20222860826
53%

Ran 15 Dec 2025 06:36AM UTC

Jobs 1

Files 417

Run time 1min

Badge

Embed ▾

Committed 15 Dec 2025 06:35AM UTC coverage: 53.731% (+0.02%) from 53.708%

Build # 20222860826

Build Type

push

github

Committed by

web-flow

Commit Message

MB-69641: Add support for retrieval of multi-vector documents (#2258)

- KNN search can now be performed on documents that contain a field with
multiple vectors (e.g., vectors in JSON object arrays or arrays of
vectors). For such documents, we select only the best-scoring vector and
discard the rest, assigning the document the similarity score of that
best vector.
- Added unit tests to verify this behavior.
- Fixed a bug where using a filter query that marked every document as
ineligible in a multi-vector query could corrupt the internal state of
the KNN collector.
- Fixed a bug where a KNN query was incorrectly marked as non-fieldable
due to not adhering to the `FieldableQuery` interface.
- Requires: https://github.com/blevesearch/zapx/pull/348

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Abhinav Dangeti <abhinav@couchbase.com>

Run Details

19098 of 35544 relevant lines covered (53.73%)

0.6 hits per line

Uncovered Existing Lines

Lines	Coverage	∆	File
2	92.29	1.54%	index/scorch/introducer.go

Jobs

ID	Job ID	Ran	Files	Coverage
1	20222860826.1	15 Dec 2025 06:36AM UTC	417	53.73	GitHub Action Run

blevesearch / bleve / 20222860826
53%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Uncovered Existing Lines

Jobs

Source Files on build 20222860826

blevesearch / bleve / 20222860826 53%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Uncovered Existing Lines

Jobs

Source Files on build 20222860826

blevesearch / bleve / 20222860826
53%

README BADGES
x