This is so both internal code and service code can save data to sub-keys without the parent keys needing to exist.
A doc-string is now set to the data property denoting some keys as reserved as well as their typing and meaning.
This also fixes a bug introduced in v3.3.3 where it will fail to download tracks without the "hls" key in the data property. This can happen when manually making Audio tracks using the HLS descriptor, and not putting any of the hls data the HLS class sets in to_tracks().
We can remove explicit dependency on language-data and marisa-trie because langcodes v3.3.0 now depends on language-data 1.2.0 and language-data 1.2.0 now depends on marisa-trie 1.1.0.
Now it will no longer print the full traceback for errors caused by a missing binary file. Other errors still include it and now explicitly label them as unexpected. CalledProcessError handling is now merged with all non-environment related errors and explicitly mentions that a binary call failed.
The function now located at core/binaries should only be used by services to find a specific binary not listed in there already, or if the name of the binary needed by your service differs.
This simplifies and centralizes all definitions on where these binaries can be found to a singular reference, making it easier to modify, edit, and improve.
Some HLS playlists can have extremely limited information so to accommodate this we need to make the Subtitle track support having almost no information. This isn't ideal but it's really the only solution.
Some HLS playlists can have extremely limited information so to accommodate this we need to make the Audio track support having almost no information. This isn't ideal but it's really the only solution.
Some HLS playlists can have extremely limited information so to accommodate this we need to make the Video track support having almost no information. This isn't ideal but it's really the only solution.
For some reason moving the download speed calculation code from the requests() function to the download() function makes it actually multi-threaded instead of sequential downloads.
The previously named --workers which is now --downloads specified how many tracks to download, not how many threads/workers are used per-download.
It defaults to nothing, which each downloader then has their own defaults. All current downloaders though currently default to `min(32, (os.cpu_count() or 1) + 4)`, which is also the default for `ThreadPoolExecutor` in general.
This also brings a side effect of changing DASH and HLS's forced max_workers of 16 to now a more appropriate default but more importantly actually configurable. You can set a default in your config under `dl.workers`.
For some reason curl-impersonate (curl_cffi project) does not set the certificate-authority bundle for proxies, which to be fair is for some reason seperated into two curl-options.
Doing this change as well as removing the https->http scheme enforcement on proxies, fixes HTTPS proxies on the curl-impersonate downloaders. I also simplified the seperate http and https proxy definitions to the `all` definition which was not originally supported but does seem to be supported as of v0.6.2.
I tested this on NordVPN proxies which are explicitly HTTPS-only and it does work.
The fonts must be within the /devine/fonts folder. This folder location can be changed in the config. If a font is missing it will warn the user and continue.
Closes#82
This allows requests to open and save/cache up to *max_workers* amount of TCP connections. In most situations it will still only save and re-use one TCP Connection since it always tries to re-use the connection if one is available.
However, in situations where downloads are from more than 10 Host/Port combinations (the default pool connections/maxsize) then this will improve download speeds.
The pool_maxsize value here isn't actually doing much. It should have also been applied to pool_connections. What we realistically needed was just pool_block to prevent opening too much connections (causing a warning). The default pool_connections=10 and pool_maxsize=10 is fine. The downloader doesn't currently use this value.
This allows you to override the whole track name instead of just prefixing before the script/territory. If you want no track name at all, you can set the track name to an empty string.
The script "Zzzz" (placeholder?) and territory "ZZ" (placeholder?) are not used. The script/territory values are only used if available and if necessary. I.e., fr-CA will use "Canada" but fr-FR will NOT use "France", it will be blank.
It was using the wrong instance, leaving the convert() method to seemingly default to str() for the returned chosen value types (or something, I don't really see why this works).
The downloader property must be a Callable of the same signature as the aria2c, curl_impersonate, and requests downloader functions. You can pass it these functions by importing, or a custom function of a matching signature.
Note: It will still override the chosen downloader and use a fallback one in the case of using aria2c downloader but the download uses the HTTP Range header.
Closes#70
From aria2c's changelog (2007-09-02):
```
Now *.aria2 contorol file is first saved to *.aria2__temp and if it is successful, then renamed to *.aria2.
This prevents *.aria2 file from being truncated or corrupted when file system becomes out of space.
```
It seems something went wrong in 1.37.0 resulting in these files sometimes not being renamed back to `.aria2` and then being left there for good. The fix for devine would be to simply detect `.aria2__temp` and delete them once all segments finish downloading. My only worry here is the root cause for why it has failed to rename. Did the download actually complete without error? According to aria2c's RPC, no errors occurred. There's no way to add support for Aria2(c) 1.37.0 without this sort of change as the files to seem to download correctly regardless of the file not being renamed and then deleted.
Fixes#71
This reduces the amount of connections being made by quite a bit for playlists that constantly change keys, or have new key data for every single segment (e.g., Pluto sometimes).
It also allows you to pass headers and cookies, while still also being able to supply a proxy.
