2339: deny warnings in CI r=irevoire a=MarinPostma
Add `RUSTFLAGS= -D warnings` to the CI so all warnings are treated as hard errors.
Co-authored-by: ad hoc <postma.marin@protonmail.com>
2355: feat(http): add analytics on typo tolerance setting r=curquiza a=MarinPostma
Add analytics for the typo settings.
Also make settings analytics return null for settings that are not set, instead of a default value, as seen with `@gmourier`
Co-authored-by: ad hoc <postma.marin@protonmail.com>
Co-authored-by: Guillaume Mourier <guillaume@meilisearch.com>
2356: fix dumps bug r=MarinPostma a=MarinPostma
move the db version check after the dump import
fix#2353
Co-authored-by: ad hoc <postma.marin@protonmail.com>
2347: Change Nelson path r=curquiza a=curquiza
Nelson is now on the Meilisearch orga side
Co-authored-by: Clémentine Urquizar <clementine@meilisearch.com>
2336: Move permissive-json-pointer in the meilisearch repository r=Kerollmops a=irevoire
Move the permissive-json-pointer crate in the meilisearch repository.
Co-authored-by: Tamo <tamo@meilisearch.com>
2316: Add version flag r=Kerollmops a=sanders41
# Pull Request
## What does this PR do?
Fixes#2315
## PR checklist
Please check if your PR fulfills the following requirements:
- [x] Does this PR fix an existing issue?
- [x] Have you read the contributing guidelines?
- [x] 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: Paul Sanders <psanders1@gmail.com>
2313: fix(search): remove the back and forth between the IndexMap and the serde_json::Map r=irevoire a=irevoire
This is ok because we're using the preserve_order feature in serde_json which is already internally using an IndexMap.
See https://github.com/meilisearch/meilisearch/pull/2298#discussion_r845228412_
Co-authored-by: Tamo <tamo@meilisearch.com>
2178: Refacto docker r=irevoire a=irevoire
closes#2166 and #2085
-----------
I noticed many people had issues with the default configuration of our Dockerfile.
Some examples:
- #2166: If you use ubuntu and mount your `data.ms` in a volume (as shown in the [doc](https://docs.meilisearch.com/learn/getting_started/installation.html#download-and-launch)), you can't run meilisearch
- #2085: Here, meilisearch was not able to erase the `data.ms` when loading a dump because it's the mount point
Currently, we don't show how to use the snapshot and dumps with docker in the documentation. And it's quite hard to do:
- You either send a big command to meilisearch to change the dump-path, snapshot-path and db-path a single directory and then mount that one
- Or you mount three volumes
- And there were other issues on the slack community
I think this PR solve the problem.
Now the image contains the `meilisearch` binary in the `/bin` directory, so it's easy to find and always in the `PATH`.
It creates a `data` directory and moves the working-dir in it.
So now you can find the `dumps`, `snapshots` and `data.ms` directory in `/data`.
Here is the new command to run meilisearch with a volume:
```
docker run -it --rm -v $PWD/meili_data:/data -p 7700:7700 getmeili/meilisearch:latest
```
And if you need to import a dump or a snapshot, you don't need to restart your container and mount another volume. You can directly hit the `POST /dumps` route and then run:
```
docker run -it --rm -v $PWD/meili_data:/data -p 7700:7700 getmeili/meilisearch:latest meilisearch --import-dump dumps/20220217-152115159.dump
```
-------
You can already try this PR with the following docker image:
```
getmeili/meilisearch:test-docker-v0.26.0
```
If you want to use the v0.25.2 I created another image;
```
getmeili/meilisearch:test-docker-v0.25.2
```
------
If you're using helm I created a branch [here](https://github.com/meilisearch/meilisearch-kubernetes/tree/test-docker-v0.26.0) that use the v0.26.0 image with the good volume 👍
If you use this conf with the v0.25.2, it should also work.
Co-authored-by: Tamo <tamo@meilisearch.com>
2298: Nested fields r=irevoire a=irevoire
There are a few things that I want to fix _AFTER_ merging this PR.
For the following RCs.
## Stop the useless conversion
In the `search.rs` I convert a `Document` to a `Value`, and then the `Value` to a `Document` and then back to a `Value` etc. I should stop doing all these conversion and stick to one format.
Probably by merging my `permissive-json-pointer` crate into meilisearch.
That would also give me the opportunity to work directly with obkvs and stops deserializing fields I don't need.
## Add more test specific to the nested
Everything seems to works but I should write tests to double check that the nested works well with the `formatted` field.
## See how I could stop iterating on hashmap and instead fill them correctly
This is related to milli. I really often needs to iterate over hashmap to see if a field is a subset of another field. I could probably generate a structure containing all the possible key values.
ie. the user say `doggo` is an attribute to retrieve. Instead of iterating on all the attributes to retrieve to check if `doggo.name` is a subset of `doggo`. I should insert `doggo.name` in the attributes to retrieve map.
Co-authored-by: Tamo <tamo@meilisearch.com>
2304: chore(bors): comments clippy out r=curquiza a=irevoire
There is currently an issue with clippy that stops us from merging PRs.
https://github.com/rust-lang/rust-clippy/issues/8662#issuecomment-1093899755
We can't use clippy in the CI while that's not merged
Co-authored-by: Tamo <tamo@meilisearch.com>
