aboutsummaryrefslogtreecommitdiffstats
path: root/searchcommon
Commit message (Collapse)AuthorAgeFilesLines
* swappable -> pagedHenning Baldersheim2021-08-122-6/+6
|
* swapable -> swappableHenning Baldersheim2021-08-122-23/+23
|
* Add swapable attribute option.Henning Baldersheim2021-08-122-0/+6
|
* Update _maxDeadBytesRatio to reflect current default.Henning Baldersheim2021-05-031-1/+1
|
* Support both case sensitive, and case-insensitive search in non-fast-search ↵Henning Baldersheim2021-04-222-0/+7
| | | | attributes.
* Track changed default max dead bytes ratio.Tor Egge2021-04-091-1/+1
|
* Add CompactionStrategy methods to determine if compaction should be performed.Tor Egge2021-04-091-0/+14
|
* Wire in match configHenning Baldersheim2021-04-083-11/+42
|
* Use BTREE/HASH instead of ORDERED/UNORDERED.Henning Baldersheim2021-03-233-27/+31
|
* Add class commentHenning Baldersheim2021-03-111-0/+3
|
* Add dictionary configHenning Baldersheim2021-03-118-21/+94
|
* string.h does not need alloc.hHenning Baldersheim2021-02-211-0/+1
|
* Improve handling of changed alloc config.Tor Egge2021-02-025-0/+41
|
* Adjust constructor parameter name.Tor Egge2021-02-021-2/+2
|
* Add noexcept specifiers to non-throwing constructors and operators.Tor Egge2020-12-052-18/+18
|
* no need to copy by value hereArne Juul2020-11-181-1/+1
|
* - Move utility methods from proton::attribute to search::attribute.Henning Baldersheim2020-11-084-1/+55
| | | | - Compute isUpdateableInmemoryOnly once.
* Add noexcept as indicated by -WnoeexceptHenning Baldersheim2020-10-071-2/+2
|
* add "Hamming" distance metric for ANNArne Juul2020-09-041-1/+1
|
* Consolidate search for GTest.Tor Egge2020-07-071-1/+0
|
* Use find_package to find gtest library.Tor Egge2020-06-291-1/+2
|
* add "InnerProduct" distance metricArne Juul2020-06-251-1/+1
|
* Wire flag to enable multi-threaded indexing for a hnsw index.Geir Storli2020-06-191-5/+5
|
* Test that attribute writer can handle put in two phases.Geir Storli2020-06-172-8/+14
|
* Implement initial support for two-phase puts in attribute writer.Geir Storli2020-06-173-6/+51
| | | | This is only turned on for tensor attributes with a hnsw index that allows multi-threaded indexing.
* Revert "Two phase put in attribute writer"Harald Musum2020-06-155-65/+14
|
* Test that attribute writer can handle put in two phases.Geir Storli2020-06-152-8/+14
|
* Implement initial support for two-phase puts in attribute writer.Geir Storli2020-06-153-6/+51
| | | | This is only turned on for tensor attributes with a hnsw index that allows multi-threaded indexing.
* Since there might be a large variation in number of entries, let us not take ↵Henning Baldersheim2020-05-151-7/+6
| | | | a fixed cost up front.
* Support specifying a distance metric for nearest neighbor search when not ↵Geir Storli2020-05-123-0/+12
| | | | | | having a hnsw index. This also changes the syntax in the sd file to specifying the distance metric in the attribute tag.
* Ignore hnsw index attribute save file if major parameters are changed.Tor Egge2020-04-241-0/+4
|
* add distance metric to HnswIndexParamsArne Juul2020-03-231-4/+11
|
* add enum for selecting distance metricArne Juul2020-03-231-0/+9
|
* Add hnsw index params to config for attribute vector.Geir Storli2020-02-203-6/+53
|
* Add ReadableAttributeVector accessor to IAttributeManagerTor Brede Vekterli2020-01-291-3/+3
| | | | | | | | | | Provides a unified interface for fetching both regular as well as imported attributes. Exposing `ReadableAttributeVector` instead of raw `AttributeVector` instances enforces that all access is done via appropriate acquired read guards. Refactor document selection processing code to use the new interface in order to prepare for imported field support in selections.
* Inline frequently called methodHenning Baldersheim2020-01-232-12/+11
|
* Add noexcept to move constructors and operatorsHenning Baldersheim2020-01-214-14/+31
|
* Revert "Revert "Balder/add executeinfo""Henning Baldersheim2020-01-031-2/+5
|
* Revert "Balder/add executeinfo"Arnstein Ressem2020-01-031-5/+2
|
* Merge pull request #11619 from vespa-engine/balder/add-executeinfoHenning Baldersheim2020-01-021-2/+5
|\ | | | | Balder/add executeinfo
| * Add ExecutionInfo to propagate more information than just strictness to ↵Henning Baldersheim2019-12-311-2/+5
| | | | | | | | | | | | | | fetchPostings. The estimated hitrate that each iterator will will enable smarter query evaluation. Especially will it be possible to figure out if static up front cost can be avoided.
* | GC unused codeHenning Baldersheim2020-01-021-2/+0
|/
* Drop timestamp.hHenning Baldersheim2019-12-161-1/+0
|
* Drop all of TimeStamp too.Henning Baldersheim2019-12-161-15/+11
|
* GC unused timestamp in schema.Henning Baldersheim2019-12-163-133/+25
|
* Rename QueryTermBase -> QueryTermUCS4 and move to separate h and cpp files.Geir Storli2019-11-281-2/+2
|
* Remove "enable enumerated save" flag from attribute vectors as this should ↵Geir Storli2019-08-191-0/+9
| | | | | | be always on. For production code this has been on for several years already.
* Use proper scope for enum constants.Tor Egge2019-07-301-5/+5
|
* Rename experimentalpostinglistformat -> interleavedfeatures in indexschema.def.Geir Storli2019-06-245-16/+16
|
* Move RcuVector and relevant support classes to vespalibTor Brede Vekterli2019-05-231-0/+5
| | | | | | | | | | | | | Having RCU support available across all our C++ modules open up new opportunities for optimizations. This changes the following: - `RcuVector` moved from `searchlib` to `vespalib` - `MemoryUsage` moved from `searchlib` to `vespalib` - Introduce a simplified, more generic `GrowStrategy` in `vespalib` used by the moved `RcuVector` which does not have any notion of documents. Existing searchlib `GrowStrategy` gets a simple function to convert to this generic strategy.