Commit Graph

218 Commits

Author SHA1 Message Date
Kerollmops
c79e82c62a
Move back to the hnsw crate
This reverts commit 7a4b6c065482f988b01298642f4c18775503f92f.
2023-06-27 12:32:39 +02:00
Kerollmops
268a9ef416
Move to the hgg crate 2023-06-27 12:32:38 +02:00
Clément Renault
4571e512d2
Store the vectors in an HNSW in LMDB 2023-06-27 12:32:38 +02:00
Clément Renault
34349faeae
Create a new _vector extractor 2023-06-27 12:32:37 +02:00
meili-bors[bot]
45636d315c
Merge #3670
3670: Fix addition deletion bug r=irevoire a=irevoire

The first commit of this PR is a revert of https://github.com/meilisearch/meilisearch/pull/3667. It re-enable the auto-batching of addition and deletion of tasks. No new changes have been introduced outside of `milli`. So all the changes you see on the autobatcher have actually already been reviewed.

It fixes https://github.com/meilisearch/meilisearch/issues/3440.

### What was happening?

The issue was that the `external_documents_ids` generated in the `transform` were used in a very strange way that wasn’t compatible with the deletion of documents.
Instead of doing a clear merge between the external document IDs of the DB and the one returned by the transform + writing it on disk, we were doing some weird tricks with the soft-deleted to avoid writing the fst on disk as much as possible.
The new algorithm may be a bit slower but is way more straightforward and doesn’t change depending on if the soft deletion was used or not. Here is a list of the changes introduced:
1. We now do a clear distinction between the `new_external_documents_ids` coming from the transform and only held on RAM and the `external_documents_ids` coming from the DB.
2. The `new_external_documents_ids` (coming out of the transform) are now represented as an `fst`. We don't need to struggle with the hard, soft distinction + the soft_deleted => That's easier to understand
3. When indexing documents, we merge the `external_documents_ids` coming from the DB and the `new_external_documents_ids` coming from the transform.

### Other things introduced in this  PR

Since we constantly have to write small, very specialized fuzzers for this kind of bug, we decided to push the one used to reproduce this bug.
It's not perfect, but it's easy to improve in the future.
It'll also run for as long as possible on every merge on the main branch.

Co-authored-by: Tamo <tamo@meilisearch.com>
Co-authored-by: Loïc Lecrenier <loic.lecrenier@icloud.com>
2023-06-19 09:09:30 +00:00
Tamo
6c6387d05e
move the fuzzer to its own crate 2023-05-29 12:27:39 +02:00
Tamo
602ad98cb8 improve the way we handle the fsts 2023-05-22 11:15:14 +02:00
Tamo
4391cba6ca
fix the addition + deletion bug 2023-05-17 18:28:57 +02:00
Kerollmops
1a79fd0c3c
Use the new heed v0.12.6 2023-05-15 11:42:30 +02:00
Kerollmops
c4a40e7110
Use the writemap flag to reduce the memory usage 2023-05-15 10:15:33 +02:00
Jakub Jirutka
13f1277637 Allow to disable specialized tokenizations (again)
In PR #2773, I added the `chinese`, `hebrew`, `japanese` and `thai`
feature flags to allow melisearch to be built without huge specialed
tokenizations that took up 90% of the melisearch binary size.
Unfortunately, due to some recent changes, this doesn't work anymore.
The problem lies in excessive use of the `default` feature flag, which
infects the dependency graph.

Instead of adding `default-features = false` here and there, it's easier
and more future-proof to not declare `default` in `milli` and
`meilisearch-types`. I've renamed it to `all-tokenizers`, which also
makes it a bit clearer what it's about.
2023-05-04 15:45:40 +02:00
Louis Dureuil
90bc230820
Merge remote-tracking branch 'origin/main' into search-refactor
Conflicts | resolution
----------|-----------
Cargo.lock | added mimalloc
Cargo.toml |  took origin/main version
milli/src/search/criteria/exactness.rs | deleted after checking it was only clippy changes
milli/src/search/query_tree.rs | deleted after checking it was only clippy changes
2023-05-03 12:19:06 +02:00
ManyTheFish
249053e514 Update feature flags 2023-04-26 14:59:25 +02:00
ManyTheFish
ff2cf2a5ae Update charabia in milli 2023-04-26 14:56:54 +02:00
Kerollmops
a109802d45
Upgrade the incompatible versions of the dependencies 2023-04-24 17:50:57 +02:00
Kerollmops
47b66e49b8
Upgrade the compatible versions of the dependencies 2023-04-24 17:50:52 +02:00
Loïc Lecrenier
6c659dc12f Use MiMalloc in milli tests 2023-03-20 09:41:37 +01:00
bors[bot]
39407885c2
Merge #3347
3347: Enhance language detection r=irevoire a=ManyTheFish

## Summary

Some completely unrelated Languages can share the same characters, in Meilisearch we detect the Languages using `whatlang`, which works well on large texts but fails on small search queries leading to a bad segmentation and normalization of the query.

