436: Speed up the word prefix databases computation time r=Kerollmops a=Kerollmops
This PR depends on the fixes done in #431 and must be merged after it.
In this PR we will bring the `WordPrefixPairProximityDocids`, `WordPrefixDocids` and, `WordPrefixPositionDocids` update structures to a new era, a better era, where computing the word prefix pair proximities costs much fewer CPU cycles, an era where this update structure can use the, previously computed, set of new word docids from the newly indexed batch of documents.
---
The `WordPrefixPairProximityDocids` is an update structure, which means that it is an object that we feed with some parameters and which modifies the LMDB database of an index when asked for. This structure specifically computes the list of word prefix pair proximities, which correspond to a list of pairs of words associated with a proximity (the distance between both words) where the second word is not a word but a prefix e.g. `s`, `se`, `a`. This word prefix pair proximity is associated with the list of documents ids which contains the pair of words and prefix at the given proximity.
The origin of the performances issue that this struct brings is related to the fact that it starts its job from the beginning, it clears the LMDB database before rewriting everything from scratch, using the other LMDB databases to achieve that. I hope you understand that this is absolutely not an optimized way of doing things.
Co-authored-by: Clément Renault <clement@meilisearch.com>
Co-authored-by: Kerollmops <clement@meilisearch.com>
returned metaimprove document addition returned metaimprove document
addition returned metaimprove document addition returned metaimprove
document addition returned metaimprove document addition returned
metaimprove document addition returned meta
Instead of using an arbitrary limit we encode the absolute position in a u32
using one strong u16 for the field id and a weak u16 for the relative position in the attribute.
322: Geosearch r=ManyTheFish a=irevoire
This PR introduces [basic geo-search functionalities](https://github.com/meilisearch/specifications/pull/59), it makes the engine able to index, filter and, sort by geo-point. We decided to use [the rstar library](https://docs.rs/rstar) and to save the points in [an RTree](https://docs.rs/rstar/0.9.1/rstar/struct.RTree.html) that we de/serialize in the index database [by using serde](https://serde.rs/) with [bincode](https://docs.rs/bincode). This is not an efficient way to query this tree as it will consume a lot of CPU and memory when a search is made, but at least it is an easy first way to do so.
### What we will have to do on the indexing part:
- [x] Index the `_geo` fields from the documents.
- [x] Create a new module with an extractor in the `extract` module that takes the `obkv_documents` and retrieves the latitude and longitude coordinates, outputting them in a `grenad::Reader` for further process.
- [x] Call the extractor in the `extract::extract_documents_data` function and send the result to the `TypedChunk` module.
- [x] Get the `grenad::Reader` in the `typed_chunk::write_typed_chunk_into_index` function and store all the points in the `rtree`
- [x] Delete the documents from the `RTree` when deleting documents from the database. All this can be done in the `delete_documents.rs` file by getting the data structure and removing the points from it, inserting it back after the modification.
- [x] Clearing the `RTree` entirely when we clear the documents from the database, everything happens in the `clear_documents.rs` file.
- [x] save a Roaring bitmap of all documents containing the `_geo` field
### What we will have to do on the query part:
- [x] Filter the documents at a certain distance around a point, this is done by [collecting the documents from the searched point](https://docs.rs/rstar/0.9.1/rstar/struct.RTree.html#method.nearest_neighbor_iter) while they are in range.
- [x] We must introduce new `geoLowerThan` and `geoGreaterThan` variants to the `Operator` filter enum.
- [x] Implement the `negative` method on both variants where the `geoGreaterThan` variant is implemented by executing the `geoLowerThan` and removing the results found from the whole list of geo faceted documents.
- [x] Add the `_geoRadius` function in the pest parser.
- [x] Introduce a `_geo` ascending ranking function that takes a point in parameter, ~~this function must keep the iterator on the `RTree` and make it peekable~~ This was not possible for now, we had to collect the whole iterator. Only the documents that are part of the candidates must be sent too!
- [x] This ascending ranking rule will only be active if the search is set up with the `_geoPoint` parameter that indicates the center point of the ascending ranking rule.
