sql: generate stats for multi-column inverted index columns

This commit allows stats to be created for columns in multi-column
inverted indexes.

Also, it fixes a bug the caused histograms to not be collected for
non-inverted columns referenced in partial inverted index predicates.

Release note (bug fix): Statistics are now correctly generated for
columns in multi-column inverted indexes and columns referenced in
partial inverted index predicates.

pkg/sql/create_stats.go

@@ -377,9 +377,10 @@ func createStatsDefaultColumns(
 
 	// Add column stats for each secondary index.
 	for _, idx := range desc.PublicNonPrimaryIndexes() {
-		isInverted := idx.GetType() == descpb.IndexDescriptor_INVERTED
+		for j, n := 0, idx.NumColumns(); j < n; j++ {
+			// Only the last column of an inverted index is an inverted column.
+			isInverted := idx.GetType() == descpb.IndexDescriptor_INVERTED && j == n-1
 
-		for j := 0; j < idx.NumColumns(); j++ {
 			// Generate stats for each indexed column.
 			addIndexColumnStatsIfNotExists(idx.GetColumnID(j), isInverted)
 
@@ -420,6 +421,11 @@ func createStatsDefaultColumns(
 
 			// Generate stats for each column individually.
 			for _, colID := range colIDs.Ordered() {
+				col, err := desc.FindColumnByID(colID)
+				if err != nil {
+					return nil, err
+				}
+				isInverted := colinfo.ColumnTypeIsInvertedIndexable(col.Type)
 				addIndexColumnStatsIfNotExists(colID, isInverted)
 			}
 		}

pkg/sql/distsql_plan_stats.go

@@ -131,7 +131,7 @@ func (dsp *DistSQLPlanner) createStatsPlan(
 			// with different configurations). See #50655.
 			col := s.columns[0]
 			for _, index := range desc.PublicNonPrimaryIndexes() {
-				if index.GetType() == descpb.IndexDescriptor_INVERTED && index.GetColumnID(0) == col {
+				if index.GetType() == descpb.IndexDescriptor_INVERTED && index.InvertedColumnID() == col {
 					spec.Index = index.IndexDesc()
 					break
 				}

pkg/sql/logictest/testdata/logic_test/distsql_stats

@@ -798,8 +798,16 @@ upper_bound  range_rows  distinct_range_rows  equal_rows
 # Test that columns referenced in partial index predicates always have
 # histograms collected for them, with up to 200 buckets.
 statement ok
-CREATE TABLE partial (a INT, b INT, c INT, INDEX (a) WHERE b > 0 OR c > 0);
-INSERT INTO partial VALUES (1, 1, 1), (2, 2, 2), (3, 3, 3);
+CREATE TABLE partial (
+  a INT,
+  b INT,
+  c INT,
+  d INT,
+  j JSON,
+  INDEX (a) WHERE b > 0 OR c > 0,
+  INVERTED INDEX (j) WHERE d = 10
+);
+INSERT INTO partial VALUES (1, 1, 1, 1, '{"a": "b"}'), (2, 2, 2, 10, '{"c": "d"}'), (3, 3, 3, 1, '{"e": "f"}');
 CREATE STATISTICS s FROM partial
 
 query TTIIB colnames,rowsort
@@ -818,6 +826,8 @@ statistics_name  column_names  row_count  null_count  has_histogram
 s                {a}           3          0           true
 s                {b}           3          0           true
 s                {c}           3          0           true
+s                {d}           3          0           true
+s                {j}           3          0           true
 s                {rowid}       3          0           true
 
 let $hist_id_1
@@ -960,6 +970,31 @@ upper_bound
 '\x42fd4700000000000000000000000000000000bcc00000000000003ffbecde5da115a83ff661bdc396bcdc'  0           0                    1
 '\x42fd5ad4000000000000000000000000000000bcc00000000000003ffbecde5da115a83ff661bdc396bcdc'  0           0                    1
 
+# Stats for multi-column inverted indexes.
+statement ok
+CREATE TABLE multi_col (
+  id INT PRIMARY KEY,
+  s STRING,
+  j JSON,
+  INVERTED INDEX (s, j)
+);
+INSERT INTO multi_col VALUES (1, 'foo', '{"a": "b"}');
+CREATE STATISTICS s FROM multi_col;
+
+query TB colnames
+SELECT
+  column_names,
+  histogram_id IS NOT NULL AS has_histogram
+FROM
+  [SHOW STATISTICS FOR TABLE multi_col]
+ORDER BY
+  column_names::STRING, created
+----
+column_names  has_histogram
+{id}          true
+{j}           true
+{s}           true
+
 # Regression test for #56356. Histograms on all-null columns should not cause
 # an error.
 statement ok