Commit Graph

363 Commits

Author SHA1 Message Date
Paul Masurel
b5f3dcdc8b TermInfo contain the end_offset of the postings.
We slice the ReadOnlySource tightly.
2020-11-06 15:18:51 +09:00
Paul Masurel
b2dfacdc70 Fixed bench and removed unnecessary public symbols. 2020-11-05 16:12:03 +09:00
Paul Masurel
581c2bb718 Minor refactoring postings serializers options. 2020-11-03 15:47:25 +09:00
Paul Masurel
c3f44d38f3 Moving HasLen (#910) 2020-10-13 10:19:30 +09:00
Paul Masurel
01b4aa9adc Refactoring dir (#905) 2020-10-11 22:22:56 +09:00
Paul Masurel
c23a03ad81 Large API Change in the Directory API. (#901)
Tantivy used to assume that all files could be somehow memory mapped. After this change, Directory return a `FileSlice` that can be reduced and eventually read into an `OwnedBytes` object. Long and blocking io operation are still required by they do not span over the entire file.
2020-10-08 16:36:51 +09:00
Paul Masurel
96f946d4c3 Raultang master (#879)
* add support for indexed bytes fast field

* remove backup code file

* refine test cases

* Simplified unit test. Renamed it as it is testing the storable part. Not the indexed part.

* Small refactoring and added unit test. If multivalued we only retain the first FAST value.

Co-authored-by: Raul <raul.tang.lc@gmail.com>
2020-10-01 18:03:18 +09:00
Paul Masurel
70bae7ce4c Removing Term Vec allocation (#881) 2020-09-08 23:11:00 +09:00
Paul Masurel
4ce9517a82 fix unit test for bench. remove scoref64 feature. fixed test for lz4 feature. 2020-09-08 07:35:00 +09:00
Paul Masurel
439d6956a9 Returning Result in some of the API (#880)
* Returning Result in some of the API

* Introducing `.writer_for_test(..)`
2020-09-07 15:52:34 +09:00
Paul Masurel
2481c87be8 Block wand (#856) 2020-08-19 22:36:36 +09:00
Paul Masurel
68fe406924 Removed asserts (#850) 2020-07-16 12:24:55 +09:00
Paul Masurel
f71b04acb0 Bugfix. (#849)
go_to_first_doc was typically calling seek with a target smaller than
doc.

Since SegmentPostings typically do a linear search on the full block,
regardless of the current position, it could have our segment postings
go backward.
2020-07-16 10:57:51 +09:00
Paul Masurel
c0f5645cd9 Move for_each functions from Scorer to Weight. (#836)
* Move for_each functions from Scorer to Weight.

* Specialized foreach / foreach_pruning for union of termscorer.
2020-06-01 11:31:18 +09:00
Paul Masurel
cbff874e43 Change the loading of blocks. 2020-05-27 16:36:50 +09:00
Paul Masurel
baf015fc57 Simplification of the segment postings seek implementation. (#834) 2020-05-27 08:49:47 +09:00
Paul Masurel
7275ebdf3c Skiprefactoring skipabsolute (#831)
Simplification of the way we handle positions.
2020-05-25 09:51:23 +09:00
Paul Masurel
e25284bafe Major change in the DocSet/Scorer API (#824)
- Change in the DocSet and Scorer API. (@fulmicoton). 
A freshly created DocSet point directly to their first doc. A sentinel value called TERMINATED marks the end of a DocSet.
`.advance()` returns the new DocId. `Scorer::skip(target)` has been replaced by `Scorer::seek(target)` and returns the resulting DocId.
As a result, iterating through DocSet now looks as follows
```rust
let mut doc = docset.doc();
while doc != TERMINATED {
   // ...
   doc = docset.advance();
}
```
The change made it possible to greatly simplify a lot of the docset's code.
- Misc internal optimization and introduction of the `Scorer::for_each_pruning` function. (@fulmicoton)
2020-05-16 16:33:36 +09:00
Paul Masurel
1e5ebdbf3c Format and remove useless import (#819) 2020-04-27 11:56:49 +09:00
Paul Masurel
1748602691 ignore -> compile_fail 2020-03-02 09:59:48 +09:00
Paul Masurel
ae14022bf0 Removed use::Result. (#771) 2020-01-31 18:47:02 +09:00
Audun Halland
f85d0a522a Optimize TermDictionary::empty by precomputed data source (#767) 2020-01-30 10:04:58 +09:00
Paul Masurel
7b21b3f25a Refactoring around Field (#673)
* Refactoring around Field

Removing the contract about the order of the field, and the
field id allocation.

