1465 Commits

Author SHA1 Message Date
Louis Dureuil
ef084ef042
SmallBitmap: Consistently panic on incoherent universe lengths 2023-03-29 08:45:38 +02:00
Louis Dureuil
3524bd1257
SmallBitmap: Add documentation 2023-03-29 08:44:11 +02:00
Tamo
a50b058557 update the geoBoundingBox feature
Now instead of using the (top_left, bottom_right) corners of the bounding box it s using the (top_right, bottom_left) corners.
2023-03-28 18:26:18 +02:00
Louis Dureuil
d4f6216966
Resolve rule time sort criteria 2023-03-28 16:42:02 +02:00
Louis Dureuil
77acafe534
Resolve search time sort criteria for placeholder search 2023-03-28 16:41:03 +02:00
Louis Dureuil
abb19d368d
Initialize query time ranking rule for query search 2023-03-28 12:40:52 +02:00
Louis Dureuil
b4a52a622e
BoxRankingRule 2023-03-28 12:39:42 +02:00
Louis Dureuil
9b83b1deb0
Expose SearchLogger trait 2023-03-27 17:49:18 +02:00
Louis Dureuil
e9eb271499
Remove empty attribute_rule mod 2023-03-27 11:08:03 +02:00
Louis Dureuil
3281a88d08
SmallBitmap: don't expose internal items 2023-03-27 11:04:43 +02:00
Louis Dureuil
5a644054ab
Removed unused search impl 2023-03-27 11:04:27 +02:00
Louis Dureuil
16fefd364e
Add TODO notes 2023-03-27 11:04:04 +02:00
Gregory Conrad
e7994cdeb3 feat: check to see if the PK changed before erroring out
Previously, if the primary key was set and a Settings update contained
a primary key, an error would be returned.
However, this error is not needed if the new PK == the current PK.
This commit just checks to see if the PK actually changes
before raising an error.
2023-03-26 12:18:39 -04:00
Loïc Lecrenier
00bad8c716 Add comments suggesting performance improvements 2023-03-23 10:18:24 +01:00
Loïc Lecrenier
862714a18b Remove criterion_implementation_strategy param of Search 2023-03-23 09:44:12 +01:00
Loïc Lecrenier
d18ebe4f3a Remove more warnings 2023-03-23 09:41:18 +01:00
Loïc Lecrenier
7169d85115 Remove old query_tree code and make clippy happy 2023-03-23 09:39:16 +01:00
Loïc Lecrenier
f5f5f03ec0 Remove old criteria code 2023-03-23 09:35:53 +01:00
Loïc Lecrenier
9b2653427d Split position DB into fid and relative position DB 2023-03-23 09:22:01 +01:00
Loïc Lecrenier
56b7209f26 Make clippy happy 2023-03-23 09:16:17 +01:00
Loïc Lecrenier
9b1f439a91 WIP 2023-03-23 09:12:35 +01:00
Loïc Lecrenier
a86aeba411 WIP 2023-03-22 14:43:08 +01:00
Loïc Lecrenier
384fdc2df4 Fix two bugs in proximity ranking rule 2023-03-21 11:43:25 +01:00
Loïc Lecrenier
83e5b4ed0d Compute edges of proximity graph lazily 2023-03-21 10:44:40 +01:00
Loïc Lecrenier
272cd7ebbd Small cleanup 2023-03-20 13:39:19 +01:00
Loïc Lecrenier
c63c7377e6 Switch order of MappedInterner generic params 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
5b50e49522 cargo fmt 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
65474c8de5 Update new sort ranking rule after rebasing 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
fbb1ba3de0 Cargo fmt 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
a59ca28e2c Add forgotten file 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
825f742000 Simplify graph-based ranking rule impl 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
dd491320e5 Simplify graph-based ranking rule impl 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
c6ff97a220 Rewrite the dead-ends cache to detect more dead-ends 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
49240c367a Fix bug in cost of typo conditions 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
1e6e624078 Fix bug in SmallBitmap 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
8b4e07e1a3 WIP 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
2853009987 Renaming Edge -> Condition 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
aa59c3bc2c Replace EdgeCondition with an Option<..> + other code cleanup 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
7b1d8f4c6d Make PathSet strongly typed 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
a49ddec9df Prune the query graph after executing a ranking rule 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
05fe856e6e Merge forward and backward proximity conditions in proximity graph 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
c0cdaf9f53 Fix bug in the proximity ranking rule for queries with ngrams 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
e9cf58d584 Refactor of the Interner 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
31628c5cd4 Merge Phrase and WordDerivations into one structure 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
3004e281d7 Support ngram typos + splitwords and splitwords+synonyms in proximity 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
14e8d0aaa2 Rename lifetime 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
1c58cf8426 Intern ranking rule graph edge conditions as well 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
5155fd2bf1 Reorganise initialisation of ranking rules + rename PathsMap -> PathSet 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
9ec9c204d3 Small code cleanup 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
78b9304d52 Implement distinct attribute 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
0465ba4a05 Intern more values 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
2099991dd1 Continue documenting and cleaning up the code 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
c232cdabf5 Add documentation 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
4e266211bf Small code reorganisation 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
57fa689131 Cargo fmt 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
10626dddfc Add a few more optimisations to new search algorithms 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
9051065c22 Apply a few optimisations for graph-based ranking rules 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
e8c76cf7bf Intern all strings and phrases in the search logic 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
3f1729a17f Update new search test 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
cab2b6bcda Fix: computation of initial universe, code organisation 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
c4979a2fda Fix code visibility issue + unimplemented detail in proximity rule 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
23931f8a4f Fix small bug in visual logger of search algo 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
aa414565bb Fix proximity graph edge builder to include all proximities 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
1db152046e WIP on split words and synonyms support 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
c27ea2677f Rewrite cheapest path algorithm and empty path cache
It is now much simpler and has much better performance.
2023-03-20 09:41:56 +01:00
Loïc Lecrenier
caa1e1b923 Add typo ranking rule to new search impl 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
71f18e4379 Add sort ranking rule to new search impl 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
600e3dd1c5 Remove warnings 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
362eb0de86 Add support for filters 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
998d46ac10 Add support for search offset and limit 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
6c85c0d95e Fix more bugs + visual empty path cache logging 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
0e1fbbf7c6 Fix bugs in query graph's "remove word" and "cheapest paths" algos 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
6806640ef0 Fix d2 description of paths map 2023-03-20 09:41:56 +01:00
Loïc Lecrenier
173e37584c Improve the visual/detailed search logger 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
6ba4d5e987 Add a search logger 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
dd12d44134 Support swapped word pairs in new proximity ranking rule impl 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
c8e251bf24 Remove noise in codebase 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
a938fbde4a Use a cache when resolving the query graph 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
dcf3f1d18a Remove EdgeIndex and NodeIndex types, prefer u32 instead 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
66d0c63694 Add some documentation and use bitmaps instead of hashmaps when possible 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
132191360b Introduce the sort ranking rule working with the new search structures 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
345c99d5bd Introduce the words ranking rule working with the new search structures 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
89d696c1e3 Introduce the proximity ranking rule as a graph-based ranking rule 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
c645853529 Introduce a generic graph-based ranking rule 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
a70ab8b072 Introduce a function to find the K shortest paths in a graph 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
48aae76b15 Introduce a function to find the docids of a set of paths in a graph 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
23bf572dea Introduce cache structures used with ranking rule graphs 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
864f6410ed Introduce a structure to represent a set of graph paths efficiently 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
c9bf6bb2fa Introduce a structure to implement ranking rules with graph algorithms 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
46249ea901 Implement a function to find a QueryGraph's docids 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
ce0d1e0e13 Introduce a common way to manage the coordination between ranking rules 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
5065d8b0c1 Introduce a DatabaseCache to memorize the addresses of LMDB values 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
a83007c013 Introduce structure to represent search queries as graphs 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
79e0a6dd4e Introduce a new search module, eventually meant to replace the old one
The code here does not compile, because I am merely splitting one giant
commit into smaller ones where each commit explains a single file.
2023-03-20 09:41:55 +01:00
Loïc Lecrenier
2d88089129 Remove unused term matching strategies 2023-03-20 09:41:55 +01:00
Loïc Lecrenier
6c659dc12f Use MiMalloc in milli tests 2023-03-20 09:41:37 +01:00
Clément Renault
1a9c58a7ab
Fix a bug with the new flattening rules 2023-03-15 16:56:44 +01:00
Clément Renault
64571c8288
Improve the testing of the filters 2023-03-15 14:57:17 +01:00
Clément Renault
ea016d97af
Implementing an IS EMPTY filter 2023-03-15 14:12:34 +01:00
Tamo
0f33a65468 makes kero happy 2023-03-13 16:51:11 +01:00
bors[bot]
fb1260ee88
Merge #3568 #3569
3568: CI: Fix `publish-aarch64` job that still uses ubuntu-18.04 r=Kerollmops a=curquiza

