ci: Refactor sqllogictests (#96)

Xuanwo · web-flow · commit ceb313182c02 · 2025-12-28T22:38:16.000+08:00
* ci: Refactor sqllogictests

* ci: Fix MinIO sqllogictest failures

* test: Stop using removed lance_scan table function
diff --git a/AGENTS.md b/AGENTS.md
@@ -106,10 +106,12 @@ Uses DuckDB's `sqllogictest` format in `test/sql/`:
 - `require lance`: Load the extension
 
 Key test files:
-- `test/sql/lance.test` (scan, pushdown, explain diagnostics, replacement scan)
-- `test/sql/knn.test` (KNN API validation)
-- `test/sql/search.test` (vector/FTS/hybrid search surface)
-- `test/sql/lance_s3_minio.test` (S3 scan/search via DuckDB secrets, gated by `LANCE_TEST_S3=1`)
+- `test/sql/scan_smoke.test` (scan smoke + replacement scan + error handling)
+- `test/sql/scan_limit_offset_sampling.test` (LIMIT/OFFSET and TABLESAMPLE pushdown)
+- `test/sql/scan_filter_pushdown.test` (filter/projection pushdown + explain diagnostics)
+- `test/sql/search_functions.test` (vector/FTS/hybrid search surface)
+- `test/sql/index_ddl.test` (index DDL + index-backed query paths)
+- `test/sql/s3_scan_minio.test` (S3 scan/search via DuckDB secrets, gated by `LANCE_TEST_S3=1`)
 
 ## Common Issues
 
diff --git a/test/sql/README.md b/test/sql/README.md
@@ -0,0 +1,34 @@
+# SQLLogicTest Conventions
+
+This directory contains SQLLogicTest suites for the `lance` DuckDB extension.
+
+## File Naming
+
+Tests are grouped by prefix to keep related coverage discoverable:
+
+- `scan_*`: Scan/replacement scan behavior and rowid semantics
+- `pushdown_*`: Filter/predicate pushdown and related diagnostics
+- `search_*`: `lance_*_search` and `lance_fts` SQL surface
+- `index_*`: Index DDL and index-backed query paths
+- `optimizer_*`: Optimizer/statistics integration
+- `namespace_*`: `ATTACH ... (TYPE LANCE)` namespace mapping and table discovery
+- `dml_*`: Write-path SQL (COPY/INSERT/UPDATE/DELETE/TRUNCATE/DROP/ALTER)
+- `s3_*`: End-to-end S3 tests (MinIO), gated by `require-env`
+- `tpch.test`: TPC-H correctness suite
+- `bench_*`: Larger correctness suites and fixtures (e.g., BigANN)
+
+## Header and Layout
+
+Each file starts with:
+
+- `# name: ...`
+- `# description: ...`
+- `# group: [sql]`
+
+For environment-gated tests, use this order:
+
+1. `require-env ...`
+2. `test-env ...` (separate each entry with an empty line)
+3. `require lance`
+
+Prefer small, focused files with section comments for setup, correctness, and diagnostics.
diff --git a/test/sql/bench_bigann_tiny.test b/test/sql/bench_bigann_tiny.test
@@ -1,4 +1,4 @@
-# name: test/sql/bigann_tiny.test
+# name: test/sql/bench_bigann_tiny.test
 # description: Big-ANN style fixture coverage (exact + prefilter + index path)
 # group: [sql]
 
diff --git a/test/sql/dml_alter_table_schema_evolution.test b/test/sql/dml_alter_table_schema_evolution.test
@@ -1,4 +1,4 @@
-# name: test/sql/alter_table_schema_evolution.test
+# name: test/sql/dml_alter_table_schema_evolution.test
 # description: ALTER TABLE schema evolution for attached Lance tables
 # group: [sql]
 
diff --git a/test/sql/dml_copy_to.test b/test/sql/dml_copy_to.test
@@ -1,4 +1,4 @@
-# name: test/sql/lance_write.test
+# name: test/sql/dml_copy_to.test
 # description: COPY TO (FORMAT lance) writes datasets
 # group: [sql]
 
diff --git a/test/sql/dml_delete.test b/test/sql/dml_delete.test
@@ -1,4 +1,4 @@
-# name: test/sql/lance_delete.test
+# name: test/sql/dml_delete.test
 # description: DELETE FROM ... WHERE ... deletes rows in Lance datasets
 # group: [sql]
 
diff --git a/test/sql/dml_drop_table.test b/test/sql/dml_drop_table.test
@@ -1,4 +1,4 @@
-# name: test/sql/drop_table.test
+# name: test/sql/dml_drop_table.test
 # description: DROP TABLE deletes Lance datasets in attached directory namespaces
 # group: [sql]
 
diff --git a/test/sql/dml_insert.test b/test/sql/dml_insert.test
@@ -1,4 +1,4 @@
-# name: test/sql/lance_insert.test
+# name: test/sql/dml_insert.test
 # description: INSERT INTO appends to Lance datasets
 # group: [sql]
 
