2280: Bump the milli dependency to 0.24.1 r=curquiza a=Kerollmops
We had issues with lindera recently, it was unable to download the official dictionaries from Google Drive and this was causing issues with our CIs (and other users' CIs too). The maintainer changed the source to download the dictionaries to get it from Sourceforge and it is much better and stable now.
This PR bumps the milli dependency to the latest version which includes the latest version of the tokenizer which, itself, includes the latest version of lindera, I advise that we rebase the currently opened pull requests to include this PR when it is merged on main.
Co-authored-by: Kerollmops <clement@meilisearch.com>
2263: Upgrade the config of the issue management r=curquiza a=curquiza
To reduce the feature feedback in the meilisearch repo
Co-authored-by: Clémentine Urquizar - curqui <clementine@meilisearch.com>
2253: refactor authentication key extraction r=ManyTheFish a=MarinPostma
I am concerned that the part of the code that performs the key prefix extraction from the jwt token migh be misused in the future. Since this is a critical part of the code, I moved it into it's own function. Since we deserialized the payload twice anyway, I reordered the verifications, and we now use the data from the validated token.
Co-authored-by: ad hoc <postma.marin@protonmail.com>
2264: Import milli from meilisearch-lib r=Kerollmops a=Kerollmops
This PR directly imports the milli dependency used in _meilisearch-http_ from _meilisearch-lib_. I can't import _meilisearch-lib_ in _meiliserach-auth_ and that _meilisearch-auth_ can't use the milli exported by _meilisearch-lib_ 😞
Co-authored-by: Kerollmops <clement@meilisearch.com>
2245: Add test to validate cli r=irevoire a=MarinPostma
followup on #2242 and #2243
Add a test to make sure the cli is valid, and add a CI task to run the tests in debug to make sure we hit debug assertions.
FYI `@curquiza,` because of CI changes
Co-authored-by: ad hoc <postma.marin@protonmail.com>
2244: chore(all): bump milli r=curquiza a=MarinPostma
continues the work initiated by `@psvnlsaikumar` in #2228
Co-authored-by: Sai Kumar <psvnlsaikumar@gmail.com>
2243: bug(http): fix panic on startup r=MarinPostma a=MarinPostma
this seems to fix#2242
I am not sure why this doesn't reproduce on v0.26.0, so we should remain vigilant.
`@curquiza` FYI
Co-authored-by: ad hoc <postma.marin@protonmail.com>
2238: cargo: use resolver 2 r=MarinPostma a=happysalada
# Pull Request
## What does this PR do?
use resolver 2 from cargo.
This enables mainly to propagate the `--no-default-features` flag to workspace crates.
I mistakenly thought before that it was enough to have edition 2021 enabled. However it turns out that for virtual workspaces, this needs to be explicitely defined.
https://doc.rust-lang.org/edition-guide/rust-2021/default-cargo-resolver.html
This will also change a little how your dependencies are compiled. See https://blog.rust-lang.org/2021/03/25/Rust-1.51.0.html#cargos-new-feature-resolver for more details.
Just to give a bit more context, this is for usage in nixos. I have tried to do the upgrade today with the latest version, and the no default features flag is just ignored.
Let me know if you need more details of course.
## PR checklist
Please check if your PR fulfills the following requirements:
- [ ] Does this PR fix an existing issue?
- [ ] Have you read the contributing guidelines?
- [ ] Have you made sure that the title is accurate and descriptive of the changes?
Thank you so much for contributing to Meilisearch!
Co-authored-by: happysalada <raphael@megzari.com>
2233: Change CI name for publishing binaries r=Kerollmops a=curquiza
Minor change regarding the CI job names. Should not impact the usage.
Co-authored-by: Clémentine Urquizar - curqui <clementine@meilisearch.com>
2204: Fix blocking auth r=Kerollmops a=MarinPostma
Fix auth blocking runtime
I have decided to remove async code from `meilisearch-auth` and let `meilisearch-http` handle that.
Because Actix polls the extractor futures concurrently, I have made a wrapper extractor that forces the errors from the futures to be returned sequentially (though is still polls them sequentially).
close#2201
Co-authored-by: ad hoc <postma.marin@protonmail.com>