Fixes #3563 

Main change
- add the usage of the `ubuntu-18.04` container instead of the native `ubuntu-18.04` of GitHub actions: I had to install docker in the container.

Small additional changes
- remove useless `fail-fast` and unused/irrelevant matrix inputs (`build`, `linker`, `os`, `use-cross`...)
- Remove useless step in job

Proof of work with this CI triggered on this current branch: https://github.com/meilisearch/meilisearch/actions/runs/4366233882

3569: Enhance Japanese language detection r=dureuill a=ManyTheFish

# Pull Request

This PR is a prototype and can be tested by downloading [the dedicated docker image](https://hub.docker.com/layers/getmeili/meilisearch/prototype-better-language-detection-0/images/sha256-a12847de00e21a71ab797879fd09777dadcb0881f65b5f810e7d1ed434d116ef?context=explore):

```bash
$ docker pull getmeili/meilisearch:prototype-better-language-detection-0
```

## Context
Some Languages are harder to detect than others, this miss-detection leads to bad tokenization making some words or even documents completely unsearchable. Japanese is the main Language affected and can be detected as Chinese which has a completely different way of tokenization.

A [first iteration has been implemented for v1.1.0](https://github.com/meilisearch/meilisearch/pull/3347) but is an insufficient enhancement to make Japanese work. This first implementation was detecting the Language during the indexing to avoid bad detections during the search.
Unfortunately, some documents (shorter ones) can be wrongly detected as Chinese running bad tokenization for these documents and making possible the detection of Chinese during the search because it has been detected during the indexing.

For instance, a Japanese document `{"id": 1, "name": "東京スカパラダイスオーケストラ"}` is detected as Japanese during indexing, during the search the query `東京` will be detected as Japanese because only Japanese documents have been detected during indexing despite the fact that v1.0.2 would detect it as Chinese.
However if in the dataset there is at least one document containing a field with only Kanjis like:
_A document with only 1 field containing only Kanjis:_
```json
{
 "id":4,
 "name": "東京特許許可局"
}
```
_A document with 1 field containing only Kanjis and 1 field containing several Japanese characters:_
```json
{
 "id":105,
 "name": "東京特許許可局",
 "desc": "日経平均株価は26日 に約8カ月ぶりに2万4000円の心理的な節目を上回った。株高を支える材料のひとつは、自民党総裁選で3選を決めた安倍晋三首相の経済政策への期待だ。恩恵が見込まれるとされる人材サービスや建設株の一角が買われている。ただ思惑が先行して資金が集まっている面 は否めない。実際に政策効果を取り込む企業はどこか、なお未知数だ。"
}
```

Then, in both cases, the field `name` will be detected as Chinese during indexing allowing the search to detect Chinese in queries. Therefore,  the query `東京` will be detected as Chinese and only the two last documents will be retrieved by Meilisearch.

## Technical Approach

The current PR partially fixes these issues by:
1) Adding a check over potential miss-detections and rerunning the extraction of the document forcing the tokenization over the main Languages detected in it.
 >  1) run a first extraction allowing the tokenizer to detect any Language in any Script
 >  2) generate a distribution of tokens by Script and Languages (`script_language`)
 >  3) if for a Script we have a token distribution of one of the Language that is under the threshold, then we rerun the extraction forbidding the tokenizer to detect the marginal Languages
 >  4) the tokenizer will fall back on the other available Languages to tokenize the text. For example, if the Chinese were marginally detected compared to the Japanese on the CJ script, then the second extraction will force Japanese tokenization for CJ text in the document. however, the text on another script like Latin will not be impacted by this restriction.

2) Adding a filtering threshold during the search over Languages that have been marginally detected in documents

## Limits
This PR introduces 2 arbitrary thresholds:
1) during the indexing, a Language is considered miss-detected if the number of detected tokens of this Language is under 10% of the tokens detected in the same Script (Japanese and Chinese are 2 different Languages sharing the "same" script "CJK").
2) during the search, a Language is considered marginal if less than 5% of documents are detected as this Language.