-----------
- On Meilisearch part: We must introduce a new concept, returning the documents with a new `_geoDistance` field when it passed by the `_geo` ranking rule, this has never been done before. We could maybe just do it afterward when the documents have been retrieved from the database, computing the distance from the `_geoPoint` and all of the documents to be returned.
Co-authored-by: Irevoire <tamo@meilisearch.com>
Co-authored-by: cvermand <33010418+bidoubiwa@users.noreply.github.com>
Co-authored-by: Tamo <tamo@meilisearch.com>
291: Fix a bug about zero bytes in the inputs r=irevoire a=Kerollmops
Ok, good news, after a little session of debugging with `@irevoire` we found out that the bug seems to be related to zeroes in the input update. The engine wasn't designed to accept those. The chosen solution is to update the tokenizer to remove those zeroes. We are waiting on https://github.com/meilisearch/tokenizer/pull/52 to be merged and a new version to be released.
It is not an undefined behavior, I repeat: it is a "normal" bug 🎉👏
----
This PR tries to fix a bug where we use LMDB in the wrong way, leading to panic due to an undefined behavior on the Rust side. I thought [we fixed it in a previous PR](https://github.com/meilisearch/milli/pull/264) but we found out that _a similar_ bug was still present. `@bb` found a way to trigger this bug and helped us find the origin of it.
As I don't have a minimal reproducible example of this bug I bet on the unsafe `put_current` calls when we index new documents as the bug was trigger after a big indexation on a clean database, thus not triggering a deletion update. I only replaced the unsafe `put_current` with two safe calls to `get`/`put`.
I hope it helps and fixes the bug, only `@bb` can help us check that. I am not even sure how I can create a custom Docker image and expose it for testing purposes.
<details>
<summary>The backtrace leading us to a panic in grenad.</summary>
```
meilisearch_1 | thread 'tokio-runtime-worker' panicked at 'assertion failed: key > &last_key', /root/.cargo/git/checkouts/grenad-e2cb77f65d31bb02/3adcb26/src/block_builder.rs:38:17
meilisearch_1 | stack backtrace:
meilisearch_1 | 0: rust_begin_unwind
meilisearch_1 | at ./rustc/53cb7b09b00cbea8754ffb78e7e3cb521cb8af4b/library/std/src/panicking.rs:493:5
meilisearch_1 | 1: core::panicking::panic_fmt
meilisearch_1 | at ./rustc/53cb7b09b00cbea8754ffb78e7e3cb521cb8af4b/library/core/src/panicking.rs:92:14
meilisearch_1 | 2: core::panicking::panic
meilisearch_1 | at ./rustc/53cb7b09b00cbea8754ffb78e7e3cb521cb8af4b/library/core/src/panicking.rs:50:5
meilisearch_1 | 3: grenad::block_builder::BlockBuilder::insert
meilisearch_1 | at ./root/.cargo/git/checkouts/grenad-e2cb77f65d31bb02/3adcb26/src/block_builder.rs:38:17
meilisearch_1 | 4: grenad::writer::Writer<W>::insert
meilisearch_1 | at ./root/.cargo/git/checkouts/grenad-e2cb77f65d31bb02/3adcb26/src/writer.rs:92:12
meilisearch_1 | 5: milli::update::words_level_positions::write_level_entry
meilisearch_1 | at ./root/.cargo/git/checkouts/milli-00376cd5db949a15/007fec2/milli/src/update/words_level_positions.rs:262:5
meilisearch_1 | 6: milli::update::words_level_positions::compute_positions_levels
meilisearch_1 | at ./root/.cargo/git/checkouts/milli-00376cd5db949a15/007fec2/milli/src/update/words_level_positions.rs:211:13
meilisearch_1 | 7: milli::update::words_level_positions::WordsLevelPositions::execute
meilisearch_1 | at ./root/.cargo/git/checkouts/milli-00376cd5db949a15/007fec2/milli/src/update/words_level_positions.rs:65:23
meilisearch_1 | 8: milli::update::index_documents::IndexDocuments::execute_raw
