snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2024-11-22 18:17:39 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

Merge pull request #58 from meilisearch/actor-index-controller

Browse Source 
actor index controller
This commit is contained in:
marin 2021-03-15 18:25:35 +01:00 committed by GitHub
commit 0c17b166df
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
52 changed files with 3934 additions and 2759 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

768
Cargo.lock generated
View File

File diff suppressed because it is too large Load Diff

33
meilisearch-error/src/lib.rs
View File

@ -81,7 +81,6 @@ pub enum Code {
}
impl Code {
/// ascociate a `Code` variant to the actual ErrCode
fn err_code(&self) -> ErrCode {
use Code::*;
@ -94,17 +93,23 @@ impl Code {
// thrown when requesting an unexisting index
IndexNotFound => ErrCode::invalid("index_not_found", StatusCode::NOT_FOUND),
InvalidIndexUid => ErrCode::invalid("invalid_index_uid", StatusCode::BAD_REQUEST),
OpenIndex => ErrCode::internal("index_not_accessible", StatusCode::INTERNAL_SERVER_ERROR),
OpenIndex => {
ErrCode::internal("index_not_accessible", StatusCode::INTERNAL_SERVER_ERROR)
}
// invalid state error
InvalidState => ErrCode::internal("invalid_state", StatusCode::INTERNAL_SERVER_ERROR),
// thrown when no primary key has been set
MissingPrimaryKey => ErrCode::invalid("missing_primary_key", StatusCode::BAD_REQUEST),
// error thrown when trying to set an already existing primary key
PrimaryKeyAlreadyPresent => ErrCode::invalid("primary_key_already_present", StatusCode::BAD_REQUEST),
PrimaryKeyAlreadyPresent => {
ErrCode::invalid("primary_key_already_present", StatusCode::BAD_REQUEST)
}
// invalid document
MaxFieldsLimitExceeded => ErrCode::invalid("max_fields_limit_exceeded", StatusCode::BAD_REQUEST),
MaxFieldsLimitExceeded => {
ErrCode::invalid("max_fields_limit_exceeded", StatusCode::BAD_REQUEST)
}
MissingDocumentId => ErrCode::invalid("missing_document_id", StatusCode::BAD_REQUEST),
// error related to facets
@ -117,16 +122,26 @@ impl Code {
DocumentNotFound => ErrCode::invalid("document_not_found", StatusCode::NOT_FOUND),
Internal => ErrCode::internal("internal", StatusCode::INTERNAL_SERVER_ERROR),
InvalidToken => ErrCode::authentication("invalid_token", StatusCode::FORBIDDEN),
MissingAuthorizationHeader => ErrCode::authentication("missing_authorization_header", StatusCode::UNAUTHORIZED),
MissingAuthorizationHeader => {
ErrCode::authentication("missing_authorization_header", StatusCode::UNAUTHORIZED)
}
NotFound => ErrCode::invalid("not_found", StatusCode::NOT_FOUND),
PayloadTooLarge => ErrCode::invalid("payload_too_large", StatusCode::PAYLOAD_TOO_LARGE),
RetrieveDocument => ErrCode::internal("unretrievable_document", StatusCode::BAD_REQUEST),
RetrieveDocument => {
ErrCode::internal("unretrievable_document", StatusCode::BAD_REQUEST)
}
SearchDocuments => ErrCode::internal("search_error", StatusCode::BAD_REQUEST),
UnsupportedMediaType => ErrCode::invalid("unsupported_media_type", StatusCode::UNSUPPORTED_MEDIA_TYPE),
UnsupportedMediaType => {
ErrCode::invalid("unsupported_media_type", StatusCode::UNSUPPORTED_MEDIA_TYPE)
}
// error related to dump
DumpAlreadyInProgress => ErrCode::invalid("dump_already_in_progress", StatusCode::CONFLICT),
DumpProcessFailed => ErrCode::internal("dump_process_failed", StatusCode::INTERNAL_SERVER_ERROR),
DumpAlreadyInProgress => {
ErrCode::invalid("dump_already_in_progress", StatusCode::CONFLICT)
}
DumpProcessFailed => {
ErrCode::internal("dump_process_failed", StatusCode::INTERNAL_SERVER_ERROR)
}
}
}

489
meilisearch-http/Cargo.lock generated
View File

@ -1,5 +1,7 @@
# This file is automatically @generated by Cargo.
# It is not intended for manual editing.
version = 3
[[package]]
name = "actix-codec"
version = "0.3.0"
@ -12,8 +14,24 @@ dependencies = [
"futures-sink",
"log",
"pin-project 0.4.27",
"tokio",
"tokio-util",
"tokio 0.2.24",
"tokio-util 0.3.1",
]
[[package]]
name = "actix-codec"
version = "0.4.0-beta.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "90673465c6187bd0829116b02be465dc0195a74d7719f76ffff0effef934a92e"
dependencies = [
"bitflags",
"bytes 1.0.1",
"futures-core",
"futures-sink",
"log",
"pin-project-lite 0.2.0",
"tokio 1.2.0",
"tokio-util 0.6.3",
]
[[package]]
@ -22,27 +40,22 @@ version = "2.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "177837a10863f15ba8d3ae3ec12fac1099099529ed20083a27fdfe247381d0dc"
dependencies = [
"actix-codec",
"actix-rt",
"actix-service",
"actix-utils",
"actix-codec 0.3.0",
"actix-rt 1.1.1",
"actix-service 1.0.6",
"actix-utils 2.0.0",
"derive_more",
"either",
"futures-util",
"http",
"log",
"rustls",
"tokio-rustls",
"trust-dns-proto",
"trust-dns-resolver",
"webpki",
]
[[package]]
name = "actix-cors"
version = "0.5.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3f3a3d5493dbc9b8769fe88c030d057ef8d2edc5728e5e26267780e8fc5db0be"
version = "0.5.4"
dependencies = [
"actix-web",
"derive_more",
@ -58,28 +71,25 @@ version = "2.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "452299e87817ae5673910e53c243484ca38be3828db819b6011736fc6982e874"
dependencies = [
"actix-codec",
"actix-codec 0.3.0",
"actix-connect",
"actix-rt",
"actix-service",
"actix-rt 1.1.1",
"actix-service 1.0.6",
"actix-threadpool",
"actix-tls",
"actix-utils",
"actix-utils 2.0.0",
"base64 0.13.0",
"bitflags",
"brotli2",
"bytes 0.5.6",
"cookie",
"copyless",
"derive_more",
"either",
"encoding_rs",
"flate2",
"futures-channel",
"futures-core",
"futures-util",
"fxhash",
"h2",
"h2 0.2.7",
"http",
"httparse",
"indexmap",
@ -100,6 +110,51 @@ dependencies = [
"time 0.2.23",
]
[[package]]
name = "actix-http"
version = "3.0.0-beta.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8a01f9e0681608afa887d4269a0857ac4226f09ba5ceda25939e8391c9da610a"
dependencies = [
"actix-codec 0.4.0-beta.1",
"actix-rt 2.1.0",
"actix-service 2.0.0-beta.4",
"actix-tls",
"actix-utils 3.0.0-beta.2",
"ahash 0.7.0",
"base64 0.13.0",
"bitflags",
"brotli2",
"bytes 1.0.1",
"bytestring",
"cfg-if 1.0.0",
"cookie",
"derive_more",
"encoding_rs",
"flate2",
"futures-core",
"futures-util",
"h2 0.3.1",
"http",
"httparse",
"itoa",
"language-tags",
"log",
"mime",
"once_cell",
"percent-encoding",
"pin-project 1.0.2",
"rand 0.8.3",
"regex",
"serde",
"serde_json",
"serde_urlencoded",
"sha-1 0.9.2",
"smallvec",
"time 0.2.23",
"tokio 1.2.0",
]
[[package]]
name = "actix-macros"
version = "0.1.3"
@ -111,10 +166,20 @@ dependencies = [
]
[[package]]
name = "actix-router"
version = "0.2.5"
name = "actix-macros"
version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "bbd1f7dbda1645bf7da33554db60891755f6c01c1b2169e2f4c492098d30c235"
checksum = "dbcb2b608f0accc2f5bcf3dd872194ce13d94ee45b571487035864cf966b04ef"
dependencies = [
"quote",
"syn",
]
[[package]]
name = "actix-router"
version = "0.2.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2ad299af73649e1fc893e333ccf86f377751eb95ff875d095131574c6f43452c"
dependencies = [
"bytestring",
"http",
@ -129,33 +194,42 @@ version = "1.1.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "143fcc2912e0d1de2bcf4e2f720d2a60c28652ab4179685a1ee159e0fb3db227"
dependencies = [
"actix-macros",
"actix-macros 0.1.3",
"actix-threadpool",
"copyless",
"futures-channel",
"futures-util",
"smallvec",
"tokio",
"tokio 0.2.24",
]
[[package]]
name = "actix-rt"
version = "2.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0b4e57bc1a3915e71526d128baf4323700bd1580bc676239e2298a4c5b001f18"
dependencies = [
"actix-macros 0.2.0",
"futures-core",
"tokio 1.2.0",
]
[[package]]
name = "actix-server"
version = "1.0.4"
version = "2.0.0-beta.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "45407e6e672ca24784baa667c5d32ef109ccdd8d5e0b5ebb9ef8a67f4dfb708e"
checksum = "a99198727204a48f82559c18e4b0ba3197b97d5f4576a32bdbef371f3b4599c1"
dependencies = [
"actix-codec",
"actix-rt",
"actix-service",
"actix-utils",
"futures-channel",
"futures-util",
"actix-codec 0.4.0-beta.1",
"actix-rt 2.1.0",
"actix-service 2.0.0-beta.4",
"actix-utils 3.0.0-beta.2",
"futures-core",
"log",
"mio",
"mio-uds",
"mio 0.7.9",
"num_cpus",
"slab",
"socket2",
"tokio 1.2.0",
]
[[package]]
@ -169,17 +243,13 @@ dependencies = [
]
[[package]]
name = "actix-testing"
version = "1.0.1"
name = "actix-service"
version = "2.0.0-beta.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "47239ca38799ab74ee6a8a94d1ce857014b2ac36f242f70f3f75a66f691e791c"
checksum = "ca9756f4d32984ac454ae3155a276f6be69b424197bd3f0ca3c87cde72f41d63"
dependencies = [
"actix-macros",
"actix-rt",
"actix-server",
"actix-service",
"log",
"socket2",
"futures-core",
"pin-project-lite 0.2.0",
]
[[package]]
@ -199,18 +269,21 @@ dependencies = [
[[package]]
name = "actix-tls"
version = "2.0.0"
version = "3.0.0-beta.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "24789b7d7361cf5503a504ebe1c10806896f61e96eca9a7350e23001aca715fb"
checksum = "d2b1455e3f7a26d40cfc1080b571f41e8165e5a88e937ed579f7a4b3d55b0370"
dependencies = [
"actix-codec",
"actix-service",
"actix-utils",
"futures-util",
"rustls",
"tokio-rustls",
"webpki",
"webpki-roots",
"actix-codec 0.4.0-beta.1",
"actix-rt 2.1.0",
"actix-service 2.0.0-beta.4",
"actix-utils 3.0.0-beta.2",
"derive_more",
"futures-core",
"http",
"log",
"tokio-rustls 0.22.0",
"tokio-util 0.6.3",
"webpki-roots 0.21.0",
]
[[package]]
@ -219,9 +292,9 @@ version = "2.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2e9022dec56632d1d7979e59af14f0597a28a830a9c1c7fec8b2327eb9f16b5a"
dependencies = [
"actix-codec",
"actix-rt",
"actix-service",
"actix-codec 0.3.0",
"actix-rt 1.1.1",
"actix-service 1.0.6",
"bitflags",
"bytes 0.5.6",
"either",
@ -234,50 +307,63 @@ dependencies = [
]
[[package]]
name = "actix-web"
version = "3.3.2"
name = "actix-utils"
version = "3.0.0-beta.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e641d4a172e7faa0862241a20ff4f1f5ab0ab7c279f00c2d4587b77483477b86"
checksum = "458795e09a29bc5557604f9ff6f32236fd0ee457d631672e4ec8f6a0103bb292"
dependencies = [
"actix-codec",
"actix-http",
"actix-macros",
"actix-codec 0.4.0-beta.1",
"actix-rt 2.1.0",
"actix-service 2.0.0-beta.4",
"futures-core",
"futures-sink",
"log",
"pin-project-lite 0.2.0",
]
[[package]]
name = "actix-web"
version = "4.0.0-beta.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1d95e50c9e32e8456220b5804867de76e97a86ab8c38b51c9edcccc0f0fddca7"
dependencies = [
"actix-codec 0.4.0-beta.1",
"actix-http 3.0.0-beta.4",
"actix-macros 0.2.0",
"actix-router",
"actix-rt",
"actix-rt 2.1.0",
"actix-server",
"actix-service",
"actix-testing",
"actix-threadpool",
"actix-service 2.0.0-beta.4",
"actix-tls",
"actix-utils",
"actix-utils 3.0.0-beta.2",
"actix-web-codegen",
"ahash 0.7.0",
"awc",
"bytes 0.5.6",
"bytes 1.0.1",
"derive_more",
"either",
"encoding_rs",
"futures-channel",
"futures-core",
"futures-util",
"fxhash",
"log",
"mime",
"pin-project 1.0.2",
"regex",
"rustls",
"rustls 0.19.0",
"serde",
"serde_json",
"serde_urlencoded",
"smallvec",
"socket2",
"time 0.2.23",
"tinyvec",
"url",
]
[[package]]
name = "actix-web-codegen"
version = "0.4.0"
version = "0.5.0-beta.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ad26f77093333e0e7c6ffe54ebe3582d908a104e448723eec6d43d08b07143fb"
checksum = "7f138ac357a674c3b480ddb7bbd894b13c1b6e8927d728bc9ea5e17eee2f8fc9"
dependencies = [
"proc-macro2",
"quote",
@ -305,6 +391,17 @@ version = "0.3.8"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e8fd72866655d1904d6b0997d0b07ba561047d070fbe29de039031c641b61217"
[[package]]
name = "ahash"
version = "0.7.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "efa60d2eadd8b12a996add391db32bd1153eac697ba4869660c0016353611426"
dependencies = [
"getrandom 0.2.2",
"once_cell",
"version_check",
]
[[package]]
name = "aho-corasick"
version = "0.7.15"
@ -332,7 +429,7 @@ checksum = "afddf7f520a80dbf76e6f50a35bca42a2331ef227a28b3b6dc5c2e2338d114b1"
[[package]]
name = "assert-json-diff"
version = "1.0.1"
source = "git+https://github.com/qdequele/assert-json-diff#9012a0c8866d0f2db0ef9a6242e4a19d1e8c67e4"
source = "git+https://github.com/qdequele/assert-json-diff?branch=master#9012a0c8866d0f2db0ef9a6242e4a19d1e8c67e4"
dependencies = [
"serde",
"serde_json",
@ -348,7 +445,28 @@ dependencies = [
"futures-core",
"memchr",
"pin-project-lite 0.1.11",
"tokio",
"tokio 0.2.24",
]
[[package]]
name = "async-stream"
version = "0.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3670df70cbc01729f901f94c887814b3c68db038aad1329a418bae178bc5295c"
dependencies = [
"async-stream-impl",
"futures-core",
]
[[package]]
name = "async-stream-impl"
version = "0.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a3548b8efc9f8e8a5a0a2808c5bd8451a9031b9e5b879a79590304ae928b0a70"
dependencies = [
"proc-macro2",
"quote",
"syn",
]
[[package]]
@ -381,24 +499,26 @@ checksum = "cdb031dd78e28731d87d56cc8ffef4a8f36ca26c38fe2de700543e627f8a464a"
[[package]]
name = "awc"
version = "2.0.3"
version = "3.0.0-beta.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b381e490e7b0cfc37ebc54079b0413d8093ef43d14a4e4747083f7fa47a9e691"
checksum = "09aecd8728f6491a62b27454ea4b36fb7e50faf32928b0369b644e402c651f4e"
dependencies = [
"actix-codec",
"actix-http",
"actix-rt",
"actix-service",
"actix-codec 0.4.0-beta.1",
"actix-http 3.0.0-beta.4",
"actix-rt 2.1.0",
"actix-service 2.0.0-beta.4",
"base64 0.13.0",
"bytes 0.5.6",
"bytes 1.0.1",
"cfg-if 1.0.0",
"derive_more",
"futures-core",
"itoa",
"log",
"mime",
"percent-encoding",
"rand 0.7.3",
"rustls",
"pin-project-lite 0.2.0",
"rand 0.8.3",
"rustls 0.19.0",
"serde",
"serde_json",
"serde_urlencoded",
@ -569,12 +689,18 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e0dcbc35f504eb6fc275a6d20e4ebcda18cf50d40ba6fabff8c711fa16cb3b16"
[[package]]
name = "bytestring"
version = "0.1.5"
name = "bytes"
version = "1.0.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "fc7c05fa5172da78a62d9949d662d2ac89d4cc7355d7b49adee5163f1fb3f363"
checksum = "b700ce4376041dcd0a327fd0097c41095743c4c8af8887265942faf1100bd040"
[[package]]
name = "bytestring"
version = "1.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "90706ba19e97b90786e19dc0d5e2abd80008d99d4c0c5d1ad0b5e72cec7c494d"
dependencies = [
"bytes 0.5.6",
"bytes 1.0.1",
]
[[package]]
@ -1177,19 +1303,38 @@ dependencies = [
"http",
"indexmap",
"slab",
"tokio",
"tokio-util",
"tokio 0.2.24",
"tokio-util 0.3.1",
"tracing",
"tracing-futures",
]
[[package]]
name = "h2"
version = "0.3.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d832b01df74254fe364568d6ddc294443f61cbec82816b60904303af87efae78"
dependencies = [
"bytes 1.0.1",
"fnv",
"futures-core",
"futures-sink",
"futures-util",
"http",
"indexmap",
"slab",
"tokio 1.2.0",
"tokio-util 0.6.3",
"tracing",
]
[[package]]
name = "hashbrown"
version = "0.7.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "96282e96bfcd3da0d3aa9938bedf1e50df3269b6db08b4876d2da0bb1a0841cf"
dependencies = [
"ahash",
"ahash 0.3.8",
"autocfg",
]
@ -1268,11 +1413,11 @@ dependencies = [
[[package]]
name = "http"
version = "0.2.2"
version = "0.2.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "84129d298a6d57d246960ff8eb831ca4af3f96d29e2e28848dae275408658e26"
checksum = "7245cd7449cc792608c3c8a9eaf69bd4eabbabf802713748fd739c98b82f0747"
dependencies = [
"bytes 0.5.6",
"bytes 1.0.1",
"fnv",
"itoa",
]
@ -1330,7 +1475,7 @@ dependencies = [
"futures-channel",
"futures-core",
"futures-util",
"h2",
"h2 0.2.7",
"http",
"http-body",
"httparse",
@ -1338,7 +1483,7 @@ dependencies = [
"itoa",
"pin-project 1.0.2",
"socket2",
"tokio",
"tokio 0.2.24",
"tower-service",
"tracing",
"want",
@ -1354,9 +1499,9 @@ dependencies = [
"futures-util",
"hyper",
"log",
"rustls",
"tokio",
"tokio-rustls",
"rustls 0.18.1",
"tokio 0.2.24",
"tokio-rustls 0.14.1",
"webpki",
]
@ -1543,9 +1688,9 @@ dependencies = [
[[package]]
name = "libc"
version = "0.2.81"
version = "0.2.86"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1482821306169ec4d07f6aca392a4681f66c75c9918aa49641a2595db64053cb"
checksum = "b7282d924be3275cec7f6756ff4121987bc6481325397dde6ba3e7802b1a8b1c"
[[package]]
name = "linked-hash-map"
@ -1619,7 +1764,7 @@ checksum = "7ffc5c5338469d4d3ea17d269fa8ea3512ad247247c30bd2df69e68309ed0a08"
name = "meilisearch-error"
version = "0.19.0"
dependencies = [
"actix-http",
"actix-http 2.2.0",
]
[[package]]
@ -1627,13 +1772,15 @@ name = "meilisearch-http"
version = "0.17.0"
dependencies = [
"actix-cors",
"actix-http",
"actix-rt",
"actix-service",
"actix-http 3.0.0-beta.4",
"actix-rt 2.1.0",
"actix-service 2.0.0-beta.4",
"actix-web",
"anyhow",
"assert-json-diff",
"async-compression",
"async-stream",
"async-trait",
"byte-unit",
"bytes 0.6.0",
"chrono",
@ -1662,7 +1809,7 @@ dependencies = [
"rand 0.7.3",
"rayon",
"regex",
"rustls",
"rustls 0.19.0",
"sentry",
"serde",
"serde_json",
@ -1674,7 +1821,8 @@ dependencies = [
"tar",
"tempdir",
"tempfile",
"tokio",
"thiserror",
"tokio 1.2.0",
"uuid",
"vergen",
]
@ -1723,7 +1871,7 @@ dependencies = [
[[package]]
name = "milli"
version = "0.1.0"
source = "git+https://github.com/meilisearch/milli.git?rev=8dcb3e0#8dcb3e0c41965c96ae718ae85c45004cf94c6e94"
source = "git+https://github.com/meilisearch/milli.git?rev=794fce7#794fce7bff3e3461a7f3954fd97f58f8232e5a8e"
dependencies = [
"anyhow",
"bstr",
@ -1806,14 +1954,15 @@ dependencies = [
]
[[package]]
name = "mio-named-pipes"
version = "0.1.7"
name = "mio"
version = "0.7.9"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0840c1c50fd55e521b247f949c241c9997709f23bd7f023b9762cd561e935656"
checksum = "a5dede4e2065b3842b8b0af444119f3aa331cc7cc2dd20388bfb0f5d5a38823a"
dependencies = [
"libc",
"log",
"mio",
"miow 0.3.6",
"ntapi",
"winapi 0.3.9",
]
@ -1825,7 +1974,7 @@ checksum = "afcb699eb26d4332647cc848492bbc15eafb26f08d0304550d5aa1f612e066f0"
dependencies = [
"iovec",
"libc",
"mio",
"mio 0.6.23",
]
[[package]]
@ -1873,6 +2022,15 @@ dependencies = [
"libc",
]
[[package]]
name = "ntapi"
version = "0.3.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3f6bb902e437b6d86e03cce10a7e2af662292c5dfef23b65899ea3ac9354ad44"
dependencies = [
"winapi 0.3.9",
]
[[package]]
name = "num-integer"
version = "0.1.44"
@ -2437,17 +2595,17 @@ dependencies = [
"mime_guess",
"percent-encoding",
"pin-project-lite 0.2.0",
"rustls",
"rustls 0.18.1",
"serde",
"serde_json",
"serde_urlencoded",
"tokio",
"tokio-rustls",
"tokio 0.2.24",
"tokio-rustls 0.14.1",
"url",
"wasm-bindgen",
"wasm-bindgen-futures",
"web-sys",
"webpki-roots",
"webpki-roots 0.20.0",
"winreg 0.7.0",
]
@ -2521,6 +2679,19 @@ dependencies = [
"webpki",
]
[[package]]
name = "rustls"
version = "0.19.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "064fd21ff87c6e87ed4506e68beb42459caa4a0e2eb144932e6776768556980b"
dependencies = [
"base64 0.13.0",
"log",
"ring",
"sct",
"webpki",
]
[[package]]
name = "ryu"
version = "1.0.5"
@ -2947,18 +3118,18 @@ dependencies = [
[[package]]
name = "thiserror"
version = "1.0.22"
version = "1.0.24"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0e9ae34b84616eedaaf1e9dd6026dbe00dcafa92aa0c8077cb69df1fcfe5e53e"
checksum = "e0f4a65597094d4483ddaed134f409b2cb7c1beccf25201a9f73c719254fa98e"
dependencies = [
"thiserror-impl",
]
[[package]]
name = "thiserror-impl"
version = "1.0.22"
version = "1.0.24"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9ba20f23e85b10754cd195504aebf6a27e2e6cbe28c17778a0c930724628dd56"
checksum = "7765189610d8241a44529806d6fd1f2e0a08734313a35d5b3a556f92b381f3c0"
dependencies = [
"proc-macro2",
"quote",
@ -3060,22 +3231,40 @@ dependencies = [
"lazy_static",
"libc",
"memchr",
"mio",
"mio-named-pipes",
"mio 0.6.23",
"mio-uds",
"num_cpus",
"pin-project-lite 0.1.11",
"signal-hook-registry",
"slab",
"winapi 0.3.9",
]
[[package]]
name = "tokio"
version = "1.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e8190d04c665ea9e6b6a0dc45523ade572c088d2e6566244c1122671dbf4ae3a"
dependencies = [
"autocfg",
"bytes 1.0.1",
"libc",
"memchr",
"mio 0.7.9",
"num_cpus",
"once_cell",
"parking_lot",
"pin-project-lite 0.2.0",
"signal-hook-registry",
"tokio-macros",
"winapi 0.3.9",
]
[[package]]
name = "tokio-macros"
version = "0.2.6"
version = "1.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e44da00bfc73a25f814cd8d7e57a68a5c31b74b3152a0a1d1f590c97ed06265a"
checksum = "caf7b11a536f46a809a8a9f0bb4237020f70ecbf115b842360afb127ea2fda57"
dependencies = [
"proc-macro2",
"quote",
@ -3089,8 +3278,19 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e12831b255bcfa39dc0436b01e19fea231a37db570686c06ee72c423479f889a"
dependencies = [
"futures-core",
"rustls",
"tokio",
"rustls 0.18.1",
"tokio 0.2.24",
"webpki",
]
[[package]]
name = "tokio-rustls"
version = "0.22.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "bc6844de72e57df1980054b38be3a9f4702aba4858be64dd700181a8a6d0e1b6"
dependencies = [
"rustls 0.19.0",
"tokio 1.2.0",
"webpki",
]
@ -3105,7 +3305,21 @@ dependencies = [
"futures-sink",
"log",
"pin-project-lite 0.1.11",
"tokio",
"tokio 0.2.24",
]
[[package]]
name = "tokio-util"
version = "0.6.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ebb7cb2f00c5ae8df755b252306272cd1790d39728363936e01827e11f0b017b"
dependencies = [
"bytes 1.0.1",
"futures-core",
"futures-sink",
"log",
"pin-project-lite 0.2.0",
"tokio 1.2.0",
]
[[package]]
@ -3161,7 +3375,7 @@ dependencies = [
"rand 0.7.3",
"smallvec",
"thiserror",
"tokio",
"tokio 0.2.24",
"url",
]
@ -3181,7 +3395,7 @@ dependencies = [
"resolv-conf",
"smallvec",
"thiserror",
"tokio",
"tokio 0.2.24",
"trust-dns-proto",
]
@ -3433,6 +3647,15 @@ dependencies = [
"webpki",
]
[[package]]
name = "webpki-roots"
version = "0.21.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "82015b7e0b8bad8185994674a13a93306bea76cf5a16c5a181382fd3a5ec2376"
dependencies = [
"webpki",
]
[[package]]
name = "whatlang"
version = "0.9.0"

40
meilisearch-http/Cargo.toml
View File

@ -13,17 +13,20 @@ path = "src/main.rs"
vergen = "3.1.0"
[dependencies]
actix-cors = "0.5.3"
actix-http = "2"
actix-rt = "1"
actix-service = "1.0.6"
actix-web = { version = "3.3.2", features = ["rustls"] }
anyhow = "1.0.38"
actix-cors = { git = "https://github.com/MarinPostma/actix-extras.git", rev = "8f7b1fd" }
actix-http = { version = "3.0.0-beta.4", features = ["cookies"] }
actix-service = "2.0.0-beta.4"
actix-web = { version = "4.0.0-beta.4", features = ["rustls", "cookies"] }
anyhow = "1.0.36"
async-compression = { version = "0.3.6", features = ["gzip", "tokio-02"] }
async-stream = "0.3.0"
async-trait = "0.1.42"
byte-unit = { version = "4.0.9", default-features = false, features = ["std"] }
bytes = "0.6.0"
chrono = { version = "0.4.19", features = ["serde"] }
crossbeam-channel = "0.5.0"
dashmap = "4.0.2"
either = "1.6.1"
env_logger = "0.8.2"
flate2 = "1.0.19"
fst = "0.4.5"
@ -33,18 +36,20 @@ grenad = { git = "https://github.com/Kerollmops/grenad.git", rev = "3adcb26" }
heed = "0.10.6"
http = "0.2.1"
indexmap = { version = "1.3.2", features = ["serde-1"] }
itertools = "0.10.0"
log = "0.4.8"
main_error = "0.1.0"
meilisearch-error = { path = "../meilisearch-error" }
meilisearch-tokenizer = { git = "https://github.com/meilisearch/Tokenizer.git", branch = "main" }
memmap = "0.7.0"
milli = { git = "https://github.com/meilisearch/milli.git", rev = "f190d5f" }
milli = { git = "https://github.com/meilisearch/milli.git", rev = "b7b23cd" }
mime = "0.3.16"
once_cell = "1.5.2"
parking_lot = "0.11.1"
rand = "0.7.3"
rayon = "1.5.0"
regex = "1.4.2"
rustls = "0.18"
rustls = "0.19"
serde = { version = "1.0", features = ["derive"] }
serde_json = { version = "1.0.59", features = ["preserve_order"] }
sha2 = "0.9.1"
@ -53,15 +58,22 @@ slice-group-by = "0.2.6"
structopt = "0.3.20"
tar = "0.4.29"
tempfile = "3.1.0"
tokio = { version = "0.2", features = ["full"] }
dashmap = "4.0.2"
thiserror = "1.0.24"
tokio = { version = "1", features = ["full"] }
uuid = "0.8.2"
itertools = "0.10.0"
either = "1.6.1"
[dependencies.sentry]
default-features = false
features = ["with_client_implementation", "with_panic", "with_failure", "with_device_info", "with_rust_info", "with_reqwest_transport", "with_rustls", "with_env_logger"]
features = [
"with_client_implementation",
"with_panic",
"with_failure",
"with_device_info",
"with_rust_info",
"with_reqwest_transport",
"with_rustls",
"with_env_logger"
]
optional = true
version = "0.18.1"
@ -70,7 +82,7 @@ version = "0.18.1"
serde_url_params = "0.2.0"
tempdir = "0.3.7"
assert-json-diff = { branch = "master", git = "https://github.com/qdequele/assert-json-diff" }
tokio = { version = "0.2", features = ["macros", "time"] }
actix-rt = "2.1.0"
urlencoding = "1.1.1"
[features]