This PR only partially fixes these issues:
-  the query `東京` now find Japanese documents if less than 5% of documents are detected as Chinese.
-  the document with the id `105` containing the Japanese field `desc` but the miss-detected field `name` is now completely detected and tokenized as Japanese and is found with the query `東京`.
-  the document with the id `4` no longer breaks the search Language detection but continues to be detected as a Chinese document and can't be found during the search.

## Related issue
Fixes #3565

## Possible future enhancements
- Change or contribute to the Library used to detect the Language
  - the related issue on Whatlang: https://github.com/greyblake/whatlang-rs/issues/122

Co-authored-by: curquiza <clementine@meilisearch.com>
Co-authored-by: ManyTheFish <many@meilisearch.com>
Co-authored-by: Many the fish <many@meilisearch.com>
2023-03-09 15:34:35 +00:00
ManyTheFish
2f8eb4f54a last PR fixes 2023-03-09 15:34:36 +01:00
Clément Renault
175e8a8495
Fix a diacritic issue 2023-03-09 14:57:47 +01:00
Clément Renault
df48ac8803
Add one more test for the NULL operator 2023-03-09 13:53:37 +01:00
Clément Renault
ff86073288
Add a snapshot for the NULL facet database 2023-03-09 13:32:27 +01:00
Clément Renault
0ad53784e7
Create a new struct to reduce the type complexity 2023-03-09 13:21:21 +01:00
Clément Renault
e064c52544
Rename an internal facet deletion method 2023-03-09 13:08:02 +01:00
Clément Renault
e106b16148
Fix a typo in a variable
Co-authored-by: Louis Dureuil <louis@meilisearch.com>

aaa
2023-03-09 13:08:02 +01:00
Tamo
eddefb0e0f refactor the error type of the milli::document thing
silence a warning
2023-03-09 13:03:14 +01:00
ManyTheFish
5deea631ea fix clippy too many arguments 2023-03-09 11:19:13 +01:00
Tamo
c5f22be6e1 add boolean support for csv documents 2023-03-09 11:12:49 +01:00
ManyTheFish
b4b859ec8c Fix typos 2023-03-09 10:58:35 +01:00
Clément Renault
7dc04747fd
Make clippy happy 2023-03-08 17:37:08 +01:00
Clément Renault
7c0cd7172d
Introduce the NULL and NOT value NULL operator 2023-03-08 17:14:34 +01:00
Clément Renault
43ff236df8
Write the NULL facet values in the database 2023-03-08 16:49:53 +01:00
Clément Renault
19ab4d1a15
Classify the NULL fields values in the facet extractor 2023-03-08 16:49:31 +01:00
Clément Renault
9287858997
Introduce a new facet_id_is_null_docids database in the index 2023-03-08 16:14:00 +01:00
ManyTheFish
24c0775c67 Change indexing threshold 2023-03-08 12:36:04 +01:00
ManyTheFish
3092cf0448 Fix clippy errors 2023-03-08 10:53:42 +01:00
ManyTheFish
37d4551e8e Add a threshold filtering the Languages allowed to be detected at search time 2023-03-07 19:38:01 +01:00
ManyTheFish
da48506f15 Rerun extraction when language detection might have failed 2023-03-07 18:35:26 +01:00
bors[bot]
4f1ccbc495
Merge #3525
3525: Fix phrase search containing stop words r=ManyTheFish a=ManyTheFish

# Summary
A search with a phrase containing only stop words was returning an HTTP error 500,
this PR filters the phrase containing only stop words dropping them before the search starts, a query with a phrase containing only stop words now behaves like a placeholder search.

fixes https://github.com/meilisearch/meilisearch/issues/3521

related v1.0.2 PR on milli: https://github.com/meilisearch/milli/pull/779



Co-authored-by: ManyTheFish <many@meilisearch.com>
2023-03-02 10:55:37 +00:00
ManyTheFish
37489fd495 Return an internal error in the case of matching word is invalid 2023-03-01 19:05:16 +01:00
Louis Dureuil
5822764be9
Skip computing index budget in tests 2023-02-23 11:23:39 +01:00
bors[bot]
ac5a1e4c4b
Merge #3423
3423: Add min and max facet stats r=dureuill a=dureuill

# Pull Request

## Related issue
Fixes #3426

## What does this PR do?

### User standpoint

- When using a `facets` parameter in search, the facets that have numeric values are displayed in a new section of the response called `facetStats` that contains, per facet, the numeric min and max value of the hits returned by the search.

<details>
<summary>
Sample request/response
</summary>

```json
❯ curl \
  -X POST 'http://localhost:7700/indexes/meteorites/search?facets=mass' \
  -H 'Content-Type: application/json' \
  --data-binary '{ "q": "LL6", "facets":["mass", "recclass"], "limit": 5 }' | jsonxf
{
  "hits": [
    {
      "name": "Niger (LL6)",
      "id": "16975",
      "nametype": "Valid",
      "recclass": "LL6",
      "mass": 3.3,
      "fall": "Fell"
    },
    {
      "name": "Appley Bridge",
      "id": "2318",
      "nametype": "Valid",
      "recclass": "LL6",
      "mass": 15000,
      "fall": "Fell",
      "_geo": {
        "lat": 53.58333,
        "lng": -2.71667
      }
    },
    {
      "name": "Athens",
      "id": "4885",
      "nametype": "Valid",
      "recclass": "LL6",
      "mass": 265,
      "fall": "Fell",
      "_geo": {
        "lat": 34.75,
        "lng": -87.0
      }
    },
    {
      "name": "Bandong",
      "id": "4935",
      "nametype": "Valid",
      "recclass": "LL6",
      "mass": 11500,
      "fall": "Fell",
      "_geo": {
        "lat": -6.91667,
        "lng": 107.6
      }
    },
    {
      "name": "Benguerir",
      "id": "30443",
      "nametype": "Valid",
      "recclass": "LL6",
      "mass": 25000,
      "fall": "Fell",
      "_geo": {
        "lat": 32.25,
        "lng": -8.15
      }
    }
  ],
  "query": "LL6",
  "processingTimeMs": 15,
  "limit": 5,
  "offset": 0,
  "estimatedTotalHits": 42,
  "facetDistribution": {
    "mass": {
      "110000": 1,
      "11500": 1,
      "1161": 1,
      "12000": 1,
      "1215.5": 1,
      "127000": 1,
      "15000": 1,
      "1676": 1,
      "1700": 1,
      "1710.5": 1,
      "18000": 1,
      "19000": 1,
      "220000": 1,
      "2220": 1,
      "22300": 1,
      "25000": 2,
      "265": 1,
      "271000": 1,
      "2840": 1,
      "3.3": 1,
      "3000": 1,
      "303": 1,
      "32000": 1,
      "34000": 1,
      "36.1": 1,
      "45000": 1,
      "460": 1,
      "478": 1,
      "483": 1,
      "5500": 2,
      "600": 1,
      "6000": 1,
      "67.8": 1,
      "678": 1,
      "680.5": 1,
      "6930": 1,
      "8": 1,
      "8300": 1,
      "840": 1,
      "8400": 1
    },
    "recclass": {
      "L/LL6": 3,
      "LL6": 39
    }
  },
  "facetStats": {
    "mass": {
      "min": 3.3,
      "max": 271000.0
    }
  }
}
```

