleted it in my version! Also, do you think it's a good or bad idea to add progress bars? I wasn't sure, but if you'd like it, I'm very glad to add it. |
|
Please do! ❤️ It’d be nice to have the freaky auto-package installation that’s in the original, as well – the script is meant to be as friendly as possible to non-technical users (…while still being a command-line tool), which is why that unusual feature was in there. Saves me/you/whoever from reports about Pip not being on someone's PATH, too. I’d prefer it to Oh, and so as not to surprise users who upgrade (and again, to be user-friendly), could you make the verbose mode the default, and add a |
|
Wilco, I'll ditch the requirements.txt for automatic package installation (I'll find out the best way to do that) and make verbose the default. |
|
You are a saint. ❤️ |
|
Ooh, one more thing I just noticed: Downloading images should be default, to match the expectation set by the existing version. So perhaps make that |
|
I'll make downloading images the default. I had changed it because I don't like downloading images haha, but it makes more sense to have continuity, yes. |
|
Feel free to add an environment variable or something if you want it default on your end, of course! (Might be overcomplicating it, though. Do as you see fit!) |
|
For some reason it this fork doesn't accept whole link as parameter, says album not found.
If I only put album name, it actually does download but progress not shown at all. |
Thanks for letting me know, I'll investigate that. |
Tried tqdm, but it was FAR too slow, created no-deps implementation.
|
I believe I have addressed everything you brought up! Let me know if you spot anything else. |
4 participants
Migrate the codebase to Python 3.7+, remove all Python 2 compatibility shims, and modernize core logic and documentation.
Strip surrounding quotes from the output path argument before directory creation -> Fixes #55
Drop reliance on a missing #pageContent container and parse directly from the full soup -> Fixes #71
Sanitize and normalize output paths with pathlib.Path.mkdir(parents=True), removing illegal Windows characters (e.g., colons) -> Fixes #72
Switched to chunked, streaming downloads with a 30 s timeout for large files -> Fixes #75
Detect “No such album” via a text search instead of relying on find('p').string -> Fixes #81
Extract the album ID from a full KHInsider URL before CLI parsing -> Fixes #82
Updated README with clear, step-by-step installation and usage instructions for beginners -> Fixes #85
Ensure the output directory (and any missing parent folders) is created before writing each file -> Fixes #89
Guard against a missing
songlisttable before iterating rows -> Fixes #90Added requirements.txt -> Fixes #95
Removed the Python 2 unicodePrint wrapper and now use native print with UTF-8 support -> Fixes #96
Guard against a missing
songlisttable before iterating rows -> Fixes #100Includes a user agent string spoofer to circumvent their new anti-web-scraping. Still fully robots.txt compliant. ->Fixes #100 | Fixes #101 | Fixes #104 | Fixes #105