Repository navigation
tests: run 13 vector tests on the native build through get_metadata() - #168
Merged
Merged
Conversation
…A-JVM tests run on both builds The pq, id_property, quantization, store_vectors_in_graph, add_hierarchy and cache tests, the two TypeIndex checks of the approximate searches and the two max_connections default tests read a Java index through _java_index or the JPype classes. get_metadata() returns the same values on the native build, and an index SQL creates without maxConnections shows the engine default on both. Their NATIVE_NA_JVM_TESTS entries go. test_encoding_param stays (encoding is not in get_metadata()); the WAL tests stay (no public read of the WAL setting on either build). Against the 26.11.1.dev0 dev wheels (old pair and the new dd45b53 pair) all 13 rewritten tests pass on native and JVM. Closes #167 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #167
Thirteen vector tests were skipped on the native build (
NATIVE_NA_JVM_TESTS) only because they read the index configuration from the Java index behind the wrapper. They now read it throughVectorIndex.get_metadata()/get_quantization(), which return the same values on both builds, and their list entries are removed.test_vector.py: PQ parameters, custom id property, and the two approximate-search tests (the TypeIndex check is nowindex_nameagainstbucket_index_name, plusget_quantization() == "PRODUCT").test_vector_params_verification.py: the four quantization types,store_vectors_in_graph,add_hierarchy, the per-index cache settings, and the twomax_connectionsdefault tests (the engine default is read from an index SQL creates withoutmaxConnections, instead ofGlobalConfigurationthrough JPype).toString()are gone; each setting has one direct assert, andstore_vectors_in_graphandadd_hierarchygained a negative control (a default index reads backFalse).test_encoding_param(encodingis not inget_metadata()on either build), the tests about the JVM itself, and the two WAL tests (no public read of the WAL setting on either build).Verification: both whole files against the 26.11.1.dev0 dev wheels, on the old pair and on a pair built from main
dd45b53947: JVM 60 passed; native passes all 13 rewritten tests. One unrelated native failure shows on those local dev wheels only,test_lsm_vector_search_with_filter(str(record.get_identity())of an@ridthat the enginee353330c5csends as an RID object, wrapped twice); native CI builds against a newer engine and passes it, and the one-line fix is #170.🤖 Generated with Claude Code