* Update delete_queue.rs

* Update field.rs
2019-10-25 09:06:44 +09:00
Paul Masurel
2f867aad17 Fix bench (#663)
* fmt

* Fixing bench compilation
2019-10-04 17:07:49 +09:00
Paul Masurel
5c6580eb15 fmt (#661) 2019-10-04 12:10:01 +09:00
petr-tik
2ec19b21ae Remove unnecessary duplicate methods (#650)
Closes #649

Spotted by @imor
2019-09-09 06:36:04 +09:00
Raminder Singh
141f5a93f7 Using FnvHashMap for mapping UnorderedTermId to TermOrdinal. Fixes #507 (#647)
* Using FnvHashMap for mapping UnorderedTermId to TermOrdinal. Fixes #507

* Fixed cargo fmt errors
2019-09-07 19:40:21 +09:00
fdb-hiroshima
6eb4e08636 add support for float (#603)
* add basic support for float

as for i64, they are mapped to u64 for indexing
query parser don't work yet

* Update value.rs

* implement support for float in query parser

* Update README.md
2019-07-27 17:57:33 +09:00
Paul Masurel
5095e6b010 Introduce a small refactoring of the sgment writer. (#596) 2019-07-17 08:32:29 +09:00
Paul Masurel
4867be3d3b Kompass master (#590)
* Use once_cell in place of lazy_static

* Minor changes
2019-07-10 19:24:54 +09:00
Paul Masurel
462774b15c Tiqb feature/2018 (#583)
* rust 2018

* Added CHANGELOG comment
2019-07-01 10:01:46 +09:00
Paul Masurel
66b4615e4e Issue/542 (#543)
* Closes 542.

Fast fields are all loaded when the segment reader is created.
2019-05-05 13:52:43 +09:00
Paul Masurel
31b22c5acc Added logging when token is dropped. (#538) 2019-04-26 09:23:28 +09:00
Paul Masurel
96a4f503ec Closes #526 (#535) 2019-04-24 20:59:48 +09:00
Paul Masurel
b7c2d0de97 Clippy2 (#534)
* Clippy comments

Clippy complaints that about the cast of &[u32] to a *const __m128i,
because of the lack of alignment constraints.

This commit passes the OutputBuffer object (which enforces proper
    alignment) instead of `&[u32]`.

* Clippy. Block alignment

* Code simplification

* Added comment. Code simplification

* Removed the extraneous freq block len hack.
2019-04-24 12:31:32 +09:00
Paul Masurel
a228825462 Clippy comments (#532)
Clippy complaints that about the cast of &[u32] to a *const __m128i,
because of the lack of alignment constraints.

This commit passes the OutputBuffer object (which enforces proper
    alignment) instead of `&[u32]`.
2019-04-23 09:54:02 +09:00
Paul Masurel
acd29b535d Fix comment 2019-04-02 10:05:14 +09:00
Paul Masurel
a8cc5208f1 Linear simd (#519)
* linear simd search within block
2019-03-20 22:10:05 +09:00
Paul Masurel
83eb0d0cb7 Disabling tests on Android 2019-03-20 10:24:17 +09:00
Paul Masurel
663dd89c05 Feature/reader (#517)
Adding IndexReader to the API. Making it possible to watch for changes.

* Closes #500
2019-03-20 08:39:22 +09:00
barrotsteindev
a934577168 WIP: date field (#487)
* initial version, still a work in progress

* remove redudant or

* add chrono::DateTime and index i64

* add more tests

* fix tests

* pass DateTime by ptr

* remove println!

* document query_parser rfc 3339 date support

* added some more docs about implementation to schema.rs

* enforce DateTime is UTC, and re-export chrono

* added DateField to changelog

* fixed conflict

* use INDEXED instead of INT_INDEXED for date fields
2019-03-15 22:10:37 +09:00
Paul Masurel
94f1885334 Issue/513 (#514)
* Closes #513

* Clean up and doc

* Updated changelog
2019-03-07 09:39:30 +09:00
Paul Masurel
e67883138d Cargo fmt 2019-03-06 10:31:00 +09:00
Paul Masurel
9acadb3756 Code cleaning 2019-02-26 10:50:36 +09:00
Paul Masurel
774fcecf23 cargo fmt 2019-02-26 10:44:59 +09:00
Paul Masurel
27c9fa6028 Jannickj prove bug with facets (#508)
* prove bug with facets

* Closing #505

Introduce a term id in the TermHashMap
2019-02-25 22:33:17 +09:00
Paul Masurel
b422f9c389 Partially addresses #500 (#502)
Using `tantivy_fst`. Storing `Weak<Mmap>` in the Mmap cache.
2019-02-23 10:33:59 +09:00
Paul Masurel
50ed6fb534 Code cleanup
Fixed compilation without the mmap directory
2019-02-05 12:39:30 +01:00
Paul Masurel
259ce567d1 Using linear search 2019-01-29 15:59:24 +01:00
Paul Masurel
4c93b096eb Rustfmt 2019-01-29 11:45:30 +01:00