</details>

## PR checklist
Please check if your PR fulfills the following requirements:
- [ ] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [ ] 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: Louis Dureuil <louis@meilisearch.com>
2023-02-22 13:06:43 +00:00
ManyTheFish
900bae3d9d keep phrases that has at least one word 2023-02-21 18:16:51 +01:00
ManyTheFish
bbecab8948 fix clippy 2023-02-21 10:18:44 +01:00
ManyTheFish
8aa808d51b Merge branch 'main' into enhance-language-detection 2023-02-20 18:14:34 +01:00
bors[bot]
b08a49a16e
Merge #3319 #3470
3319: Transparently resize indexes on MaxDatabaseSizeReached errors r=Kerollmops a=dureuill

# Pull Request

## Related issue
Related to https://github.com/meilisearch/meilisearch/discussions/3280, depends on https://github.com/meilisearch/milli/pull/760

## What does this PR do?

### User standpoint

- Meilisearch no longer fails tasks that encounter the `milli::UserError(MaxDatabaseSizeReached)` error.
- Instead, these tasks are retried after increasing the maximum size allocated to the index where the failure occurred.

### Implementation standpoint

- Add `Batch::index_uid` to get the `index_uid` of a batch of task if there is one
- `IndexMapper::create_or_open_index` now takes an additional `size` argument that allows to (re)open indexes with a size different from the base `IndexScheduler::index_size` field
- `IndexScheduler::tick` now returns a `Result<TickOutcome>` instead of a `Result<usize>`. This offers more explicit control over what the behavior should be wrt the next tick.
- Add `IndexStatus::BeingResized` that contains a handle that a thread can use to await for the resize operation to complete and the index to be available again.
- Add `IndexMapper::resize_index` to increase the size of an index.
- In `IndexScheduler::tick`, intercept task batches that failed due to `MaxDatabaseSizeReached` and resize the index that caused the error, then request a new tick that will eventually handle the still enqueued task.

## Testing the PR

The following diff can be applied to this branch to make testing the PR easier:

<details>


```diff
diff --git a/index-scheduler/src/index_mapper.rs b/index-scheduler/src/index_mapper.rs
index 553ab45a..022b2f00 100644
--- a/index-scheduler/src/index_mapper.rs
+++ b/index-scheduler/src/index_mapper.rs
`@@` -228,13 +228,15 `@@` impl IndexMapper {
 
         drop(lock);
 
+        std:🧵:sleep_ms(2000);
+
         let current_size = index.map_size()?;
         let closing_event = index.prepare_for_closing();
-        log::info!("Resizing index {} from {} to {} bytes", name, current_size, current_size * 2);
+        log::error!("Resizing index {} from {} to {} bytes", name, current_size, current_size * 2);
 
         closing_event.wait();
 
-        log::info!("Resized index {} from {} to {} bytes", name, current_size, current_size * 2);
+        log::error!("Resized index {} from {} to {} bytes", name, current_size, current_size * 2);
 
         let index_path = self.base_path.join(uuid.to_string());
         let index = self.create_or_open_index(&index_path, None, 2 * current_size)?;
`@@` -268,8 +270,10 `@@` impl IndexMapper {
             match index {
                 Some(Available(index)) => break index,
                 Some(BeingResized(ref resize_operation)) => {
+                    log::error!("waiting for resize end");
                     // Deadlock: no lock taken while doing this operation.
                     resize_operation.wait();
+                    log::error!("trying our luck again!");
                     continue;
                 }
                 Some(BeingDeleted) => return Err(Error::IndexNotFound(name.to_string())),
diff --git a/index-scheduler/src/lib.rs b/index-scheduler/src/lib.rs
index 11b17d05..242dc095 100644
--- a/index-scheduler/src/lib.rs
+++ b/index-scheduler/src/lib.rs
`@@` -908,6 +908,7 `@@` impl IndexScheduler {
     ///
     /// Returns the number of processed tasks.
     fn tick(&self) -> Result<TickOutcome> {
+        log::error!("ticking!");
         #[cfg(test)]
         {
             *self.run_loop_iteration.write().unwrap() += 1;
diff --git a/meilisearch/src/main.rs b/meilisearch/src/main.rs
index 050c825a..63f312f6 100644
--- a/meilisearch/src/main.rs
+++ b/meilisearch/src/main.rs
`@@` -25,7 +25,7 `@@` fn setup(opt: &Opt) -> anyhow::Result<()> {
 
 #[actix_web::main]
 async fn main() -> anyhow::Result<()> {
-    let (opt, config_read_from) = Opt::try_build()?;
+    let (mut opt, config_read_from) = Opt::try_build()?;
 
     setup(&opt)?;
 
`@@` -56,6 +56,8 `@@` We generated a secure master key for you (you can safely copy this token):
         _ => (),
     }
 
+    opt.max_index_size = byte_unit::Byte::from_str("1MB").unwrap();
+
     let (index_scheduler, auth_controller) = setup_meilisearch(&opt)?;
 
     #[cfg(all(not(debug_assertions), feature = "analytics"))]
```
</details>

Mainly, these debug changes do the following:

- Set the default index size to 1MiB so that index resizes are initially frequent
- Turn some logs from info to error so that they can be displayed with `--log-level ERROR` (hiding the other infos)
- Add a long sleep between the beginning and the end of the resize so that we can observe the `BeingResized` index status (otherwise it would never come up in my tests)

## Open questions

- Is the growth factor of x2 the correct solution? For a `Vec` in memory it makes sense, but here we're manipulating quantities that are potentially in the order of 500GiBs. For bigger indexes it may make more sense to add at most e.g. 100GiB on each resize operation, avoiding big steps like 500GiB -> 1TiB.