@@ -70,4 +70,3 @@ SELECT sum(id) FROM ns.main.lance_insert_append;
 
 statement ok
 DETACH ns;
-
diff --git a/test/sql/dml_truncate.test b/test/sql/dml_truncate.test
@@ -1,4 +1,4 @@
-# name: test/sql/truncate_table.test
+# name: test/sql/dml_truncate.test
 # description: TRUNCATE TABLE for Lance namespace tables
 # group: [sql]
 
diff --git a/test/sql/dml_update.test b/test/sql/dml_update.test
@@ -1,4 +1,4 @@
-# name: test/sql/lance_update.test
+# name: test/sql/dml_update.test
 # description: UPDATE ... SET ... WHERE updates Lance datasets
 # group: [sql]
 
diff --git a/test/sql/lance.test b/test/sql/lance.test
diff --git a/test/sql/namespace_attach.test b/test/sql/namespace_attach.test
@@ -1,5 +1,5 @@
-# name: test/sql/attach_namespace.test
-# description: Test ATTACH TYPE LANCE namespace mapping
+# name: test/sql/namespace_attach.test
+# description: ATTACH TYPE LANCE namespace mapping
 # group: [sql]
 
 require lance
diff --git a/test/sql/namespace_create_table.test b/test/sql/namespace_create_table.test
@@ -1,4 +1,4 @@
-# name: test/sql/attach_namespace_create_table.test
+# name: test/sql/namespace_create_table.test
 # description: CREATE TABLE in attached LANCE directory namespace
 # group: [sql]
 
diff --git a/test/sql/namespace_dynamic_discovery.test b/test/sql/namespace_dynamic_discovery.test
@@ -1,4 +1,4 @@
-# name: test/sql/attach_namespace_dynamic_discovery.test
+# name: test/sql/namespace_dynamic_discovery.test
 # description: Dynamic table discovery in attached LANCE directory namespace
 # group: [sql]
 
diff --git a/test/sql/namespace_rest_attach.test b/test/sql/namespace_rest_attach.test
@@ -1,11 +1,13 @@
-# name: test/sql/lance_namespace_rest.test
+# name: test/sql/namespace_rest_attach.test
 # description: End-to-end Lance REST namespace attach (requires external service)
 # group: [sql]
 
 require-env LANCE_TEST_NAMESPACE 1
 
 test-env LANCE_NAMESPACE_ENDPOINT http://127.0.0.1:2333
+
 test-env LANCE_NAMESPACE_ID default
+
 test-env LANCE_NAMESPACE_TABLE default.table
 
 require lance
diff --git a/test/sql/optimizer_statistics_cardinality.test b/test/sql/optimizer_statistics_cardinality.test
@@ -1,4 +1,4 @@
-# name: test/sql/statistics_cardinality.test
+# name: test/sql/optimizer_statistics_cardinality.test
 # description: Optimizer statistics hooks (cardinality/partition stats)
 # group: [sql]
 
diff --git a/test/sql/pushdown_filter_ir_types.test b/test/sql/pushdown_filter_ir_types.test
@@ -1,4 +1,4 @@
-# name: test/sql/filter_ir_types.test
+# name: test/sql/pushdown_filter_ir_types.test
 # description: Filter IR supports timestamps/decimals/struct fields
 # group: [sql]
 
diff --git a/test/sql/pushdown_is_distinct_from.test b/test/sql/pushdown_is_distinct_from.test
@@ -1,4 +1,4 @@
-# name: test/sql/distinct_from_pushdown.test
+# name: test/sql/pushdown_is_distinct_from.test
 # description: Filter IR supports IS DISTINCT FROM / IS NOT DISTINCT FROM
 # group: [sql]
 
diff --git a/test/sql/pushdown_like.test b/test/sql/pushdown_like.test
@@ -1,4 +1,4 @@
-# name: test/sql/like_pushdown.test
+# name: test/sql/pushdown_like.test
 # description: LIKE/ILIKE filter pushdown
 # group: [sql]
 
diff --git a/test/sql/pushdown_regex.test b/test/sql/pushdown_regex.test
@@ -1,4 +1,4 @@
-# name: test/sql/regex_pushdown.test
+# name: test/sql/pushdown_regex.test
 # description: regexp_matches/SIMILAR TO filter pushdown
 # group: [sql]
 
diff --git a/test/sql/s3_drop_table_minio.test b/test/sql/s3_drop_table_minio.test
@@ -1,4 +1,4 @@
-# name: test/sql/drop_table_s3_minio.test
+# name: test/sql/s3_drop_table_minio.test
 # description: DROP TABLE deletes Lance datasets in S3 directory namespaces (MinIO)
 # group: [sql]
 