98
meilisearch-http/src/data/mod.rs
View File

@ -1,16 +1,15 @@
mod search;
pub mod search;
mod updates;
pub use search::{SearchQuery, SearchResult, DEFAULT_SEARCH_LIMIT};
use std::fs::create_dir_all;
use std::ops::Deref;
use std::sync::Arc;
use sha2::Digest;
use anyhow::bail;
use crate::index_controller::{IndexController, LocalIndexController, IndexMetadata, Settings, IndexSettings};
use crate::index::Settings;
use crate::index_controller::IndexController;
use crate::index_controller::{IndexMetadata, IndexSettings};
use crate::option::Opt;
#[derive(Clone)]
@ -26,9 +25,8 @@ impl Deref for Data {
}
}
#[derive(Clone)]
pub struct DataInner {
pub index_controller: Arc<LocalIndexController>,
pub index_controller: IndexController,
pub api_keys: ApiKeys,
options: Opt,
}
@ -60,15 +58,11 @@ impl ApiKeys {
impl Data {
pub fn new(options: Opt) -> anyhow::Result<Data> {
let path = options.db_path.clone();
let indexer_opts = options.indexer_options.clone();
create_dir_all(&path)?;
let index_controller = LocalIndexController::new(
&path,
indexer_opts,
options.max_mdb_size.get_bytes(),
options.max_udb_size.get_bytes(),
)?;
let index_controller = Arc::new(index_controller);
let index_size = options.max_mdb_size.get_bytes() as usize;
let update_store_size = options.max_udb_size.get_bytes() as usize;
let index_controller = IndexController::new(&path, index_size, update_store_size)?;
let mut api_keys = ApiKeys {
master: options.clone().master_key,
@ -78,70 +72,39 @@ impl Data {
api_keys.generate_missing_api_keys();
let inner = DataInner { index_controller, options, api_keys };
let inner = DataInner {
index_controller,
options,
api_keys,
};
let inner = Arc::new(inner);
Ok(Data { inner })
}
pub fn settings<S: AsRef<str>>(&self, index_uid: S) -> anyhow::Result<Settings> {
let index = self.index_controller
.index(&index_uid)?
.ok_or_else(|| anyhow::anyhow!("Index {} does not exist.", index_uid.as_ref()))?;
let txn = index.read_txn()?;
let displayed_attributes = index
.displayed_fields(&txn)?
.map(|fields| fields.into_iter().map(String::from).collect())
.unwrap_or_else(|| vec!["*".to_string()]);
let searchable_attributes = index
.searchable_fields(&txn)?
.map(|fields| fields.into_iter().map(String::from).collect())
.unwrap_or_else(|| vec!["*".to_string()]);
let faceted_attributes = index
.faceted_fields(&txn)?
.into_iter()
.map(|(k, v)| (k, v.to_string()))
.collect();
let criteria = index
.criteria(&txn)?
.into_iter()
.map(|v| v.to_string())
.collect();
Ok(Settings {
displayed_attributes: Some(Some(displayed_attributes)),
searchable_attributes: Some(Some(searchable_attributes)),
faceted_attributes: Some(Some(faceted_attributes)),
ranking_rules: Some(Some(criteria)),
})
pub async fn settings(&self, uid: String) -> anyhow::Result<Settings> {
self.index_controller.settings(uid).await
}
pub fn list_indexes(&self) -> anyhow::Result<Vec<IndexMetadata>> {
self.index_controller.list_indexes()
pub async fn list_indexes(&self) -> anyhow::Result<Vec<IndexMetadata>> {
self.index_controller.list_indexes().await
}
pub fn index(&self, name: impl AsRef<str>) -> anyhow::Result<Option<IndexMetadata>> {
Ok(self
.list_indexes()?
.into_iter()
.find(|i| i.uid == name.as_ref()))
pub async fn index(&self, uid: String) -> anyhow::Result<IndexMetadata> {
self.index_controller.get_index(uid).await
}
pub fn create_index(&self, name: impl AsRef<str>, primary_key: Option<impl AsRef<str>>) -> anyhow::Result<IndexMetadata> {
if !is_index_uid_valid(name.as_ref()) {
bail!("invalid index uid: {:?}", name.as_ref())
}
pub async fn create_index(
&self,
uid: String,
primary_key: Option<String>,
) -> anyhow::Result<IndexMetadata> {
let settings = IndexSettings {
name: Some(name.as_ref().to_string()),
primary_key: primary_key.map(|s| s.as_ref().to_string()),
uid: Some(uid),
primary_key,
};
let meta = self.index_controller.create_index(settings)?;
let meta = self.index_controller.create_index(settings).await?;
Ok(meta)
}
@ -155,8 +118,3 @@ impl Data {
&self.api_keys
}
}
fn is_index_uid_valid(uid: &str) -> bool {
uid.chars().all(|x| x.is_ascii_alphanumeric() || x == '-' || x == '_')
}

354
meilisearch-http/src/data/search.rs
View File

@ -1,353 +1,37 @@
use std::collections::{HashSet, BTreeMap};
use std::mem;
use std::time::Instant;
use anyhow::{bail, Context};
use either::Either;
use heed::RoTxn;
use meilisearch_tokenizer::{Analyzer, AnalyzerConfig};
use milli::{FacetCondition, Index, MatchingWords, facet::FacetValue, obkv_to_json};
use serde::{Deserialize, Serialize};
use serde_json::{Map, Value};
use super::Data;
use crate::index_controller::IndexController;
pub const DEFAULT_SEARCH_LIMIT: usize = 20;
const fn default_search_limit() -> usize {
DEFAULT_SEARCH_LIMIT
}
#[derive(Deserialize)]
#[serde(rename_all = "camelCase", deny_unknown_fields)]
#[allow(dead_code)]
pub struct SearchQuery {
pub q: Option<String>,
pub offset: Option<usize>,
#[serde(default = "default_search_limit")]
pub limit: usize,
pub attributes_to_retrieve: Option<Vec<String>>,
pub attributes_to_crop: Option<Vec<String>>,
pub crop_length: Option<usize>,
pub attributes_to_highlight: Option<HashSet<String>>,
pub filters: Option<String>,
pub matches: Option<bool>,
pub facet_filters: Option<Value>,
pub facet_distributions: Option<Vec<String>>,
}
impl SearchQuery {
pub fn perform(&self, index: impl AsRef<Index>) -> anyhow::Result<SearchResult> {
let index = index.as_ref();
let before_search = Instant::now();
let rtxn = index.read_txn()?;
let mut search = index.search(&rtxn);
if let Some(ref query) = self.q {
search.query(query);
}
search.limit(self.limit);
search.offset(self.offset.unwrap_or_default());
if let Some(ref facets) = self.facet_filters {
if let Some(facets) = parse_facets(facets, index, &rtxn)? {
search.facet_condition(facets);
}
}
let milli::SearchResult {
documents_ids,
matching_words,
candidates,
} = search.execute()?;
let mut documents = Vec::new();
let fields_ids_map = index.fields_ids_map(&rtxn)?;
let displayed_fields_ids = index.displayed_fields_ids(&rtxn)?;
let attributes_to_retrieve_ids = match self.attributes_to_retrieve {
Some(ref attrs) if attrs.iter().any(|f| f == "*") => None,
Some(ref attrs) => attrs
.iter()
.filter_map(|f| fields_ids_map.id(f))
.collect::<Vec<_>>()
.into(),
None => None,
};
let displayed_fields_ids = match (displayed_fields_ids, attributes_to_retrieve_ids) {
(_, Some(ids)) => ids,
(Some(ids), None) => ids,
(None, None) => fields_ids_map.iter().map(|(id, _)| id).collect(),
};
let stop_words = fst::Set::default();
let highlighter = Highlighter::new(&stop_words);
for (_id, obkv) in index.documents(&rtxn, documents_ids)? {
let mut object = obkv_to_json(&displayed_fields_ids, &fields_ids_map, obkv)?;
if let Some(ref attributes_to_highlight) = self.attributes_to_highlight {
highlighter.highlight_record(&mut object, &matching_words, attributes_to_highlight);
}
documents.push(object);
}
let nb_hits = candidates.len();
let facet_distributions = match self.facet_distributions {
Some(ref fields) => {
let mut facet_distribution = index.facets_distribution(&rtxn);
if fields.iter().all(|f| f != "*") {
facet_distribution.facets(fields);
}
Some(facet_distribution.candidates(candidates).execute()?)
}
None => None,
};
Ok(SearchResult {
hits: documents,
nb_hits,
query: self.q.clone().unwrap_or_default(),
limit: self.limit,
offset: self.offset.unwrap_or_default(),
processing_time_ms: before_search.elapsed().as_millis(),
facet_distributions,
})
}
}
#[derive(Serialize)]
#[serde(rename_all = "camelCase")]
pub struct SearchResult {
hits: Vec<Map<String, Value>>,
nb_hits: u64,
query: String,
limit: usize,
offset: usize,
processing_time_ms: u128,
#[serde(skip_serializing_if = "Option::is_none")]
facet_distributions: Option<BTreeMap<String, BTreeMap<FacetValue, u64>>>,
}
struct Highlighter<'a, A> {
analyzer: Analyzer<'a, A>,
}
impl<'a, A: AsRef<[u8]>> Highlighter<'a, A> {
fn new(stop_words: &'a fst::Set<A>) -> Self {
let analyzer = Analyzer::new(AnalyzerConfig::default_with_stopwords(stop_words));
Self { analyzer }
}
fn highlight_value(&self, value: Value, words_to_highlight: &MatchingWords) -> Value {
match value {
Value::Null => Value::Null,
Value::Bool(boolean) => Value::Bool(boolean),
Value::Number(number) => Value::Number(number),
Value::String(old_string) => {
let mut string = String::new();
let analyzed = self.analyzer.analyze(&old_string);
for (word, token) in analyzed.reconstruct() {
if token.is_word() {
let to_highlight = words_to_highlight.matches(token.text());
if to_highlight {
string.push_str("<mark>")
}
string.push_str(word);
if to_highlight {
string.push_str("</mark>")
}
} else {
string.push_str(word);
}
}
Value::String(string)
}
Value::Array(values) => Value::Array(
values
.into_iter()
.map(|v| self.highlight_value(v, words_to_highlight))
.collect(),
),
Value::Object(object) => Value::Object(
object
.into_iter()
.map(|(k, v)| (k, self.highlight_value(v, words_to_highlight)))
.collect(),
),
}
}
fn highlight_record(
&self,
object: &mut Map<String, Value>,
words_to_highlight: &MatchingWords,
attributes_to_highlight: &HashSet<String>,
) {
// TODO do we need to create a string for element that are not and needs to be highlight?
for (key, value) in object.iter_mut() {
if attributes_to_highlight.contains(key) {
let old_value = mem::take(value);
*value = self.highlight_value(old_value, words_to_highlight);
}
}
}
}
use crate::index::{SearchQuery, SearchResult};
impl Data {
pub fn search<S: AsRef<str>>(
pub async fn search(
&self,
index: S,
index: String,
search_query: SearchQuery,
) -> anyhow::Result<SearchResult> {
match self.index_controller.index(&index)? {
Some(index) => Ok(search_query.perform(index)?),
None => bail!("index {:?} doesn't exists", index.as_ref()),
}
self.index_controller.search(index, search_query).await
}
pub async fn retrieve_documents<S>(
pub async fn retrieve_documents(
&self,
index: impl AsRef<str> + Send + Sync + 'static,
index: String,
offset: usize,
limit: usize,
attributes_to_retrieve: Option<Vec<S>>,
) -> anyhow::Result<Vec<Map<String, Value>>>
where
S: AsRef<str> + Send + Sync + 'static,
{
let index_controller = self.index_controller.clone();
let documents: anyhow::Result<_> = tokio::task::spawn_blocking(move || {
let index = index_controller
.index(&index)?
.with_context(|| format!("Index {:?} doesn't exist", index.as_ref()))?;
let txn = index.read_txn()?;
let fields_ids_map = index.fields_ids_map(&txn)?;
let attributes_to_retrieve_ids = match attributes_to_retrieve {
Some(attrs) => attrs
.iter()
.filter_map(|f| fields_ids_map.id(f.as_ref()))
.collect::<Vec<_>>(),
None => fields_ids_map.iter().map(|(id, _)| id).collect(),
};
let iter = index.documents.range(&txn, &(..))?.skip(offset).take(limit);
let mut documents = Vec::new();
for entry in iter {
let (_id, obkv) = entry?;
let object = obkv_to_json(&attributes_to_retrieve_ids, &fields_ids_map, obkv)?;
documents.push(object);
}
Ok(documents)
})
.await?;
documents
attributes_to_retrieve: Option<Vec<String>>,
) -> anyhow::Result<Vec<Map<String, Value>>> {
self.index_controller
.documents(index, offset, limit, attributes_to_retrieve)
.await
}
pub async fn retrieve_document<S>(
pub async fn retrieve_document(
&self,
index: impl AsRef<str> + Sync + Send + 'static,
document_id: impl AsRef<str> + Sync + Send + 'static,
attributes_to_retrieve: Option<Vec<S>>,
) -> anyhow::Result<Map<String, Value>>
where
S: AsRef<str> + Sync + Send + 'static,
{
let index_controller = self.index_controller.clone();
let document: anyhow::Result<_> = tokio::task::spawn_blocking(move || {
let index = index_controller
.index(&index)?
.with_context(|| format!("Index {:?} doesn't exist", index.as_ref()))?;
let txn = index.read_txn()?;
let fields_ids_map = index.fields_ids_map(&txn)?;
let attributes_to_retrieve_ids = match attributes_to_retrieve {
Some(attrs) => attrs
.iter()
.filter_map(|f| fields_ids_map.id(f.as_ref()))
.collect::<Vec<_>>(),
None => fields_ids_map.iter().map(|(id, _)| id).collect(),
};
let internal_id = index
.external_documents_ids(&txn)?
.get(document_id.as_ref().as_bytes())
.with_context(|| format!("Document with id {} not found", document_id.as_ref()))?;
let document = index
.documents(&txn, std::iter::once(internal_id))?
.into_iter()
.next()
.map(|(_, d)| d);
match document {
Some(document) => Ok(obkv_to_json(
&attributes_to_retrieve_ids,
&fields_ids_map,
document,
)?),
None => bail!("Document with id {} not found", document_id.as_ref()),
}
})
.await?;
document
index: String,
document_id: String,
attributes_to_retrieve: Option<Vec<String>>,
) -> anyhow::Result<Map<String, Value>> {
self.index_controller
.document(index, document_id, attributes_to_retrieve)
.await
}
}
fn parse_facets_array(
txn: &RoTxn,
index: &Index,
arr: &Vec<Value>,
) -> anyhow::Result<Option<FacetCondition>> {
let mut ands = Vec::new();
for value in arr {
match value {
Value::String(s) => ands.push(Either::Right(s.clone())),
Value::Array(arr) => {
let mut ors = Vec::new();
for value in arr {
match value {
Value::String(s) => ors.push(s.clone()),
v => bail!("Invalid facet expression, expected String, found: {:?}", v),
}
}
ands.push(Either::Left(ors));
}
v => bail!(
"Invalid facet expression, expected String or [String], found: {:?}",
v
),
}
}
FacetCondition::from_array(txn, index, ands)
}
fn parse_facets(
facets: &Value,
index: &Index,
txn: &RoTxn,
) -> anyhow::Result<Option<FacetCondition>> {
match facets {
// Disabled for now
//Value::String(expr) => Ok(Some(FacetCondition::from_str(txn, index, expr)?)),
Value::Array(arr) => parse_facets_array(txn, index, arr),
v => bail!(
"Invalid facet expression, expected Array, found: {:?}",
v
),
}
}

125
meilisearch-http/src/data/updates.rs
View File

@ -1,123 +1,80 @@
use std::ops::Deref;
use anyhow::bail;
use async_compression::tokio_02::write::GzipEncoder;
use futures_util::stream::StreamExt;
use actix_web::web::Payload;
use milli::update::{IndexDocumentsMethod, UpdateFormat};
use tokio::io::AsyncWriteExt;
use super::{Data, is_index_uid_valid};
use crate::index_controller::{UpdateStatus, IndexController, Settings, IndexSettings, IndexMetadata};
use super::Data;
use crate::index::Settings;
use crate::index_controller::{IndexMetadata, IndexSettings, UpdateStatus};
impl Data {
pub async fn add_documents<B, E>(
pub async fn add_documents(
&self,
index: impl AsRef<str> + Send + Sync + 'static,
index: String,
method: IndexDocumentsMethod,
format: UpdateFormat,
mut stream: impl futures::Stream<Item=Result<B, E>> + Unpin,
stream: Payload,
primary_key: Option<String>,
) -> anyhow::Result<UpdateStatus>
where
B: Deref<Target = [u8]>,
E: std::error::Error + Send + Sync + 'static,
{
if !is_index_uid_valid(index.as_ref()) {
bail!("invalid index uid: {:?}", index.as_ref())
}
let file = tokio::task::spawn_blocking(tempfile::tempfile).await?;
let file = tokio::fs::File::from_std(file?);
let mut encoder = GzipEncoder::new(file);
let mut empty_update = true;
while let Some(result) = stream.next().await {
empty_update = false;
let bytes = &*result?;
encoder.write_all(&bytes[..]).await?;
}
encoder.shutdown().await?;
let mut file = encoder.into_inner();
file.sync_all().await?;
let file = file.into_std().await;
let index_controller = self.index_controller.clone();
let update = tokio::task::spawn_blocking(move ||{
let mmap;
let bytes = if empty_update {
&[][..]
} else {
mmap = unsafe { memmap::Mmap::map(&file)? };
&mmap
};
index_controller.add_documents(index, method, format, &bytes, primary_key)
}).await??;
Ok(update.into())
) -> anyhow::Result<UpdateStatus> {
let update_status = self
.index_controller
.add_documents(index, method, format, stream, primary_key)
.await?;
Ok(update_status)
}
pub async fn update_settings(
&self,
index: impl AsRef<str> + Send + Sync + 'static,
index: String,
settings: Settings,
create: bool,
) -> anyhow::Result<UpdateStatus> {
if !is_index_uid_valid(index.as_ref()) {
bail!("invalid index uid: {:?}", index.as_ref())
}
let index_controller = self.index_controller.clone();
let update = tokio::task::spawn_blocking(move || index_controller.update_settings(index, settings, create)).await??;
Ok(update.into())
let update = self
.index_controller
.update_settings(index, settings, create)
.await?;
Ok(update)
}
pub async fn clear_documents(
&self,
index: impl AsRef<str> + Sync + Send + 'static,
) -> anyhow::Result<UpdateStatus> {
let index_controller = self.index_controller.clone();
let update = tokio::task::spawn_blocking(move || index_controller.clear_documents(index)).await??;
Ok(update.into())
pub async fn clear_documents(&self, index: String) -> anyhow::Result<UpdateStatus> {
let update = self.index_controller.clear_documents(index).await?;
Ok(update)
}
pub async fn delete_documents(
&self,
index: impl AsRef<str> + Sync + Send + 'static,
index: String,
document_ids: Vec<String>,
) -> anyhow::Result<UpdateStatus> {
let index_controller = self.index_controller.clone();
let update = tokio::task::spawn_blocking(move || index_controller.delete_documents(index, document_ids)).await??;
Ok(update.into())
let update = self
.index_controller
.delete_documents(index, document_ids)
.await?;
Ok(update)
}
pub async fn delete_index(
&self,
index: impl AsRef<str> + Send + Sync + 'static,
) -> anyhow::Result<()> {
let index_controller = self.index_controller.clone();
tokio::task::spawn_blocking(move || { index_controller.delete_index(index) }).await??;
pub async fn delete_index(&self, index: String) -> anyhow::Result<()> {
self.index_controller.delete_index(index).await?;
Ok(())
}
#[inline]
pub fn get_update_status(&self, index: impl AsRef<str>, uid: u64) -> anyhow::Result<Option<UpdateStatus>> {
self.index_controller.update_status(index, uid)
pub async fn get_update_status(&self, index: String, uid: u64) -> anyhow::Result<UpdateStatus> {
self.index_controller.update_status(index, uid).await
}
pub fn get_updates_status(&self, index: impl AsRef<str>) -> anyhow::Result<Vec<UpdateStatus>> {
self.index_controller.all_update_status(index)
pub async fn get_updates_status(&self, index: String) -> anyhow::Result<Vec<UpdateStatus>> {
self.index_controller.all_update_status(index).await
}
pub fn update_index(
pub async fn update_index(
&self,
name: impl AsRef<str>,
primary_key: Option<impl AsRef<str>>,
new_name: Option<impl AsRef<str>>
uid: String,
primary_key: Option<String>,
new_uid: Option<String>,
) -> anyhow::Result<IndexMetadata> {
let settings = IndexSettings {
name: new_name.map(|s| s.as_ref().to_string()),
primary_key: primary_key.map(|s| s.as_ref().to_string()),
uid: new_uid,
primary_key,
};
self.index_controller.update_index(name, settings)
self.index_controller.update_index(uid, settings).await
}
}

52
meilisearch-http/src/error.rs
View File

@ -1,13 +1,13 @@
use std::error;
use std::fmt;
use actix_http::ResponseBuilder;
use actix_web as aweb;
use actix_web::dev::HttpResponseBuilder;
use actix_web::error::{JsonPayloadError, QueryPayloadError};
use actix_web::http::Error as HttpError;
use actix_web::http::StatusCode;
use serde::ser::{Serialize, Serializer, SerializeStruct};
use meilisearch_error::{ErrorCode, Code};
use meilisearch_error::{Code, ErrorCode};
use serde::ser::{Serialize, SerializeStruct, Serializer};
#[derive(Debug)]
pub struct ResponseError {
@ -31,19 +31,25 @@ impl fmt::Display for ResponseError {
// TODO: remove this when implementing actual error handling
impl From<anyhow::Error> for ResponseError {
fn from(other: anyhow::Error) -> ResponseError {
ResponseError { inner: Box::new(Error::NotFound(other.to_string())) }
ResponseError {
inner: Box::new(Error::NotFound(other.to_string())),
}
}
}
impl From<Error> for ResponseError {
fn from(error: Error) -> ResponseError {
ResponseError { inner: Box::new(error) }
ResponseError {
inner: Box::new(error),
}
}
}
impl From<FacetCountError> for ResponseError {
fn from(err: FacetCountError) -> ResponseError {
ResponseError { inner: Box::new(err) }
ResponseError {
inner: Box::new(err),
}
}
}
@ -66,7 +72,7 @@ impl Serialize for ResponseError {
impl aweb::error::ResponseError for ResponseError {
fn error_response(&self) -> aweb::HttpResponse {
ResponseBuilder::new(self.status_code()).json(&self)
HttpResponseBuilder::new(self.status_code()).json(&self)
}
fn status_code(&self) -> StatusCode {
@ -129,7 +135,10 @@ impl ErrorCode for Error {
pub enum FacetCountError {
AttributeNotSet(String),
SyntaxError(String),
UnexpectedToken { found: String, expected: &'static [&'static str] },
UnexpectedToken {
found: String,
expected: &'static [&'static str],
},
NoFacetSet,
}
@ -142,7 +151,10 @@ impl ErrorCode for FacetCountError {
}
impl FacetCountError {
pub fn unexpected_token(found: impl ToString, expected: &'static [&'static str]) -> FacetCountError {
pub fn unexpected_token(
found: impl ToString,
expected: &'static [&'static str],
) -> FacetCountError {
let found = found.to_string();
FacetCountError::UnexpectedToken { expected, found }
}
@ -161,7 +173,9 @@ impl fmt::Display for FacetCountError {
match self {
AttributeNotSet(attr) => write!(f, "Attribute {} is not set as facet", attr),
SyntaxError(msg) => write!(f, "Syntax error: {}", msg),
UnexpectedToken { expected, found } => write!(f, "Unexpected {} found, expected {:?}", found, expected),
UnexpectedToken { expected, found } => {
write!(f, "Unexpected {} found, expected {:?}", found, expected)
}
NoFacetSet => write!(f, "Can't perform facet count, as no facet is set"),
}
}
@ -260,8 +274,8 @@ impl From<std::io::Error> for Error {
}
}
impl From<actix_http::Error> for Error {
fn from(err: actix_http::Error) -> Error {
impl From<HttpError> for Error {
fn from(err: HttpError) -> Error {
Error::Internal(err.to_string())
}
}
@ -275,10 +289,14 @@ impl From<serde_json::error::Error> for Error {
impl From<JsonPayloadError> for Error {
fn from(err: JsonPayloadError) -> Error {
match err {
JsonPayloadError::Deserialize(err) => Error::BadRequest(format!("Invalid JSON: {}", err)),
JsonPayloadError::Deserialize(err) => {
Error::BadRequest(format!("Invalid JSON: {}", err))
}
JsonPayloadError::Overflow => Error::PayloadTooLarge,
JsonPayloadError::ContentType => Error::UnsupportedMediaType,
JsonPayloadError::Payload(err) => Error::BadRequest(format!("Problem while decoding the request: {}", err)),
JsonPayloadError::Payload(err) => {
Error::BadRequest(format!("Problem while decoding the request: {}", err))
}
}
}
}
@ -286,7 +304,9 @@ impl From<JsonPayloadError> for Error {
impl From<QueryPayloadError> for Error {
fn from(err: QueryPayloadError) -> Error {
match err {
QueryPayloadError::Deserialize(err) => Error::BadRequest(format!("Invalid query parameters: {}", err)),
QueryPayloadError::Deserialize(err) => {
Error::BadRequest(format!("Invalid query parameters: {}", err))
}
}
}
}

39
meilisearch-http/src/helpers/authentication.rs
View File

@ -3,27 +3,26 @@ use std::pin::Pin;
use std::rc::Rc;
use std::task::{Context, Poll};
use actix_service::{Service, Transform};
use actix_web::{dev::ServiceRequest, dev::ServiceResponse, web};
use actix_web::dev::{Service, ServiceRequest, ServiceResponse, Transform};
use actix_web::web;
use futures::future::{err, ok, Future, Ready};
use crate::error::{Error, ResponseError};
use crate::Data;
#[derive(Clone)]
#[derive(Clone, Copy)]
pub enum Authentication {
Public,
Private,
Admin,
}
impl<S: 'static, B> Transform<S> for Authentication
impl<S: 'static, B> Transform<S, ServiceRequest> for Authentication
where
S: Service<Request = ServiceRequest, Response = ServiceResponse<B>, Error = actix_web::Error>,
S: Service<ServiceRequest, Response = ServiceResponse<B>, Error = actix_web::Error>,
S::Future: 'static,
B: 'static,
{
type Request = ServiceRequest;
type Response = ServiceResponse<B>;
type Error = actix_web::Error;
type InitError = ();
@ -32,7 +31,7 @@ where
fn new_transform(&self, service: S) -> Self::Future {
ok(LoggingMiddleware {
acl: self.clone(),
acl: *self,
service: Rc::new(RefCell::new(service)),
})
}
@ -44,23 +43,22 @@ pub struct LoggingMiddleware<S> {
}
#[allow(clippy::type_complexity)]
impl<S, B> Service for LoggingMiddleware<S>
impl<S, B> Service<ServiceRequest> for LoggingMiddleware<S>
where
S: Service<Request = ServiceRequest, Response = ServiceResponse<B>, Error = actix_web::Error> + 'static,
S: Service<ServiceRequest, Response = ServiceResponse<B>, Error = actix_web::Error> + 'static,
S::Future: 'static,
B: 'static,
{
type Request = ServiceRequest;
type Response = ServiceResponse<B>;
type Error = actix_web::Error;
type Future = Pin<Box<dyn Future<Output = Result<Self::Response, Self::Error>>>>;
fn poll_ready(&mut self, cx: &mut Context) -> Poll<Result<(), Self::Error>> {
fn poll_ready(&self, cx: &mut Context) -> Poll<Result<(), Self::Error>> {
self.service.poll_ready(cx)
}
fn call(&mut self, req: ServiceRequest) -> Self::Future {
let mut svc = self.service.clone();
fn call(&self, req: ServiceRequest) -> Self::Future {
let svc = self.service.clone();
// This unwrap is left because this error should never appear. If that's the case, then
// it means that actix-web has an issue or someone changes the type `Data`.
let data = req.app_data::<web::Data<Data>>().unwrap();
@ -72,10 +70,16 @@ where
let auth_header = match req.headers().get("X-Meili-API-Key") {
Some(auth) => match auth.to_str() {
Ok(auth) => auth,
Err(_) => return Box::pin(err(ResponseError::from(Error::MissingAuthorizationHeader).into())),
Err(_) => {
return Box::pin(err(
ResponseError::from(Error::MissingAuthorizationHeader).into()
))
}
},
None => {
return Box::pin(err(ResponseError::from(Error::MissingAuthorizationHeader).into()));
return Box::pin(err(
ResponseError::from(Error::MissingAuthorizationHeader).into()
));
}
};
@ -95,9 +99,10 @@ where
if authenticated {
Box::pin(svc.call(req))
} else {
Box::pin(err(
ResponseError::from(Error::InvalidToken(auth_header.to_string())).into()
Box::pin(err(ResponseError::from(Error::InvalidToken(
auth_header.to_string(),
))
.into()))
}
}
}