## PR checklist
Please check if your PR fulfills the following requirements:
- [ ] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [ ] 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!


3470: Autobatch addition and deletion r=irevoire a=irevoire

This PR adds the capability to meilisearch to batch document addition and deletion together.

Fix https://github.com/meilisearch/meilisearch/issues/3440

--------------

Things to check before merging;

- [x] What happens if we delete multiple time the same documents -> add a test
- [x] If a documentDeletion gets batched with a documentAddition but the index doesn't exist yet? It should not work

Co-authored-by: Louis Dureuil <louis@meilisearch.com>
Co-authored-by: Tamo <tamo@meilisearch.com>
2023-02-20 15:00:19 +00:00
Many the fish
119e6d8811
Update milli/src/search/mod.rs
Co-authored-by: Tamo <tamo@meilisearch.com>
2023-02-20 15:33:10 +01:00
Louis Dureuil
eb28d4c525
add facet test 2023-02-20 13:52:28 +01:00
Louis Dureuil
9ac981d025
Remove some clippy type complexity warns by deboxing iters 2023-02-20 13:52:27 +01:00
Louis Dureuil
74859ecd61
Add min and max facet stats 2023-02-20 13:52:27 +01:00
Louis Dureuil
8ae441a4db
Update usage of iterators 2023-02-20 13:52:27 +01:00
Louis Dureuil
042d86cbb3
facet sort ascending/descending now also return the values 2023-02-20 13:52:27 +01:00
Tamo
18796d6e6a Consider null as a valid geo object 2023-02-20 13:45:51 +01:00
Tamo
895ab2906c apply review suggestions 2023-02-16 18:42:47 +01:00
bors[bot]
143e3cf948
Merge #3490
3490: Fix attributes set candidates r=curquiza a=ManyTheFish

# Pull Request

Fix attributes set candidates for v1.1.0

## details

The attribute criterion was not returning the remaining candidates when its internal algorithm was been exhausted.
We had a loss of candidates by the attribute criterion leading to the bug reported in the issue linked below.
After some investigation, it seems that it was the only criterion that had this behavior.

We are now returning the remaining candidates instead of an empty bitmap.

## Related issue

Fixes #3483
PR on milli for v1.0.1: https://github.com/meilisearch/milli/pull/777


Co-authored-by: ManyTheFish <many@meilisearch.com>
2023-02-15 17:38:07 +00:00
bors[bot]
91ce8a5e67
Merge #3492
3492: Bump deserr r=Kerollmops a=irevoire

Bump deserr to the latest version;
- We now use the default actix-web extractors that deserr provides (which were copy/pasted from meilisearch)
- We also use the default `JsonError` message provided by deserr instead of defining our own in meilisearch
- Finally, we get the new `did you mean?` error message. Fix #3493

Co-authored-by: Tamo <tamo@meilisearch.com>
2023-02-15 10:05:05 +00:00
Tamo
8fb7b1d10f
bump deserr 2023-02-14 20:04:30 +01:00
Tamo
74dcfe9676
Fix a bug when you update a document that was already present in the db, deleted and then inserted again in the same transform 2023-02-14 19:09:40 +01:00
Tamo
1b1703a609
make a small optimization to merge obkvs a little bit faster 2023-02-14 18:32:41 +01:00
Tamo
fb5e4957a6
fix and test the early exit in case a grenad ends with a deletion 2023-02-14 18:23:57 +01:00
Tamo
8de3c9f737
Update milli/src/update/index_documents/transform.rs
Co-authored-by: Clément Renault <clement@meilisearch.com>
2023-02-14 17:57:14 +01:00
Tamo
43a19d0709
document the operation enum + the grenads 2023-02-14 17:55:26 +01:00
Filip Bachul
a53536836b fmt 2023-02-14 17:04:22 +01:00
Filip Bachul
d7ad39ad77 fix: clippy error 2023-02-14 00:15:35 +01:00
Filip Bachul
849de089d2 add thiserror for AscDescError 2023-02-14 00:15:35 +01:00
filip
7f25007d31 Update milli/src/asc_desc.rs
Co-authored-by: Tamo <irevoire@protonmail.ch>
2023-02-14 00:15:35 +01:00
Filip Bachul
c810af3ebf implement From<ParseGeoError> for AscDescError 2023-02-14 00:15:35 +01:00
Filip Bachul
c0b77773ba fmt asc_desc 2023-02-14 00:15:35 +01:00
Filip Bachul
7481559e8b move BadGeo to FilterError 2023-02-14 00:15:35 +01:00
Filip Bachul
83c765ce6c implement From<ParseGeoError> for FilterError 2023-02-14 00:15:35 +01:00
Filip Bachul
4c91037602 use ParseGeoError in sort parser 2023-02-14 00:15:35 +01:00
Filip Bachul
825923f6fc export ParseGeoError 2023-02-14 00:15:35 +01:00
Filip Bachul
e405702733 chore: introduce new error ParseGeoError type 2023-02-14 00:15:35 +01:00
ManyTheFish
6fa877efb0 Fix attributes set candidates 2023-02-13 17:49:52 +01:00
Tamo
746b31c1ce
makes clippy happy 2023-02-09 12:23:01 +01:00
Tamo
93db755d57
add a test to ensure we handle correctly a deletion of multiple time the same document 2023-02-08 21:03:34 +01:00
Tamo
93f130a400
fix all warnings 2023-02-08 20:57:35 +01:00
Tamo
421a9cf05e
provide a new method on the transform to remove documents 2023-02-08 16:06:09 +01:00
Tamo
8f64fba1ce
rewrite the current transform to handle a new byte specifying the kind of operation it's merging 2023-02-08 12:53:38 +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
Tamo
42114325cd
Apply suggestions from code review
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
2023-02-06 18:07:00 +01:00
Tamo
7a38fe624f
throw an error if the top left corner is found below the bottom right corner 2023-02-06 17:50:47 +01:00
Tamo
1b005f697d
update the syntax of the geoboundingbox filter to uses brackets instead of parens around lat and lng 2023-02-06 16:50:27 +01:00
Kerollmops
fbec48f56e
Merge remote-tracking branch 'milli/main' into bring-v1-changes 2023-02-06 16:48:10 +01:00
Tamo
3ebc99473f
Apply suggestions from code review
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
2023-02-06 13:29:37 +01:00
Tamo
d27007005e
comments the geoboundingbox + forbid the usage of the lexeme method which could introduce bugs 2023-02-06 11:36:49 +01:00
Tamo
fcb09ccc3d
add tests on the geoBoundingBox 2023-02-02 18:19:56 +01:00
Louis Dureuil
ae8660e585
Add Token::original_span rather than making Token::span pub 2023-02-02 15:03:34 +01:00
Guillaume Mourier
b297b5deb0
cargo fmt 2023-02-02 12:34:49 +01:00
Guillaume Mourier
0d71c80ba6
add tests 2023-02-02 12:31:27 +01:00
Guillaume Mourier
65a3086cf1
fix test 2023-02-02 12:27:58 +01:00
Guillaume Mourier
426d63b01b
Update insta test suite 2023-02-02 12:27:56 +01:00
Guillaume Mourier
b078477d80
Add error handling and earth lap collision with bounding box 2023-02-02 12:17:38 +01:00
ManyTheFish
0bc1a18f52 Use Languages list detected during indexing at search time 2023-02-01 18:57:43 +01:00
ManyTheFish
643d99e0f9 Add expectancy test 2023-02-01 18:39:54 +01:00
ManyTheFish
064158e4e2 Update test 2023-02-01 15:34:01 +01:00
ManyTheFish
77d32d0ee8 Fix codec deserialization 2023-02-01 15:26:26 +01:00
Loïc Lecrenier
a2690ea8d4 Reduce incremental indexing time of words_prefix_position_docids DB
This database can easily contain millions of entries. Thus, iterating
over it can be very expensive.

