aboutsummaryrefslogtreecommitdiffstats
path: root/document
Commit message (Collapse)AuthorAgeFilesLines
...
* | Add test for proper setup and fix dirty retyping of ReferenceDataType target ↵Henning Baldersheim2018-09-123-0/+66
| | | | | | | | to also do renaming.
* | Never give up.Henning Baldersheim2018-09-111-5/+0
|/
* Add toStringHenning Baldersheim2018-09-112-0/+17
|
* Add explicit test to equals and compareTo and avoid * importsHenning Baldersheim2018-09-112-2/+58
|
* Due to some ugly inheritance you must use values with great care. In equals ↵Henning Baldersheim2018-09-103-7/+35
| | | | it must use entrySet() due to some very ugly inheritance.
* Also prune away ClearValue updates if possible.Henning Baldersheim2018-09-102-0/+54
|
* Use the cheaper equals instead of potentially expensive compareTo.Henning Baldersheim2018-09-101-1/+1
|
* Before compareTo on a hashmap makes sense it must be sorted.Henning Baldersheim2018-09-103-23/+84
|
* Switch to new javacc maven plugin:gjoranv2018-08-274-13/+10
| | | | | | | | - config model - document - documentapi - indexinglanguage. - searchlib
* Add example in comment for remove field path update class.Tor Egge2018-08-161-0/+9
|
* Stop iterating field path when variable value is out of array boundary.Tor Egge2018-08-151-1/+2
|
* Iterate over result list in reverse order to avoid staleTor Egge2018-08-152-2/+22
| | | | array indexes when removing array elements.
* Reverse iteration over selection result in an attempt toTor Egge2018-08-153-1/+32
| | | | | | | | | | | avoid stale array indexes being used as part of applying a remove field path update. This change assumes that variables stored in result list have increasing values. Add unit test for removal of multiple array elements using remove field path update.
* Stop iterating field path when variable value is out of array boundary.Tor Egge2018-08-151-0/+2
|
* Remove whitespaceHenning Baldersheim2018-08-1240-103/+103
|
* Remove extra whitespaceHenning Baldersheim2018-08-125-5/+5
|
* Avoid ambiguity in gcc 8.2Henning Baldersheim2018-08-111-1/+1
|
* Pass stringref by valueHenning Baldersheim2018-08-1131-39/+39
|
* Pass stringref by valueHenning Baldersheim2018-08-1054-170/+166
|
* Move try/catch statements to parseDocumentSelection helper function.Tor Egge2018-08-061-13/+15
|
* Consider parse error as invalid selection expression.Tor Egge2018-08-062-7/+26
|
* Don't string wrap refs that are guaranteed to be zero-terminatedTor Brede Vekterli2018-07-265-7/+7
|
* Remove stringref::c_str()Tor Brede Vekterli2018-07-2527-53/+66
| | | | | | | | | | | | | The expected semantics of c_str() (a null-terminated string) cannot be satisfied with a string reference, so remove the function entirely to prevent people from using it in buggy ways. Replaces c_str() with data() in places where it is presumed safe, otherwise constructs temporary string instances. Certain callsites have been de-stringref'd in favor of regular strings, in particular where C APIs have been transitively called. The vast majority of these were called with string parameters anyway, so should not cause much extra allocation.
* Merge pull request #6451 from ↵Henning Baldersheim2018-07-243-3/+39
|\ | | | | | | | | vespa-engine/vekterli/ignore-array-element-update-for-out-of-bounds-index Ignore array element update with index out of bounds
| * Ignore array element update with index out of boundsTor Brede Vekterli2018-07-233-3/+39
| | | | | | | | | | | | Only affects "element match" field value updates. Silently ignoring the update matches the semantics of semantically identical field path updates.
* | Merge pull request #6444 from vespa-engine/bratseth/java-model-inferenceJon Bratseth2018-07-231-5/+8
|\ \ | |/ |/| Bratseth/java model inference
| * Model inference initial commitJon Bratseth2018-06-281-5/+8
| |
* | Remove a restriction that is not necessary since FieldUpdates are applied first.Henning Baldersheim2018-07-201-2/+0
| |
* | Update javadoc.Henning Baldersheim2018-07-201-10/+10
| | | | | | | | Improve error message and produce it in one place.
* | Add a prune method that can be used to remove updates that will have no ↵Henning Baldersheim2018-07-202-0/+90
| | | | | | | | | | | | effect when applied. Currently will only affect ASSIGN updates.
* | Fix author tag for SimonBjørn Christian Seime2018-07-0538-38/+38
| |
* | Explicit UTF-8 charsetOlli Virtanen2018-06-281-1/+6
| |
* | Minor warning cleanupOlli Virtanen2018-06-271-4/+0
| |
* | Include .toJson() in Document analogously to .toXml()Olli Virtanen2018-06-272-5/+59
|/
* some more cleanupArne Juul2018-06-203-7/+7
|
* add and use HDR_ABORT insteadArne Juul2018-06-201-2/+3
|
* Merge branch 'master' into arnej/use-log-abortArne Juul2018-06-2039-806/+641
|\
| * Reuse common codeHenning Baldersheim2018-06-151-4/+2
| |
| * Only deserialize the docId on the initial deserialize of the header.Henning Baldersheim2018-06-142-76/+69
| | | | | | | | Just skip and verify on subsequent invocations.
| * Update commentsHenning Baldersheim2018-06-145-17/+7
| |
| * Update after comments from @vekterliHenning Baldersheim2018-06-1415-65/+26
| |
| * ByteBuffer -> nbostreamHenning Baldersheim2018-06-149-36/+30
| |
| * Remove unused versionHenning Baldersheim2018-06-1411-32/+19
| |
| * Remove unused version from apiHenning Baldersheim2018-06-1418-50/+34
| |
| * ByteBuffer -> nbostreamHenning Baldersheim2018-06-1418-120/+72
| |
| * Merge pull request #6185 from ↵Henning Baldersheim2018-06-148-352/+399
| |\ | | | | | | | | | | | | vespa-engine/balder/balder/prepare-lazy-deserialize-rebased-2 Balder/balder/prepare lazy deserialize rebased 2
| | * Address somment form @vekterliHenning Baldersheim2018-06-141-12/+22
| | |
| | * 1 - Use a backing buffer for the DocumentUpdate that always is source of truth.Henning Baldersheim2018-06-128-349/+386
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 2 - Use this buffer for re-serialization. 3 - Make deserialization lazy where possible. Currently lazy on replay and when arriving over the storageapi. Still needs to eager over documentapi. 4 - Deserialize eagerly in the persistence thread since those are plentyfull and not bottlenecked, instead of in the single master thread. Use real repo.
| * | Merge pull request #6188 from vespa-engine/balder/cache-the-default-doctypeGeir Storli2018-06-135-167/+98
| |\ \ | | | | | | | | Balder/cache the default doctype
| | * | Reduce code visibility.Henning Baldersheim2018-06-135-45/+23
| | | | | | | | | | | | | | | | Avoid do very frequent hash lookup for something that is always present.