4
meilisearch-http/src/helpers/compression.rs
View File

@ -1,9 +1,9 @@
use flate2::Compression;
use flate2::read::GzDecoder;
use flate2::write::GzEncoder;
use flate2::Compression;
use std::fs::{create_dir_all, File};
use std::path::Path;
use tar::{Builder, Archive};
use tar::{Archive, Builder};
use crate::error::Error;

2
meilisearch-http/src/helpers/mod.rs
View File

@ -1,6 +1,4 @@
pub mod authentication;
pub mod normalize_path;
pub mod compression;
pub use authentication::Authentication;
pub use normalize_path::NormalizePath;

86
meilisearch-http/src/helpers/normalize_path.rs
View File

@ -1,86 +0,0 @@
/// From https://docs.rs/actix-web/3.0.0-alpha.2/src/actix_web/middleware/normalize.rs.html#34
use actix_http::Error;
use actix_service::{Service, Transform};
use actix_web::{
dev::ServiceRequest,
dev::ServiceResponse,
http::uri::{PathAndQuery, Uri},
};
use futures::future::{ok, Ready};
use regex::Regex;
use std::task::{Context, Poll};
pub struct NormalizePath;
impl<S, B> Transform<S> for NormalizePath
where
S: Service<Request = ServiceRequest, Response = ServiceResponse<B>, Error = Error>,
S::Future: 'static,
{
type Request = ServiceRequest;
type Response = ServiceResponse<B>;
type Error = Error;
type InitError = ();
type Transform = NormalizePathNormalization<S>;
type Future = Ready<Result<Self::Transform, Self::InitError>>;
fn new_transform(&self, service: S) -> Self::Future {
ok(NormalizePathNormalization {
service,
merge_slash: Regex::new("//+").unwrap(),
})
}
}
pub struct NormalizePathNormalization<S> {
service: S,
merge_slash: Regex,
}
impl<S, B> Service for NormalizePathNormalization<S>
where
S: Service<Request = ServiceRequest, Response = ServiceResponse<B>, Error = Error>,
S::Future: 'static,
{
type Request = ServiceRequest;
type Response = ServiceResponse<B>;
type Error = Error;
type Future = S::Future;
fn poll_ready(&mut self, cx: &mut Context<'_>) -> Poll<Result<(), Self::Error>> {
self.service.poll_ready(cx)
}
fn call(&mut self, mut req: ServiceRequest) -> Self::Future {
let head = req.head_mut();
// always add trailing slash, might be an extra one
let path = head.uri.path().to_string() + "/";
if self.merge_slash.find(&path).is_some() {
// normalize multiple /'s to one /
let path = self.merge_slash.replace_all(&path, "/");
let path = if path.len() > 1 {
path.trim_end_matches('/')
} else {
&path
};
let mut parts = head.uri.clone().into_parts();
let pq = parts.path_and_query.as_ref().unwrap();
let path = if let Some(q) = pq.query() {
bytes::Bytes::from(format!("{}?{}", path, q))
} else {
bytes::Bytes::copy_from_slice(path.as_bytes())
};
parts.path_and_query = Some(PathAndQuery::from_maybe_shared(path).unwrap());
let uri = Uri::from_parts(parts).unwrap();
req.match_info_mut().get_mut().update(&uri);
req.head_mut().uri = uri;
}
self.service.call(req)
}
}

138
meilisearch-http/src/index/mod.rs Normal file
View File

@ -0,0 +1,138 @@
mod search;
mod updates;
use std::collections::HashSet;
use std::ops::Deref;
use std::sync::Arc;
use anyhow::{bail, Context};
use milli::obkv_to_json;
use serde_json::{Map, Value};
pub use search::{SearchQuery, SearchResult, DEFAULT_SEARCH_LIMIT};
pub use updates::{Facets, Settings, UpdateResult};
pub type Document = Map<String, Value>;
#[derive(Clone)]
pub struct Index(pub Arc<milli::Index>);
impl Deref for Index {
type Target = milli::Index;
fn deref(&self) -> &Self::Target {
self.0.as_ref()
}
}
impl Index {
pub fn settings(&self) -> anyhow::Result<Settings> {
let txn = self.read_txn()?;
let displayed_attributes = self
.displayed_fields(&txn)?
.map(|fields| fields.into_iter().map(String::from).collect())
.unwrap_or_else(|| vec!["*".to_string()]);
let searchable_attributes = self
.searchable_fields(&txn)?
.map(|fields| fields.into_iter().map(String::from).collect())
.unwrap_or_else(|| vec!["*".to_string()]);
let faceted_attributes = self
.faceted_fields(&txn)?
.into_iter()
.map(|(k, v)| (k, v.to_string()))
.collect();
let criteria = self
.criteria(&txn)?
.into_iter()
.map(|c| c.to_string())
.collect();
Ok(Settings {
displayed_attributes: Some(Some(displayed_attributes)),
searchable_attributes: Some(Some(searchable_attributes)),
attributes_for_faceting: Some(Some(faceted_attributes)),
ranking_rules: Some(Some(criteria)),
})
}
pub fn retrieve_documents<S: AsRef<str>>(
&self,
offset: usize,
limit: usize,
attributes_to_retrieve: Option<Vec<S>>,
) -> anyhow::Result<Vec<Map<String, Value>>> {
let txn = self.read_txn()?;
let fields_ids_map = self.fields_ids_map(&txn)?;
let fields_to_display =
self.fields_to_display(&txn, attributes_to_retrieve, &fields_ids_map)?;
let iter = self.documents.range(&txn, &(..))?.skip(offset).take(limit);
let mut documents = Vec::new();
for entry in iter {
let (_id, obkv) = entry?;
let object = obkv_to_json(&fields_to_display, &fields_ids_map, obkv)?;
documents.push(object);
}
Ok(documents)
}
pub fn retrieve_document<S: AsRef<str>>(
&self,
doc_id: String,
attributes_to_retrieve: Option<Vec<S>>,
) -> anyhow::Result<Map<String, Value>> {
let txn = self.read_txn()?;
let fields_ids_map = self.fields_ids_map(&txn)?;
let fields_to_display =
self.fields_to_display(&txn, attributes_to_retrieve, &fields_ids_map)?;
let internal_id = self
.external_documents_ids(&txn)?
.get(doc_id.as_bytes())
.with_context(|| format!("Document with id {} not found", doc_id))?;
let document = self
.documents(&txn, std::iter::once(internal_id))?
.into_iter()
.next()
.map(|(_, d)| d);
match document {
Some(document) => Ok(obkv_to_json(&fields_to_display, &fields_ids_map, document)?),
None => bail!("Document with id {} not found", doc_id),
}
}
fn fields_to_display<S: AsRef<str>>(
&self,
txn: &heed::RoTxn,
attributes_to_retrieve: Option<Vec<S>>,
fields_ids_map: &milli::FieldsIdsMap,
) -> anyhow::Result<Vec<u8>> {
let mut displayed_fields_ids = match self.displayed_fields_ids(&txn)? {
Some(ids) => ids.into_iter().collect::<Vec<_>>(),
None => fields_ids_map.iter().map(|(id, _)| id).collect(),
};
let attributes_to_retrieve_ids = match attributes_to_retrieve {
Some(attrs) => attrs
.iter()
.filter_map(|f| fields_ids_map.id(f.as_ref()))
.collect::<HashSet<_>>(),
None => fields_ids_map.iter().map(|(id, _)| id).collect(),
};
displayed_fields_ids.retain(|fid| attributes_to_retrieve_ids.contains(fid));
Ok(displayed_fields_ids)
}
}

228
meilisearch-http/src/index/search.rs Normal file
View File

@ -0,0 +1,228 @@
use std::collections::{BTreeMap, HashSet};
use std::mem;
use std::time::Instant;
use anyhow::bail;
use either::Either;
use heed::RoTxn;
use meilisearch_tokenizer::{Analyzer, AnalyzerConfig};
use milli::{facet::FacetValue, FacetCondition, MatchingWords};
use serde::{Deserialize, Serialize};
use serde_json::{Map, Value};
use super::Index;
pub const DEFAULT_SEARCH_LIMIT: usize = 20;
const fn default_search_limit() -> usize {
DEFAULT_SEARCH_LIMIT
}
#[derive(Deserialize)]
#[serde(rename_all = "camelCase", deny_unknown_fields)]
#[allow(dead_code)]
pub struct SearchQuery {
pub q: Option<String>,
pub offset: Option<usize>,
#[serde(default = "default_search_limit")]
pub limit: usize,
pub attributes_to_retrieve: Option<Vec<String>>,
pub attributes_to_crop: Option<Vec<String>>,
pub crop_length: Option<usize>,
pub attributes_to_highlight: Option<HashSet<String>>,
pub filters: Option<String>,
pub matches: Option<bool>,
pub facet_filters: Option<Value>,
pub facet_distributions: Option<Vec<String>>,
}
#[derive(Serialize)]
#[serde(rename_all = "camelCase")]
pub struct SearchResult {
pub hits: Vec<Map<String, Value>>,
pub nb_hits: u64,
pub query: String,
pub limit: usize,
pub offset: usize,
pub processing_time_ms: u128,
#[serde(skip_serializing_if = "Option::is_none")]
pub facet_distributions: Option<BTreeMap<String, BTreeMap<FacetValue, u64>>>,
}
impl Index {
pub fn perform_search(&self, query: SearchQuery) -> anyhow::Result<SearchResult> {
let before_search = Instant::now();
let rtxn = self.read_txn()?;
let mut search = self.search(&rtxn);
if let Some(ref query) = query.q {
search.query(query);
}
search.limit(query.limit);
search.offset(query.offset.unwrap_or_default());
if let Some(ref facets) = query.facet_filters {
if let Some(facets) = parse_facets(facets, self, &rtxn)? {
search.facet_condition(facets);
}
}
let milli::SearchResult {
documents_ids,
matching_words,
candidates,
..
} = search.execute()?;
let mut documents = Vec::new();
let fields_ids_map = self.fields_ids_map(&rtxn).unwrap();
let fields_to_display =
self.fields_to_display(&rtxn, query.attributes_to_retrieve, &fields_ids_map)?;
let stop_words = fst::Set::default();
let highlighter = Highlighter::new(&stop_words);
for (_id, obkv) in self.documents(&rtxn, documents_ids)? {
let mut object =
milli::obkv_to_json(&fields_to_display, &fields_ids_map, obkv).unwrap();
if let Some(ref attributes_to_highlight) = query.attributes_to_highlight {
highlighter.highlight_record(&mut object, &matching_words, attributes_to_highlight);
}
documents.push(object);
}
let nb_hits = candidates.len();
let facet_distributions = match query.facet_distributions {
Some(ref fields) => {
let mut facet_distribution = self.facets_distribution(&rtxn);
if fields.iter().all(|f| f != "*") {
facet_distribution.facets(fields);
}
Some(facet_distribution.candidates(candidates).execute()?)
}
None => None,
};
let result = SearchResult {
hits: documents,
nb_hits,
query: query.q.clone().unwrap_or_default(),
limit: query.limit,
offset: query.offset.unwrap_or_default(),
processing_time_ms: before_search.elapsed().as_millis(),
facet_distributions,
};
Ok(result)
}
}
fn parse_facets_array(
txn: &RoTxn,
index: &Index,
arr: &[Value],
) -> anyhow::Result<Option<FacetCondition>> {
let mut ands = Vec::new();
for value in arr {
match value {
Value::String(s) => ands.push(Either::Right(s.clone())),
Value::Array(arr) => {
let mut ors = Vec::new();
for value in arr {
match value {
Value::String(s) => ors.push(s.clone()),
v => bail!("Invalid facet expression, expected String, found: {:?}", v),
}
}
ands.push(Either::Left(ors));
}
v => bail!(
"Invalid facet expression, expected String or [String], found: {:?}",
v
),
}
}
FacetCondition::from_array(txn, &index.0, ands)
}
pub struct Highlighter<'a, A> {
analyzer: Analyzer<'a, A>,
}
impl<'a, A: AsRef<[u8]>> Highlighter<'a, A> {
pub fn new(stop_words: &'a fst::Set<A>) -> Self {
let analyzer = Analyzer::new(AnalyzerConfig::default_with_stopwords(stop_words));
Self { analyzer }
}
pub fn highlight_value(&self, value: Value, words_to_highlight: &MatchingWords) -> Value {
match value {
Value::Null => Value::Null,
Value::Bool(boolean) => Value::Bool(boolean),
Value::Number(number) => Value::Number(number),
Value::String(old_string) => {
let mut string = String::new();
let analyzed = self.analyzer.analyze(&old_string);
for (word, token) in analyzed.reconstruct() {
if token.is_word() {
let to_highlight = words_to_highlight.matches(token.text());
if to_highlight {
string.push_str("<mark>")
}
string.push_str(word);
if to_highlight {
string.push_str("</mark>")
}
} else {
string.push_str(word);
}
}
Value::String(string)
}
Value::Array(values) => Value::Array(
values
.into_iter()
.map(|v| self.highlight_value(v, words_to_highlight))
.collect(),
),
Value::Object(object) => Value::Object(
object
.into_iter()
.map(|(k, v)| (k, self.highlight_value(v, words_to_highlight)))
.collect(),
),
}
}
pub fn highlight_record(
&self,
object: &mut Map<String, Value>,
words_to_highlight: &MatchingWords,
attributes_to_highlight: &HashSet<String>,
) {
// TODO do we need to create a string for element that are not and needs to be highlight?
for (key, value) in object.iter_mut() {
if attributes_to_highlight.contains(key) {
let old_value = mem::take(value);
*value = self.highlight_value(old_value, words_to_highlight);
}
}
}
}
fn parse_facets(
facets: &Value,
index: &Index,
txn: &RoTxn,
) -> anyhow::Result<Option<FacetCondition>> {
match facets {
// Disabled for now
//Value::String(expr) => Ok(Some(FacetCondition::from_str(txn, index, expr)?)),
Value::Array(arr) => parse_facets_array(txn, index, arr),
v => bail!("Invalid facet expression, expected Array, found: {:?}", v),
}
}

227
meilisearch-http/src/index/updates.rs Normal file
View File

@ -0,0 +1,227 @@
use std::collections::HashMap;
use std::io;
use std::num::NonZeroUsize;
use flate2::read::GzDecoder;
use log::info;
use milli::update::{DocumentAdditionResult, IndexDocumentsMethod, UpdateBuilder, UpdateFormat};
use serde::{de::Deserializer, Deserialize, Serialize};
use super::Index;
#[derive(Debug, Clone, Serialize, Deserialize)]
pub enum UpdateResult {
DocumentsAddition(DocumentAdditionResult),
DocumentDeletion { deleted: u64 },
Other,
}
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
#[serde(rename_all = "camelCase")]
pub struct Settings {
#[serde(
default,
deserialize_with = "deserialize_some",
skip_serializing_if = "Option::is_none"
)]
pub displayed_attributes: Option<Option<Vec<String>>>,
#[serde(
default,
deserialize_with = "deserialize_some",
skip_serializing_if = "Option::is_none"
)]
pub searchable_attributes: Option<Option<Vec<String>>>,
#[serde(default)]
pub attributes_for_faceting: Option<Option<HashMap<String, String>>>,
#[serde(
default,
deserialize_with = "deserialize_some",
skip_serializing_if = "Option::is_none"
)]
pub ranking_rules: Option<Option<Vec<String>>>,
}
impl Settings {
pub fn cleared() -> Self {
Self {
displayed_attributes: Some(None),
searchable_attributes: Some(None),
attributes_for_faceting: Some(None),
ranking_rules: Some(None),
}
}
}
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
#[serde(rename_all = "camelCase")]
pub struct Facets {
pub level_group_size: Option<NonZeroUsize>,
pub min_level_size: Option<NonZeroUsize>,
}
fn deserialize_some<'de, T, D>(deserializer: D) -> Result<Option<T>, D::Error>
where
T: Deserialize<'de>,
D: Deserializer<'de>,
{
Deserialize::deserialize(deserializer).map(Some)
}
impl Index {
pub fn update_documents(
&self,
format: UpdateFormat,
method: IndexDocumentsMethod,
content: impl io::Read,
update_builder: UpdateBuilder,
primary_key: Option<&str>,
) -> anyhow::Result<UpdateResult> {
info!("performing document addition");
// We must use the write transaction of the update here.
let mut wtxn = self.write_txn()?;
// Set the primary key if not set already, ignore if already set.
if let (None, Some(ref primary_key)) = (self.primary_key(&wtxn)?, primary_key) {
self.put_primary_key(&mut wtxn, primary_key)?;
}
let mut builder = update_builder.index_documents(&mut wtxn, self);
builder.update_format(format);
builder.index_documents_method(method);
let gzipped = false;
let reader = if gzipped {
Box::new(GzDecoder::new(content))
} else {
Box::new(content) as Box<dyn io::Read>
};
let result = builder.execute(reader, |indexing_step, update_id| {
info!("update {}: {:?}", update_id, indexing_step)
});
info!("document addition done: {:?}", result);
result.and_then(|addition_result| {
wtxn.commit()
.and(Ok(UpdateResult::DocumentsAddition(addition_result)))
.map_err(Into::into)
})
}
pub fn clear_documents(&self, update_builder: UpdateBuilder) -> anyhow::Result<UpdateResult> {
// We must use the write transaction of the update here.
let mut wtxn = self.write_txn()?;
let builder = update_builder.clear_documents(&mut wtxn, self);
match builder.execute() {
Ok(_count) => wtxn
.commit()
.and(Ok(UpdateResult::Other))
.map_err(Into::into),
Err(e) => Err(e),
}
}
pub fn update_settings(
&self,
settings: &Settings,
update_builder: UpdateBuilder,
) -> anyhow::Result<UpdateResult> {
// We must use the write transaction of the update here.
let mut wtxn = self.write_txn()?;
let mut builder = update_builder.settings(&mut wtxn, self);
// We transpose the settings JSON struct into a real setting update.
if let Some(ref names) = settings.searchable_attributes {
match names {
Some(names) => builder.set_searchable_fields(names.clone()),
None => builder.reset_searchable_fields(),
}
}
// We transpose the settings JSON struct into a real setting update.
if let Some(ref names) = settings.displayed_attributes {
match names {
Some(names) => builder.set_displayed_fields(names.clone()),
None => builder.reset_displayed_fields(),
}
}
// We transpose the settings JSON struct into a real setting update.
if let Some(ref facet_types) = settings.attributes_for_faceting {
let facet_types = facet_types.clone().unwrap_or_else(HashMap::new);
builder.set_faceted_fields(facet_types);
}
// We transpose the settings JSON struct into a real setting update.
if let Some(ref criteria) = settings.ranking_rules {
match criteria {
Some(criteria) => builder.set_criteria(criteria.clone()),
None => builder.reset_criteria(),
}
}
let result = builder
.execute(|indexing_step, update_id| info!("update {}: {:?}", update_id, indexing_step));
match result {
Ok(()) => wtxn
.commit()
.and(Ok(UpdateResult::Other))
.map_err(Into::into),
Err(e) => Err(e),
}
}
pub fn update_facets(
&self,
levels: &Facets,
update_builder: UpdateBuilder,
) -> anyhow::Result<UpdateResult> {
// We must use the write transaction of the update here.
let mut wtxn = self.write_txn()?;
let mut builder = update_builder.facets(&mut wtxn, self);
if let Some(value) = levels.level_group_size {
builder.level_group_size(value);
}
if let Some(value) = levels.min_level_size {
builder.min_level_size(value);
}
match builder.execute() {
Ok(()) => wtxn
.commit()
.and(Ok(UpdateResult::Other))
.map_err(Into::into),
Err(e) => Err(e),
}
}
pub fn delete_documents(
&self,
document_ids: impl io::Read,
update_builder: UpdateBuilder,
) -> anyhow::Result<UpdateResult> {
let ids: Vec<String> = serde_json::from_reader(document_ids)?;
let mut txn = self.write_txn()?;
let mut builder = update_builder.delete_documents(&mut txn, self)?;
// We ignore unexisting document ids
ids.iter().for_each(|id| {
builder.delete_external_id(id);
});
match builder.execute() {
Ok(deleted) => txn
.commit()
.and(Ok(UpdateResult::DocumentDeletion { deleted }))
.map_err(Into::into),
Err(e) => Err(e),
}
}
}

612
meilisearch-http/src/index_controller/index_actor.rs Normal file
View File

@ -0,0 +1,612 @@
use std::collections::HashMap;
use std::fs::{create_dir_all, File};
use std::future::Future;
use std::path::{Path, PathBuf};
use std::sync::Arc;
use async_stream::stream;
use chrono::{DateTime, Utc};
use futures::pin_mut;
use futures::stream::StreamExt;
use heed::EnvOpenOptions;
use log::debug;
use serde::{Deserialize, Serialize};
use thiserror::Error;
use tokio::fs::remove_dir_all;
use tokio::sync::{mpsc, oneshot, RwLock};
use tokio::task::spawn_blocking;
use uuid::Uuid;
use super::update_handler::UpdateHandler;
use super::{get_arc_ownership_blocking, IndexSettings};
use crate::index::UpdateResult as UResult;
use crate::index::{Document, Index, SearchQuery, SearchResult, Settings};
use crate::index_controller::{
updates::{Failed, Processed, Processing},
UpdateMeta,
};
use crate::option::IndexerOpts;
pub type Result<T> = std::result::Result<T, IndexError>;
type AsyncMap<K, V> = Arc<RwLock<HashMap<K, V>>>;
type UpdateResult = std::result::Result<Processed<UpdateMeta, UResult>, Failed<UpdateMeta, String>>;
#[derive(Debug, Serialize, Deserialize, Clone)]
#[serde(rename_all = "camelCase")]
pub struct IndexMeta {
created_at: DateTime<Utc>,
updated_at: DateTime<Utc>,
primary_key: Option<String>,
}
impl IndexMeta {
fn new(index: &Index) -> Result<Self> {
let txn = index.read_txn()?;
Self::new_txn(index, &txn)
}
fn new_txn(index: &Index, txn: &heed::RoTxn) -> Result<Self> {
let created_at = index.created_at(&txn)?;
let updated_at = index.updated_at(&txn)?;
let primary_key = index.primary_key(&txn)?.map(String::from);
Ok(Self {
primary_key,
updated_at,
created_at,
})
}
}
enum IndexMsg {
CreateIndex {
uuid: Uuid,
primary_key: Option<String>,
ret: oneshot::Sender<Result<IndexMeta>>,
},
Update {
meta: Processing<UpdateMeta>,
data: std::fs::File,
ret: oneshot::Sender<Result<UpdateResult>>,
},
Search {
uuid: Uuid,
query: SearchQuery,
ret: oneshot::Sender<anyhow::Result<SearchResult>>,
},
Settings {
uuid: Uuid,
ret: oneshot::Sender<Result<Settings>>,
},
Documents {
uuid: Uuid,
attributes_to_retrieve: Option<Vec<String>>,
offset: usize,
limit: usize,
ret: oneshot::Sender<Result<Vec<Document>>>,
},
Document {
uuid: Uuid,
attributes_to_retrieve: Option<Vec<String>>,
doc_id: String,
ret: oneshot::Sender<Result<Document>>,
},
Delete {
uuid: Uuid,
ret: oneshot::Sender<Result<()>>,
},
GetMeta {
uuid: Uuid,
ret: oneshot::Sender<Result<IndexMeta>>,
},
UpdateIndex {
uuid: Uuid,
index_settings: IndexSettings,
ret: oneshot::Sender<Result<IndexMeta>>,
},
}
struct IndexActor<S> {
read_receiver: Option<mpsc::Receiver<IndexMsg>>,
write_receiver: Option<mpsc::Receiver<IndexMsg>>,
update_handler: Arc<UpdateHandler>,
store: S,
}
#[derive(Error, Debug)]
pub enum IndexError {
#[error("error with index: {0}")]
Error(#[from] anyhow::Error),
#[error("index already exists")]
IndexAlreadyExists,
#[error("Index doesn't exists")]
UnexistingIndex,
#[error("Heed error: {0}")]
HeedError(#[from] heed::Error),
#[error("Existing primary key")]
ExistingPrimaryKey,
}
#[async_trait::async_trait]
trait IndexStore {
async fn create(&self, uuid: Uuid, primary_key: Option<String>) -> Result<Index>;
async fn get(&self, uuid: Uuid) -> Result<Option<Index>>;
async fn delete(&self, uuid: Uuid) -> Result<Option<Index>>;
}
impl<S: IndexStore + Sync + Send> IndexActor<S> {
fn new(
read_receiver: mpsc::Receiver<IndexMsg>,
write_receiver: mpsc::Receiver<IndexMsg>,
store: S,
) -> Result<Self> {
let options = IndexerOpts::default();
let update_handler = UpdateHandler::new(&options).map_err(IndexError::Error)?;
let update_handler = Arc::new(update_handler);
let read_receiver = Some(read_receiver);
let write_receiver = Some(write_receiver);
Ok(Self {
read_receiver,
write_receiver,
store,
update_handler,
})
}
/// `run` poll the write_receiver and read_receiver concurrently, but while messages send
/// through the read channel are processed concurrently, the messages sent through the write
/// channel are processed one at a time.
async fn run(mut self) {
let mut read_receiver = self
.read_receiver
.take()
.expect("Index Actor must have a inbox at this point.");
let read_stream = stream! {
loop {
match read_receiver.recv().await {
Some(msg) => yield msg,
None => break,
}
}
};
let mut write_receiver = self
.write_receiver
.take()
.expect("Index Actor must have a inbox at this point.");
let write_stream = stream! {
loop {
match write_receiver.recv().await {
Some(msg) => yield msg,
None => break,
}
}
};
pin_mut!(write_stream);
pin_mut!(read_stream);
let fut1 = read_stream.for_each_concurrent(Some(10), |msg| self.handle_message(msg));
let fut2 = write_stream.for_each_concurrent(Some(1), |msg| self.handle_message(msg));
let fut1: Box<dyn Future<Output = ()> + Unpin + Send> = Box::new(fut1);
let fut2: Box<dyn Future<Output = ()> + Unpin + Send> = Box::new(fut2);
tokio::join!(fut1, fut2);
}
async fn handle_message(&self, msg: IndexMsg) {
use IndexMsg::*;
match msg {
CreateIndex {
uuid,
primary_key,
ret,
} => {
let _ = ret.send(self.handle_create_index(uuid, primary_key).await);
}
Update { ret, meta, data } => {
let _ = ret.send(self.handle_update(meta, data).await);
}
Search { ret, query, uuid } => {
let _ = ret.send(self.handle_search(uuid, query).await);
}
Settings { ret, uuid } => {
let _ = ret.send(self.handle_settings(uuid).await);
}
Documents {
ret,
uuid,
attributes_to_retrieve,
offset,
limit,
} => {
let _ = ret.send(
self.handle_fetch_documents(uuid, offset, limit, attributes_to_retrieve)
.await,
);
}
Document {
uuid,
attributes_to_retrieve,
doc_id,
ret,
} => {
let _ = ret.send(
self.handle_fetch_document(uuid, doc_id, attributes_to_retrieve)
.await,
);
}
Delete { uuid, ret } => {
let _ = ret.send(self.handle_delete(uuid).await);
}
GetMeta { uuid, ret } => {
let _ = ret.send(self.handle_get_meta(uuid).await);
}
UpdateIndex {
uuid,
index_settings,
ret,
} => {
let _ = ret.send(self.handle_update_index(uuid, index_settings).await);
}
}
}
async fn handle_search(&self, uuid: Uuid, query: SearchQuery) -> anyhow::Result<SearchResult> {
let index = self
.store
.get(uuid)
.await?
.ok_or(IndexError::UnexistingIndex)?;
spawn_blocking(move || index.perform_search(query)).await?
}
async fn handle_create_index(
&self,
uuid: Uuid,
primary_key: Option<String>,
) -> Result<IndexMeta> {
let index = self.store.create(uuid, primary_key).await?;
let meta = spawn_blocking(move || IndexMeta::new(&index))
.await
.map_err(|e| IndexError::Error(e.into()))??;
Ok(meta)
}
async fn handle_update(
&self,
meta: Processing<UpdateMeta>,
data: File,
) -> Result<UpdateResult> {
debug!("Processing update {}", meta.id());
let uuid = meta.index_uuid();
let update_handler = self.update_handler.clone();
let index = match self.store.get(*uuid).await? {
Some(index) => index,
None => self.store.create(*uuid, None).await?,
};
spawn_blocking(move || update_handler.handle_update(meta, data, index))
.await
.map_err(|e| IndexError::Error(e.into()))
}
async fn handle_settings(&self, uuid: Uuid) -> Result<Settings> {
let index = self
.store
.get(uuid)
.await?
.ok_or(IndexError::UnexistingIndex)?;
spawn_blocking(move || index.settings().map_err(IndexError::Error))
.await
.map_err(|e| IndexError::Error(e.into()))?
}
async fn handle_fetch_documents(
&self,
uuid: Uuid,
offset: usize,
limit: usize,
attributes_to_retrieve: Option<Vec<String>>,
) -> Result<Vec<Document>> {
let index = self
.store
.get(uuid)
.await?
.ok_or(IndexError::UnexistingIndex)?;
spawn_blocking(move || {
index
.retrieve_documents(offset, limit, attributes_to_retrieve)
.map_err(IndexError::Error)
})
.await
.map_err(|e| IndexError::Error(e.into()))?
}
async fn handle_fetch_document(
&self,
uuid: Uuid,
doc_id: String,
attributes_to_retrieve: Option<Vec<String>>,
) -> Result<Document> {
let index = self
.store
.get(uuid)
.await?
.ok_or(IndexError::UnexistingIndex)?;
spawn_blocking(move || {
index
.retrieve_document(doc_id, attributes_to_retrieve)
.map_err(IndexError::Error)
})
.await
.map_err(|e| IndexError::Error(e.into()))?
}
async fn handle_delete(&self, uuid: Uuid) -> Result<()> {
let index = self.store.delete(uuid).await?;
if let Some(index) = index {
tokio::task::spawn(async move {
let index = index.0;
let store = get_arc_ownership_blocking(index).await;
spawn_blocking(move || {
store.prepare_for_closing().wait();
debug!("Index closed");
});
});
}
Ok(())
}
async fn handle_get_meta(&self, uuid: Uuid) -> Result<IndexMeta> {
match self.store.get(uuid).await? {
Some(index) => {
let meta = spawn_blocking(move || IndexMeta::new(&index))
.await
.map_err(|e| IndexError::Error(e.into()))??;
Ok(meta)
}
None => Err(IndexError::UnexistingIndex),
}
}
async fn handle_update_index(
&self,
uuid: Uuid,
index_settings: IndexSettings,
) -> Result<IndexMeta> {
let index = self
.store
.get(uuid)
.await?
.ok_or(IndexError::UnexistingIndex)?;
spawn_blocking(move || match index_settings.primary_key {
Some(ref primary_key) => {
let mut txn = index.write_txn()?;
if index.primary_key(&txn)?.is_some() {
return Err(IndexError::ExistingPrimaryKey);
}
index.put_primary_key(&mut txn, primary_key)?;
let meta = IndexMeta::new_txn(&index, &txn)?;
txn.commit()?;
Ok(meta)
}
None => {
let meta = IndexMeta::new(&index)?;
Ok(meta)
}
})
.await
.map_err(|e| IndexError::Error(e.into()))?
}
}
#[derive(Clone)]
pub struct IndexActorHandle {
read_sender: mpsc::Sender<IndexMsg>,
write_sender: mpsc::Sender<IndexMsg>,
}
impl IndexActorHandle {
pub fn new(path: impl AsRef<Path>, index_size: usize) -> anyhow::Result<Self> {
let (read_sender, read_receiver) = mpsc::channel(100);
let (write_sender, write_receiver) = mpsc::channel(100);
let store = HeedIndexStore::new(path, index_size);
let actor = IndexActor::new(read_receiver, write_receiver, store)?;
tokio::task::spawn(actor.run());
Ok(Self {
read_sender,
write_sender,
})
}
pub async fn create_index(&self, uuid: Uuid, primary_key: Option<String>) -> Result<IndexMeta> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::CreateIndex {
ret,
uuid,
primary_key,
};
let _ = self.read_sender.send(msg).await;
receiver.await.expect("IndexActor has been killed")
}
pub async fn update(
&self,
meta: Processing<UpdateMeta>,
data: std::fs::File,
) -> anyhow::Result<UpdateResult> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::Update { ret, meta, data };
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
pub async fn search(&self, uuid: Uuid, query: SearchQuery) -> Result<SearchResult> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::Search { uuid, query, ret };
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
pub async fn settings(&self, uuid: Uuid) -> Result<Settings> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::Settings { uuid, ret };
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
pub async fn documents(
&self,
uuid: Uuid,
offset: usize,
limit: usize,
attributes_to_retrieve: Option<Vec<String>>,
) -> Result<Vec<Document>> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::Documents {
uuid,
ret,
offset,
attributes_to_retrieve,
limit,
};
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
pub async fn document(
&self,
uuid: Uuid,
doc_id: String,
attributes_to_retrieve: Option<Vec<String>>,
) -> Result<Document> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::Document {
uuid,
ret,
doc_id,
attributes_to_retrieve,
};
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
pub async fn delete(&self, uuid: Uuid) -> Result<()> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::Delete { uuid, ret };
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
pub async fn get_index_meta(&self, uuid: Uuid) -> Result<IndexMeta> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::GetMeta { uuid, ret };
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
pub async fn update_index(
&self,
uuid: Uuid,
index_settings: IndexSettings,
) -> Result<IndexMeta> {
let (ret, receiver) = oneshot::channel();
let msg = IndexMsg::UpdateIndex {
uuid,
index_settings,
ret,
};
let _ = self.read_sender.send(msg).await;
Ok(receiver.await.expect("IndexActor has been killed")?)
}
}
struct HeedIndexStore {
index_store: AsyncMap<Uuid, Index>,
path: PathBuf,
index_size: usize,
}
impl HeedIndexStore {
fn new(path: impl AsRef<Path>, index_size: usize) -> Self {
let path = path.as_ref().join("indexes/");
let index_store = Arc::new(RwLock::new(HashMap::new()));
Self {
index_store,
path,
index_size,
}
}
}
#[async_trait::async_trait]
impl IndexStore for HeedIndexStore {
async fn create(&self, uuid: Uuid, primary_key: Option<String>) -> Result<Index> {
let path = self.path.join(format!("index-{}", uuid));
if path.exists() {
return Err(IndexError::IndexAlreadyExists);
}
let index_size = self.index_size;
let index = spawn_blocking(move || -> Result<Index> {
let index = open_index(&path, index_size)?;
if let Some(primary_key) = primary_key {
let mut txn = index.write_txn()?;
index.put_primary_key(&mut txn, &primary_key)?;
txn.commit()?;
}
Ok(index)
})
.await
.map_err(|e| IndexError::Error(e.into()))??;
self.index_store.write().await.insert(uuid, index.clone());
Ok(index)
}
async fn get(&self, uuid: Uuid) -> Result<Option<Index>> {
let guard = self.index_store.read().await;
match guard.get(&uuid) {
Some(index) => Ok(Some(index.clone())),
None => {
// drop the guard here so we can perform the write after without deadlocking;
drop(guard);
let path = self.path.join(format!("index-{}", uuid));
if !path.exists() {
return Ok(None);
}
let index_size = self.index_size;
let index = spawn_blocking(move || open_index(path, index_size))
.await
.map_err(|e| IndexError::Error(e.into()))??;
self.index_store.write().await.insert(uuid, index.clone());
Ok(Some(index))
}
}
}
async fn delete(&self, uuid: Uuid) -> Result<Option<Index>> {
let db_path = self.path.join(format!("index-{}", uuid));
remove_dir_all(db_path)
.await
.map_err(|e| IndexError::Error(e.into()))?;
let index = self.index_store.write().await.remove(&uuid);
Ok(index)
}
}
fn open_index(path: impl AsRef<Path>, size: usize) -> Result<Index> {
create_dir_all(&path).map_err(|e| IndexError::Error(e.into()))?;
let mut options = EnvOpenOptions::new();
options.map_size(size);
let index = milli::Index::new(options, &path).map_err(IndexError::Error)?;
Ok(Index(Arc::new(index)))
}