For regular `documentAdditionOrUpdate` tasks, `del_prefix_fst_words`
will always be empty. Thus, we can save a significant amount of time
by adding this `if !del_prefix_fst_words.is_empty()` condition.

The code's behaviour remains completely unchanged.
2023-01-31 11:42:24 +01:00
f3r10
2922c5c899 Fix code format 2023-01-31 11:28:05 +01:00
f3r10
7681be5367 Format code 2023-01-31 11:28:05 +01:00
f3r10
50bc156257 Fix tests 2023-01-31 11:28:05 +01:00
f3r10
d8207356f4 Skip script,language insertion if language is undetected 2023-01-31 11:28:05 +01:00
f3r10
2d58b28f43 Improve script language codec 2023-01-31 11:28:05 +01:00
f3r10
fd60a39f1c Format code 2023-01-31 11:28:05 +01:00
f3r10
369c05732e Add test checking if from script_language_docids database were removed
deleted docids
2023-01-31 11:28:05 +01:00
f3r10
34d04f3d3f Filter from script_language_docids database soft deleted documents 2023-01-31 11:28:05 +01:00
f3r10
a27f329e3a Add tests for checking that detected script and language associated with document(s) were stored during indexing 2023-01-31 11:28:05 +01:00
f3r10
b216ddba63 Delete and clear data from the new database 2023-01-31 11:28:05 +01:00
f3r10
d97fb6117e Extract and index data 2023-01-31 11:28:05 +01:00
f3r10
c45d1e3610 Create a new database on index and add a specialized codec for it 2023-01-31 11:28:05 +01:00
Louis Dureuil
20f05efb3c
clippy: needless_lifetimes 2023-01-31 11:12:59 +01:00
Louis Dureuil
cbf029f64c
clippy: --fix 2023-01-31 11:12:59 +01:00
Louis Dureuil
3296cf7ae6
clippy: remove needless lifetimes 2023-01-31 09:32:40 +01:00
Louis Dureuil
89675e5f15
clippy: Replace seek 0 by rewind 2023-01-31 09:32:40 +01:00
Tamo
de3c4f1986 throw an error on unknown fields specified in the _geo field 2023-01-24 12:23:24 +01:00
bors[bot]
3521a3a0b2
Merge #763
763: Fixes error message when lat and lng are unparseable r=loiclec a=ahlner

# Pull Request

## Related issue
Fixes partially [#3007](https://github.com/meilisearch/meilisearch/issues/3007)

## What does this PR do?
- Changes function validate_geo_from_json to return a BadLatitudeAndLongitude if lat or lng is a string and not parseable to f64
- implemented some unittests
- Derived PartialEq for GeoError to use assert_eq! in tests

## PR checklist
Please check if your PR fulfills the following requirements:
- [x] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [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: Philipp Ahlner <philipp@ahlner.com>
2023-01-19 15:15:46 +00:00
Philipp Ahlner
f5ca421227
Superfluous test removed 2023-01-19 15:39:21 +01:00
Louis Dureuil
4fd6fd9bef
Indicate filterable attributes when the user set a non filterable attribute in facet distributions 2023-01-19 12:25:18 +01:00
Philipp Ahlner
a2cd7214f0
Fixes error message when lat/lng are unparseable 2023-01-19 10:10:26 +01:00
ManyTheFish
d1fc42b53a Use compatibility decomposition normalizer in facets 2023-01-18 15:02:13 +01:00
Philipp Ahlner
497187083b
Add test for bug #3007: Wrong error message
Adds a test for #3007: Wrong error message when lat and lng are
unparseable
2023-01-18 13:24:26 +01:00
Clément Renault
1d507c84b2
Fix the formatting 2023-01-17 18:25:55 +01:00
Clément Renault
1b78231e18
Make clippy happy 2023-01-17 18:25:54 +01:00
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
bors[bot]
302d6cccd7
Merge #761
761: Integrate deserr r=irevoire a=loiclec

1. `Setting<T>` now implements `DeserializeFromValue`
2. The settings now store ranking rules as strongly typed `Criterion` instead of `String`, since the validation of the ranking rules will be done on meilisearch's side from now on


Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
2023-01-11 14:35:15 +00:00
bors[bot]
21b7d709ad
Merge #759
759: Change primary key inference error messages r=Kerollmops a=dureuill

