aboutsummaryrefslogtreecommitdiffstats
path: root/searchlib/src
Commit message (Collapse)AuthorAgeFilesLines
...
* | | DeinlineHenning Baldersheim2018-04-203-1/+12
| |/ |/|
* | Assert that attribute can be down-casted to tensor attribute.Geir Storli2018-04-191-1/+13
| |
* | Add type-safe down-cast to ITensorAttribute in IAttributeVector and use this ↵Geir Storli2018-04-1912-11/+36
| | | | | | | | instead of dynamic_cast.
* | Change ImportedTensorAttributeVector to no longer implement ITensorAttribute ↵Geir Storli2018-04-192-32/+4
| | | | | | | | | | | | API. All read access happens through short-lived ImportedTensorAttributeVectorReadGuard.
* | Add ReadableAttributeVector interface used to create a short-lived read ↵Geir Storli2018-04-1917-26/+133
| | | | | | | | | | | | guard over an attribute vector. Let AttributeVector and ImportedAttributeVector implement this interface.
* | Remove useReadGuard template argument from fixtures as read guard is always ↵Geir Storli2018-04-191-104/+18
| | | | | | | | used now.
* | Change ImportedAttributeVector to no longer implement IAttributeVector API.Geir Storli2018-04-1914-261/+159
|/ | | | All read access happens through short lived ImportedAttributeVectorReadGuard.
* Use ISearchContext instead of AttributeVector::SearchContext in ↵Geir Storli2018-04-183-17/+12
| | | | ImportedSearchContext.
* Explicit default destructor.Henning Baldersheim2018-04-162-7/+5
|
* Use more descriptive type name (EC -> EncodeContext).Tor Egge2018-04-131-2/+2
|
* Let L4SkipCheck inherit from L3SkipCheck (pagedict4 validation).Tor Egge2018-04-131-6/+3
|
* Add comments.Tor Egge2018-04-131-15/+22
| | | | Use more descriptive variable names.
* Remove dead code.Tor Egge2018-04-131-250/+0
|
* Rename test function name.Tor Egge2018-04-131-2/+2
|
* Replace some assert statements with test macros.Tor Egge2018-04-131-5/+4
|
* Simplify pagedict4 unit tests by moving ownership of encode and decodeTor Egge2018-04-1313-91/+147
| | | | contexts to shared test classes.
* Validate skip info when reading dictionary.Tor Egge2018-04-132-25/+306
|
* Handle being at end of dictionary page after reading counts.Tor Egge2018-04-131-1/+1
|
* Add unit test for pagedict4 reading where dictionary page is filled by a ↵Tor Egge2018-04-132-0/+214
| | | | single count.
* Factor out classes for read/write access to memory based pagedict4 structuresTor Egge2018-04-138-158/+271
| | | | used by unit tests.
* Update word number before setting up dictionary page and dictionaryTor Egge2018-04-131-7/+9
| | | | | | | sparse page for read. Setup dictionary sparse page for read before setting up dictionary page for read.
* Eliminate dead code.Tor Egge2018-04-131-10/+0
|
* Adjust comments.Tor Egge2018-04-132-3/+3
|
* allow proton to drop sort data when sending query resultsHåvard Pettersen2018-04-112-1/+10
|
* allow not unpacking the whitelist iteratorHåvard Pettersen2018-04-102-3/+5
|
* Merge pull request #5487 from ↵Henning Baldersheim2018-04-091-3/+3
|\ | | | | | | | | vespa-engine/toregge/use-active-lids-bitvector-for-whitelisting Use active lids bitvector for whitelisting.
| * Use active lids bitvector for whitelisting.Tor Egge2018-04-061-3/+3
| | | | | | | | This replaces nested small numeric attribute vector for blaclisting.
* | Nonfunctional changesJon Bratseth2018-04-062-1/+3
|/
* Use forward declared DocumentTypeRepo.Tor Egge2018-04-042-1/+1
|
* Stop using nested typedef for shared pointer to const DocumentTypeRepo.Tor Egge2018-04-041-2/+2
|
* Add newlineHenning Baldersheim2018-04-031-0/+1
|
* Remove extra white spaceHenning Baldersheim2018-04-031-1/+0
|
* make the test pass again as the code was already correct.Henning Baldersheim2018-04-031-1/+1
|
* Add missing newlineHenning Baldersheim2018-04-032-3/+2
|
* Nested namespace and default destructorsHenning Baldersheim2018-04-0326-164/+75
|
* Add test to prove that the filter aspect is correctly handled also for the ↵Henning Baldersheim2018-04-031-1/+22
| | | | DirectAttributeBlueprint
* Cleanup formatting.Henning Baldersheim2018-03-241-32/+18
|
* Rewrite test to new test syntaxHenning Baldersheim2018-03-241-61/+25
|
* Nonfunctional changes onlyJon Bratseth2018-03-191-2/+4
|
* Prefix macro names by modelName to avoid name conflictsJon Bratseth2018-03-1320-79/+75
|
* Prefix constants by model nameJon Bratseth2018-03-1311-46/+95
| | | | | | | | | Large constants are cross rank profiles. This avoids name conflicts when multiple models are used. It is not strictly necessary because the user can always disambiguate when chosing names, but there is a scenario where conflicts are plausible and leaving this to users is inconvenient: Multiple versions of the "same" model are tested in different rank profiles.
* Merge pull request #5267 from vespa-engine/lesters/tensorflow-broadcastingJon Bratseth2018-03-096-30/+218
|\ | | | | Make TensorFlow import joins compatible with broadcasting
| * Add test for tensorflow broadcastingLester Solbakken2018-03-085-5/+127
| |
| * Fix size comparisonLester Solbakken2018-03-081-1/+1
| |
| * Make TensorFlow import joins compatible with broadcastingLester Solbakken2018-03-081-25/+91
| |
* | Add TensorFlow variable converterJon Bratseth2018-03-084-10/+95
| |
* | Log OrderedTensorType of imported TensorFlow variablesJon Bratseth2018-03-081-3/+28
| |
* | OrderedTensorType.to/from specJon Bratseth2018-03-082-0/+39
| |
* | Refactor onlyJon Bratseth2018-03-082-24/+24
|/
* Merge pull request #5241 from vespa-engine/bratseth/tf-constants-in-parent-docLester Solbakken2018-03-0710-43/+53
|\ | | | | Bratseth/tf constants in parent doc