607
meilisearch-http/src/index_controller/local_index_controller/index_store.rs
View File

@ -1,607 +0,0 @@
use std::fs::{create_dir_all, remove_dir_all};
use std::path::{Path, PathBuf};
use std::sync::Arc;
use std::time::Duration;
use anyhow::{bail, Context};
use chrono::{DateTime, Utc};
use dashmap::{mapref::entry::Entry, DashMap};
use heed::{
types::{ByteSlice, SerdeJson, Str},
Database, Env, EnvOpenOptions, RoTxn, RwTxn,
};
use log::{error, info};
use milli::Index;
use rayon::ThreadPool;
use serde::{Deserialize, Serialize};
use uuid::Uuid;
use super::update_handler::UpdateHandler;
use super::{UpdateMeta, UpdateResult};
use crate::option::IndexerOpts;
type UpdateStore = super::update_store::UpdateStore<UpdateMeta, UpdateResult, String>;
#[derive(Serialize, Deserialize, Debug, PartialEq)]
pub struct IndexMeta {
update_store_size: u64,
index_store_size: u64,
pub uuid: Uuid,
pub created_at: DateTime<Utc>,
pub updated_at: DateTime<Utc>,
}
impl IndexMeta {
fn open(
&self,
path: impl AsRef<Path>,
thread_pool: Arc<ThreadPool>,
indexer_options: &IndexerOpts,
) -> anyhow::Result<(Arc<Index>, Arc<UpdateStore>)> {
let update_path = make_update_db_path(&path, &self.uuid);
let index_path = make_index_db_path(&path, &self.uuid);
create_dir_all(&update_path)?;
create_dir_all(&index_path)?;
let mut options = EnvOpenOptions::new();
options.map_size(self.index_store_size as usize);
let index = Arc::new(Index::new(options, index_path)?);
let mut options = EnvOpenOptions::new();
options.map_size(self.update_store_size as usize);
let handler = UpdateHandler::new(indexer_options, index.clone(), thread_pool)?;
let update_store = UpdateStore::open(options, update_path, handler)?;
Ok((index, update_store))
}
}
pub struct IndexStore {
env: Env,
name_to_uuid: Database<Str, ByteSlice>,
uuid_to_index: DashMap<Uuid, (Arc<Index>, Arc<UpdateStore>)>,
uuid_to_index_meta: Database<ByteSlice, SerdeJson<IndexMeta>>,
thread_pool: Arc<ThreadPool>,
indexer_options: IndexerOpts,
}
impl IndexStore {
pub fn new(path: impl AsRef<Path>, indexer_options: IndexerOpts) -> anyhow::Result<Self> {
let env = EnvOpenOptions::new()
.map_size(4096 * 100)
.max_dbs(2)
.open(path)?;
let uuid_to_index = DashMap::new();
let name_to_uuid = open_or_create_database(&env, Some("name_to_uid"))?;
let uuid_to_index_meta = open_or_create_database(&env, Some("uid_to_index_db"))?;
let thread_pool = rayon::ThreadPoolBuilder::new()
.num_threads(indexer_options.indexing_jobs.unwrap_or(0))
.build()?;
let thread_pool = Arc::new(thread_pool);
Ok(Self {
env,
name_to_uuid,
uuid_to_index,
uuid_to_index_meta,
thread_pool,
indexer_options,
})
}
pub fn delete(&self, index_uid: impl AsRef<str>) -> anyhow::Result<()> {
// we remove the references to the index from the index map so it is not accessible anymore
let mut txn = self.env.write_txn()?;
let uuid = self
.index_uuid(&txn, &index_uid)?
.with_context(|| format!("Index {:?} doesn't exist", index_uid.as_ref()))?;
self.name_to_uuid.delete(&mut txn, index_uid.as_ref())?;
self.uuid_to_index_meta.delete(&mut txn, uuid.as_bytes())?;
txn.commit()?;
// If the index was loaded (i.e it is present in the uuid_to_index map), then we need to
// close it. The process goes as follow:
//
// 1) We want to remove any pending updates from the store.
// 2) We try to get ownership on the update store so we can close it. It may take a
// couple of tries, but since the update store event loop only has a weak reference to
// itself, and we are the only other function holding a reference to it otherwise, we will
// get it eventually.
// 3) We request a closing of the update store.
// 4) We can take ownership on the index, and close it.
// 5) We remove all the files from the file system.
let index_uid = index_uid.as_ref().to_string();
let path = self.env.path().to_owned();
if let Some((_, (index, updates))) = self.uuid_to_index.remove(&uuid) {
std::thread::spawn(move || {
info!("Preparing for {:?} deletion.", index_uid);
// this error is non fatal, but may delay the deletion.
if let Err(e) = updates.abort_pendings() {
error!(
"error aborting pending updates when deleting index {:?}: {}",
index_uid, e
);
}
let updates = get_arc_ownership_blocking(updates);
let close_event = updates.prepare_for_closing();
close_event.wait();
info!("closed update store for {:?}", index_uid);
let index = get_arc_ownership_blocking(index);
let close_event = index.prepare_for_closing();
close_event.wait();
let update_path = make_update_db_path(&path, &uuid);
let index_path = make_index_db_path(&path, &uuid);
if let Err(e) = remove_dir_all(index_path) {
error!("error removing index {:?}: {}", index_uid, e);
}
if let Err(e) = remove_dir_all(update_path) {
error!("error removing index {:?}: {}", index_uid, e);
}
info!("index {:?} deleted.", index_uid);
});
}
Ok(())
}
fn index_uuid(&self, txn: &RoTxn, name: impl AsRef<str>) -> anyhow::Result<Option<Uuid>> {
match self.name_to_uuid.get(txn, name.as_ref())? {
Some(bytes) => {
let uuid = Uuid::from_slice(bytes)?;
Ok(Some(uuid))
}
None => Ok(None),
}
}
fn retrieve_index(
&self,
txn: &RoTxn,
uid: Uuid,
) -> anyhow::Result<Option<(Arc<Index>, Arc<UpdateStore>)>> {
match self.uuid_to_index.entry(uid.clone()) {
Entry::Vacant(entry) => match self.uuid_to_index_meta.get(txn, uid.as_bytes())? {
Some(meta) => {
let path = self.env.path();
let (index, updates) =
meta.open(path, self.thread_pool.clone(), &self.indexer_options)?;
entry.insert((index.clone(), updates.clone()));
Ok(Some((index, updates)))
}
None => Ok(None),
},
Entry::Occupied(entry) => {
let (index, updates) = entry.get();
Ok(Some((index.clone(), updates.clone())))
}
}
}
fn get_index_txn(
&self,
txn: &RoTxn,
name: impl AsRef<str>,
) -> anyhow::Result<Option<(Arc<Index>, Arc<UpdateStore>)>> {
match self.index_uuid(&txn, name)? {
Some(uid) => self.retrieve_index(&txn, uid),
None => Ok(None),
}
}
pub fn index(
&self,
name: impl AsRef<str>,
) -> anyhow::Result<Option<(Arc<Index>, Arc<UpdateStore>)>> {
let txn = self.env.read_txn()?;
self.get_index_txn(&txn, name)
}
/// Use this function to perform an update on an index.
/// This function also puts a lock on what index is allowed to perform an update.
pub fn update_index<F, T>(&self, name: impl AsRef<str>, f: F) -> anyhow::Result<(T, IndexMeta)>
where
F: FnOnce(&Index) -> anyhow::Result<T>,
{
let mut txn = self.env.write_txn()?;
let (index, _) = self
.get_index_txn(&txn, &name)?
.with_context(|| format!("Index {:?} doesn't exist", name.as_ref()))?;
let result = f(index.as_ref());
match result {
Ok(ret) => {
let meta = self.update_meta(&mut txn, name, |meta| meta.updated_at = Utc::now())?;
txn.commit()?;
Ok((ret, meta))
}
Err(e) => Err(e),
}
}
pub fn index_with_meta(
&self,
name: impl AsRef<str>,
) -> anyhow::Result<Option<(Arc<Index>, IndexMeta)>> {
let txn = self.env.read_txn()?;
let uuid = self.index_uuid(&txn, &name)?;
match uuid {
Some(uuid) => {
let meta = self
.uuid_to_index_meta
.get(&txn, uuid.as_bytes())?
.with_context(|| {
format!("unable to retrieve metadata for index {:?}", name.as_ref())
})?;
let (index, _) = self
.retrieve_index(&txn, uuid)?
.with_context(|| format!("unable to retrieve index {:?}", name.as_ref()))?;
Ok(Some((index, meta)))
}
None => Ok(None),
}
}
fn update_meta<F>(
&self,
txn: &mut RwTxn,
name: impl AsRef<str>,
f: F,
) -> anyhow::Result<IndexMeta>
where
F: FnOnce(&mut IndexMeta),
{
let uuid = self
.index_uuid(txn, &name)?
.with_context(|| format!("Index {:?} doesn't exist", name.as_ref()))?;
let mut meta = self
.uuid_to_index_meta
.get(txn, uuid.as_bytes())?
.with_context(|| format!("couldn't retrieve metadata for index {:?}", name.as_ref()))?;
f(&mut meta);
self.uuid_to_index_meta.put(txn, uuid.as_bytes(), &meta)?;
Ok(meta)
}
pub fn get_or_create_index(
&self,
name: impl AsRef<str>,
update_size: u64,
index_size: u64,
) -> anyhow::Result<(Arc<Index>, Arc<UpdateStore>)> {
let mut txn = self.env.write_txn()?;
match self.get_index_txn(&txn, name.as_ref())? {
Some(res) => Ok(res),
None => {
let uuid = Uuid::new_v4();
let (index, updates, _) =
self.create_index_txn(&mut txn, uuid, name, update_size, index_size)?;
// If we fail to commit the transaction, we must delete the database from the
// file-system.
if let Err(e) = txn.commit() {
self.clean_db(uuid);
return Err(e)?;
}
Ok((index, updates))
}
}
}
// Remove all the files and data associated with a db uuid.
fn clean_db(&self, uuid: Uuid) {
let update_db_path = make_update_db_path(self.env.path(), &uuid);
let index_db_path = make_index_db_path(self.env.path(), &uuid);
remove_dir_all(update_db_path).expect("Failed to clean database");
remove_dir_all(index_db_path).expect("Failed to clean database");
self.uuid_to_index.remove(&uuid);
}
fn create_index_txn(
&self,
txn: &mut RwTxn,
uuid: Uuid,
name: impl AsRef<str>,
update_store_size: u64,
index_store_size: u64,
) -> anyhow::Result<(Arc<Index>, Arc<UpdateStore>, IndexMeta)> {
let created_at = Utc::now();
let updated_at = created_at;
let meta = IndexMeta {
update_store_size,
index_store_size,
uuid: uuid.clone(),
created_at,
updated_at,
};
self.name_to_uuid.put(txn, name.as_ref(), uuid.as_bytes())?;
self.uuid_to_index_meta.put(txn, uuid.as_bytes(), &meta)?;
let path = self.env.path();
let (index, update_store) =
match meta.open(path, self.thread_pool.clone(), &self.indexer_options) {
Ok(res) => res,
Err(e) => {
self.clean_db(uuid);
return Err(e);
}
};
self.uuid_to_index
.insert(uuid, (index.clone(), update_store.clone()));
Ok((index, update_store, meta))
}
/// Same as `get_or_create`, but returns an error if the index already exists.
pub fn create_index(
&self,
name: impl AsRef<str>,
update_size: u64,
index_size: u64,
) -> anyhow::Result<(Arc<Index>, Arc<UpdateStore>, IndexMeta)> {
let uuid = Uuid::new_v4();
let mut txn = self.env.write_txn()?;
if self.name_to_uuid.get(&txn, name.as_ref())?.is_some() {
bail!("index {:?} already exists", name.as_ref())
}
let result = self.create_index_txn(&mut txn, uuid, name, update_size, index_size)?;
// If we fail to commit the transaction, we must delete the database from the
// file-system.
if let Err(e) = txn.commit() {
self.clean_db(uuid);
return Err(e)?;
}
Ok(result)
}
/// Returns each index associated with its metadata:
/// (index_name, IndexMeta, primary_key)
/// This method will force all the indexes to be loaded.
pub fn list_indexes(&self) -> anyhow::Result<Vec<(String, IndexMeta, Option<String>)>> {
let txn = self.env.read_txn()?;
let metas = self.name_to_uuid.iter(&txn)?.filter_map(|entry| {
entry
.map_err(|e| {
error!("error decoding entry while listing indexes: {}", e);
e
})
.ok()
});
let mut indexes = Vec::new();
for (name, uuid) in metas {
// get index to retrieve primary key
let (index, _) = self
.get_index_txn(&txn, name)?
.with_context(|| format!("could not load index {:?}", name))?;
let primary_key = index.primary_key(&index.read_txn()?)?.map(String::from);
// retieve meta
let meta = self
.uuid_to_index_meta
.get(&txn, &uuid)?
.with_context(|| format!("could not retieve meta for index {:?}", name))?;
indexes.push((name.to_owned(), meta, primary_key));
}
Ok(indexes)
}
}
// Loops on an arc to get ownership on the wrapped value. This method sleeps 100ms before retrying.
fn get_arc_ownership_blocking<T>(mut item: Arc<T>) -> T {
loop {
match Arc::try_unwrap(item) {
Ok(item) => return item,
Err(item_arc) => {
item = item_arc;
std::thread::sleep(Duration::from_millis(100));
continue;
}
}
}
}
fn open_or_create_database<K: 'static, V: 'static>(
env: &Env,
name: Option<&str>,
) -> anyhow::Result<Database<K, V>> {
match env.open_database::<K, V>(name)? {
Some(db) => Ok(db),
None => Ok(env.create_database::<K, V>(name)?),
}
}
fn make_update_db_path(path: impl AsRef<Path>, uuid: &Uuid) -> PathBuf {
let mut path = path.as_ref().to_path_buf();
path.push(format!("update{}", uuid));
path
}
fn make_index_db_path(path: impl AsRef<Path>, uuid: &Uuid) -> PathBuf {
let mut path = path.as_ref().to_path_buf();
path.push(format!("index{}", uuid));
path
}
#[cfg(test)]
mod test {
use super::*;
use std::path::PathBuf;
#[test]
fn test_make_update_db_path() {
let uuid = Uuid::new_v4();
assert_eq!(
make_update_db_path("/home", &uuid),
PathBuf::from(format!("/home/update{}", uuid))
);
}
#[test]
fn test_make_index_db_path() {
let uuid = Uuid::new_v4();
assert_eq!(
make_index_db_path("/home", &uuid),
PathBuf::from(format!("/home/index{}", uuid))
);
}
mod index_store {
use super::*;
#[test]
fn test_index_uuid() {
let temp = tempfile::tempdir().unwrap();
let store = IndexStore::new(temp, IndexerOpts::default()).unwrap();
let name = "foobar";
let txn = store.env.read_txn().unwrap();
// name is not found if the uuid in not present in the db
assert!(store.index_uuid(&txn, &name).unwrap().is_none());
drop(txn);
// insert an uuid in the the name_to_uuid_db:
let uuid = Uuid::new_v4();
let mut txn = store.env.write_txn().unwrap();
store
.name_to_uuid
.put(&mut txn, &name, uuid.as_bytes())
.unwrap();
txn.commit().unwrap();
// check that the uuid is there
let txn = store.env.read_txn().unwrap();
assert_eq!(store.index_uuid(&txn, &name).unwrap(), Some(uuid));
}
#[test]
fn test_retrieve_index() {
let temp = tempfile::tempdir().unwrap();
let store = IndexStore::new(temp, IndexerOpts::default()).unwrap();
let uuid = Uuid::new_v4();
let txn = store.env.read_txn().unwrap();
assert!(store.retrieve_index(&txn, uuid).unwrap().is_none());
let created_at = Utc::now();
let updated_at = created_at;
let meta = IndexMeta {
update_store_size: 4096 * 100,
index_store_size: 4096 * 100,
uuid: uuid.clone(),
created_at,
updated_at,
};
let mut txn = store.env.write_txn().unwrap();
store
.uuid_to_index_meta
.put(&mut txn, uuid.as_bytes(), &meta)
.unwrap();
txn.commit().unwrap();
// the index cache should be empty
assert!(store.uuid_to_index.is_empty());
let txn = store.env.read_txn().unwrap();
assert!(store.retrieve_index(&txn, uuid).unwrap().is_some());
assert_eq!(store.uuid_to_index.len(), 1);
}
#[test]
fn test_index() {
let temp = tempfile::tempdir().unwrap();
let store = IndexStore::new(temp, IndexerOpts::default()).unwrap();
let name = "foobar";
assert!(store.index(&name).unwrap().is_none());
let created_at = Utc::now();
let updated_at = created_at;
let uuid = Uuid::new_v4();
let meta = IndexMeta {
update_store_size: 4096 * 100,
index_store_size: 4096 * 100,
uuid: uuid.clone(),
created_at,
updated_at,
};
let mut txn = store.env.write_txn().unwrap();
store
.name_to_uuid
.put(&mut txn, &name, uuid.as_bytes())
.unwrap();
store
.uuid_to_index_meta
.put(&mut txn, uuid.as_bytes(), &meta)
.unwrap();
txn.commit().unwrap();
assert!(store.index(&name).unwrap().is_some());
}
#[test]
fn test_get_or_create_index() {
let temp = tempfile::tempdir().unwrap();
let store = IndexStore::new(temp, IndexerOpts::default()).unwrap();
let name = "foobar";
let update_store_size = 4096 * 100;
let index_store_size = 4096 * 100;
store
.get_or_create_index(&name, update_store_size, index_store_size)
.unwrap();
let txn = store.env.read_txn().unwrap();
let uuid = store.name_to_uuid.get(&txn, &name).unwrap();
assert_eq!(store.uuid_to_index.len(), 1);
assert!(uuid.is_some());
let uuid = Uuid::from_slice(uuid.unwrap()).unwrap();
let meta = store
.uuid_to_index_meta
.get(&txn, uuid.as_bytes())
.unwrap()
.unwrap();
assert_eq!(meta.update_store_size, update_store_size);
assert_eq!(meta.index_store_size, index_store_size);
assert_eq!(meta.uuid, uuid);
}
#[test]
fn test_create_index() {
let temp = tempfile::tempdir().unwrap();
let store = IndexStore::new(temp, IndexerOpts::default()).unwrap();
let name = "foobar";
let update_store_size = 4096 * 100;
let index_store_size = 4096 * 100;
let uuid = Uuid::new_v4();
let mut txn = store.env.write_txn().unwrap();
store
.create_index_txn(&mut txn, uuid, name, update_store_size, index_store_size)
.unwrap();
let uuid = store.name_to_uuid.get(&txn, &name).unwrap();
assert_eq!(store.uuid_to_index.len(), 1);
assert!(uuid.is_some());
let uuid = Uuid::from_slice(uuid.unwrap()).unwrap();
let meta = store
.uuid_to_index_meta
.get(&txn, uuid.as_bytes())
.unwrap()
.unwrap();
assert_eq!(meta.update_store_size, update_store_size);
assert_eq!(meta.index_store_size, index_store_size);
assert_eq!(meta.uuid, uuid);
}
}
}

243
meilisearch-http/src/index_controller/local_index_controller/mod.rs
View File

@ -1,243 +0,0 @@
mod update_store;
mod index_store;
mod update_handler;
use std::path::Path;
use std::sync::Arc;
use anyhow::{bail, Context, anyhow};
use itertools::Itertools;
use milli::Index;
use crate::option::IndexerOpts;
use index_store::IndexStore;
use super::IndexController;
use super::updates::UpdateStatus;
use super::{UpdateMeta, UpdateResult, IndexMetadata, IndexSettings};
pub struct LocalIndexController {
indexes: IndexStore,
update_db_size: u64,
index_db_size: u64,
}
impl LocalIndexController {
pub fn new(
path: impl AsRef<Path>,
opt: IndexerOpts,
index_db_size: u64,
update_db_size: u64,
) -> anyhow::Result<Self> {
let indexes = IndexStore::new(path, opt)?;
Ok(Self { indexes, index_db_size, update_db_size })
}
}
impl IndexController for LocalIndexController {
fn add_documents<S: AsRef<str>>(
&self,
index: S,
method: milli::update::IndexDocumentsMethod,
format: milli::update::UpdateFormat,
data: &[u8],
primary_key: Option<String>,
) -> anyhow::Result<UpdateStatus<UpdateMeta, UpdateResult, String>> {
let (_, update_store) = self.indexes.get_or_create_index(&index, self.update_db_size, self.index_db_size)?;
let meta = UpdateMeta::DocumentsAddition { method, format, primary_key };
let pending = update_store.register_update(meta, data)?;
Ok(pending.into())
}
fn update_settings<S: AsRef<str>>(
&self,
index: S,
settings: super::Settings,
create: bool,
) -> anyhow::Result<UpdateStatus<UpdateMeta, UpdateResult, String>> {
let (_, update_store) = if create {
self.indexes.get_or_create_index(&index, self.update_db_size, self.index_db_size)?
} else {
self.indexes.index(&index)?.ok_or_else(|| anyhow!("Index {:?} doesn't exist", index.as_ref()))?
};
let meta = UpdateMeta::Settings(settings);
let pending = update_store.register_update(meta, &[])?;
Ok(pending.into())
}
fn create_index(&self, index_settings: IndexSettings) -> anyhow::Result<IndexMetadata> {
let index_name = index_settings.name.context("Missing name for index")?;
let (index, _, meta) = self.indexes.create_index(&index_name, self.update_db_size, self.index_db_size)?;
if let Some(ref primary_key) = index_settings.primary_key {
if let Err(e) = update_primary_key(index, primary_key).context("error creating index") {
// TODO: creating index could not be completed, delete everything.
Err(e)?
}
}
let meta = IndexMetadata {
uid: index_name,
uuid: meta.uuid.clone(),
created_at: meta.created_at,
updated_at: meta.created_at,
primary_key: index_settings.primary_key,
};
Ok(meta)
}
fn delete_index<S: AsRef<str>>(&self, index_uid: S) -> anyhow::Result<()> {
self.indexes.delete(index_uid)
}
fn swap_indices<S1: AsRef<str>, S2: AsRef<str>>(&self, _index1_uid: S1, _index2_uid: S2) -> anyhow::Result<()> {
todo!()
}
fn index(&self, name: impl AsRef<str>) -> anyhow::Result<Option<Arc<Index>>> {
let index = self.indexes.index(name)?.map(|(i, _)| i);
Ok(index)
}
fn update_status(&self, index: impl AsRef<str>, id: u64) -> anyhow::Result<Option<UpdateStatus<UpdateMeta, UpdateResult, String>>> {
match self.indexes.index(&index)? {
Some((_, update_store)) => Ok(update_store.meta(id)?),
None => bail!("index {:?} doesn't exist", index.as_ref()),
}
}
fn all_update_status(&self, index: impl AsRef<str>) -> anyhow::Result<Vec<UpdateStatus<UpdateMeta, UpdateResult, String>>> {
match self.indexes.index(&index)? {
Some((_, update_store)) => {
let updates = update_store.iter_metas(|processing, processed, aborted, pending, failed| {
let processing_id = processing
.as_ref()
.map(|p| p.id());
Ok(processing
.map(UpdateStatus::from)
.into_iter()
.chain(pending.
filter_map(Result::ok)
// If an update is processing, filter out this update from the pending
// updates.
.filter(|(_, u)| processing_id
.map_or(true, |id| id != u.id()))
.map(|(_, u)| UpdateStatus::from(u)))
.chain(aborted.filter_map(Result::ok).map(|(_, u)| UpdateStatus::from(u)))
.chain(processed.filter_map(Result::ok).map(|(_, u)| UpdateStatus::from(u)))
.chain(failed.filter_map(Result::ok).map(|(_, u)| UpdateStatus::from(u)))
.sorted_by(|a, b| a.id().cmp(&b.id()))
.collect())
})?;
Ok(updates)
}
None => bail!("index {} doesn't exist.", index.as_ref()),
}
}
fn list_indexes(&self) -> anyhow::Result<Vec<IndexMetadata>> {
let metas = self.indexes.list_indexes()?;
let mut output_meta = Vec::new();
for (uid, meta, primary_key) in metas {
let created_at = meta.created_at;
let uuid = meta.uuid;
let updated_at = self
.all_update_status(&uid)?
.iter()
.filter_map(|u| u.processed().map(|u| u.processed_at))
.max()
.unwrap_or(created_at);
let index_meta = IndexMetadata {
uid,
created_at,
updated_at,
uuid,
primary_key,
};
output_meta.push(index_meta);
}
Ok(output_meta)
}
fn update_index(&self, uid: impl AsRef<str>, index_settings: IndexSettings) -> anyhow::Result<IndexMetadata> {
if index_settings.name.is_some() {
bail!("can't udpate an index name.")
}
let (primary_key, meta) = match index_settings.primary_key {
Some(ref primary_key) => {
self.indexes
.update_index(&uid, |index| {
let mut txn = index.write_txn()?;
if index.primary_key(&txn)?.is_some() {
bail!("primary key already exists.")
}
index.put_primary_key(&mut txn, primary_key)?;
txn.commit()?;
Ok(Some(primary_key.clone()))
})?
},
None => {
let (index, meta) = self.indexes
.index_with_meta(&uid)?
.with_context(|| format!("index {:?} doesn't exist.", uid.as_ref()))?;
let primary_key = index
.primary_key(&index.read_txn()?)?
.map(String::from);
(primary_key, meta)
},
};
Ok(IndexMetadata {
uid: uid.as_ref().to_string(),
uuid: meta.uuid.clone(),
created_at: meta.created_at,
updated_at: meta.updated_at,
primary_key,
})
}
fn clear_documents(&self, index: impl AsRef<str>) -> anyhow::Result<super::UpdateStatus> {
let (_, update_store) = self.indexes.index(&index)?
.with_context(|| format!("Index {:?} doesn't exist", index.as_ref()))?;
let meta = UpdateMeta::ClearDocuments;
let pending = update_store.register_update(meta, &[])?;
Ok(pending.into())
}
fn delete_documents(&self, index: impl AsRef<str>, document_ids: Vec<String>) -> anyhow::Result<super::UpdateStatus> {
let (_, update_store) = self.indexes.index(&index)?
.with_context(|| format!("Index {:?} doesn't exist", index.as_ref()))?;
let meta = UpdateMeta::DeleteDocuments;
let content = serde_json::to_vec(&document_ids)?;
let pending = update_store.register_update(meta, &content)?;
Ok(pending.into())
}
}
fn update_primary_key(index: impl AsRef<Index>, primary_key: impl AsRef<str>) -> anyhow::Result<()> {
let index = index.as_ref();
let mut txn = index.write_txn()?;
if index.primary_key(&txn)?.is_some() {
bail!("primary key already set.")
}
index.put_primary_key(&mut txn, primary_key.as_ref())?;
txn.commit()?;
Ok(())
}
#[cfg(test)]
mod test {
use super::*;
use tempfile::tempdir;
use crate::make_index_controller_tests;
make_index_controller_tests!({
let options = IndexerOpts::default();
let path = tempdir().unwrap();
let size = 4096 * 100;
LocalIndexController::new(path, options, size, size).unwrap()
});
}

