Sync: BioPortal v6.8.0 release #18
Draft
alexskr wants to merge 240 commits into
Draft
Conversation
…gies_api into feature/user-admin-endpoints
Feature: Ontology Admin Endpoints for Logs and Submission Management
…ests Replace non-idiomatic assertions with Minitest equivalents: - assert(x == y) -> assert_equal - .eql?() -> assert_equal - assert !x.empty? -> assert_operator x.length, :>, 0 - submissions.all? -> submissions.each for better error messages Extract repeated contact validation into helper method to reduce duplication. Fix intermittent test failures by sorting array keys before comparison to handle non-deterministic ordering. refs #197
- Replace class variables (@@) with class instance variables using accessors - Add attr_accessor definitions for shared test state - Modernize assertions to Minitest style (assert_equal, assert_includes, assert_nil) - Improve code formatting and readability
…dition - Replace random .sample(3) with deterministic .first(3) - Use dynamic assertion for keep_ids count instead of hardcoded value - Capture process_id before use for clarity - Simplify polling loop condition to check only for 'done' or 'errors' - Clean up assertion message formatting for consistency
…n/deletion - Replace `assert 201, ...` with `assert_equal 201, ...` to properly compare response status. - Add GET checks after slice creation to confirm the new slice exists. - Remove redundant manual reset of LinkedData.settings.enable_security. - Update DELETE test to expect 204 (No Content) instead of 201. - Add GET checks after deletion to verify the slice was actually removed.
- Add explicit flunk when status_payload contains 'errors' to prevent false positives from incomplete or invalid responses - Ensure test fails immediately on backend error instead of silently skipping assertions
Restore LinkedData.settings.enable_slices to its original value after tests.
…com:ncbo/ontologies_api into refactor/tests-stability-and-minitest-style
… user helpers This change refactors `TestOntologySubmissionsController` to eliminate suite-level globals and class-variable state in favor of per-test setup and shared helpers: - Replace `before_suite` + `@@acronym`/`@@name`/`@@file_params` with per-test `setup` that generates a unique ontology acronym and creates an ontology administered by a shared test user. - Introduce a suite-local `USERNAME = "test_user"` and use `ensure_user(USERNAME)` to guarantee the user exists without races. - Normalize assertion style: - Use `assert_equal(expected, actual, message)` with parentheses - Remove `msg=` pseudo-named args - Add `.sort` where appropriate when comparing unordered keys - Replace direct references to `@@acronym`/`@@name` with instance vars (`@acronym`, `@name`). New helper methods (in the base test class) used by this suite: - `create_user(username, email:, password:)` — always creates a user - `ensure_user(username, email:, password:)` — idempotently finds or creates - `delete_user(username)` — removes a user if present Notes: - `delete_ontologies_and_submissions` now runs in `setup` to guarantee a clean slate per test. `after_all` also cleans up the shared user via `delete_user(USERNAME)`. If the extra cleanup proves redundant, we can drop `before_all`/`after_all` and keep only the per-test cleanup.
…st-style Refactor/tests stability and minitest style
Develop -> Master merge, release v7.3.1
Apply boost=sum(ontologyRank,1) on all three edismax paths via the common parameter section. Multiplier is (1 + rank), bounded in [1, 2]; unranked ontologies are unaffected. Rank now participates in Solr scoring before pagination, addressing #230. Bump goo, ncbo_cron, and ontologies_linked_data gem pins to the matching fix/incorrect-search-result-ranking branches. The new boost references the ontologyRank Solr field which is only declared in the ontologies_linked_data branch's schema. The post-hoc Ruby tiebreaker remains in place for this commit and will be removed in the next.
Solr is now the single source of result ordering via the boost added in the previous commit. The in-page Ruby tiebreaker is redundant and would mask any Solr-side scoring regression. The doc[:ontology_rank] attachment and Ontology.rank fetch upstream become unused; cleaned up in the next commit.
The post-hoc Ruby sort was the sole consumer of these. The line-293 skip-list entry (filter_attrs_by_language) was a defensive exclusion to prevent ontology_rank from being treated as a language-suffixed attribute during response serialization, also dead now.
Seeds four ontologies with synthetic BioPortal ranks (1.0, 0.7, 0.4, 0.1) into a clean Solr index, queries /search?q=melanoma, and asserts the returned acronyms come back in descending rank order. With Solr's boost=sum(ontologyRank,1) participating before pagination, the ordering reflects ontology rank from the very first page returned by Solr. Adds two minimal OWL fixtures (search_rank_melanoma_upper.owl with "Melanoma", search_rank_melanoma_lower.owl with "melanoma") that exercise the string_ci prefLabelExact field type and guarantee a case-insensitive match for the lowercase query.
The previous test method bundled two distinct invariants: the Phase 1 schema fix (string_ci prefLabelExact / synonymExact for case-insensitive exact match) and the Phase 2 mechanism (Solr-side boost producing rank-ordered results before pagination). Splitting them gives more diagnostic information when something regresses and makes each test's intent explicit. - test_schema_uses_case_insensitive_string_ci_exact_fields seeds only RANKHIGH (one ontology with mixed-case "Melanoma" prefLabel) and asserts the schema field types plus a lowercase prefLabelExact match. - test_search_rank_orders_results_via_solr_side_boost seeds all four ranked ontologies and asserts pagesize=3 returns the three highest ranks in order, with RANKLOW counted in totalCount but excluded from page 1. Shared setup is factored into a private with_ranked_melanoma_ontologies helper. Each test method now has a docstring explaining the mechanism it exercises (Phase 1 vs Phase 2) and the coverage gap the existing fixtures don't address (the original #230 cross-Solr-page failure mode would need 50+ matching docs to reproduce). Both tests verified against the live setup; original combined test's asserted behavior is preserved.
…ch-result-ranking # Conflicts: # Gemfile.lock
Fix: incorrect search result rankings
The prefLabelExact/synonymExact fields now use the string_ci_exact field type (case-insensitive exact match + binary scoring via omitTermFreqAndPositions). The schema assertions still expected the old string_ci literal, failing CI on develop after the type change merged. Case-insensitivity is preserved, so the functional check is unchanged.
Fix: malformed Solr query in Annotator (#233)
Fix: restore production access log output
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
3 participants
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.
Sync BioPortal v6.8.0 release