meilisearch_1 | at ./root/.cargo/git/checkouts/milli-00376cd5db949a15/007fec2/milli/src/update/index_documents/mod.rs:831:9
meilisearch_1 | 9: milli::update::index_documents::IndexDocuments::execute
meilisearch_1 | at ./root/.cargo/git/checkouts/milli-00376cd5db949a15/007fec2/milli/src/update/index_documents/mod.rs:372:9
meilisearch_1 | 10: meilisearch_http::index::updates::<impl meilisearch_http::index::Index>::update_documents_txn
meilisearch_1 | at ./meilisearch/meilisearch-http/src/index/updates.rs:225:30
meilisearch_1 | 11: meilisearch_http::index::updates::<impl meilisearch_http::index::Index>::update_documents
meilisearch_1 | at ./meilisearch/meilisearch-http/src/index/updates.rs:183:22
meilisearch_1 | 12: meilisearch_http::index::update_handler::UpdateHandler::handle_update
meilisearch_1 | at ./meilisearch/meilisearch-http/src/index/update_handler.rs:75:18
meilisearch_1 | 13: meilisearch_http::index_controller::index_actor::actor::IndexActor<S>::handle_update::{{closure}}::{{closure}}
meilisearch_1 | at ./meilisearch/meilisearch-http/src/index_controller/index_actor/actor.rs:174:35
meilisearch_1 | 14: <tokio::runtime::blocking::task::BlockingTask<T> as core::future::future::Future>::poll
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/blocking/task.rs:42:21
meilisearch_1 | 15: tokio::runtime::task::core::CoreStage<T>::poll::{{closure}}
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/core.rs:243:17
meilisearch_1 | 16: tokio::loom::std::unsafe_cell::UnsafeCell<T>::with_mut
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/loom/std/unsafe_cell.rs:14:9
meilisearch_1 | 17: tokio::runtime::task::core::CoreStage<T>::poll
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/core.rs:233:13
meilisearch_1 | 18: tokio::runtime::task::harness::poll_future::{{closure}}
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/harness.rs:427:23
meilisearch_1 | 19: <std::panic::AssertUnwindSafe<F> as core::ops::function::FnOnce<()>>::call_once
meilisearch_1 | at ./rustc/53cb7b09b00cbea8754ffb78e7e3cb521cb8af4b/library/std/src/panic.rs:344:9
meilisearch_1 | 20: std::panicking::try::do_call
meilisearch_1 | at ./rustc/53cb7b09b00cbea8754ffb78e7e3cb521cb8af4b/library/std/src/panicking.rs:379:40
meilisearch_1 | 21: std::panicking::try
meilisearch_1 | at ./rustc/53cb7b09b00cbea8754ffb78e7e3cb521cb8af4b/library/std/src/panicking.rs:343:19
meilisearch_1 | 22: std::panic::catch_unwind
meilisearch_1 | at ./rustc/53cb7b09b00cbea8754ffb78e7e3cb521cb8af4b/library/std/src/panic.rs:431:14
meilisearch_1 | 23: tokio::runtime::task::harness::poll_future
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/harness.rs:414:19
meilisearch_1 | 24: tokio::runtime::task::harness::Harness<T,S>::poll_inner
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/harness.rs:89:9
meilisearch_1 | 25: tokio::runtime::task::harness::Harness<T,S>::poll
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/harness.rs:59:15
meilisearch_1 | 26: tokio::runtime::task::raw::RawTask::poll
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/raw.rs:66:18
meilisearch_1 | 27: tokio::runtime::task::Notified<S>::run
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/task/mod.rs:171:9
meilisearch_1 | 28: tokio::runtime::blocking::pool::Inner::run
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/blocking/pool.rs:265:17
meilisearch_1 | 29: tokio::runtime::blocking::pool::Spawner::spawn_thread::{{closure}}
meilisearch_1 | at ./root/.cargo/registry/src/github.com-1ecc6299db9ec823/tokio-1.7.1/src/runtime/blocking/pool.rs:245:17
meilisearch_1 | note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
```
</details>
Co-authored-by: Kerollmops <clement@meilisearch.com>