255
meilisearch-http/src/index_controller/local_index_controller/update_handler.rs
View File

@ -1,255 +0,0 @@
use std::collections::HashMap;
use std::io;
use std::sync::Arc;
use anyhow::Result;
use flate2::read::GzDecoder;
use grenad::CompressionType;
use log::info;
use milli::update::{IndexDocumentsMethod, UpdateBuilder, UpdateFormat};
use milli::Index;
use rayon::ThreadPool;
use super::update_store::HandleUpdate;
use crate::index_controller::updates::{Failed, Processed, Processing};
use crate::index_controller::{Facets, Settings, UpdateMeta, UpdateResult};
use crate::option::IndexerOpts;
pub struct UpdateHandler {
index: Arc<Index>,
max_nb_chunks: Option<usize>,
chunk_compression_level: Option<u32>,
thread_pool: Arc<ThreadPool>,
log_frequency: usize,
max_memory: usize,
linked_hash_map_size: usize,
chunk_compression_type: CompressionType,
chunk_fusing_shrink_size: u64,
}
impl UpdateHandler {
pub fn new(
opt: &IndexerOpts,
index: Arc<Index>,
thread_pool: Arc<ThreadPool>,
) -> anyhow::Result<Self> {
Ok(Self {
index,
max_nb_chunks: opt.max_nb_chunks,
chunk_compression_level: opt.chunk_compression_level,
thread_pool,
log_frequency: opt.log_every_n,
max_memory: opt.max_memory.get_bytes() as usize,
linked_hash_map_size: opt.linked_hash_map_size,
chunk_compression_type: opt.chunk_compression_type,
chunk_fusing_shrink_size: opt.chunk_fusing_shrink_size.get_bytes(),
})
}
fn update_buidler(&self, update_id: u64) -> UpdateBuilder {
// We prepare the update by using the update builder.
let mut update_builder = UpdateBuilder::new(update_id);
if let Some(max_nb_chunks) = self.max_nb_chunks {
update_builder.max_nb_chunks(max_nb_chunks);
}
if let Some(chunk_compression_level) = self.chunk_compression_level {
update_builder.chunk_compression_level(chunk_compression_level);
}
update_builder.thread_pool(&self.thread_pool);
update_builder.log_every_n(self.log_frequency);
update_builder.max_memory(self.max_memory);
update_builder.linked_hash_map_size(self.linked_hash_map_size);
update_builder.chunk_compression_type(self.chunk_compression_type);
update_builder.chunk_fusing_shrink_size(self.chunk_fusing_shrink_size);
update_builder
}
fn update_documents(
&self,
format: UpdateFormat,
method: IndexDocumentsMethod,
content: &[u8],
update_builder: UpdateBuilder,
primary_key: Option<&str>,
) -> anyhow::Result<UpdateResult> {
// We must use the write transaction of the update here.
let mut wtxn = self.index.write_txn()?;
// Set the primary key if not set already, ignore if already set.
match (self.index.primary_key(&wtxn)?, primary_key) {
(None, Some(ref primary_key)) => {
self.index.put_primary_key(&mut wtxn, primary_key)?;
}
_ => (),
}
let mut builder = update_builder.index_documents(&mut wtxn, &self.index);
builder.update_format(format);
builder.index_documents_method(method);
let gzipped = true;
let reader = if gzipped && !content.is_empty() {
Box::new(GzDecoder::new(content))
} else {
Box::new(content) as Box<dyn io::Read>
};
let result = builder.execute(reader, |indexing_step, update_id| {
info!("update {}: {:?}", update_id, indexing_step)
});
match result {
Ok(addition_result) => wtxn
.commit()
.and(Ok(UpdateResult::DocumentsAddition(addition_result)))
.map_err(Into::into),
Err(e) => Err(e.into()),
}
}
fn clear_documents(&self, update_builder: UpdateBuilder) -> anyhow::Result<UpdateResult> {
// We must use the write transaction of the update here.
let mut wtxn = self.index.write_txn()?;
let builder = update_builder.clear_documents(&mut wtxn, &self.index);
match builder.execute() {
Ok(_count) => wtxn
.commit()
.and(Ok(UpdateResult::Other))
.map_err(Into::into),
Err(e) => Err(e.into()),
}
}
fn update_settings(
&self,
settings: &Settings,
update_builder: UpdateBuilder,
) -> anyhow::Result<UpdateResult> {
// We must use the write transaction of the update here.
let mut wtxn = self.index.write_txn()?;
let mut builder = update_builder.settings(&mut wtxn, &self.index);
// We transpose the settings JSON struct into a real setting update.
if let Some(ref names) = settings.searchable_attributes {
match names {
Some(names) => builder.set_searchable_fields(names.clone()),
None => builder.reset_searchable_fields(),
}
}
// We transpose the settings JSON struct into a real setting update.
if let Some(ref names) = settings.displayed_attributes {
match names {
Some(names) => builder.set_displayed_fields(names.clone()),
None => builder.reset_displayed_fields(),
}
}
// We transpose the settings JSON struct into a real setting update.
if let Some(ref facet_types) = settings.faceted_attributes {
let facet_types = facet_types.clone().unwrap_or_else(|| HashMap::new());
builder.set_faceted_fields(facet_types);
}
// We transpose the settings JSON struct into a real setting update.
if let Some(ref criteria) = settings.ranking_rules {
match criteria {
Some(criteria) => builder.set_criteria(criteria.clone()),
None => builder.reset_criteria(),
}
}
let result = builder
.execute(|indexing_step, update_id| info!("update {}: {:?}", update_id, indexing_step));
match result {
Ok(()) => wtxn
.commit()
.and(Ok(UpdateResult::Other))
.map_err(Into::into),
Err(e) => Err(e.into()),
}
}
fn update_facets(
&self,
levels: &Facets,
update_builder: UpdateBuilder,
) -> anyhow::Result<UpdateResult> {
// We must use the write transaction of the update here.
let mut wtxn = self.index.write_txn()?;
let mut builder = update_builder.facets(&mut wtxn, &self.index);
if let Some(value) = levels.level_group_size {
builder.level_group_size(value);
}
if let Some(value) = levels.min_level_size {
builder.min_level_size(value);
}
match builder.execute() {
Ok(()) => wtxn
.commit()
.and(Ok(UpdateResult::Other))
.map_err(Into::into),
Err(e) => Err(e.into()),
}
}
fn delete_documents(
&self,
document_ids: &[u8],
update_builder: UpdateBuilder,
) -> anyhow::Result<UpdateResult> {
let ids: Vec<String> = serde_json::from_slice(document_ids)?;
let mut txn = self.index.write_txn()?;
let mut builder = update_builder.delete_documents(&mut txn, &self.index)?;
// We ignore unexisting document ids
ids.iter().for_each(|id| { builder.delete_external_id(id); });
match builder.execute() {
Ok(deleted) => txn
.commit()
.and(Ok(UpdateResult::DocumentDeletion { deleted }))
.map_err(Into::into),
Err(e) => Err(e.into())
}
}
}
impl HandleUpdate<UpdateMeta, UpdateResult, String> for UpdateHandler {
fn handle_update(
&mut self,
meta: Processing<UpdateMeta>,
content: &[u8],
) -> Result<Processed<UpdateMeta, UpdateResult>, Failed<UpdateMeta, String>> {
use UpdateMeta::*;
let update_id = meta.id();
let update_builder = self.update_buidler(update_id);
let result = match meta.meta() {
DocumentsAddition {
method,
format,
primary_key,
} => self.update_documents(
*format,
*method,
content,
update_builder,
primary_key.as_deref(),
),
ClearDocuments => self.clear_documents(update_builder),
DeleteDocuments => self.delete_documents(content, update_builder),
Settings(settings) => self.update_settings(settings, update_builder),
Facets(levels) => self.update_facets(levels, update_builder),
};
match result {
Ok(result) => Ok(meta.process(result)),
Err(e) => Err(meta.fail(e.to_string())),
}
}
}

470
meilisearch-http/src/index_controller/mod.rs
View File

@ -1,31 +1,34 @@
mod local_index_controller;
mod index_actor;
mod update_actor;
mod update_handler;
mod update_store;
mod updates;
mod uuid_resolver;
pub use local_index_controller::LocalIndexController;
use std::collections::HashMap;
use std::num::NonZeroUsize;
use std::path::Path;
use std::sync::Arc;
use std::time::Duration;
use anyhow::Result;
use chrono::{DateTime, Utc};
use milli::Index;
use milli::update::{IndexDocumentsMethod, UpdateFormat, DocumentAdditionResult};
use serde::{Serialize, Deserialize, de::Deserializer};
use uuid::Uuid;
use actix_web::web::{Bytes, Payload};
use anyhow::bail;
use futures::stream::StreamExt;
use milli::update::{IndexDocumentsMethod, UpdateFormat};
use serde::{Deserialize, Serialize};
use tokio::sync::mpsc;
use tokio::time::sleep;
pub use updates::{Processed, Processing, Failed};
use crate::index::{Document, SearchQuery, SearchResult};
use crate::index::{Facets, Settings, UpdateResult};
pub use updates::{Failed, Processed, Processing};
pub type UpdateStatus = updates::UpdateStatus<UpdateMeta, UpdateResult, String>;
#[derive(Debug, Serialize, Deserialize, Clone)]
#[serde(rename_all = "camelCase")]
pub struct IndexMetadata {
pub uid: String,
uuid: Uuid,
created_at: DateTime<Utc>,
updated_at: DateTime<Utc>,
primary_key: Option<String>,
uid: String,
#[serde(flatten)]
meta: index_actor::IndexMeta,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
@ -42,241 +45,248 @@ pub enum UpdateMeta {
Facets(Facets),
}
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
#[serde(rename_all = "camelCase")]
pub struct Facets {
pub level_group_size: Option<NonZeroUsize>,
pub min_level_size: Option<NonZeroUsize>,
}
fn deserialize_some<'de, T, D>(deserializer: D) -> Result<Option<T>, D::Error>
where T: Deserialize<'de>,
D: Deserializer<'de>
{
Deserialize::deserialize(deserializer).map(Some)
}
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
#[serde(deny_unknown_fields)]
#[serde(rename_all = "camelCase")]
pub struct Settings {
#[serde(
default,
deserialize_with = "deserialize_some",
skip_serializing_if = "Option::is_none",
)]
pub displayed_attributes: Option<Option<Vec<String>>>,
#[serde(
default,
deserialize_with = "deserialize_some",
skip_serializing_if = "Option::is_none",
)]
pub searchable_attributes: Option<Option<Vec<String>>>,
#[serde(default)]
pub faceted_attributes: Option<Option<HashMap<String, String>>>,
#[serde(
default,
deserialize_with = "deserialize_some",
skip_serializing_if = "Option::is_none",
)]
pub ranking_rules: Option<Option<Vec<String>>>,
}
impl Settings {
pub fn cleared() -> Self {
Self {
displayed_attributes: Some(None),
searchable_attributes: Some(None),
faceted_attributes: Some(None),
ranking_rules: Some(None),
}
}
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub enum UpdateResult {
DocumentsAddition(DocumentAdditionResult),
DocumentDeletion { deleted: usize },
Other,
}
#[derive(Clone, Debug)]
pub struct IndexSettings {
pub name: Option<String>,
pub uid: Option<String>,
pub primary_key: Option<String>,
}
/// The `IndexController` is in charge of the access to the underlying indices. It splits the logic
/// for read access which is provided thanks to an handle to the index, and write access which must
/// be provided. This allows the implementer to define the behaviour of write accesses to the
/// indices, and abstract the scheduling of the updates. The implementer must be able to provide an
/// instance of `IndexStore`
pub trait IndexController {
pub struct IndexController {
uuid_resolver: uuid_resolver::UuidResolverHandle,
index_handle: index_actor::IndexActorHandle,
update_handle: update_actor::UpdateActorHandle<Bytes>,
}
/*
* Write operations
*
* Logic for the write operation need to be provided by the implementer, since they can be made
* asynchronous thanks to an update_store for example.
*
* */
impl IndexController {
pub fn new(
path: impl AsRef<Path>,
index_size: usize,
update_store_size: usize,
) -> anyhow::Result<Self> {
let uuid_resolver = uuid_resolver::UuidResolverHandle::new(&path)?;
let index_actor = index_actor::IndexActorHandle::new(&path, index_size)?;
let update_handle =
update_actor::UpdateActorHandle::new(index_actor.clone(), &path, update_store_size)?;
Ok(Self {
uuid_resolver,
index_handle: index_actor,
update_handle,
})
}
/// Perform document addition on the database. If the provided index does not exist, it will be
/// created when the addition is applied to the index.
fn add_documents<S: AsRef<str>>(
pub async fn add_documents(
&self,
index: S,
method: IndexDocumentsMethod,
format: UpdateFormat,
data: &[u8],
uid: String,
method: milli::update::IndexDocumentsMethod,
format: milli::update::UpdateFormat,
mut payload: Payload,
primary_key: Option<String>,
) -> anyhow::Result<UpdateStatus>;
) -> anyhow::Result<UpdateStatus> {
let uuid = self.uuid_resolver.get_or_create(uid).await?;
let meta = UpdateMeta::DocumentsAddition {
method,
format,
primary_key,
};
let (sender, receiver) = mpsc::channel(10);
/// Clear all documents in the given index.
fn clear_documents(&self, index: impl AsRef<str>) -> anyhow::Result<UpdateStatus>;
// It is necessary to spawn a local task to senf the payload to the update handle to
// prevent dead_locking between the update_handle::update that waits for the update to be
// registered and the update_actor that waits for the the payload to be sent to it.
tokio::task::spawn_local(async move {
while let Some(bytes) = payload.next().await {
match bytes {
Ok(bytes) => {
let _ = sender.send(Ok(bytes)).await;
}
Err(e) => {
let error: Box<dyn std::error::Error + Sync + Send + 'static> = Box::new(e);
let _ = sender.send(Err(error)).await;
}
}
}
});
/// Delete all documents in `document_ids`.
fn delete_documents(&self, index: impl AsRef<str>, document_ids: Vec<String>) -> anyhow::Result<UpdateStatus>;
// This must be done *AFTER* spawning the task.
let status = self.update_handle.update(meta, receiver, uuid).await?;
Ok(status)
}
/// Updates an index settings. If the index does not exist, it will be created when the update
/// is applied to the index. `create` specifies whether an index should be created if not
/// existing.
fn update_settings<S: AsRef<str>>(&self, index_uid: S, settings: Settings, create: bool) -> anyhow::Result<UpdateStatus>;
pub async fn clear_documents(&self, uid: String) -> anyhow::Result<UpdateStatus> {
let uuid = self.uuid_resolver.resolve(uid).await?;
let meta = UpdateMeta::ClearDocuments;
let (_, receiver) = mpsc::channel(1);
let status = self.update_handle.update(meta, receiver, uuid).await?;
Ok(status)
}
/// Create an index with the given `index_uid`.
fn create_index(&self, index_settings: IndexSettings) -> Result<IndexMetadata>;
/// Delete index with the given `index_uid`, attempting to close it beforehand.
fn delete_index<S: AsRef<str>>(&self, index_uid: S) -> Result<()>;
/// Swap two indexes, concretely, it simply swaps the index the names point to.
fn swap_indices<S1: AsRef<str>, S2: AsRef<str>>(&self, index1_uid: S1, index2_uid: S2) -> Result<()>;
/// Apply an update to the given index. This method can be called when an update is ready to be
/// processed
fn handle_update<S: AsRef<str>>(
pub async fn delete_documents(
&self,
_index: S,
_update_id: u64,
_meta: Processing<UpdateMeta>,
_content: &[u8]
) -> Result<Processed<UpdateMeta, UpdateResult>, Failed<UpdateMeta, String>> {
todo!()
uid: String,
document_ids: Vec<String>,
) -> anyhow::Result<UpdateStatus> {
let uuid = self.uuid_resolver.resolve(uid).await?;
let meta = UpdateMeta::DeleteDocuments;
let (sender, receiver) = mpsc::channel(10);
tokio::task::spawn(async move {
let json = serde_json::to_vec(&document_ids).unwrap();
let bytes = Bytes::from(json);
let _ = sender.send(Ok(bytes)).await;
});
let status = self.update_handle.update(meta, receiver, uuid).await?;
Ok(status)
}
/// Returns, if it exists, the `Index` with the povided name.
fn index(&self, name: impl AsRef<str>) -> anyhow::Result<Option<Arc<Index>>>;
pub async fn update_settings(
&self,
uid: String,
settings: Settings,
create: bool,
) -> anyhow::Result<UpdateStatus> {
let uuid = if create {
let uuid = self.uuid_resolver.get_or_create(uid).await?;
// We need to create the index upfront, since it would otherwise only be created when
// the update is processed. This would make calls to GET index to fail until the update
// is complete. Since this is get or create, we ignore the error when the index already
// exists.
match self.index_handle.create_index(uuid, None).await {
Ok(_) | Err(index_actor::IndexError::IndexAlreadyExists) => (),
Err(e) => return Err(e.into()),
}
uuid
} else {
self.uuid_resolver.resolve(uid).await?
};
let meta = UpdateMeta::Settings(settings);
// Nothing so send, drop the sender right away, as not to block the update actor.
let (_, receiver) = mpsc::channel(1);
/// Returns the udpate status an update
fn update_status(&self, index: impl AsRef<str>, id: u64) -> anyhow::Result<Option<UpdateStatus>>;
let status = self.update_handle.update(meta, receiver, uuid).await?;
Ok(status)
}
/// Returns all the udpate status for an index
fn all_update_status(&self, index: impl AsRef<str>) -> anyhow::Result<Vec<UpdateStatus>>;
pub async fn create_index(
&self,
index_settings: IndexSettings,
) -> anyhow::Result<IndexMetadata> {
let IndexSettings { uid, primary_key } = index_settings;
let uid = uid.ok_or_else(|| anyhow::anyhow!("Can't create an index without a uid."))?;
let uuid = self.uuid_resolver.create(uid.clone()).await?;
let meta = self.index_handle.create_index(uuid, primary_key).await?;
let _ = self.update_handle.create(uuid).await?;
let meta = IndexMetadata { uid, meta };
/// List all the indexes
fn list_indexes(&self) -> anyhow::Result<Vec<IndexMetadata>>;
Ok(meta)
}
fn update_index(&self, name: impl AsRef<str>, index_settings: IndexSettings) -> anyhow::Result<IndexMetadata>;
pub async fn delete_index(&self, uid: String) -> anyhow::Result<()> {
let uuid = self.uuid_resolver.delete(uid).await?;
self.update_handle.delete(uuid).await?;
self.index_handle.delete(uuid).await?;
Ok(())
}
pub async fn update_status(&self, uid: String, id: u64) -> anyhow::Result<UpdateStatus> {
let uuid = self.uuid_resolver.resolve(uid).await?;
let result = self.update_handle.update_status(uuid, id).await?;
Ok(result)
}
pub async fn all_update_status(&self, uid: String) -> anyhow::Result<Vec<UpdateStatus>> {
let uuid = self.uuid_resolver.resolve(uid).await?;
let result = self.update_handle.get_all_updates_status(uuid).await?;
Ok(result)
}
pub async fn list_indexes(&self) -> anyhow::Result<Vec<IndexMetadata>> {
let uuids = self.uuid_resolver.list().await?;
let mut ret = Vec::new();
for (uid, uuid) in uuids {
let meta = self.index_handle.get_index_meta(uuid).await?;
let meta = IndexMetadata { uid, meta };
ret.push(meta);
}
Ok(ret)
}
pub async fn settings(&self, uid: String) -> anyhow::Result<Settings> {
let uuid = self.uuid_resolver.resolve(uid.clone()).await?;
let settings = self.index_handle.settings(uuid).await?;
Ok(settings)
}
pub async fn documents(
&self,
uid: String,
offset: usize,
limit: usize,
attributes_to_retrieve: Option<Vec<String>>,
) -> anyhow::Result<Vec<Document>> {
let uuid = self.uuid_resolver.resolve(uid.clone()).await?;
let documents = self
.index_handle
.documents(uuid, offset, limit, attributes_to_retrieve)
.await?;
Ok(documents)
}
pub async fn document(
&self,
uid: String,
doc_id: String,
attributes_to_retrieve: Option<Vec<String>>,
) -> anyhow::Result<Document> {
let uuid = self.uuid_resolver.resolve(uid.clone()).await?;
let document = self
.index_handle
.document(uuid, doc_id, attributes_to_retrieve)
.await?;
Ok(document)
}
pub async fn update_index(
&self,
uid: String,
index_settings: IndexSettings,
) -> anyhow::Result<IndexMetadata> {
if index_settings.uid.is_some() {
bail!("Can't change the index uid.")
}
let uuid = self.uuid_resolver.resolve(uid.clone()).await?;
let meta = self.index_handle.update_index(uuid, index_settings).await?;
let meta = IndexMetadata { uid, meta };
Ok(meta)
}
pub async fn search(&self, uid: String, query: SearchQuery) -> anyhow::Result<SearchResult> {
let uuid = self.uuid_resolver.resolve(uid).await?;
let result = self.index_handle.search(uuid, query).await?;
Ok(result)
}
pub async fn get_index(&self, uid: String) -> anyhow::Result<IndexMetadata> {
let uuid = self.uuid_resolver.resolve(uid.clone()).await?;
let meta = self.index_handle.get_index_meta(uuid).await?;
let meta = IndexMetadata { uid, meta };
Ok(meta)
}
}
#[cfg(test)]
#[macro_use]
pub(crate) mod test {
use super::*;
#[macro_export]
macro_rules! make_index_controller_tests {
($controller_buider:block) => {
#[test]
fn test_create_and_list_indexes() {
crate::index_controller::test::create_and_list_indexes($controller_buider);
pub async fn get_arc_ownership_blocking<T>(mut item: Arc<T>) -> T {
loop {
match Arc::try_unwrap(item) {
Ok(item) => return item,
Err(item_arc) => {
item = item_arc;
sleep(Duration::from_millis(100)).await;
continue;
}
#[test]
fn test_create_index_with_no_name_is_error() {
crate::index_controller::test::create_index_with_no_name_is_error($controller_buider);
}
#[test]
fn test_update_index() {
crate::index_controller::test::update_index($controller_buider);
}
};
}
pub(crate) fn create_and_list_indexes(controller: impl IndexController) {
let settings1 = IndexSettings {
name: Some(String::from("test_index")),
primary_key: None,
};
let settings2 = IndexSettings {
name: Some(String::from("test_index2")),
primary_key: Some(String::from("foo")),
};
controller.create_index(settings1).unwrap();
controller.create_index(settings2).unwrap();
let indexes = controller.list_indexes().unwrap();
assert_eq!(indexes.len(), 2);
assert_eq!(indexes[0].uid, "test_index");
assert_eq!(indexes[1].uid, "test_index2");
assert_eq!(indexes[1].primary_key.clone().unwrap(), "foo");
}
pub(crate) fn create_index_with_no_name_is_error(controller: impl IndexController) {
let settings = IndexSettings {
name: None,
primary_key: None,
};
assert!(controller.create_index(settings).is_err());
}
pub(crate) fn update_index(controller: impl IndexController) {
let settings = IndexSettings {
name: Some(String::from("test")),
primary_key: None,
};
assert!(controller.create_index(settings).is_ok());
// perform empty update returns index meta unchanged
let settings = IndexSettings {
name: None,
primary_key: None,
};
let result = controller.update_index("test", settings).unwrap();
assert_eq!(result.uid, "test");
assert_eq!(result.created_at, result.updated_at);
assert!(result.primary_key.is_none());
// Changing the name trigger an error
let settings = IndexSettings {
name: Some(String::from("bar")),
primary_key: None,
};
assert!(controller.update_index("test", settings).is_err());
// Update primary key
let settings = IndexSettings {
name: None,
primary_key: Some(String::from("foo")),
};
let result = controller.update_index("test", settings.clone()).unwrap();
assert_eq!(result.uid, "test");
assert!(result.created_at < result.updated_at);
assert_eq!(result.primary_key.unwrap(), "foo");
// setting the primary key again is an error
assert!(controller.update_index("test", settings).is_err());
}
}
}

