summaryrefslogtreecommitdiffstats
path: root/searchcore/src
Commit message (Collapse)AuthorAgeFilesLines
* Push handling of NearestNeighborTerm down to AttributeBlueprintFactory.Geir Storli2019-11-192-8/+2
|
* Add skeleton for NearestNeighborTerm in C++.Geir Storli2019-11-197-0/+15
|
* Specify UTC explicitHenning Baldersheim2019-11-1830-268/+180
|
* Merge pull request #11327 from ↵Geir Storli2019-11-185-313/+12
|\ | | | | | | | | vespa-engine/balder/fdispatch-is-really-gone-rebased Balder/fdispatch is really gone rebased
| * Do not let the lingering transportthread overrides prevent good code health.Henning Baldersheim2019-11-182-0/+7
| |
| * Now fdispatch reliques can goHenning Baldersheim2019-11-165-320/+12
| |
* | Add missing whitespaceHenning Baldersheim2019-11-151-5/+5
| |
* | Add typesafe SteadyTimeStamp.Henning Baldersheim2019-11-1526-89/+93
| |
* | Use ClockSteady.Henning Baldersheim2019-11-154-55/+29
| |
* | Use ClockSteady and return TimeStamp instead of untyped int64_t without unit.Henning Baldersheim2019-11-151-4/+1
| |
* | Use steady_clockHenning Baldersheim2019-11-153-12/+9
| |
* | Use a steady_clock for the stop watch.Henning Baldersheim2019-11-151-1/+1
| |
* | Use steady clock to ensure monotonically increasing clock.Henning Baldersheim2019-11-147-26/+23
|/
* Ensure we don't resurrect stale position field values from the docstoreTor Brede Vekterli2019-11-071-0/+2
|
* Let Z-curve attribute be authoritative for position field existenceTor Brede Vekterli2019-11-072-9/+22
| | | | | | | | Avoids inconsistencies caused by concrete position struct fields not being created or cleared by partial updates, only the Z-curve attribute fields. This fixes #11208
* Merge pull request #11086 from ↵Tor Brede Vekterli2019-10-251-1/+3
|\ | | | | | | | | vespa-engine/vekterli/improve-multi-value-attribute-update-complexity Make multi value attribute inserts O(n) instead of O(n^2)
| * Make multi value attribute inserts O(n) instead of O(n^2)Tor Brede Vekterli2019-10-241-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previous code had O(n^2) complexity for * Weighted set appends, removes and weight updates * Array element removes (in-place erasing) New code moves all weighted set operations to a temporary hash map from key to weight, building a value vector at the very end. For arrays it defers all removals to the very end, amortizing cost down to linear time by tracking element removal times and avoiding push_backs instead of performing explicit element erasures. This fixes #11069
* | GC no longer used ports.Henning Baldersheim2019-10-241-4/+0
|/
* Extend AttributeAspectDelayer to handle matchedattributeelementsfilterTor Egge2019-10-222-6/+68
| | | | summary map override.
* Avoid dependency on weighted set ordering in unit testsTor Brede Vekterli2019-10-181-2/+8
| | | | | Weighted sets are not guaranteed to be ordered, but thus far they have been in practice due to implementation details of the update logic.
* Add spam logging for get() and remove(), and log all input params for put().Geir Storli2019-10-111-4/+9
|
* improve fake searchable attributesHåvard Pettersen2019-10-082-21/+61
| | | | | | | | | | | | | blueprint/search iterator now exposes a more functional attribute search context fake attribute searches will now unpack a single entry containing the sum of all matched weights improve matcher test for same element matching by also using the attribute element iterator added tests for identifying matching elements in docsum request
* Use summary executor for compactLidSpace() on docstore.Tor Egge2019-10-082-2/+3
|
* Reduce timing sensitivity of feedview unit test.Tor Egge2019-10-071-4/+4
|
* Merge pull request #10875 from ↵Geir Storli2019-10-041-0/+21
|\ | | | | | | | | vespa-engine/havardpe/find-matching-elements-for-stand-alone-attribute-terms find matching elements for attribute terms searching in struct subfields
| * find matching elements for attribute terms searching in struct subfieldsHåvard Pettersen2019-10-041-0/+21
| |
* | Add wiring of fill_matching_elements() to matcher.Tor Egge2019-10-031-1/+4
|/
* Merge pull request #10850 from ↵Geir Storli2019-10-035-11/+53
|\ | | | | | | | | vespa-engine/havardpe/collect-same-element-matches-from-query-state collect matching elements by using SameElement blueprints
| * collect matching elements by using SameElement blueprintsHåvard Pettersen2019-10-025-11/+53
| |
* | Change signature of fill_matching_elements to take a const reference toTor Egge2019-10-022-2/+2
|/ | | | StructFieldMapper.
* GC unused filesHenning Baldersheim2019-09-3029-680/+0
|
* Remove some more transport related stuff.Henning Baldersheim2019-09-307-38/+25
|
* Merge pull request #10804 from vespa-engine/balder/remove-leftover-fs4-codeHenning Baldersheim2019-09-304-25/+3
|\ | | | | Balder/remove leftover fs4 code
| * Remove unused error codeHenning Baldersheim2019-09-301-1/+0
| |
| * GC some unused code related to fs4 protocol that has now gone missing.Henning Baldersheim2019-09-273-3/+1
| |
| * Remove TransportServerHenning Baldersheim2019-09-272-21/+2
| |
* | Merge pull request #10782 from vespa-engine/arnej/catch-network-exceptionsArne H Juul2019-09-301-3/+6
|\ \ | | | | | | catch exceptions in network setup
| * | catch exceptions in network setupArne Juul2019-09-241-3/+6
| | |
* | | Merge pull request #10794 from ↵Geir Storli2019-09-272-0/+8
|\ \ \ | |_|/ |/| | | | | | | | vespa-engine/toregge/add-filtering-support-to-attribute-combiner Add filtering support to attribute combiner dynamic field writer
| * | Move ownership of MatchingElements to GetDocsumsState.Tor Egge2019-09-272-7/+4
| | |
| * | Add filtering support to attribute combiner dynamic field writer.Tor Egge2019-09-252-1/+12
| |/
* / Remove fdispatchHenning Baldersheim2019-09-2668-10302/+0
|/
* introduce separate docsum matcher classHåvard Pettersen2019-09-237-120/+233
|
* added function identifying which elements matchedHåvard Pettersen2019-09-192-0/+35
| | | | | only a skeleton for now; outlines the interface between the summary generator and the matcher.
* Adjust defaults for lid space compaction job.Geir Storli2019-09-171-3/+3
| | | | | Run check for lid space compaction more frequently and reduce time for when to consider a remove batch operation as ongoing.
* Avoid shadowing of global variables.Tor Egge2019-09-161-0/+2
|
* Remove redundant move.Tor Egge2019-09-161-1/+1
|
* Style changes and function renaming in enum store.Geir Storli2019-09-131-3/+3
|
* implement and test unpacking iterators optimizerHåvard Pettersen2019-09-123-5/+484
|
* added empty skeleton for unpacking iterators optimizerHåvard Pettersen2019-09-126-15/+67
|