# Pull Request

## Related issue
Milli part of https://github.com/meilisearch/meilisearch/issues/3301

## What does this PR do?
- Change error message strings

## PR checklist
Please check if your PR fulfills the following requirements:
- [x] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [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: Louis Dureuil <louis@meilisearch.com>
2023-01-11 14:04:25 +00:00
Loïc Lecrenier
02fd06ea0b Integrate deserr 2023-01-11 13:56:47 +01:00
Louis Dureuil
00746b32c0
Add Index::map_size 2023-01-10 11:16:51 +01:00
Louis Dureuil
be9786bed9
Change primary key inference error messages 2023-01-05 10:40:09 +01:00
bors[bot]
c3f4835e8e
Merge #733
733: Avoid a prefix-related worst-case scenario in the proximity criterion r=loiclec a=loiclec

# Pull Request

## Related issue
Somewhat fixes (until merged into meilisearch) https://github.com/meilisearch/meilisearch/issues/3118

## What does this PR do?
When a query ends with a word and a prefix, such as:
```
word pr
```
Then we first determine whether `pre` *could possibly* be in the proximity prefix database before querying it. There are then three possibilities:

1. `pr` is not in any prefix cache because it is not the prefix of many words. We don't query the proximity prefix database. Instead, we list all the word derivations of `pre` through the FST and query the regular proximity databases.

2. `pr` is in the prefix cache but cannot be found in the proximity prefix databases. **In this case, we partially disable the proximity ranking rule for the pair `word pre`.** This is done as follows:
   1. Only find the documents where `word` is in proximity to `pre` **exactly** (no derivations)
   2. Otherwise, assume that their proximity in all the documents in which they coexist is >= 8

3. `pr` is in the prefix cache and can be found in the proximity prefix databases. In this case we simply query the proximity prefix databases.

Note that if a prefix is longer than 2 bytes, then it cannot be in the proximity prefix databases. Also, proximities larger than 4 are not present in these databases either. Therefore, the impact on relevancy is:

1. For common prefixes of one or two letters: we no longer distinguish between proximities from 4 to 8
2. For common prefixes of more than two letters: we no longer distinguish between any proximities
3. For uncommon prefixes: nothing changes

Regarding (1), it means that these two documents would be considered equally relevant according to the proximity rule for the query `heard pr` (IF `pr` is the prefix of more than 200 words in the dataset):
```json
[
    { "text": "I heard there is a faster proximity criterion" },
    { "text": "I heard there is a faster but less relevant proximity criterion" }
]
```

Regarding (2), it means that two documents would be considered equally relevant according to the proximity rule for the query "faster pro":
```json
[
    { "text": "I heard there is a faster but less relevant proximity criterion" }
    { "text": "I heard there is a faster proximity criterion" },
]
```
But the following document would be considered more relevant than the two documents above:
```json
{ "text": "I heard there is a faster swimmer who is competing in the pro section of the competition " }
```

Note, however, that this change of behaviour only occurs when using the set-based version of the proximity criterion. In cases where there are fewer than 1000 candidate documents when the proximity criterion is called, this PR does not change anything. 

---

## Performance

I couldn't use the existing search benchmarks to measure the impact of the PR, but I did some manual tests with the `songs` benchmark dataset.   

```
1. 10x 'a': 
	- 640ms ⟹ 630ms                  = no significant difference
2. 10x 'b':
	- set-based: 4.47s ⟹ 7.42        = bad, ~2x regression
	- dynamic: 1s ⟹ 870 ms           = no significant difference
3. 'Someone I l':
	- set-based: 250ms ⟹ 12 ms       = very good, x20 speedup
	- dynamic: 21ms ⟹ 11 ms          = good, x2 speedup 
4. 'billie e':
	- set-based: 623ms ⟹ 2ms         = very good, x300 speedup 
	- dynamic: ~4ms ⟹ 4ms            = no difference
5. 'billie ei':
	- set-based: 57ms ⟹ 20ms         = good, ~2x speedup
	- dynamic: ~4ms ⟹ ~2ms.          = no significant difference
6. 'i am getting o' 
	- set-based: 300ms ⟹ 60ms        = very good, 5x speedup
	- dynamic: 30ms ⟹ 6ms            = very good, 5x speedup
7. 'prologue 1 a 1:
	- set-based: 3.36s ⟹ 120ms       = very good, 30x speedup
	- dynamic: 200ms ⟹ 30ms          = very good, 6x speedup
8. 'prologue 1 a 10':
	- set-based: 590ms ⟹ 18ms        = very good, 30x speedup 
	- dynamic: 82ms ⟹ 35ms           = good, ~2x speedup
```

Performance is often significantly better, but there is also one regression in the set-based implementation with the query `b b b b b b b b b b`.

Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
2023-01-04 09:00:50 +00:00
bors[bot]
49f58b2c47
Merge #732
732: Interpret synonyms as phrases r=loiclec a=loiclec

# Pull Request

## Related issue
Fixes (when merged into meilisearch) https://github.com/meilisearch/meilisearch/issues/3125

## What does this PR do?
We now map multi-word synonyms to phrases instead of loose words. Such that the request:
```
btw I am going to nyc soon
```
is interpreted as (when the synonym interpretation is chosen for both `btw` and `nyc`):
```
"by the way" I am going to "New York City" soon
```
instead of:
```
by the way I am going to New York City soon
```

This prevents queries containing multi-word synonyms to exceed to word length limit and degrade the search performance.

In terms of relevancy, there is a debate to have. I personally think this could be considered an improvement, since it would be strange for a user to search for:
```
good DIY project
```
and have a result such as:
```
{
    "text": "whether it is a good project to do, you'll have to decide for yourself"
}
```
However, for synonyms such as `NYC -> New York City`, then we will stop matching documents where `New York` is separated from `City`. This is however solvable by adding an additional mapping: `NYC -> New York`.

## Performance

With the old behaviour, some long search requests making heavy uses of synonyms could take minutes to be executed. This is no longer the case, these search requests now take an average amount of time to be resolved.

Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
2023-01-04 08:34:18 +00:00
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
bors[bot]
9519e60f97
Merge #709
709: Optimise the `ExactWords` sub-criterion within `Exactness` r=loiclec a=loiclec

# Pull Request

## Related issue
Fixes (partially) https://github.com/meilisearch/meilisearch/issues/3116

## What does this PR do?
1. Reduces the algorithmic complexity of finding the documents containing N exact words from something that is exponential to something that is polynomial.
2. Cache intermediary results between different calls to the `exactness` criterion.