369
meilisearch-http/src/index_controller/update_actor.rs Normal file
View File

@ -0,0 +1,369 @@
use std::collections::{hash_map::Entry, HashMap};
use std::fs::{create_dir_all, remove_dir_all};
use std::path::{Path, PathBuf};
use std::sync::Arc;
use super::index_actor::IndexActorHandle;
use log::info;
use thiserror::Error;
use tokio::fs::File;
use tokio::io::AsyncWriteExt;
use tokio::sync::{mpsc, oneshot, RwLock};
use uuid::Uuid;
use super::get_arc_ownership_blocking;
use crate::index::UpdateResult;
use crate::index_controller::{UpdateMeta, UpdateStatus};
pub type Result<T> = std::result::Result<T, UpdateError>;
type UpdateStore = super::update_store::UpdateStore<UpdateMeta, UpdateResult, String>;
type PayloadData<D> = std::result::Result<D, Box<dyn std::error::Error + Sync + Send + 'static>>;
#[derive(Debug, Error)]
pub enum UpdateError {
#[error("error with update: {0}")]
Error(Box<dyn std::error::Error + Sync + Send + 'static>),
#[error("Index {0} doesn't exist.")]
UnexistingIndex(Uuid),
#[error("Update {0} doesn't exist.")]
UnexistingUpdate(u64),
}
enum UpdateMsg<D> {
Update {
uuid: Uuid,
meta: UpdateMeta,
data: mpsc::Receiver<PayloadData<D>>,
ret: oneshot::Sender<Result<UpdateStatus>>,
},
ListUpdates {
uuid: Uuid,
ret: oneshot::Sender<Result<Vec<UpdateStatus>>>,
},
GetUpdate {
uuid: Uuid,
ret: oneshot::Sender<Result<UpdateStatus>>,
id: u64,
},
Delete {
uuid: Uuid,
ret: oneshot::Sender<Result<()>>,
},
Create {
uuid: Uuid,
ret: oneshot::Sender<Result<()>>,
},
}
struct UpdateActor<D, S> {
path: PathBuf,
store: S,
inbox: mpsc::Receiver<UpdateMsg<D>>,
}
#[async_trait::async_trait]
trait UpdateStoreStore {
async fn get_or_create(&self, uuid: Uuid) -> Result<Arc<UpdateStore>>;
async fn delete(&self, uuid: Uuid) -> Result<Option<Arc<UpdateStore>>>;
async fn get(&self, uuid: Uuid) -> Result<Option<Arc<UpdateStore>>>;
}
impl<D, S> UpdateActor<D, S>
where
D: AsRef<[u8]> + Sized + 'static,
S: UpdateStoreStore,
{
fn new(
store: S,
inbox: mpsc::Receiver<UpdateMsg<D>>,
path: impl AsRef<Path>,
) -> anyhow::Result<Self> {
let path = path.as_ref().to_owned().join("update_files");
create_dir_all(&path)?;
assert!(path.exists());
Ok(Self { store, inbox, path })
}
async fn run(mut self) {
use UpdateMsg::*;
info!("Started update actor.");
loop {
match self.inbox.recv().await {
Some(Update {
uuid,
meta,
data,
ret,
}) => {
let _ = ret.send(self.handle_update(uuid, meta, data).await);
}
Some(ListUpdates { uuid, ret }) => {
let _ = ret.send(self.handle_list_updates(uuid).await);
}
Some(GetUpdate { uuid, ret, id }) => {
let _ = ret.send(self.handle_get_update(uuid, id).await);
}
Some(Delete { uuid, ret }) => {
let _ = ret.send(self.handle_delete(uuid).await);
}
Some(Create { uuid, ret }) => {
let _ = ret.send(self.handle_create(uuid).await);
}
None => break,
}
}
}
async fn handle_update(
&self,
uuid: Uuid,
meta: UpdateMeta,
mut payload: mpsc::Receiver<PayloadData<D>>,
) -> Result<UpdateStatus> {
let update_store = self.store.get_or_create(uuid).await?;
let update_file_id = uuid::Uuid::new_v4();
let path = self.path.join(format!("update_{}", update_file_id));
let mut file = File::create(&path)
.await
.map_err(|e| UpdateError::Error(Box::new(e)))?;
while let Some(bytes) = payload.recv().await {
match bytes {
Ok(bytes) => {
file.write_all(bytes.as_ref())
.await
.map_err(|e| UpdateError::Error(Box::new(e)))?;
}
Err(e) => {
return Err(UpdateError::Error(e));
}
}
}
file.flush()
.await
.map_err(|e| UpdateError::Error(Box::new(e)))?;
tokio::task::spawn_blocking(move || {
update_store
.register_update(meta, path, uuid)
.map(UpdateStatus::Pending)
.map_err(|e| UpdateError::Error(Box::new(e)))
})
.await
.map_err(|e| UpdateError::Error(Box::new(e)))?
}
async fn handle_list_updates(&self, uuid: Uuid) -> Result<Vec<UpdateStatus>> {
let update_store = self.store.get(uuid).await?;
tokio::task::spawn_blocking(move || {
let result = update_store
.ok_or(UpdateError::UnexistingIndex(uuid))?
.list()
.map_err(|e| UpdateError::Error(e.into()))?;
Ok(result)
})
.await
.map_err(|e| UpdateError::Error(Box::new(e)))?
}
async fn handle_get_update(&self, uuid: Uuid, id: u64) -> Result<UpdateStatus> {
let store = self
.store
.get(uuid)
.await?
.ok_or(UpdateError::UnexistingIndex(uuid))?;
let result = store
.meta(id)
.map_err(|e| UpdateError::Error(Box::new(e)))?
.ok_or(UpdateError::UnexistingUpdate(id))?;
Ok(result)
}
async fn handle_delete(&self, uuid: Uuid) -> Result<()> {
let store = self.store.delete(uuid).await?;
if let Some(store) = store {
tokio::task::spawn(async move {
let store = get_arc_ownership_blocking(store).await;
tokio::task::spawn_blocking(move || {
store.prepare_for_closing().wait();
info!("Update store {} was closed.", uuid);
});
});
}
Ok(())
}
async fn handle_create(&self, uuid: Uuid) -> Result<()> {
let _ = self.store.get_or_create(uuid).await?;
Ok(())
}
}
#[derive(Clone)]
pub struct UpdateActorHandle<D> {
sender: mpsc::Sender<UpdateMsg<D>>,
}
impl<D> UpdateActorHandle<D>
where
D: AsRef<[u8]> + Sized + 'static + Sync + Send,
{
pub fn new(
index_handle: IndexActorHandle,
path: impl AsRef<Path>,
update_store_size: usize,
) -> anyhow::Result<Self> {
let path = path.as_ref().to_owned().join("updates");
let (sender, receiver) = mpsc::channel(100);
let store = MapUpdateStoreStore::new(index_handle, &path, update_store_size);
let actor = UpdateActor::new(store, receiver, path)?;
tokio::task::spawn(actor.run());
Ok(Self { sender })
}
pub async fn update(
&self,
meta: UpdateMeta,
data: mpsc::Receiver<PayloadData<D>>,
uuid: Uuid,
) -> Result<UpdateStatus> {
let (ret, receiver) = oneshot::channel();
let msg = UpdateMsg::Update {
uuid,
data,
meta,
ret,
};
let _ = self.sender.send(msg).await;
receiver.await.expect("update actor killed.")
}
pub async fn get_all_updates_status(&self, uuid: Uuid) -> Result<Vec<UpdateStatus>> {
let (ret, receiver) = oneshot::channel();
let msg = UpdateMsg::ListUpdates { uuid, ret };
let _ = self.sender.send(msg).await;
receiver.await.expect("update actor killed.")
}
pub async fn update_status(&self, uuid: Uuid, id: u64) -> Result<UpdateStatus> {
let (ret, receiver) = oneshot::channel();
let msg = UpdateMsg::GetUpdate { uuid, id, ret };
let _ = self.sender.send(msg).await;
receiver.await.expect("update actor killed.")
}
pub async fn delete(&self, uuid: Uuid) -> Result<()> {
let (ret, receiver) = oneshot::channel();
let msg = UpdateMsg::Delete { uuid, ret };
let _ = self.sender.send(msg).await;
receiver.await.expect("update actor killed.")
}
pub async fn create(&self, uuid: Uuid) -> Result<()> {
let (ret, receiver) = oneshot::channel();
let msg = UpdateMsg::Create { uuid, ret };
let _ = self.sender.send(msg).await;
receiver.await.expect("update actor killed.")
}
}
struct MapUpdateStoreStore {
db: Arc<RwLock<HashMap<Uuid, Arc<UpdateStore>>>>,
index_handle: IndexActorHandle,
path: PathBuf,
update_store_size: usize,
}
impl MapUpdateStoreStore {
fn new(
index_handle: IndexActorHandle,
path: impl AsRef<Path>,
update_store_size: usize,
) -> Self {
let db = Arc::new(RwLock::new(HashMap::new()));
let path = path.as_ref().to_owned();
Self {
db,
index_handle,
path,
update_store_size,
}
}
}
#[async_trait::async_trait]
impl UpdateStoreStore for MapUpdateStoreStore {
async fn get_or_create(&self, uuid: Uuid) -> Result<Arc<UpdateStore>> {
match self.db.write().await.entry(uuid) {
Entry::Vacant(e) => {
let mut options = heed::EnvOpenOptions::new();
let update_store_size = self.update_store_size;
options.map_size(update_store_size);
let path = self.path.clone().join(format!("updates-{}", e.key()));
create_dir_all(&path).unwrap();
let index_handle = self.index_handle.clone();
let store = UpdateStore::open(options, &path, move |meta, file| {
futures::executor::block_on(index_handle.update(meta, file))
})
.map_err(|e| UpdateError::Error(e.into()))?;
let store = e.insert(store);
Ok(store.clone())
}
Entry::Occupied(e) => Ok(e.get().clone()),
}
}
async fn get(&self, uuid: Uuid) -> Result<Option<Arc<UpdateStore>>> {
let guard = self.db.read().await;
match guard.get(&uuid) {
Some(uuid) => Ok(Some(uuid.clone())),
None => {
// The index is not found in the found in the loaded indexes, so we attempt to load
// it from disk. We need to acquire a write lock **before** attempting to open the
// index, because someone could be trying to open it at the same time as us.
drop(guard);
let path = self.path.clone().join(format!("updates-{}", uuid));
if path.exists() {
let mut guard = self.db.write().await;
match guard.entry(uuid) {
Entry::Vacant(entry) => {
// We can safely load the index
let index_handle = self.index_handle.clone();
let mut options = heed::EnvOpenOptions::new();
let update_store_size = self.update_store_size;
options.map_size(update_store_size);
let store = UpdateStore::open(options, &path, move |meta, file| {
futures::executor::block_on(index_handle.update(meta, file))
})
.map_err(|e| UpdateError::Error(e.into()))?;
let store = entry.insert(store);
Ok(Some(store.clone()))
}
Entry::Occupied(entry) => {
// The index was loaded while we attempted to to iter
Ok(Some(entry.get().clone()))
}
}
} else {
Ok(None)
}
}
}
}
async fn delete(&self, uuid: Uuid) -> Result<Option<Arc<UpdateStore>>> {
let store = self.db.write().await.remove(&uuid);
let path = self.path.clone().join(format!("updates-{}", uuid));
if store.is_some() || path.exists() {
remove_dir_all(path).unwrap();
}
Ok(store)
}
}

95
meilisearch-http/src/index_controller/update_handler.rs Normal file
View File

@ -0,0 +1,95 @@
use std::fs::File;
use crate::index::Index;
use anyhow::Result;
use grenad::CompressionType;
use milli::update::UpdateBuilder;
use rayon::ThreadPool;
use crate::index::UpdateResult;
use crate::index_controller::updates::{Failed, Processed, Processing};
use crate::index_controller::UpdateMeta;
use crate::option::IndexerOpts;
pub struct UpdateHandler {
max_nb_chunks: Option<usize>,
chunk_compression_level: Option<u32>,
thread_pool: ThreadPool,
log_frequency: usize,
max_memory: usize,
linked_hash_map_size: usize,
chunk_compression_type: CompressionType,
chunk_fusing_shrink_size: u64,
}
impl UpdateHandler {
pub fn new(opt: &IndexerOpts) -> anyhow::Result<Self> {
let thread_pool = rayon::ThreadPoolBuilder::new()
.num_threads(opt.indexing_jobs.unwrap_or(0))
.build()?;
Ok(Self {
max_nb_chunks: opt.max_nb_chunks,
chunk_compression_level: opt.chunk_compression_level,
thread_pool,
log_frequency: opt.log_every_n,
max_memory: opt.max_memory.get_bytes() as usize,
linked_hash_map_size: opt.linked_hash_map_size,
chunk_compression_type: opt.chunk_compression_type,
chunk_fusing_shrink_size: opt.chunk_fusing_shrink_size.get_bytes(),
})
}
fn update_buidler(&self, update_id: u64) -> UpdateBuilder {
// We prepare the update by using the update builder.
let mut update_builder = UpdateBuilder::new(update_id);
if let Some(max_nb_chunks) = self.max_nb_chunks {
update_builder.max_nb_chunks(max_nb_chunks);
}
if let Some(chunk_compression_level) = self.chunk_compression_level {
update_builder.chunk_compression_level(chunk_compression_level);
}
update_builder.thread_pool(&self.thread_pool);
update_builder.log_every_n(self.log_frequency);
update_builder.max_memory(self.max_memory);
update_builder.linked_hash_map_size(self.linked_hash_map_size);
update_builder.chunk_compression_type(self.chunk_compression_type);
update_builder.chunk_fusing_shrink_size(self.chunk_fusing_shrink_size);
update_builder
}
pub fn handle_update(
&self,
meta: Processing<UpdateMeta>,
content: File,
index: Index,
) -> Result<Processed<UpdateMeta, UpdateResult>, Failed<UpdateMeta, String>> {
use UpdateMeta::*;
let update_id = meta.id();
let update_builder = self.update_buidler(update_id);
let result = match meta.meta() {
DocumentsAddition {
method,
format,
primary_key,
} => index.update_documents(
*format,
*method,
content,
update_builder,
primary_key.as_deref(),
),
ClearDocuments => index.clear_documents(update_builder),
DeleteDocuments => index.delete_documents(content, update_builder),
Settings(settings) => index.update_settings(settings, update_builder),
Facets(levels) => index.update_facets(levels, update_builder),
};
match result {
Ok(result) => Ok(meta.process(result)),
Err(e) => Err(meta.fail(e.to_string())),
}
}
}

327
meilisearch-http/src/index_controller/local_index_controller/update_store.rs → meilisearch-http/src/index_controller/update_store.rs
View File

@ -1,10 +1,14 @@
use std::path::Path;
use std::sync::{Arc, RwLock};
use std::fs::remove_file;
use std::path::{Path, PathBuf};
use std::sync::Arc;
use crossbeam_channel::Sender;
use heed::types::{OwnedType, DecodeIgnore, SerdeJson, ByteSlice};
use heed::{EnvOpenOptions, Env, Database};
use serde::{Serialize, Deserialize};
use heed::types::{DecodeIgnore, OwnedType, SerdeJson};
use heed::{Database, Env, EnvOpenOptions};
use parking_lot::RwLock;
use serde::{Deserialize, Serialize};
use std::fs::File;
use tokio::sync::mpsc;
use uuid::Uuid;
use crate::index_controller::updates::*;
@ -14,16 +18,33 @@ type BEU64 = heed::zerocopy::U64<heed::byteorder::BE>;
pub struct UpdateStore<M, N, E> {
env: Env,
pending_meta: Database<OwnedType<BEU64>, SerdeJson<Pending<M>>>,
pending: Database<OwnedType<BEU64>, ByteSlice>,
pending: Database<OwnedType<BEU64>, SerdeJson<PathBuf>>,
processed_meta: Database<OwnedType<BEU64>, SerdeJson<Processed<M, N>>>,
failed_meta: Database<OwnedType<BEU64>, SerdeJson<Failed<M, E>>>,
aborted_meta: Database<OwnedType<BEU64>, SerdeJson<Aborted<M>>>,
processing: Arc<RwLock<Option<Processing<M>>>>,
notification_sender: Sender<()>,
notification_sender: mpsc::Sender<()>,
}
pub trait HandleUpdate<M, N, E> {
fn handle_update(&mut self, meta: Processing<M>, content: &[u8]) -> Result<Processed<M, N>, Failed<M, E>>;
fn handle_update(
&mut self,
meta: Processing<M>,
content: File,
) -> anyhow::Result<Result<Processed<M, N>, Failed<M, E>>>;
}
impl<M, N, E, F> HandleUpdate<M, N, E> for F
where
F: FnMut(Processing<M>, File) -> anyhow::Result<Result<Processed<M, N>, Failed<M, E>>>,
{
fn handle_update(
&mut self,
meta: Processing<M>,
content: File,
) -> anyhow::Result<Result<Processed<M, N>, Failed<M, E>>> {
self(meta, content)
}
}
impl<M, N, E> UpdateStore<M, N, E>
@ -35,11 +56,11 @@ where
pub fn open<P, U>(
mut options: EnvOpenOptions,
path: P,
mut update_handler: U,
update_handler: U,
) -> heed::Result<Arc<Self>>
where
P: AsRef<Path>,
U: HandleUpdate<M, N, E> + Send + 'static,
U: HandleUpdate<M, N, E> + Sync + Clone + Send + 'static,
{
options.max_dbs(5);
@ -51,7 +72,7 @@ where
let failed_meta = env.create_database(Some("failed-meta"))?;
let processing = Arc::new(RwLock::new(None));
let (notification_sender, notification_receiver) = crossbeam_channel::bounded(1);
let (notification_sender, mut notification_receiver) = mpsc::channel(10);
// Send a first notification to trigger the process.
let _ = notification_sender.send(());
@ -69,13 +90,19 @@ where
// We need a weak reference so we can take ownership on the arc later when we
// want to close the index.
let update_store_weak = Arc::downgrade(&update_store);
std::thread::spawn(move || {
tokio::task::spawn(async move {
// Block and wait for something to process.
'outer: for _ in notification_receiver {
'outer: while notification_receiver.recv().await.is_some() {
loop {
match update_store_weak.upgrade() {
Some(update_store) => {
match update_store.process_pending_update(&mut update_handler) {
let handler = update_handler.clone();
let res = tokio::task::spawn_blocking(move || {
update_store.process_pending_update(handler)
})
.await
.expect("Fatal error processing update.");
match res {
Ok(Some(_)) => (),
Ok(None) => break,
Err(e) => eprintln!("error while processing update: {}", e),
@ -97,17 +124,20 @@ where
/// Returns the new biggest id to use to store the new update.
fn new_update_id(&self, txn: &heed::RoTxn) -> heed::Result<u64> {
let last_pending = self.pending_meta
let last_pending = self
.pending_meta
.remap_data_type::<DecodeIgnore>()
.last(txn)?
.map(|(k, _)| k.get());
let last_processed = self.processed_meta
let last_processed = self
.processed_meta
.remap_data_type::<DecodeIgnore>()
.last(txn)?
.map(|(k, _)| k.get());
let last_aborted = self.aborted_meta
let last_aborted = self
.aborted_meta
.remap_data_type::<DecodeIgnore>()
.last(txn)?
.map(|(k, _)| k.get());
@ -129,7 +159,8 @@ where
pub fn register_update(
&self,
meta: M,
content: &[u8]
content: impl AsRef<Path>,
index_uuid: Uuid,
) -> heed::Result<Pending<M>> {
let mut wtxn = self.env.write_txn()?;
@ -140,23 +171,24 @@ where
let update_id = self.new_update_id(&wtxn)?;
let update_key = BEU64::new(update_id);
let meta = Pending::new(meta, update_id);
let meta = Pending::new(meta, update_id, index_uuid);
self.pending_meta.put(&mut wtxn, &update_key, &meta)?;
self.pending.put(&mut wtxn, &update_key, content)?;
self.pending
.put(&mut wtxn, &update_key, &content.as_ref().to_owned())?;
wtxn.commit()?;
if let Err(e) = self.notification_sender.try_send(()) {
assert!(!e.is_disconnected(), "update notification channel is disconnected");
}
self.notification_sender
.blocking_send(())
.expect("Update store loop exited.");
Ok(meta)
}
/// Executes the user provided function on the next pending update (the one with the lowest id).
/// This is asynchronous as it let the user process the update with a read-only txn and
/// only writing the result meta to the processed-meta store *after* it has been processed.
fn process_pending_update<U>(&self, handler: &mut U) -> heed::Result<Option<()>>
fn process_pending_update<U>(&self, mut handler: U) -> anyhow::Result<Option<()>>
where
U: HandleUpdate<M, N, E> + Send + 'static,
U: HandleUpdate<M, N, E>,
{
// Create a read transaction to be able to retrieve the pending update in order.
let rtxn = self.env.read_txn()?;
@ -166,7 +198,8 @@ where
// a reader while processing it, not a writer.
match first_meta {
Some((first_id, pending)) => {
let first_content = self.pending
let content_path = self
.pending
.get(&rtxn, &first_id)?
.expect("associated update content");
@ -174,23 +207,19 @@ where
// to the update handler. Processing store is non persistent to be able recover
// from a failure
let processing = pending.processing();
self.processing
.write()
.unwrap()
.replace(processing.clone());
self.processing.write().replace(processing.clone());
let file = File::open(&content_path)?;
// Process the pending update using the provided user function.
let result = handler.handle_update(processing, first_content);
let result = handler.handle_update(processing, file)?;
drop(rtxn);
// Once the pending update have been successfully processed
// we must remove the content from the pending and processing stores and
// write the *new* meta to the processed-meta store and commit.
let mut wtxn = self.env.write_txn()?;
self.processing
.write()
.unwrap()
.take();
self.processing.write().take();
self.pending_meta.delete(&mut wtxn, &first_id)?;
remove_file(&content_path)?;
self.pending.delete(&mut wtxn, &first_id)?;
match result {
Ok(processed) => self.processed_meta.put(&mut wtxn, &first_id, &processed)?,
@ -199,35 +228,60 @@ where
wtxn.commit()?;
Ok(Some(()))
},
None => Ok(None)
}
None => Ok(None),
}
}
/// Execute the user defined function with the meta-store iterators, the first
/// iterator is the *processed* meta one, the second the *aborted* meta one
/// and, the last is the *pending* meta one.
pub fn iter_metas<F, T>(&self, mut f: F) -> heed::Result<T>
where
F: for<'a> FnMut(
Option<Processing<M>>,
heed::RoIter<'a, OwnedType<BEU64>, SerdeJson<Processed<M, N>>>,
heed::RoIter<'a, OwnedType<BEU64>, SerdeJson<Aborted<M>>>,
heed::RoIter<'a, OwnedType<BEU64>, SerdeJson<Pending<M>>>,
heed::RoIter<'a, OwnedType<BEU64>, SerdeJson<Failed<M, E>>>,
) -> heed::Result<T>,
{
pub fn list(&self) -> anyhow::Result<Vec<UpdateStatus<M, N, E>>> {
let rtxn = self.env.read_txn()?;
let mut updates = Vec::new();
// We get the pending, processed and aborted meta iterators.
let processed_iter = self.processed_meta.iter(&rtxn)?;
let aborted_iter = self.aborted_meta.iter(&rtxn)?;
let pending_iter = self.pending_meta.iter(&rtxn)?;
let processing = self.processing.read().unwrap().clone();
let failed_iter = self.failed_meta.iter(&rtxn)?;
let processing = self.processing.read();
if let Some(ref processing) = *processing {
let update = UpdateStatus::from(processing.clone());
updates.push(update);
}
// We execute the user defined function with both iterators.
(f)(processing, processed_iter, aborted_iter, pending_iter, failed_iter)
let pending = self
.pending_meta
.iter(&rtxn)?
.filter_map(Result::ok)
.filter_map(|(_, p)| (Some(p.id()) != processing.as_ref().map(|p| p.id())).then(|| p))
.map(UpdateStatus::from);
updates.extend(pending);
let aborted = self
.aborted_meta
.iter(&rtxn)?
.filter_map(Result::ok)
.map(|(_, p)| p)
.map(UpdateStatus::from);
updates.extend(aborted);
let processed = self
.processed_meta
.iter(&rtxn)?
.filter_map(Result::ok)
.map(|(_, p)| p)
.map(UpdateStatus::from);
updates.extend(processed);
let failed = self
.failed_meta
.iter(&rtxn)?
.filter_map(Result::ok)
.map(|(_, p)| p)
.map(UpdateStatus::from);
updates.extend(failed);
updates.sort_by_key(|u| u.id());
Ok(updates)
}
/// Returns the update associated meta or `None` if the update doesn't exist.
@ -235,7 +289,7 @@ where
let rtxn = self.env.read_txn()?;
let key = BEU64::new(update_id);
if let Some(ref meta) = *self.processing.read().unwrap() {
if let Some(ref meta) = *self.processing.read() {
if meta.id() == update_id {
return Ok(Some(UpdateStatus::Processing(meta.clone())));
}
@ -319,89 +373,92 @@ where
}
}
#[cfg(test)]
mod tests {
use super::*;
use std::thread;
use std::time::{Duration, Instant};
//#[cfg(test)]
//mod tests {
//use super::*;
//use std::thread;
//use std::time::{Duration, Instant};
impl<M, N, F, E> HandleUpdate<M, N, E> for F
where F: FnMut(Processing<M>, &[u8]) -> Result<Processed<M, N>, Failed<M, E>> + Send + 'static {
fn handle_update(&mut self, meta: Processing<M>, content: &[u8]) -> Result<Processed<M, N>, Failed<M, E>> {
self(meta, content)
}
}
//#[test]
//fn simple() {
//let dir = tempfile::tempdir().unwrap();
//let mut options = EnvOpenOptions::new();
//options.map_size(4096 * 100);
//let update_store = UpdateStore::open(
//options,
//dir,
//|meta: Processing<String>, _content: &_| -> Result<_, Failed<_, ()>> {
//let new_meta = meta.meta().to_string() + " processed";
//let processed = meta.process(new_meta);
//Ok(processed)
//},
//)
//.unwrap();
#[test]
fn simple() {
let dir = tempfile::tempdir().unwrap();
let mut options = EnvOpenOptions::new();
options.map_size(4096 * 100);
let update_store = UpdateStore::open(options, dir, |meta: Processing<String>, _content: &_| -> Result<_, Failed<_, ()>> {
let new_meta = meta.meta().to_string() + " processed";
let processed = meta.process(new_meta);
Ok(processed)
}).unwrap();
//let meta = String::from("kiki");
//let update = update_store.register_update(meta, &[]).unwrap();
//thread::sleep(Duration::from_millis(100));
//let meta = update_store.meta(update.id()).unwrap().unwrap();
//if let UpdateStatus::Processed(Processed { success, .. }) = meta {
//assert_eq!(success, "kiki processed");
//} else {
//panic!()
//}
//}
let meta = String::from("kiki");
let update = update_store.register_update(meta, &[]).unwrap();
thread::sleep(Duration::from_millis(100));
let meta = update_store.meta(update.id()).unwrap().unwrap();
if let UpdateStatus::Processed(Processed { success, .. }) = meta {
assert_eq!(success, "kiki processed");
} else {
panic!()
}
}
//#[test]
//#[ignore]
//fn long_running_update() {
//let dir = tempfile::tempdir().unwrap();
//let mut options = EnvOpenOptions::new();
//options.map_size(4096 * 100);
//let update_store = UpdateStore::open(
//options,
//dir,
//|meta: Processing<String>, _content: &_| -> Result<_, Failed<_, ()>> {
//thread::sleep(Duration::from_millis(400));
//let new_meta = meta.meta().to_string() + "processed";
//let processed = meta.process(new_meta);
//Ok(processed)
//},
//)
//.unwrap();
#[test]
#[ignore]
fn long_running_update() {
let dir = tempfile::tempdir().unwrap();
let mut options = EnvOpenOptions::new();
options.map_size(4096 * 100);
let update_store = UpdateStore::open(options, dir, |meta: Processing<String>, _content:&_| -> Result<_, Failed<_, ()>> {
thread::sleep(Duration::from_millis(400));
let new_meta = meta.meta().to_string() + "processed";
let processed = meta.process(new_meta);
Ok(processed)
}).unwrap();
//let before_register = Instant::now();
let before_register = Instant::now();
//let meta = String::from("kiki");
//let update_kiki = update_store.register_update(meta, &[]).unwrap();
//assert!(before_register.elapsed() < Duration::from_millis(200));
let meta = String::from("kiki");
let update_kiki = update_store.register_update(meta, &[]).unwrap();
assert!(before_register.elapsed() < Duration::from_millis(200));
//let meta = String::from("coco");
//let update_coco = update_store.register_update(meta, &[]).unwrap();
//assert!(before_register.elapsed() < Duration::from_millis(200));
let meta = String::from("coco");
let update_coco = update_store.register_update(meta, &[]).unwrap();
assert!(before_register.elapsed() < Duration::from_millis(200));
//let meta = String::from("cucu");
//let update_cucu = update_store.register_update(meta, &[]).unwrap();
//assert!(before_register.elapsed() < Duration::from_millis(200));
let meta = String::from("cucu");
let update_cucu = update_store.register_update(meta, &[]).unwrap();
assert!(before_register.elapsed() < Duration::from_millis(200));
//thread::sleep(Duration::from_millis(400 * 3 + 100));
thread::sleep(Duration::from_millis(400 * 3 + 100));
//let meta = update_store.meta(update_kiki.id()).unwrap().unwrap();
//if let UpdateStatus::Processed(Processed { success, .. }) = meta {
//assert_eq!(success, "kiki processed");
//} else {
//panic!()
//}
let meta = update_store.meta(update_kiki.id()).unwrap().unwrap();
if let UpdateStatus::Processed(Processed { success, .. }) = meta {
assert_eq!(success, "kiki processed");
} else {
panic!()
}
//let meta = update_store.meta(update_coco.id()).unwrap().unwrap();
//if let UpdateStatus::Processed(Processed { success, .. }) = meta {
//assert_eq!(success, "coco processed");
//} else {
//panic!()
//}
let meta = update_store.meta(update_coco.id()).unwrap().unwrap();
if let UpdateStatus::Processed(Processed { success, .. }) = meta {
assert_eq!(success, "coco processed");
} else {
panic!()
}
let meta = update_store.meta(update_cucu.id()).unwrap().unwrap();
if let UpdateStatus::Processed(Processed { success, .. }) = meta {
assert_eq!(success, "cucu processed");
} else {
panic!()
}
}
}
//let meta = update_store.meta(update_cucu.id()).unwrap().unwrap();
//if let UpdateStatus::Processed(Processed { success, .. }) = meta {
//assert_eq!(success, "cucu processed");
//} else {
//panic!()
//}
//}
//}

13
meilisearch-http/src/index_controller/updates.rs
View File

@ -1,5 +1,6 @@
use chrono::{Utc, DateTime};
use serde::{Serialize, Deserialize};
use chrono::{DateTime, Utc};
use serde::{Deserialize, Serialize};
use uuid::Uuid;
#[derive(Debug, PartialEq, Eq, Hash, Serialize, Deserialize, Clone)]
#[serde(rename_all = "camelCase")]
@ -7,14 +8,16 @@ pub struct Pending<M> {
pub update_id: u64,
pub meta: M,
pub enqueued_at: DateTime<Utc>,
pub index_uuid: Uuid,
}
impl<M> Pending<M> {
pub fn new(meta: M, update_id: u64) -> Self {
pub fn new(meta: M, update_id: u64, index_uuid: Uuid) -> Self {
Self {
enqueued_at: Utc::now(),
meta,
update_id,
index_uuid,
}
}
@ -73,6 +76,10 @@ impl<M> Processing<M> {
self.from.meta()
}
pub fn index_uuid(&self) -> &Uuid {
&self.from.index_uuid
}
pub fn process<N>(self, meta: N) -> Processed<M, N> {
Processed {
success: meta,