@@ -45,7 +45,7 @@ COPY (
 ) TO 's3://${LANCE_S3_BUCKET}/${LANCE_S3_DROP_DATASET_PREFIX}' (FORMAT lance, mode 'overwrite');
 
 statement ok
-ATTACH 's3://${LANCE_S3_BUCKET}/' AS ns (TYPE LANCE);
+ATTACH 's3://${LANCE_S3_BUCKET}/' AS ns (TYPE LANCE, READ_WRITE true);
 
 query I
 SELECT count(*) FROM ns.main.drop_table_test_data;
@@ -65,4 +65,3 @@ DROP TABLE IF EXISTS ns.main.drop_table_test_data;
 
 statement ok
 DETACH ns;
-
diff --git a/test/sql/s3_scan_minio.test b/test/sql/s3_scan_minio.test
@@ -1,4 +1,4 @@
-# name: test/sql/lance_s3_minio.test
+# name: test/sql/s3_scan_minio.test
 # description: End-to-end S3 scan against MinIO using DuckDB secrets
 # group: [sql]
 
diff --git a/test/sql/s3_write_minio.test b/test/sql/s3_write_minio.test
@@ -1,4 +1,4 @@
-# name: test/sql/lance_write_s3_minio.test
+# name: test/sql/s3_write_minio.test
 # description: End-to-end S3 write against MinIO using DuckDB secrets
 # group: [sql]
 
diff --git a/test/sql/scan_filter_pushdown.test b/test/sql/scan_filter_pushdown.test
diff --git a/test/sql/scan_limit_offset_sampling.test b/test/sql/scan_limit_offset_sampling.test
diff --git a/test/sql/scan_rowid_in.test b/test/sql/scan_rowid_in.test
diff --git a/test/sql/scan_rowid_native.test b/test/sql/scan_rowid_native.test
diff --git a/test/sql/scan_smoke.test b/test/sql/scan_smoke.test
diff --git a/test/sql/search_functions.test b/test/sql/search_functions.test

Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# name: test/sql/bigann_tiny.test`
	`1`	`+# name: test/sql/bench_bigann_tiny.test`
`2`	`2`	`# description: Big-ANN style fixture coverage (exact + prefilter + index path)`
`3`	`3`	`# group: [sql]`
`4`	`4`
Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# name: test/sql/alter_table_schema_evolution.test`
	`1`	`+# name: test/sql/dml_alter_table_schema_evolution.test`
`2`	`2`	`# description: ALTER TABLE schema evolution for attached Lance tables`
`3`	`3`	`# group: [sql]`
`4`	`4`
Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# name: test/sql/lance_write.test`
	`1`	`+# name: test/sql/dml_copy_to.test`
`2`	`2`	`# description: COPY TO (FORMAT lance) writes datasets`
`3`	`3`	`# group: [sql]`
`4`	`4`
Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# name: test/sql/lance_delete.test`
	`1`	`+# name: test/sql/dml_delete.test`
`2`	`2`	`# description: DELETE FROM ... WHERE ... deletes rows in Lance datasets`
`3`	`3`	`# group: [sql]`
`4`	`4`
Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# name: test/sql/drop_table.test`
	`1`	`+# name: test/sql/dml_drop_table.test`
`2`	`2`	`# description: DROP TABLE deletes Lance datasets in attached directory namespaces`
`3`	`3`	`# group: [sql]`
`4`	`4`
Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# name: test/sql/truncate_table.test`
	`1`	`+# name: test/sql/dml_truncate.test`
`2`	`2`	`# description: TRUNCATE TABLE for Lance namespace tables`
`3`	`3`	`# group: [sql]`
`4`	`4`
Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# name: test/sql/lance_update.test`
	`1`	`+# name: test/sql/dml_update.test`
`2`	`2`	`# description: UPDATE ... SET ... WHERE updates Lance datasets`
`3`	`3`	`# group: [sql]`
`4`	`4`