## Performance Results
On the `smol_songs.csv` dataset, a request containing 10 common words now takes about 60ms instead of 5 seconds to execute. For example, this is the case with this (admittedly nonsensical) request: `Rock You Hip Hop Folk World Country Electronic Love The`.


Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
2023-01-02 12:28:30 +00:00
Loïc Lecrenier
b5df889dcb Apply review suggestions: simplify implementation of exactness criterion 2023-01-02 13:11:47 +01:00
Loïc Lecrenier
8d36570958 Add explicit criterion impl strategy to proximity search tests 2023-01-02 10:37:01 +01:00
Loïc Lecrenier
32c6062e65 Optimise exactness criterion
1. Cache some results between calls to next()
2. Compute the combinations of exact words more efficiently
2022-12-22 12:28:45 +01:00
Loïc Lecrenier
f097aafa1c Add unit test for prefix handling by the proximity criterion 2022-12-22 12:08:00 +01:00
Loïc Lecrenier
777b387dc4 Avoid a prefix-related worst-case scenario in the proximity criterion 2022-12-22 12:08:00 +01:00
Loïc Lecrenier
b0f3dc2c06 Interpret synonyms as phrases 2022-12-22 12:07:51 +01:00
Louis Dureuil
4b166bea2b
Add primary_key_inference test 2022-12-21 15:13:38 +01:00
Louis Dureuil
5943100754
Fix existing tests 2022-12-21 15:13:38 +01:00
Louis Dureuil
b24def3281
Add logging when inference took place.
Displays log message in the form:
```
[2022-12-21T09:19:42Z INFO  milli::update::index_documents::enrich] Primary key was not specified in index. Inferred to 'id'
```
2022-12-21 15:13:38 +01:00
Louis Dureuil
402dcd6b2f
Simplify primary key inference 2022-12-21 15:13:38 +01:00
Louis Dureuil
13c95d25aa
Remove uses of UserError::MissingPrimaryKey not related to inference 2022-12-21 15:13:36 +01:00
bors[bot]
a8defb585b
Merge #742
742: Add a "Criterion implementation strategy" parameter to Search r=irevoire a=loiclec

Add a parameter to search requests which determines the implementation strategy of the criteria. This can be either `set-based`, `iterative`, or `dynamic` (ie choosing between set-based or iterative at search time). See https://github.com/meilisearch/milli/issues/755 for more context about this change.


Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
2022-12-21 12:18:49 +00:00
Loïc Lecrenier
339a4b0789 Make clippy happy 2022-12-21 12:49:34 +01:00
Loïc Lecrenier
229405aeb9 Choose implementation strategy of criterion at runtime 2022-12-21 09:29:39 +01:00
Loïc Lecrenier
fc0e7382fe Fix hard-deletion of an external id that was soft-deleted 2022-12-20 15:33:31 +01:00
Tamo
69edbf9f6d
Update milli/src/update/delete_documents.rs 2022-12-19 18:23:50 +01:00
Louis Dureuil
916c23e7be
Tests: rename snapshots 2022-12-19 10:07:17 +01:00
Louis Dureuil
ad9937c755
Fix tests after adding DeletionStrategy 2022-12-19 10:07:17 +01:00
Louis Dureuil
171c942282
Soft-deletion computation no longer takes into account the mapsize
Implemented solution 2.3 from https://github.com/meilisearch/meilisearch/issues/3231#issuecomment-1348628824
2022-12-19 10:07:17 +01:00
Louis Dureuil
e2ae3b24aa
Hard or soft delete according to the deletion strategy 2022-12-19 10:00:13 +01:00
Louis Dureuil
fc7618d49b
Add DeletionStrategy 2022-12-19 09:49:58 +01:00
ManyTheFish
7f88c4ff2f Fix #1714 test 2022-12-15 18:22:28 +01:00
ManyTheFish
96d4242b93 Update charabia 2022-12-15 18:22:22 +01:00
bors[bot]
5114686394
Merge #743
743: Fix finite pagination with placeholder search r=Kerollmops a=ManyTheFish

this bug is reproducible on real datasets and is hard to isolate in a simple test.

related to: https://github.com/meilisearch/meilisearch/issues/3200

poke `@curquiza` 

Co-authored-by: ManyTheFish <many@meilisearch.com>
2022-12-15 09:31:47 +00:00
ManyTheFish
3322018c06 Fix placeholder search 2022-12-14 20:09:47 +01:00
bors[bot]
0276d5212a
Merge #728
728: Add some integration tests on the sort criterion r=ManyTheFish a=loiclec

This is simply an integration test ensuring that the sort criterion works properly. 

However, only one version of the algorithm is tested here (the iterative one). To test the version that uses the facet DB, one has to manually set the `CANDIDATES_THRESHOLD` constant to `0`. I have done that and ensured that the test still succeeds. However, in the future, we will probably want to have an option to force which algorithm is used at runtime, for testing purposes.


Co-authored-by: Loïc Lecrenier <loic.lecrenier@me.com>
2022-12-14 09:27:12 +00:00
bors[bot]
406ee31d1a
Merge #737
737: Fix typo initial candidates computation r=Kerollmops a=ManyTheFish

When `Typo` criterion was after a different criterion than `Words` and the previous criterion wasn't returning any candidates at the first iteration of the bucket sort, then the `initial_candidates` were lost.

Now, `Typo`ensure to keep the `initial_candidates` between iterations.


related to https://github.com/meilisearch/meilisearch/issues/3200#issuecomment-1345179578
related to https://github.com/meilisearch/meilisearch/issues/3228

Co-authored-by: ManyTheFish <many@meilisearch.com>
2022-12-13 10:29:28 +00:00
ManyTheFish
2d8d0af1a6 Rename short name bc by ic for initial_candidates 2022-12-13 10:56:38 +01:00
Loïc Lecrenier
be3b00350c Apply review suggestions: naming and documentation 2022-12-13 10:15:22 +01:00
ManyTheFish
80d34a4169 Fix typo initial candiddates computation 2022-12-12 19:02:48 +01:00
Loïc Lecrenier
e3ee553dcc Remove soft deleted ids from ExternalDocumentIds during document import
If the document import replaces a document using hard deletion
2022-12-12 14:16:09 +01:00
Loïc Lecrenier
bebd050961 Add new test for bug 3021 2022-12-08 19:19:40 +01:00
ManyTheFish
55724f2412 Introduce an initial candidates set that makes the difference between an exhaustive count and an estimation 2022-12-08 09:41:34 +01:00