295
meilisearch-http/src/index_controller/uuid_resolver.rs Normal file
View File

@ -0,0 +1,295 @@
use std::{fs::create_dir_all, path::Path};
use heed::{
types::{ByteSlice, Str},
Database, Env, EnvOpenOptions,
};
use log::{info, warn};
use thiserror::Error;
use tokio::sync::{mpsc, oneshot};
use uuid::Uuid;
pub type Result<T> = std::result::Result<T, UuidError>;
#[derive(Debug)]
enum UuidResolveMsg {
Resolve {
uid: String,
ret: oneshot::Sender<Result<Uuid>>,
},
GetOrCreate {
uid: String,
ret: oneshot::Sender<Result<Uuid>>,
},
Create {
uid: String,
ret: oneshot::Sender<Result<Uuid>>,
},
Delete {
uid: String,
ret: oneshot::Sender<Result<Uuid>>,
},
List {
ret: oneshot::Sender<Result<Vec<(String, Uuid)>>>,
},
}
struct UuidResolverActor<S> {
inbox: mpsc::Receiver<UuidResolveMsg>,
store: S,
}
impl<S: UuidStore> UuidResolverActor<S> {
fn new(inbox: mpsc::Receiver<UuidResolveMsg>, store: S) -> Self {
Self { inbox, store }
}
async fn run(mut self) {
use UuidResolveMsg::*;
info!("uuid resolver started");
loop {
match self.inbox.recv().await {
Some(Create { uid: name, ret }) => {
let _ = ret.send(self.handle_create(name).await);
}
Some(GetOrCreate { uid: name, ret }) => {
let _ = ret.send(self.handle_get_or_create(name).await);
}
Some(Resolve { uid: name, ret }) => {
let _ = ret.send(self.handle_resolve(name).await);
}
Some(Delete { uid: name, ret }) => {
let _ = ret.send(self.handle_delete(name).await);
}
Some(List { ret }) => {
let _ = ret.send(self.handle_list().await);
}
// all senders have been dropped, need to quit.
None => break,
}
}
warn!("exiting uuid resolver loop");
}
async fn handle_create(&self, uid: String) -> Result<Uuid> {
if !is_index_uid_valid(&uid) {
return Err(UuidError::BadlyFormatted(uid));
}
self.store.create_uuid(uid, true).await
}
async fn handle_get_or_create(&self, uid: String) -> Result<Uuid> {
if !is_index_uid_valid(&uid) {
return Err(UuidError::BadlyFormatted(uid));
}
self.store.create_uuid(uid, false).await
}
async fn handle_resolve(&self, uid: String) -> Result<Uuid> {
self.store
.get_uuid(uid.clone())
.await?
.ok_or(UuidError::UnexistingIndex(uid))
}
async fn handle_delete(&self, uid: String) -> Result<Uuid> {
self.store
.delete(uid.clone())
.await?
.ok_or(UuidError::UnexistingIndex(uid))
}
async fn handle_list(&self) -> Result<Vec<(String, Uuid)>> {
let result = self.store.list().await?;
Ok(result)
}
}
fn is_index_uid_valid(uid: &str) -> bool {
uid.chars()
.all(|x| x.is_ascii_alphanumeric() || x == '-' || x == '_')
}
#[derive(Clone)]
pub struct UuidResolverHandle {
sender: mpsc::Sender<UuidResolveMsg>,
}
impl UuidResolverHandle {
pub fn new(path: impl AsRef<Path>) -> anyhow::Result<Self> {
let (sender, reveiver) = mpsc::channel(100);
let store = HeedUuidStore::new(path)?;
let actor = UuidResolverActor::new(reveiver, store);
tokio::spawn(actor.run());
Ok(Self { sender })
}
pub async fn resolve(&self, name: String) -> anyhow::Result<Uuid> {
let (ret, receiver) = oneshot::channel();
let msg = UuidResolveMsg::Resolve { uid: name, ret };
let _ = self.sender.send(msg).await;
Ok(receiver
.await
.expect("Uuid resolver actor has been killed")?)
}
pub async fn get_or_create(&self, name: String) -> Result<Uuid> {
let (ret, receiver) = oneshot::channel();
let msg = UuidResolveMsg::GetOrCreate { uid: name, ret };
let _ = self.sender.send(msg).await;
Ok(receiver
.await
.expect("Uuid resolver actor has been killed")?)
}
pub async fn create(&self, name: String) -> anyhow::Result<Uuid> {
let (ret, receiver) = oneshot::channel();
let msg = UuidResolveMsg::Create { uid: name, ret };
let _ = self.sender.send(msg).await;
Ok(receiver
.await
.expect("Uuid resolver actor has been killed")?)
}
pub async fn delete(&self, name: String) -> anyhow::Result<Uuid> {
let (ret, receiver) = oneshot::channel();
let msg = UuidResolveMsg::Delete { uid: name, ret };
let _ = self.sender.send(msg).await;
Ok(receiver
.await
.expect("Uuid resolver actor has been killed")?)
}
pub async fn list(&self) -> anyhow::Result<Vec<(String, Uuid)>> {
let (ret, receiver) = oneshot::channel();
let msg = UuidResolveMsg::List { ret };
let _ = self.sender.send(msg).await;
Ok(receiver
.await
.expect("Uuid resolver actor has been killed")?)
}
}
#[derive(Debug, Error)]
pub enum UuidError {
#[error("Name already exist.")]
NameAlreadyExist,
#[error("Index \"{0}\" doesn't exist.")]
UnexistingIndex(String),
#[error("Error performing task: {0}")]
TokioTask(#[from] tokio::task::JoinError),
#[error("Database error: {0}")]
Heed(#[from] heed::Error),
#[error("Uuid error: {0}")]
Uuid(#[from] uuid::Error),
#[error("Badly formatted index uid: {0}")]
BadlyFormatted(String),
}
#[async_trait::async_trait]
trait UuidStore {
// Create a new entry for `name`. Return an error if `err` and the entry already exists, return
// the uuid otherwise.
async fn create_uuid(&self, uid: String, err: bool) -> Result<Uuid>;
async fn get_uuid(&self, uid: String) -> Result<Option<Uuid>>;
async fn delete(&self, uid: String) -> Result<Option<Uuid>>;
async fn list(&self) -> Result<Vec<(String, Uuid)>>;
}
struct HeedUuidStore {
env: Env,
db: Database<Str, ByteSlice>,
}
impl HeedUuidStore {
fn new(path: impl AsRef<Path>) -> anyhow::Result<Self> {
let path = path.as_ref().join("index_uuids");
create_dir_all(&path)?;
let mut options = EnvOpenOptions::new();
options.map_size(1_073_741_824); // 1GB
let env = options.open(path)?;
let db = env.create_database(None)?;
Ok(Self { env, db })
}
}
#[async_trait::async_trait]
impl UuidStore for HeedUuidStore {
async fn create_uuid(&self, name: String, err: bool) -> Result<Uuid> {
let env = self.env.clone();
let db = self.db;
tokio::task::spawn_blocking(move || {
let mut txn = env.write_txn()?;
match db.get(&txn, &name)? {
Some(uuid) => {
if err {
Err(UuidError::NameAlreadyExist)
} else {
let uuid = Uuid::from_slice(uuid)?;
Ok(uuid)
}
}
None => {
let uuid = Uuid::new_v4();
db.put(&mut txn, &name, uuid.as_bytes())?;
txn.commit()?;
Ok(uuid)
}
}
})
.await?
}
async fn get_uuid(&self, name: String) -> Result<Option<Uuid>> {
let env = self.env.clone();
let db = self.db;
tokio::task::spawn_blocking(move || {
let txn = env.read_txn()?;
match db.get(&txn, &name)? {
Some(uuid) => {
let uuid = Uuid::from_slice(uuid)?;
Ok(Some(uuid))
}
None => Ok(None),
}
})
.await?
}
async fn delete(&self, uid: String) -> Result<Option<Uuid>> {
let env = self.env.clone();
let db = self.db;
tokio::task::spawn_blocking(move || {
let mut txn = env.write_txn()?;
match db.get(&txn, &uid)? {
Some(uuid) => {
let uuid = Uuid::from_slice(uuid)?;
db.delete(&mut txn, &uid)?;
txn.commit()?;
Ok(Some(uuid))
}
None => Ok(None),
}
})
.await?
}
async fn list(&self) -> Result<Vec<(String, Uuid)>> {
let env = self.env.clone();
let db = self.db;
tokio::task::spawn_blocking(move || {
let txn = env.read_txn()?;
let mut entries = Vec::new();
for entry in db.iter(&txn)? {
let (name, uuid) = entry?;
let uuid = Uuid::from_slice(uuid)?;
entries.push((name.to_owned(), uuid))
}
Ok(entries)
})
.await?
}
}

99
meilisearch-http/src/lib.rs
View File

@ -1,60 +1,59 @@
#![allow(clippy::or_fun_call)]
pub mod data;
pub mod error;
pub mod helpers;
mod index;
mod index_controller;
pub mod option;
pub mod routes;
mod index_controller;
use actix_http::Error;
use actix_service::ServiceFactory;
use actix_web::{dev, web, App};
pub use option::Opt;
pub use self::data::Data;
use self::error::payload_error_handler;
pub use option::Opt;
pub fn create_app(
data: &Data,
enable_frontend: bool,
) -> App<
impl ServiceFactory<
Config = (),
Request = dev::ServiceRequest,
Response = dev::ServiceResponse<actix_http::body::Body>,
Error = Error,
InitError = (),
>,
actix_http::body::Body,
> {
let app = App::new()
.data(data.clone())
.app_data(
web::JsonConfig::default()
.limit(data.http_payload_size_limit())
.content_type(|_mime| true) // Accept all mime types
.error_handler(|err, _req| payload_error_handler(err).into()),
)
.app_data(
web::QueryConfig::default()
.error_handler(|err, _req| payload_error_handler(err).into())
)
.configure(routes::document::services)
.configure(routes::index::services)
.configure(routes::search::services)
.configure(routes::settings::services)
.configure(routes::stop_words::services)
.configure(routes::synonym::services)
.configure(routes::health::services)
.configure(routes::stats::services)
.configure(routes::key::services);
#[macro_export]
macro_rules! create_app {
($data:expr, $enable_frontend:expr) => {{
use actix_cors::Cors;
use actix_web::middleware::TrailingSlash;
use actix_web::App;
use actix_web::{middleware, web};
use meilisearch_http::error::payload_error_handler;
use meilisearch_http::routes::*;
let app = App::new()
.data($data.clone())
.app_data(
web::JsonConfig::default()
.limit($data.http_payload_size_limit())
.content_type(|_mime| true) // Accept all mime types
.error_handler(|err, _req| payload_error_handler(err).into()),
)
.app_data(
web::QueryConfig::default()
.error_handler(|err, _req| payload_error_handler(err).into()),
)
.configure(document::services)
.configure(index::services)
.configure(search::services)
.configure(settings::services)
.configure(stop_words::services)
.configure(synonym::services)
.configure(health::services)
.configure(stats::services)
.configure(key::services);
//.configure(routes::dump::services);
if enable_frontend {
app
.service(routes::load_html)
.service(routes::load_css)
} else {
app
}
let app = if $enable_frontend {
app.service(load_html).service(load_css)
} else {
app
};
app.wrap(
Cors::default()
.send_wildcard()
.allowed_headers(vec!["content-type", "x-meili-api-key"])
.max_age(86_400), // 24h
)
.wrap(middleware::Logger::default())
.wrap(middleware::Compress::default())
.wrap(middleware::NormalizePath::new(TrailingSlash::Trim))
}};
}

46
meilisearch-http/src/main.rs
View File

@ -1,9 +1,7 @@
use std::env;
use actix_cors::Cors;
use actix_web::{middleware, HttpServer};
use actix_web::HttpServer;
use main_error::MainError;
use meilisearch_http::helpers::NormalizePath;
use meilisearch_http::{create_app, Data, Opt};
use structopt::StructOpt;
@ -46,46 +44,49 @@ async fn main() -> Result<(), MainError> {
}
}
"development" => {
env_logger::Builder::from_env(env_logger::Env::default().default_filter_or("info")).init();
env_logger::Builder::from_env(env_logger::Env::default().default_filter_or("info"))
.init();
}
_ => unreachable!(),
}
//if let Some(path) = &opt.import_snapshot {
//snapshot::load_snapshot(&opt.db_path, path, opt.ignore_snapshot_if_db_exists, opt.ignore_missing_snapshot)?;
//snapshot::load_snapshot(&opt.db_path, path, opt.ignore_snapshot_if_db_exists, opt.ignore_missing_snapshot)?;
//}
let data = Data::new(opt.clone())?;
//if !opt.no_analytics {
//let analytics_data = data.clone();
//let analytics_opt = opt.clone();
//thread::spawn(move || analytics::analytics_sender(analytics_data, analytics_opt));
//let analytics_data = data.clone();
//let analytics_opt = opt.clone();
//thread::spawn(move || analytics::analytics_sender(analytics_data, analytics_opt));
//}
//if let Some(path) = &opt.import_dump {
//dump::import_dump(&data, path, opt.dump_batch_size)?;
//dump::import_dump(&data, path, opt.dump_batch_size)?;
//}
//if opt.schedule_snapshot {
//snapshot::schedule_snapshot(data.clone(), &opt.snapshot_dir, opt.snapshot_interval_sec.unwrap_or(86400))?;
//snapshot::schedule_snapshot(data.clone(), &opt.snapshot_dir, opt.snapshot_interval_sec.unwrap_or(86400))?;
//}
print_launch_resume(&opt, &data);
let enable_frontend = opt.env != "production";
let http_server = HttpServer::new(move || {
create_app(&data, enable_frontend)
.wrap(
Cors::default()
.send_wildcard()
.allowed_headers(vec!["content-type", "x-meili-api-key"])
.max_age(86_400) // 24h
)
.wrap(middleware::Logger::default())
.wrap(middleware::Compress::default())
.wrap(NormalizePath)
});
run_http(data, opt, enable_frontend).await?;
Ok(())
}
async fn run_http(
data: Data,
opt: Opt,
enable_frontend: bool,
) -> Result<(), Box<dyn std::error::Error>> {
let http_server = HttpServer::new(move || create_app!(&data, enable_frontend))
// Disable signals allows the server to terminate immediately when a user enter CTRL-C
.disable_signals();
if let Some(config) = opt.get_ssl_config()? {
http_server
@ -95,7 +96,6 @@ async fn main() -> Result<(), MainError> {
} else {
http_server.bind(opt.http_addr)?.run().await?;
}
Ok(())
}

10
meilisearch-http/src/option.rs
View File

@ -1,15 +1,15 @@
use std::{error, fs};
use std::io::{BufReader, Read};
use std::path::PathBuf;
use std::sync::Arc;
use std::{error, fs};
use byte_unit::Byte;
use grenad::CompressionType;
use rustls::internal::pemfile::{certs, pkcs8_private_keys, rsa_private_keys};
use rustls::{
AllowAnyAnonymousOrAuthenticatedClient, AllowAnyAuthenticatedClient, NoClientAuth,
RootCertStore,
};
use grenad::CompressionType;
use structopt::StructOpt;
#[derive(Debug, Clone, StructOpt)]
@ -99,7 +99,11 @@ pub struct Opt {
/// The Sentry DSN to use for error reporting. This defaults to the MeiliSearch Sentry project.
/// You can disable sentry all together using the `--no-sentry` flag or `MEILI_NO_SENTRY` environment variable.
#[cfg(all(not(debug_assertions), feature = "sentry"))]
#[structopt(long, env = "SENTRY_DSN", default_value = "https://5ddfa22b95f241198be2271aaf028653@sentry.io/3060337")]
#[structopt(
long,
env = "SENTRY_DSN",
default_value = "https://5ddfa22b95f241198be2271aaf028653@sentry.io/3060337"
)]
pub sentry_dsn: String,
/// Disable Sentry error reporting.

51
meilisearch-http/src/routes/document.rs
View File

@ -7,10 +7,10 @@ use milli::update::{IndexDocumentsMethod, UpdateFormat};
use serde::Deserialize;
use serde_json::Value;
use crate::Data;
use crate::error::ResponseError;
use crate::helpers::Authentication;
use crate::routes::IndexParam;
use crate::Data;
const DEFAULT_RETRIEVE_DOCUMENTS_OFFSET: usize = 0;
const DEFAULT_RETRIEVE_DOCUMENTS_LIMIT: usize = 20;
@ -19,7 +19,10 @@ macro_rules! guard_content_type {
($fn_name:ident, $guard_value:literal) => {
fn $fn_name(head: &actix_web::dev::RequestHead) -> bool {
if let Some(content_type) = head.headers.get("Content-Type") {
content_type.to_str().map(|v| v.contains($guard_value)).unwrap_or(false)
content_type
.to_str()
.map(|v| v.contains($guard_value))
.unwrap_or(false)
} else {
false
}
@ -57,7 +60,10 @@ async fn get_document(
) -> Result<HttpResponse, ResponseError> {
let index = path.index_uid.clone();
let id = path.document_id.clone();
match data.retrieve_document(index, id, None as Option<Vec<String>>).await {
match data
.retrieve_document(index, id, None as Option<Vec<String>>)
.await
{
Ok(document) => {
let json = serde_json::to_string(&document).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -76,7 +82,10 @@ async fn delete_document(
data: web::Data<Data>,
path: web::Path<DocumentParam>,
) -> Result<HttpResponse, ResponseError> {
match data.delete_documents(path.index_uid.clone(), vec![path.document_id.clone()]).await {
match data
.delete_documents(path.index_uid.clone(), vec![path.document_id.clone()])
.await
{
Ok(result) => {
let json = serde_json::to_string(&result).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -104,16 +113,17 @@ async fn get_all_documents(
let attributes_to_retrieve = params
.attributes_to_retrieve
.as_ref()
.map(|attrs| attrs
.split(",")
.map(String::from)
.collect::<Vec<_>>());
.map(|attrs| attrs.split(',').map(String::from).collect::<Vec<_>>());
match data.retrieve_documents(
path.index_uid.clone(),
params.offset.unwrap_or(DEFAULT_RETRIEVE_DOCUMENTS_OFFSET),
params.limit.unwrap_or(DEFAULT_RETRIEVE_DOCUMENTS_LIMIT),
attributes_to_retrieve).await {
match data
.retrieve_documents(
path.index_uid.clone(),
params.offset.unwrap_or(DEFAULT_RETRIEVE_DOCUMENTS_OFFSET),
params.limit.unwrap_or(DEFAULT_RETRIEVE_DOCUMENTS_LIMIT),
attributes_to_retrieve,
)
.await
{
Ok(docs) => {
let json = serde_json::to_string(&docs).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -149,7 +159,8 @@ async fn add_documents_json(
UpdateFormat::Json,
body,
params.primary_key.clone(),
).await;
)
.await;
match addition_result {
Ok(update) => {
@ -163,7 +174,6 @@ async fn add_documents_json(
}
}
/// Default route for adding documents, this should return an error and redirect to the documentation
#[post("/indexes/{index_uid}/documents", wrap = "Authentication::Private")]
async fn add_documents_default(
@ -191,7 +201,7 @@ async fn update_documents_default(
#[put(
"/indexes/{index_uid}/documents",
wrap = "Authentication::Private",
guard = "guard_json",
guard = "guard_json"
)]
async fn update_documents(
data: web::Data<Data>,
@ -206,7 +216,8 @@ async fn update_documents(
UpdateFormat::Json,
body,
params.primary_key.clone(),
).await;
)
.await;
match addition_result {
Ok(update) => {
@ -231,7 +242,11 @@ async fn delete_documents(
) -> Result<HttpResponse, ResponseError> {
let ids = body
.iter()
.map(|v| v.as_str().map(String::from).unwrap_or_else(|| v.to_string()))
.map(|v| {
v.as_str()
.map(String::from)
.unwrap_or_else(|| v.to_string())
})
.collect();
match data.delete_documents(path.index_uid.clone(), ids).await {

36
meilisearch-http/src/routes/index.rs
View File

@ -3,10 +3,10 @@ use actix_web::{web, HttpResponse};
use chrono::{DateTime, Utc};
use serde::{Deserialize, Serialize};
use crate::Data;
use crate::error::ResponseError;
use crate::helpers::Authentication;
use crate::routes::IndexParam;
use crate::Data;
pub fn services(cfg: &mut web::ServiceConfig) {
cfg.service(list_indexes)
@ -18,10 +18,9 @@ pub fn services(cfg: &mut web::ServiceConfig) {
.service(get_all_updates_status);
}
#[get("/indexes", wrap = "Authentication::Private")]
async fn list_indexes(data: web::Data<Data>) -> Result<HttpResponse, ResponseError> {
match data.list_indexes() {
match data.list_indexes().await {
Ok(indexes) => {
let json = serde_json::to_string(&indexes).unwrap();
Ok(HttpResponse::Ok().body(&json))
@ -37,13 +36,12 @@ async fn get_index(
data: web::Data<Data>,
path: web::Path<IndexParam>,
) -> Result<HttpResponse, ResponseError> {
match data.index(&path.index_uid)? {
Some(meta) => {
match data.index(path.index_uid.clone()).await {
Ok(meta) => {
let json = serde_json::to_string(&meta).unwrap();
Ok(HttpResponse::Ok().body(json))
}
None => {
let e = format!("Index {:?} doesn't exist.", path.index_uid);
Err(e) => {
Ok(HttpResponse::BadRequest().body(serde_json::json!({ "error": e.to_string() })))
}
}
@ -61,7 +59,8 @@ async fn create_index(
data: web::Data<Data>,
body: web::Json<IndexCreateRequest>,
) -> Result<HttpResponse, ResponseError> {
match data.create_index(&body.uid, body.primary_key.clone()) {
let body = body.into_inner();
match data.create_index(body.uid, body.primary_key).await {
Ok(meta) => {
let json = serde_json::to_string(&meta).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -75,7 +74,7 @@ async fn create_index(
#[derive(Debug, Deserialize)]
#[serde(rename_all = "camelCase", deny_unknown_fields)]
struct UpdateIndexRequest {
name: Option<String>,
uid: Option<String>,
primary_key: Option<String>,
}
@ -95,7 +94,11 @@ async fn update_index(
path: web::Path<IndexParam>,
body: web::Json<UpdateIndexRequest>,
) -> Result<HttpResponse, ResponseError> {
match data.update_index(&path.index_uid, body.primary_key.as_ref(), body.name.as_ref()) {
let body = body.into_inner();
match data
.update_index(path.into_inner().index_uid, body.primary_key, body.uid)
.await
{
Ok(meta) => {
let json = serde_json::to_string(&meta).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -133,16 +136,15 @@ async fn get_update_status(
data: web::Data<Data>,
path: web::Path<UpdateParam>,
) -> Result<HttpResponse, ResponseError> {
let result = data.get_update_status(&path.index_uid, path.update_id);
let params = path.into_inner();
let result = data
.get_update_status(params.index_uid, params.update_id)
.await;
match result {
Ok(Some(meta)) => {
Ok(meta) => {
let json = serde_json::to_string(&meta).unwrap();
Ok(HttpResponse::Ok().body(json))
}
Ok(None) => {
let e = format!("udpate {} for index {:?} doesn't exists.", path.update_id, path.index_uid);
Ok(HttpResponse::BadRequest().body(serde_json::json!({ "error": e.to_string() })))
}
Err(e) => {
Ok(HttpResponse::BadRequest().body(serde_json::json!({ "error": e.to_string() })))
}
@ -154,7 +156,7 @@ async fn get_all_updates_status(
data: web::Data<Data>,
path: web::Path<IndexParam>,
) -> Result<HttpResponse, ResponseError> {
let result = data.get_updates_status(&path.index_uid);
let result = data.get_updates_status(path.into_inner().index_uid).await;
match result {
Ok(metas) => {
let json = serde_json::to_string(&metas).unwrap();

4
meilisearch-http/src/routes/key.rs
View File

@ -1,6 +1,6 @@
use actix_web::get;
use actix_web::web;
use actix_web::HttpResponse;
use actix_web::get;
use serde::Serialize;
use crate::helpers::Authentication;
@ -19,7 +19,7 @@ struct KeysResponse {
#[get("/keys", wrap = "Authentication::Admin")]
async fn list(data: web::Data<Data>) -> HttpResponse {
let api_keys = data.api_keys.clone();
HttpResponse::Ok().json(KeysResponse {
HttpResponse::Ok().json(&KeysResponse {
private: api_keys.private,
public: api_keys.public,
})

20
meilisearch-http/src/routes/search.rs
View File

@ -4,9 +4,9 @@ use std::convert::{TryFrom, TryInto};
use actix_web::{get, post, web, HttpResponse};
use serde::Deserialize;
use crate::data::{SearchQuery, DEFAULT_SEARCH_LIMIT};
use crate::error::ResponseError;
use crate::helpers::Authentication;
use crate::index::{SearchQuery, DEFAULT_SEARCH_LIMIT};
use crate::routes::IndexParam;
use crate::Data;
@ -36,19 +36,19 @@ impl TryFrom<SearchQueryGet> for SearchQuery {
fn try_from(other: SearchQueryGet) -> anyhow::Result<Self> {
let attributes_to_retrieve = other
.attributes_to_retrieve
.map(|attrs| attrs.split(",").map(String::from).collect::<Vec<_>>());
.map(|attrs| attrs.split(',').map(String::from).collect::<Vec<_>>());
let attributes_to_crop = other
.attributes_to_crop
.map(|attrs| attrs.split(",").map(String::from).collect::<Vec<_>>());
.map(|attrs| attrs.split(',').map(String::from).collect::<Vec<_>>());
let attributes_to_highlight = other
.attributes_to_highlight
.map(|attrs| attrs.split(",").map(String::from).collect::<HashSet<_>>());
.map(|attrs| attrs.split(',').map(String::from).collect::<HashSet<_>>());
let facet_distributions = other
.facet_distributions
.map(|attrs| attrs.split(",").map(String::from).collect::<Vec<_>>());
.map(|attrs| attrs.split(',').map(String::from).collect::<Vec<_>>());
let facet_filters = match other.facet_filters {
Some(ref f) => Some(serde_json::from_str(f)?),
@ -80,10 +80,12 @@ async fn search_with_url_query(
let query: SearchQuery = match params.into_inner().try_into() {
Ok(q) => q,
Err(e) => {
return Ok(HttpResponse::BadRequest().body(serde_json::json!({ "error": e.to_string() })))
return Ok(
HttpResponse::BadRequest().body(serde_json::json!({ "error": e.to_string() }))
)
}
};
let search_result = data.search(&path.index_uid, query);
let search_result = data.search(path.into_inner().index_uid, query).await;
match search_result {
Ok(docs) => {
let docs = serde_json::to_string(&docs).unwrap();
@ -101,7 +103,9 @@ async fn search_with_post(
path: web::Path<IndexParam>,
params: web::Json<SearchQuery>,
) -> Result<HttpResponse, ResponseError> {
let search_result = data.search(&path.index_uid, params.into_inner());
let search_result = data
.search(path.into_inner().index_uid, params.into_inner())
.await;
match search_result {
Ok(docs) => {
let docs = serde_json::to_string(&docs).unwrap();

48
meilisearch-http/src/routes/settings/mod.rs
View File

@ -1,9 +1,9 @@
use actix_web::{web, HttpResponse, delete, get, post};
use actix_web::{delete, get, post, web, HttpResponse};
use crate::Data;
use crate::error::ResponseError;
use crate::index_controller::Settings;
use crate::helpers::Authentication;
use crate::index::Settings;
use crate::Data;
#[macro_export]
macro_rules! make_setting_route {
@ -14,14 +14,14 @@ macro_rules! make_setting_route {
use crate::data;
use crate::error::ResponseError;
use crate::helpers::Authentication;
use crate::index_controller::Settings;
use crate::index::Settings;
#[actix_web::delete($route, wrap = "Authentication::Private")]
pub async fn delete(
data: web::Data<data::Data>,
index_uid: web::Path<String>,
) -> Result<HttpResponse, ResponseError> {
use crate::index_controller::Settings;
use crate::index::Settings;
let settings = Settings {
$attr: Some(None),
..Default::default()
@ -64,7 +64,7 @@ macro_rules! make_setting_route {
data: actix_web::web::Data<data::Data>,
index_uid: actix_web::web::Path<String>,
) -> std::result::Result<HttpResponse, ResponseError> {
match data.settings(index_uid.as_ref()) {
match data.settings(index_uid.into_inner()).await {
Ok(settings) => {
let setting = settings.$attr;
let json = serde_json::to_string(&setting).unwrap();
@ -82,7 +82,7 @@ macro_rules! make_setting_route {
make_setting_route!(
"/indexes/{index_uid}/settings/attributes-for-faceting",
std::collections::HashMap<String, String>,
faceted_attributes
attributes_for_faceting
);
make_setting_route!(
@ -98,16 +98,16 @@ make_setting_route!(
);
//make_setting_route!(
//"/indexes/{index_uid}/settings/distinct-attribute",
//String,
//distinct_attribute
//"/indexes/{index_uid}/settings/distinct-attribute",
//String,
//distinct_attribute
//);
make_setting_route!(
"/indexes/{index_uid}/settings/ranking-rules",
Vec<String>,
ranking_rules
);
//make_setting_route!(
//"/indexes/{index_uid}/settings/ranking-rules",
//Vec<String>,
//ranking_rules
//);
macro_rules! create_services {
($($mod:ident),*) => {
@ -126,10 +126,9 @@ macro_rules! create_services {
}
create_services!(
faceted_attributes,
attributes_for_faceting,
displayed_attributes,
searchable_attributes,
ranking_rules
searchable_attributes
);
#[post("/indexes/{index_uid}/settings", wrap = "Authentication::Private")]
@ -138,7 +137,10 @@ async fn update_all(
index_uid: web::Path<String>,
body: web::Json<Settings>,
) -> Result<HttpResponse, ResponseError> {
match data.update_settings(index_uid.into_inner(), body.into_inner(), true).await {
match data
.update_settings(index_uid.into_inner(), body.into_inner(), true)
.await
{
Ok(update_result) => {
let json = serde_json::to_string(&update_result).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -154,7 +156,7 @@ async fn get_all(
data: web::Data<Data>,
index_uid: web::Path<String>,
) -> Result<HttpResponse, ResponseError> {
match data.settings(index_uid.as_ref()) {
match data.settings(index_uid.into_inner()).await {
Ok(settings) => {
let json = serde_json::to_string(&settings).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -171,7 +173,10 @@ async fn delete_all(
index_uid: web::Path<String>,
) -> Result<HttpResponse, ResponseError> {
let settings = Settings::cleared();
match data.update_settings(index_uid.into_inner(), settings, false).await {
match data
.update_settings(index_uid.into_inner(), settings, false)
.await
{
Ok(update_result) => {
let json = serde_json::to_string(&update_result).unwrap();
Ok(HttpResponse::Ok().body(json))
@ -181,4 +186,3 @@ async fn delete_all(
}
}
}

