jkvargas / russimp-sys

Assimp raw bindings
Other
2 stars 22 forks source link

russimp-sys v2.0.2 prebuilt package is corrupt, for aarch64-apple-darwin #46

Open fgiordana opened 6 months ago

fgiordana commented 6 months ago

When building a project with a dependency on russimp-sys v2.0.2 using the prebuilt flag, the downloaded archive seems to be corrupt.

cargo build --features prebuilt
   Compiling russimp-sys v2.0.2
error: failed to run custom build command for `russimp-sys v2.0.2`

Caused by:
  process didn't exit successfully: `/Users/fgiordana/Documents/git/russimp/target/debug/build/russimp-sys-c02ae01bc0e8e6fe/build-script-build` (exit status: 101)
  --- stdout
  cargo:rustc-link-search=native=/opt/homebrew/lib/

  --- stderr
  [/Users/fgiordana/.cargo/registry/src/index.crates.io-6f17d22bba15001f/russimp-sys-2.0.2/build.rs:146:5] ar_src_dir.join(&archive_name) = "/Users/fgiordana/Documents/git/russimp/target/debug/build/russimp-sys-155fc026083726a0/out/russimp-2.0.2-aarch64-apple-darwin-dylib.tar.gz"
  thread 'main' panicked at /Users/fgiordana/.cargo/registry/src/index.crates.io-6f17d22bba15001f/russimp-sys-2.0.2/build.rs:152:34:
  called `Result::unwrap()` on an `Err` value: Custom { kind: UnexpectedEof, error: TarError { desc: "failed to iterate over archive", io: Kind(UnexpectedEof) } }
  note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

Checking the downloaded package manually also confirms that the archive seems corrupt:

tar -xzvf /Users/fgiordana/Documents/git/russimp/target/debug/build/russimp-sys-155fc026083726a0/out/russimp-2.0.2-aarch64-apple-darwin-dylib.tar.gz
tar: Error opening archive: Unrecognized archive format
jkvargas commented 6 months ago

I believe you're correct. Tried to download the file on my own machine and indeed it seems. I guess the easier would be to produce a new version with the fix. Now, there is an issue with crate size which I don't have time to look into. But I will gladly accept fixes on russimp.

fgiordana commented 6 months ago

Ok let's see if we can fix the size issue on the other thread, and then we will probably resolve this too as a consequence