This PR now stores the Languages detected during the indexing in order to reduce the Languages list that can be detected during the search.

## Detail

- Create a 19th database mapping the scripts and the Languages detected with the documents where the Language is detected
- Fill the newly created database during indexing
- Create an allow-list with this database and pass it to Charabia
- Add a test ensuring that a Japanese request containing kanjis only is detected as Japanese and not Chinese

## Related issues
Fixes #2403
Fixes #3513

Co-authored-by: f3r10 <frledesma@outlook.com>
Co-authored-by: ManyTheFish <many@meilisearch.com>
Co-authored-by: Many the fish <many@meilisearch.com>
2023-02-21 10:52:13 +00:00
ManyTheFish
8aa808d51b Merge branch 'main' into enhance-language-detection 2023-02-20 18:14:34 +01:00
bors[bot]
1e9ac00800
Merge #3505
3505: Csv delimiter r=irevoire a=irevoire

Fixes https://github.com/meilisearch/meilisearch/issues/3442
Closes https://github.com/meilisearch/meilisearch/pull/2803
Specified in https://github.com/meilisearch/specifications/pull/221

This PR is a reimplementation of https://github.com/meilisearch/meilisearch/pull/2803, on the new engine. Thanks for your idea and initial PR `@MixusMinimax;` sorry I couldn’t update/merge your PR. Way too many changes happened on the engine in the meantime.

**Attention to reviewer**; I had to update deserr to implement the support of deserializing `char`s

-------

It introduces four new error messages;
- Invalid value in parameter csvDelimiter: expected a string of one character, but found an empty string
- Invalid value in parameter csvDelimiter: expected a string of one character, but found the following string of 5 characters: doggo
- csv delimiter must be an ascii character. Found: 🍰 
- The Content-Type application/json does not support the use of a csv delimiter. The csv delimiter can only be used with the Content-Type text/csv.

And one error code;
- `invalid_index_csv_delimiter`

The `invalid_content_type` error code is now also used when we encounter the `csvDelimiter` query parameter with a non-csv content type.

Co-authored-by: Tamo <tamo@meilisearch.com>
2023-02-20 17:01:36 +00:00
ManyTheFish
cb8d5f2d4b Update Charabia to 0.7.1 2023-02-20 14:00:31 +01:00
Tamo
8c074f5028 implements the csv delimiter without tests
Co-authored-by: Maxi Barmetler <maxi.barmetler@gmail.com>
2023-02-16 17:35:36 +01:00
Tamo
74d1a67a99 Use the workspace inheritance feature of rust 1.64 2023-02-15 13:51:07 +01:00
Tamo
a43765d454
use the pre-defined deserr extractors 2023-02-14 20:05:30 +01:00
Tamo
8fb7b1d10f
bump deserr 2023-02-14 20:04:30 +01:00
bors[bot]
c88c3637b4
Merge #3461
3461: Bring v1 changes into main r=curquiza a=Kerollmops

Also bring back changes in milli (the remote repository) into main done during the pre-release

Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
Co-authored-by: bors[bot] <26634292+bors[bot]@users.noreply.github.com>
Co-authored-by: curquiza <curquiza@users.noreply.github.com>
Co-authored-by: Tamo <tamo@meilisearch.com>
Co-authored-by: Philipp Ahlner <philipp@ahlner.com>
Co-authored-by: Kerollmops <clement@meilisearch.com>
2023-02-07 11:27:27 +00:00
Kerollmops
fbec48f56e
Merge remote-tracking branch 'milli/main' into bring-v1-changes 2023-02-06 16:48:10 +01:00
ManyTheFish
f4569b04ad Update Charabia version 2023-02-01 15:26:26 +01:00
curquiza
bffabf9cc6 Update version for the next release (v0.41.1) in Cargo.toml files 2023-01-31 09:56:22 +00:00
Tamo
55e8046551
bump milli 2023-01-24 13:52:21 +01:00
Gregory Conrad
3f69dd6450 feat: add Cargo feature for LMDB's POSIX semaphores 2023-01-19 12:08:38 -05:00
bors[bot]
1c4b1b3b2d
Merge #770
770: Update deserr v0.3.0 r=irevoire a=ManyTheFish