4
meilisearch-http/src/routes/stats.rs
View File

@ -1,8 +1,8 @@
use std::collections::{HashMap, BTreeMap};
use std::collections::{BTreeMap, HashMap};
use actix_web::get;
use actix_web::web;
use actix_web::HttpResponse;
use actix_web::get;
use chrono::{DateTime, Utc};
use serde::Serialize;

2
meilisearch-http/src/routes/stop_words.rs
View File

@ -1,5 +1,5 @@
use actix_web::{web, HttpResponse};
use actix_web::{delete, get, post};
use actix_web::{web, HttpResponse};
use std::collections::BTreeSet;
use crate::error::ResponseError;

2
meilisearch-http/src/routes/synonym.rs
View File

@ -1,7 +1,7 @@
use std::collections::BTreeMap;
use actix_web::{web, HttpResponse};
use actix_web::{delete, get, post};
use actix_web::{web, HttpResponse};
use crate::error::ResponseError;
use crate::helpers::Authentication;

36
meilisearch-http/tests/common/index.rs
View File

@ -2,7 +2,7 @@ use std::time::Duration;
use actix_web::http::StatusCode;
use serde_json::{json, Value};
use tokio::time::delay_for;
use tokio::time::sleep;
use super::service::Service;
@ -19,7 +19,10 @@ impl Index<'_> {
pub async fn load_test_set(&self) -> u64 {
let url = format!("/indexes/{}/documents", self.uid);
let (response, code) = self.service.post_str(url, include_str!("../assets/test_set.json")).await;
let (response, code) = self
.service
.post_str(url, include_str!("../assets/test_set.json"))
.await;
assert_eq!(code, 200);
let update_id = response["updateId"].as_i64().unwrap();
self.wait_update_id(update_id as u64).await;
@ -60,7 +63,11 @@ impl Index<'_> {
self.service.post(url, documents).await
}
pub async fn update_documents(&self, documents: Value, primary_key: Option<&str>) -> (Value, StatusCode) {
pub async fn update_documents(
&self,
documents: Value,
primary_key: Option<&str>,
) -> (Value, StatusCode) {
let url = match primary_key {
Some(key) => format!("/indexes/{}/documents?primaryKey={}", self.uid, key),
None => format!("/indexes/{}/documents", self.uid),
@ -73,19 +80,19 @@ impl Index<'_> {
let url = format!("/indexes/{}/updates/{}", self.uid, update_id);
for _ in 0..10 {
let (response, status_code) = self.service.get(&url).await;
assert_eq!(status_code, 200);
assert_eq!(status_code, 200, "response: {}", response);
if response["status"] == "processed" || response["status"] == "failed" {
return response;
}
delay_for(Duration::from_secs(1)).await;
sleep(Duration::from_secs(1)).await;
}
panic!("Timeout waiting for update id");
}
pub async fn get_update(&self, udpate_id: u64) -> (Value, StatusCode) {
let url = format!("/indexes/{}/updates/{}", self.uid, udpate_id);
pub async fn get_update(&self, update_id: u64) -> (Value, StatusCode) {
let url = format!("/indexes/{}/updates/{}", self.uid, update_id);
self.service.get(url).await
}
@ -95,7 +102,11 @@ impl Index<'_> {
self.service.get(url).await
}
pub async fn get_document(&self, id: u64, _options: Option<GetDocumentOptions>) -> (Value, StatusCode) {
pub async fn get_document(
&self,
id: u64,
_options: Option<GetDocumentOptions>,
) -> (Value, StatusCode) {
let url = format!("/indexes/{}/documents/{}", self.uid, id);
self.service.get(url).await
}
@ -111,7 +122,10 @@ impl Index<'_> {
}
if let Some(attributes_to_retrieve) = options.attributes_to_retrieve {
url.push_str(&format!("attributesToRetrieve={}&", attributes_to_retrieve.join(",")));
url.push_str(&format!(
"attributesToRetrieve={}&",
attributes_to_retrieve.join(",")
));
}
self.service.get(url).await
@ -129,7 +143,9 @@ impl Index<'_> {
pub async fn delete_batch(&self, ids: Vec<u64>) -> (Value, StatusCode) {
let url = format!("/indexes/{}/documents/delete-batch", self.uid);
self.service.post(url, serde_json::to_value(&ids).unwrap()).await
self.service
.post(url, serde_json::to_value(&ids).unwrap())
.await
}
pub async fn settings(&self) -> (Value, StatusCode) {

2
meilisearch-http/tests/common/mod.rs
View File

@ -2,8 +2,8 @@ mod index;
mod server;
mod service;
pub use index::{GetAllDocumentsOptions, GetDocumentOptions};
pub use server::Server;
pub use index::{GetDocumentOptions, GetAllDocumentsOptions};
/// Performs a search test on both post and get routes
#[macro_export]

14
meilisearch-http/tests/common/server.rs
View File

@ -5,22 +5,24 @@ use tempdir::TempDir;
use urlencoding::encode;
use meilisearch_http::data::Data;
use meilisearch_http::option::{Opt, IndexerOpts};
use meilisearch_http::option::{IndexerOpts, Opt};
use super::index::Index;
use super::service::Service;
pub struct Server {
pub service: Service,
// hod ownership to the tempdir while we use the server instance.
_dir: tempdir::TempDir,
}
impl Server {
pub async fn new() -> Self {
let tmp_dir = TempDir::new("meilisearch").unwrap();
let dir = TempDir::new("meilisearch").unwrap();
let opt = Opt {
db_path: tmp_dir.path().join("db"),
dumps_dir: tmp_dir.path().join("dump"),
db_path: dir.path().join("db"),
dumps_dir: dir.path().join("dump"),
dump_batch_size: 16,
http_addr: "127.0.0.1:7700".to_owned(),
master_key: None,
@ -53,9 +55,7 @@ impl Server {
let data = Data::new(opt).unwrap();
let service = Service(data);
Server {
service,
}
Server { service, _dir: dir }
}
/// Returns a view to an index. There is no guarantee that the index exists.

25
meilisearch-http/tests/common/service.rs
View File

@ -1,15 +1,14 @@
use actix_web::{http::StatusCode, test};
use serde_json::Value;
use meilisearch_http::create_app;
use meilisearch_http::data::Data;
use meilisearch_http::helpers::NormalizePath;
pub struct Service(pub Data);
impl Service {
pub async fn post(&self, url: impl AsRef<str>, body: Value) -> (Value, StatusCode) {
let mut app =
test::init_service(meilisearch_http::create_app(&self.0, true).wrap(NormalizePath)).await;
let mut app = test::init_service(create_app!(&self.0, true)).await;
let req = test::TestRequest::post()
.uri(url.as_ref())
@ -24,14 +23,17 @@ impl Service {
}
/// Send a test post request from a text body, with a `content-type:application/json` header.
pub async fn post_str(&self, url: impl AsRef<str>, body: impl AsRef<str>) -> (Value, StatusCode) {
let mut app =
test::init_service(meilisearch_http::create_app(&self.0, true).wrap(NormalizePath)).await;
pub async fn post_str(
&self,
url: impl AsRef<str>,
body: impl AsRef<str>,
) -> (Value, StatusCode) {
let mut app = test::init_service(create_app!(&self.0, true)).await;
let req = test::TestRequest::post()
.uri(url.as_ref())
.set_payload(body.as_ref().to_string())
.header("content-type", "application/json")
.insert_header(("content-type", "application/json"))
.to_request();
let res = test::call_service(&mut app, req).await;
let status_code = res.status();
@ -42,8 +44,7 @@ impl Service {
}
pub async fn get(&self, url: impl AsRef<str>) -> (Value, StatusCode) {
let mut app =
test::init_service(meilisearch_http::create_app(&self.0, true).wrap(NormalizePath)).await;
let mut app = test::init_service(create_app!(&self.0, true)).await;
let req = test::TestRequest::get().uri(url.as_ref()).to_request();
let res = test::call_service(&mut app, req).await;
@ -55,8 +56,7 @@ impl Service {
}
pub async fn put(&self, url: impl AsRef<str>, body: Value) -> (Value, StatusCode) {
let mut app =
test::init_service(meilisearch_http::create_app(&self.0, true).wrap(NormalizePath)).await;
let mut app = test::init_service(create_app!(&self.0, true)).await;
let req = test::TestRequest::put()
.uri(url.as_ref())
@ -71,8 +71,7 @@ impl Service {
}
pub async fn delete(&self, url: impl AsRef<str>) -> (Value, StatusCode) {
let mut app =
test::init_service(meilisearch_http::create_app(&self.0, true).wrap(NormalizePath)).await;
let mut app = test::init_service(create_app!(&self.0, true)).await;
let req = test::TestRequest::delete().uri(url.as_ref()).to_request();
let res = test::call_service(&mut app, req).await;

41
meilisearch-http/tests/documents/add_documents.rs
View File

@ -1,7 +1,7 @@
use serde_json::{json, Value};
use chrono::DateTime;
use serde_json::{json, Value};
use crate::common::{Server, GetAllDocumentsOptions};
use crate::common::{GetAllDocumentsOptions, Server};
#[actix_rt::test]
async fn add_documents_no_index_creation() {
@ -28,14 +28,16 @@ async fn add_documents_no_index_creation() {
let (response, code) = index.get_update(0).await;
assert_eq!(code, 200);
println!("response: {}", response);
assert_eq!(response["status"], "processed");
assert_eq!(response["updateId"], 0);
assert_eq!(response["success"]["DocumentsAddition"]["nb_documents"], 1);
let processed_at = DateTime::parse_from_rfc3339(response["processedAt"].as_str().unwrap()).unwrap();
let enqueued_at = DateTime::parse_from_rfc3339(response["enqueuedAt"].as_str().unwrap()).unwrap();
let started_processing_at = DateTime::parse_from_rfc3339(response["startedProcessingAt"].as_str().unwrap()).unwrap();
let processed_at =
DateTime::parse_from_rfc3339(response["processedAt"].as_str().unwrap()).unwrap();
let enqueued_at =
DateTime::parse_from_rfc3339(response["enqueuedAt"].as_str().unwrap()).unwrap();
let started_processing_at =
DateTime::parse_from_rfc3339(response["startedProcessingAt"].as_str().unwrap()).unwrap();
assert!(processed_at > started_processing_at);
assert!(started_processing_at > enqueued_at);
@ -72,7 +74,8 @@ async fn document_addition_with_primary_key() {
"content": "foo",
}
]);
let (_response, code) = index.add_documents(documents, Some("primary")).await; assert_eq!(code, 200);
let (_response, code) = index.add_documents(documents, Some("primary")).await;
assert_eq!(code, 200);
index.wait_update_id(0).await;
@ -98,7 +101,8 @@ async fn document_update_with_primary_key() {
"content": "foo",
}
]);
let (_response, code) = index.update_documents(documents, Some("primary")).await; assert_eq!(code, 200);
let (_response, code) = index.update_documents(documents, Some("primary")).await;
assert_eq!(code, 200);
index.wait_update_id(0).await;
@ -159,7 +163,7 @@ async fn update_documents_with_primary_key_and_primary_key_already_exists() {
assert_eq!(code, 200);
index.wait_update_id(0).await;
let (response, code) = index.get_update(0).await;
let (response, code) = index.get_update(0).await;
assert_eq!(code, 200);
assert_eq!(response["status"], "processed");
assert_eq!(response["updateId"], 0);
@ -264,7 +268,10 @@ async fn update_document() {
let (response, code) = index.get_document(1, None).await;
assert_eq!(code, 200);
assert_eq!(response.to_string(), r##"{"doc_id":1,"content":"foo","other":"bar"}"##);
assert_eq!(
response.to_string(),
r##"{"doc_id":1,"content":"foo","other":"bar"}"##
);
}
#[actix_rt::test]
@ -276,7 +283,12 @@ async fn add_larger_dataset() {
assert_eq!(code, 200);
assert_eq!(response["status"], "processed");
assert_eq!(response["success"]["DocumentsAddition"]["nb_documents"], 77);
let (response, code) = index.get_all_documents(GetAllDocumentsOptions { limit: Some(1000), ..Default::default() }).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions {
limit: Some(1000),
..Default::default()
})
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 77);
}
@ -292,7 +304,12 @@ async fn update_larger_dataset() {
assert_eq!(code, 200);
assert_eq!(response["status"], "processed");
assert_eq!(response["success"]["DocumentsAddition"]["nb_documents"], 77);
let (response, code) = index.get_all_documents(GetAllDocumentsOptions { limit: Some(1000), ..Default::default() }).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions {
limit: Some(1000),
..Default::default()
})
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 77);
}

31
meilisearch-http/tests/documents/delete_documents.rs
View File

@ -1,6 +1,6 @@
use serde_json::json;
use crate::common::{Server, GetAllDocumentsOptions};
use crate::common::{GetAllDocumentsOptions, Server};
#[actix_rt::test]
async fn delete_one_document_unexisting_index() {
@ -24,7 +24,9 @@ async fn delete_one_unexisting_document() {
async fn delete_one_document() {
let server = Server::new().await;
let index = server.index("test");
index.add_documents(json!([{ "id": 0, "content": "foobar" }]), None).await;
index
.add_documents(json!([{ "id": 0, "content": "foobar" }]), None)
.await;
index.wait_update_id(0).await;
let (_response, code) = server.index("test").delete_document(0).await;
assert_eq!(code, 200);
@ -39,20 +41,26 @@ async fn clear_all_documents_unexisting_index() {
let server = Server::new().await;
let (_response, code) = server.index("test").clear_all_documents().await;
assert_eq!(code, 400);
}
#[actix_rt::test]
async fn clear_all_documents() {
let server = Server::new().await;
let index = server.index("test");
index.add_documents(json!([{ "id": 1, "content": "foobar" }, { "id": 0, "content": "foobar" }]), None).await;
index
.add_documents(
json!([{ "id": 1, "content": "foobar" }, { "id": 0, "content": "foobar" }]),
None,
)
.await;
index.wait_update_id(0).await;
let (_response, code) = index.clear_all_documents().await;
assert_eq!(code, 200);
let _update = index.wait_update_id(1).await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions::default())
.await;
assert_eq!(code, 200);
assert!(response.as_array().unwrap().is_empty());
}
@ -67,7 +75,9 @@ async fn clear_all_documents_empty_index() {
assert_eq!(code, 200);
let _update = index.wait_update_id(0).await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions::default())
.await;
assert_eq!(code, 200);
assert!(response.as_array().unwrap().is_empty());
}
@ -89,13 +99,14 @@ async fn delete_batch() {
assert_eq!(code, 200);
let _update = index.wait_update_id(1).await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions::default())
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 1);
assert_eq!(response.as_array().unwrap()[0]["id"], 3);
}
#[actix_rt::test]
async fn delete_no_document_batch() {
let server = Server::new().await;
@ -106,7 +117,9 @@ async fn delete_no_document_batch() {
assert_eq!(code, 200);
let _update = index.wait_update_id(1).await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions::default())
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 3);
}

138
meilisearch-http/tests/documents/get_documents.rs
View File

@ -1,15 +1,14 @@
use crate::common::Server;
use crate::common::GetAllDocumentsOptions;
use crate::common::Server;
use serde_json::json;
// TODO: partial test since we are testing error, amd error is not yet fully implemented in
// transplant
#[actix_rt::test]
async fn get_unexisting_index_single_document() {
let server = Server::new().await;
let (_response, code) = server
.index("test")
.get_document(1, None)
.await;
let (_response, code) = server.index("test").get_document(1, None).await;
assert_eq!(code, 400);
}
@ -18,9 +17,7 @@ async fn get_unexisting_document() {
let server = Server::new().await;
let index = server.index("test");
index.create(None).await;
let (_response, code) = index
.get_document(1, None)
.await;
let (_response, code) = index.get_document(1, None).await;
assert_eq!(code, 400);
}
@ -38,14 +35,15 @@ async fn get_document() {
let (_, code) = index.add_documents(documents, None).await;
assert_eq!(code, 200);
index.wait_update_id(0).await;
let (response, code) = index
.get_document(0, None)
.await;
let (response, code) = index.get_document(0, None).await;
assert_eq!(code, 200);
assert_eq!(response, serde_json::json!( {
"id": 0,
"content": "foobar",
}));
assert_eq!(
response,
serde_json::json!( {
"id": 0,
"content": "foobar",
})
);
}
#[actix_rt::test]
@ -65,7 +63,9 @@ async fn get_no_documents() {
let (_, code) = index.create(None).await;
assert_eq!(code, 200);
let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions::default())
.await;
assert_eq!(code, 200);
assert!(response.as_array().unwrap().is_empty());
}
@ -76,7 +76,9 @@ async fn get_all_documents_no_options() {
let index = server.index("test");
index.load_test_set().await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions::default())
.await;
assert_eq!(code, 200);
let arr = response.as_array().unwrap();
assert_eq!(arr.len(), 20);
@ -107,7 +109,12 @@ async fn test_get_all_documents_limit() {
let index = server.index("test");
index.load_test_set().await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions { limit: Some(5), ..Default::default() }).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions {
limit: Some(5),
..Default::default()
})
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 5);
assert_eq!(response.as_array().unwrap()[0]["id"], 0);
@ -119,7 +126,12 @@ async fn test_get_all_documents_offset() {
let index = server.index("test");
index.load_test_set().await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions { offset: Some(5), ..Default::default() }).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions {
offset: Some(5),
..Default::default()
})
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 20);
assert_eq!(response.as_array().unwrap()[0]["id"], 13);
@ -131,19 +143,93 @@ async fn test_get_all_documents_attributes_to_retrieve() {
let index = server.index("test");
index.load_test_set().await;
let (response, code) = index.get_all_documents(GetAllDocumentsOptions { attributes_to_retrieve: Some(vec!["name"]), ..Default::default() }).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions {
attributes_to_retrieve: Some(vec!["name"]),
..Default::default()
})
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 20);
assert_eq!(response.as_array().unwrap()[0].as_object().unwrap().keys().count(), 1);
assert!(response.as_array().unwrap()[0].as_object().unwrap().get("name").is_some());
assert_eq!(
response.as_array().unwrap()[0]
.as_object()
.unwrap()
.keys()
.count(),
1
);
assert!(response.as_array().unwrap()[0]
.as_object()
.unwrap()
.get("name")
.is_some());
let (response, code) = index.get_all_documents(GetAllDocumentsOptions { attributes_to_retrieve: Some(vec![]), ..Default::default() }).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions {
attributes_to_retrieve: Some(vec![]),
..Default::default()
})
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 20);
assert_eq!(response.as_array().unwrap()[0].as_object().unwrap().keys().count(), 0);
assert_eq!(
response.as_array().unwrap()[0]
.as_object()
.unwrap()
.keys()
.count(),
0
);
let (response, code) = index.get_all_documents(GetAllDocumentsOptions { attributes_to_retrieve: Some(vec!["name", "tags"]), ..Default::default() }).await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions {
attributes_to_retrieve: Some(vec!["name", "tags"]),
..Default::default()
})
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 20);
assert_eq!(response.as_array().unwrap()[0].as_object().unwrap().keys().count(), 2);
assert_eq!(
response.as_array().unwrap()[0]
.as_object()
.unwrap()
.keys()
.count(),
2
);
}
#[actix_rt::test]
async fn get_documents_displayed_attributes() {
let server = Server::new().await;
let index = server.index("test");
index
.update_settings(json!({"displayedAttributes": ["gender"]}))
.await;
index.load_test_set().await;
let (response, code) = index
.get_all_documents(GetAllDocumentsOptions::default())
.await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 20);
assert_eq!(
response.as_array().unwrap()[0]
.as_object()
.unwrap()
.keys()
.count(),
1
);
assert!(response.as_array().unwrap()[0]
.as_object()
.unwrap()
.get("gender")
.is_some());
let (response, code) = index.get_document(0, None).await;
assert_eq!(code, 200);
assert_eq!(response.as_object().unwrap().keys().count(), 1);
assert!(response.as_object().unwrap().get("gender").is_some());
}

2
meilisearch-http/tests/documents/mod.rs
View File

@ -1,3 +1,3 @@
mod add_documents;
mod get_documents;
mod delete_documents;
mod get_documents;

8
meilisearch-http/tests/index/create_index.rs
View File

@ -9,12 +9,11 @@ async fn create_index_no_primary_key() {
assert_eq!(code, 200);
assert_eq!(response["uid"], "test");
assert!(response.get("uuid").is_some());
assert!(response.get("createdAt").is_some());
assert!(response.get("updatedAt").is_some());
assert_eq!(response["createdAt"], response["updatedAt"]);
assert_eq!(response["primaryKey"], Value::Null);
assert_eq!(response.as_object().unwrap().len(), 5);
assert_eq!(response.as_object().unwrap().len(), 4);
}
#[actix_rt::test]
@ -25,12 +24,11 @@ async fn create_index_with_primary_key() {
assert_eq!(code, 200);
assert_eq!(response["uid"], "test");
assert!(response.get("uuid").is_some());
assert!(response.get("createdAt").is_some());
assert!(response.get("updatedAt").is_some());
assert_eq!(response["createdAt"], response["updatedAt"]);
//assert_eq!(response["createdAt"], response["updatedAt"]);
assert_eq!(response["primaryKey"], "primary");
assert_eq!(response.as_object().unwrap().len(), 5);
assert_eq!(response.as_object().unwrap().len(), 4);
}
// TODO: partial test since we are testing error, amd error is not yet fully implemented in

14
meilisearch-http/tests/index/get_index.rs
View File

@ -13,12 +13,11 @@ async fn create_and_get_index() {
assert_eq!(code, 200);
assert_eq!(response["uid"], "test");
assert!(response.get("uuid").is_some());
assert!(response.get("createdAt").is_some());
assert!(response.get("updatedAt").is_some());
assert_eq!(response["createdAt"], response["updatedAt"]);
assert_eq!(response["primaryKey"], Value::Null);
assert_eq!(response.as_object().unwrap().len(), 5);
assert_eq!(response.as_object().unwrap().len(), 4);
}
// TODO: partial test since we are testing error, amd error is not yet fully implemented in
@ -53,7 +52,12 @@ async fn list_multiple_indexes() {
assert!(response.is_array());
let arr = response.as_array().unwrap();
assert_eq!(arr.len(), 2);
assert!(arr.iter().find(|entry| entry["uid"] == "test" && entry["primaryKey"] == Value::Null).is_some());
assert!(arr.iter().find(|entry| entry["uid"] == "test1" && entry["primaryKey"] == "key").is_some());
assert!(arr
.iter()
.find(|entry| entry["uid"] == "test" && entry["primaryKey"] == Value::Null)
.is_some());
assert!(arr
.iter()
.find(|entry| entry["uid"] == "test1" && entry["primaryKey"] == "key")
.is_some());
}

2
meilisearch-http/tests/index/mod.rs
View File

@ -1,4 +1,4 @@
mod create_index;
mod delete_index;
mod get_index;
mod update_index;
mod delete_index;

3
meilisearch-http/tests/index/update_index.rs
View File

@ -13,7 +13,6 @@ async fn update_primary_key() {
assert_eq!(code, 200);
assert_eq!(response["uid"], "test");
assert!(response.get("uuid").is_some());
assert!(response.get("createdAt").is_some());
assert!(response.get("updatedAt").is_some());
@ -22,7 +21,7 @@ async fn update_primary_key() {
assert!(created_at < updated_at);
assert_eq!(response["primaryKey"], "primary");
assert_eq!(response.as_object().unwrap().len(), 5);
assert_eq!(response.as_object().unwrap().len(), 4);
}
#[actix_rt::test]

2
meilisearch-http/tests/integration.rs
View File

@ -1,8 +1,8 @@
mod common;
mod documents;
mod index;
mod search;
mod settings;
mod documents;
mod updates;
// Tests are isolated by features in different modules to allow better readability, test

1
meilisearch-http/tests/search/mod.rs
View File

@ -1,3 +1,2 @@
// This modules contains all the test concerning search. Each particular feture of the search
// should be tested in its own module to isolate tests and keep the tests readable.

55
meilisearch-http/tests/settings/get_settings.rs
View File

@ -19,8 +19,19 @@ async fn get_settings() {
assert_eq!(settings.keys().len(), 4);
assert_eq!(settings["displayedAttributes"], json!(["*"]));
assert_eq!(settings["searchableAttributes"], json!(["*"]));
assert_eq!(settings["facetedAttributes"], json!({}));
assert_eq!(settings["rankingRules"], json!(["typo", "words", "proximity", "attribute", "wordsPosition", "exactness"]));
println!("{:?}", settings);
assert_eq!(settings["attributesForFaceting"], json!({}));
assert_eq!(
settings["rankingRules"],
json!([
"typo",
"words",
"proximity",
"attribute",
"wordsPosition",
"exactness"
])
);
}
#[actix_rt::test]
@ -35,20 +46,24 @@ async fn update_settings_unknown_field() {
async fn test_partial_update() {
let server = Server::new().await;
let index = server.index("test");
index.update_settings(json!({"displayedAttributes": ["foo"]})).await;
let (_response, _code) = index
.update_settings(json!({"displayedAttributes": ["foo"]}))
.await;
index.wait_update_id(0).await;
let (response, code) = index.settings().await;
assert_eq!(code, 200);
assert_eq!(response["displayedAttributes"],json!(["foo"]));
assert_eq!(response["searchableAttributes"],json!(["*"]));
assert_eq!(response["displayedAttributes"], json!(["foo"]));
assert_eq!(response["searchableAttributes"], json!(["*"]));
index.update_settings(json!({"searchableAttributes": ["bar"]})).await;
let (_response, _) = index
.update_settings(json!({"searchableAttributes": ["bar"]}))
.await;
index.wait_update_id(1).await;
let (response, code) = index.settings().await;
assert_eq!(code, 200);
assert_eq!(response["displayedAttributes"],json!(["foo"]));
assert_eq!(response["searchableAttributes"],json!(["bar"]));
assert_eq!(response["displayedAttributes"], json!(["foo"]));
assert_eq!(response["searchableAttributes"], json!(["bar"]));
}
#[actix_rt::test]
@ -63,20 +78,22 @@ async fn delete_settings_unexisting_index() {
async fn reset_all_settings() {
let server = Server::new().await;
let index = server.index("test");
index.update_settings(json!({"displayedAttributes": ["foo"], "searchableAttributes": ["bar"]})).await;
index
.update_settings(json!({"displayedAttributes": ["foo"], "searchableAttributes": ["bar"]}))
.await;
index.wait_update_id(0).await;
let (response, code) = index.settings().await;
assert_eq!(code, 200);
assert_eq!(response["displayedAttributes"],json!(["foo"]));
assert_eq!(response["searchableAttributes"],json!(["bar"]));
assert_eq!(response["displayedAttributes"], json!(["foo"]));
assert_eq!(response["searchableAttributes"], json!(["bar"]));
index.delete_settings().await;
index.wait_update_id(1).await;
let (response, code) = index.settings().await;
assert_eq!(code, 200);
assert_eq!(response["displayedAttributes"],json!(["*"]));
assert_eq!(response["searchableAttributes"],json!(["*"]));
assert_eq!(response["displayedAttributes"], json!(["*"]));
assert_eq!(response["searchableAttributes"], json!(["*"]));
}
#[actix_rt::test]
@ -94,7 +111,6 @@ async fn update_setting_unexisting_index_invalid_uid() {
let server = Server::new().await;
let index = server.index("test##! ");
let (_response, code) = index.update_settings(json!({})).await;
println!("response: {}", _response);
assert_eq!(code, 400);
}
@ -124,10 +140,10 @@ macro_rules! test_setting_routes {
.chars()
.map(|c| if c == '_' { '-' } else { c })
.collect::<String>());
let (_response, code) = server.service.post(url, serde_json::Value::Null).await;
assert_eq!(code, 200);
let (_response, code) = server.index("test").get().await;
assert_eq!(code, 200);
let (response, code) = server.service.post(url, serde_json::Value::Null).await;
assert_eq!(code, 200, "{}", response);
let (response, code) = server.index("test").get().await;
assert_eq!(code, 200, "{}", response);
}
#[actix_rt::test]
@ -149,4 +165,5 @@ macro_rules! test_setting_routes {
test_setting_routes!(
attributes_for_faceting,
displayed_attributes,
searchable_attributes);
searchable_attributes
);

28
meilisearch-http/tests/updates/mod.rs
View File

@ -8,7 +8,7 @@ async fn get_update_unexisting_index() {
}
#[actix_rt::test]
async fn get_unexisting_udpate_status() {
async fn get_unexisting_update_status() {
let server = Server::new().await;
let index = server.index("test");
index.create(None).await;
@ -21,13 +21,15 @@ async fn get_update_status() {
let server = Server::new().await;
let index = server.index("test");
index.create(None).await;
index.add_documents(
serde_json::json!([{
"id": 1,
"content": "foobar",
}]),
None
).await;
index
.add_documents(
serde_json::json!([{
"id": 1,
"content": "foobar",
}]),
None,
)
.await;
let (_response, code) = index.get_update(0).await;
assert_eq!(code, 200);
// TODO check resonse format, as per #48
@ -55,10 +57,12 @@ async fn list_updates() {
let server = Server::new().await;
let index = server.index("test");
index.create(None).await;
index.add_documents(
serde_json::from_str(include_str!("../assets/test_set.json")).unwrap(),
None
).await;
index
.add_documents(
serde_json::from_str(include_str!("../assets/test_set.json")).unwrap(),
None,
)
.await;
let (response, code) = index.list_updates().await;
assert_eq!(code, 200);
assert_eq!(response.as_array().unwrap().len(), 1);