related to https://github.com/meilisearch/meilisearch/issues/3391


Co-authored-by: Many the fish <many@meilisearch.com>
2023-01-19 17:05:56 +00:00
curquiza
abd65d9307 Update version for the next release (v0.40.0) in Cargo.toml files 2023-01-19 16:43:45 +00:00
Many the fish
30fc376713
Update deserr v0.3.0 2023-01-19 17:37:30 +01:00
curquiza
3f048927a0 Update version for the next release (v0.39.2) in Cargo.toml files 2023-01-19 14:29:09 +00:00
bors[bot]
0c7d1f761e
Merge #765
765: Update version for the next release (v0.39.1) in Cargo.toml files r=curquiza a=meili-bot

⚠️ This PR is automatically generated. Check the new version is the expected one before merging.

Co-authored-by: curquiza <curquiza@users.noreply.github.com>
2023-01-17 11:04:26 +00:00
curquiza
e3d30e28ef Update version for the next release (v0.39.1) in Cargo.toml files 2023-01-17 10:50:29 +00:00
bors[bot]
63af1e9f28
Merge #764
764: Update deserr to latest version r=irevoire a=loiclec

Update deserr to 0.1.5, which changes the `DeserializeFromValue` trait, getting rid of the `default()` method.


Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
2023-01-17 10:39:36 +00:00
Loïc Lecrenier
f073a86387 Update deserr to latest version 2023-01-17 11:28:19 +01:00
Kerollmops
97005dd505
Bump the milli-imported crates to v1.0.0 2023-01-16 16:29:12 +01:00
curquiza
9e32ac7cb2 Update version for the next release (v0.39.0) in Cargo.toml files 2023-01-11 15:05:06 +00:00
Loïc Lecrenier
02fd06ea0b Integrate deserr 2023-01-11 13:56:47 +01:00
bors[bot]
6a10e85707
Merge #736
736: Update charabia r=curquiza a=ManyTheFish

Update Charabia to the last version.

> We are now Romanizing Chinese characters into Pinyin.
> Note that we keep the accent because they are in fact never typed directly by the end-user, moreover, changing an accent leads to a different Chinese character, and I don't have sufficient knowledge to forecast the impact of removing accents in this context.

Co-authored-by: ManyTheFish <many@meilisearch.com>
2023-01-03 15:44:41 +00:00
curquiza
c72535531b Update version for the next release (v0.38.0) in Cargo.toml files 2022-12-19 16:35:38 +00:00
ManyTheFish
96d4242b93 Update charabia 2022-12-15 18:22:22 +01:00
curquiza
cd5aaa3a9f Update version for the next release (v0.37.0) in Cargo.toml files 2022-11-17 12:50:07 +00:00
Louis Dureuil
6dc6a5d874
Force using vendored version of LMDB
- don't use lmdb master3 branch anymore
2022-11-14 17:17:51 +01:00
Kerollmops
d00d2aab3f Update version for the next release (v0.36.0) in Cargo.toml files 2022-11-09 11:03:09 +00:00
bors[bot]
f46a8ab2e2
Merge #693
693: use the lmdb-master.3 branch r=Kerollmops a=irevoire

After investigating https://github.com/meilisearch/meilisearch/issues/3017, we found out that it was due to lmdb and that, without any code change on our side, bumping using the lmdb-master-3 branch fix our issues.

But, we’re not really confident about what changed between the `mdb.master` and `mdb.master3` branches; thus this is a temporary change, and we hope we’ll be able to move to the new version of heed asap (either before the end of the pre-release or for the next release).

--------

The bug is hard to reproduce; I can reproduce it 100% of the time on my archlinux personal computer. But on a scaleway archlinux bare-metal machine, it doesn’t reproduce. It’s flaky on our test suite, but `@loiclec` was able to write a minimal test that reproduces it every time on macOS.
Basically, what happens is when there are multiple threads opening databases in a different directory at the same time.
If there are 10 or more threads running at the same time, lmdb starts throwing the `Invalid argument (os error 22)` error for no reason, we believe.
I would like to submit an issue to lmdb, but I don’t really have the time to write a test in C without heed currently.

`@hyc,` if you want to take a look at it, here is the repo that reproduces the issue on macOS: https://github.com/irevoire/heed-bug

Co-authored-by: Irevoire <tamo@meilisearch.com>
2022-11-09 09:42:38 +00:00
Irevoire
c7711daca3
use the lmdb-master.3 branch 2022-11-08 16:28:01 +01:00