sql/opt: add tests for using canary stats rollout in makeTableStatistics

This commit adds tests for usage of canary stats rollout in
makeTableStatistics(), which is the main entry point where statistics
are selected for query optimization. This commit focuses on unit
testing the makeTableStatistics() function and does not include
end-to-end logic tests, which would require additional changes to
Builder.maybeAnnotateWithEstimates() to support EXPLAIN ANALYZE
output showing which statistics were used during planning.

To enable testing, this commit adds:
- Handler in opttester for setting the canary window storage
parameter
- Testing knob for controlling the canary fraction setting
- Three new test files covering basic canary stats, histogram canary
stats, and multi-column canary stats scenarios

Release note: None

Author: ZhouXing19

pkg/sql/opt/memo/testdata/stats/canary-stats

@@ -0,0 +1,135 @@
+exec-ddl
+CREATE TABLE sel (a INT PRIMARY KEY) WITH (canary_window = '30s')
+----
+
+# Single stats case: both canary and stable path should use the same stats.
+exec-ddl
+ALTER TABLE sel INJECT STATISTICS '[
+    {
+    	"columns": ["a"],
+      "created_at": "2018-01-01 11:00:00.00000+00:00",
+      "row_count": 20,
+      "distinct_count": 20
+    }
+	]'
+----
+
+opt stats-as-of=(2018-01-01 11:00:10.00000+00:00) canary-fraction=1.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=20]
+ ├── key: (1)
+ └── ordering: +1
+
+opt stats-as-of=(2018-01-01 11:00:10.00000+00:00) canary-fraction=0.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=20]
+ ├── key: (1)
+ └── ordering: +1
+
+
+# More than 2 stats available: should only consider the latest 2 stats as possible candidates.
+exec-ddl
+ALTER TABLE sel INJECT STATISTICS '[
+    {
+    	"columns": ["a"],
+      "created_at": "2018-01-01 11:00:40.00000+00:00",
+      "row_count": 30,
+      "distinct_count": 30
+    },
+    {
+      "columns": ["a"],
+      "created_at": "2018-01-01 11:00:00.00000+00:00",
+      "row_count": 20,
+      "distinct_count": 20
+    },
+    {
+      "columns": ["a"],
+      "created_at": "2018-01-01 10:59:40.00000+00:00",
+      "row_count": 10,
+      "distinct_count": 10
+    }
+	]'
+----
+
+opt stats-as-of=(2018-01-01 11:00:50.00000+00:00) canary-fraction=1.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=20]
+ ├── key: (1)
+ └── ordering: +1
+
+opt stats-as-of=(2018-01-01 11:00:50.00000+00:00) canary-fraction=0.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=20]
+ ├── key: (1)
+ └── ordering: +1
+
+# Optimize as of a later timestamp, so that latest stats is out of the
+# canary window, and canary stats has been promoted to stable stats.
+opt stats-as-of=(2018-01-01 11:01:30.00000+00:00) canary-fraction=1.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=30]
+ ├── key: (1)
+ └── ordering: +1
+
+opt stats-as-of=(2018-01-01 11:01:30.00000+00:00) canary-fraction=0.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=30]
+ ├── key: (1)
+ └── ordering: +1
+
+# Test time traveling of stats_as_of.
+opt stats-as-of=(2018-01-01 11:00:20.00000+00:00) canary-fraction=1.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=10]
+ ├── key: (1)
+ └── ordering: +1
+
+opt stats-as-of=(2018-01-01 11:00:20.00000+00:00) canary-fraction=0.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=10]
+ ├── key: (1)
+ └── ordering: +1
+
+# If travelling way too back in the past so that no stats are valid,
+# both paths should use the fake stats (with rows = 1000).
+opt stats-as-of=(2018-01-01 10:00:00.00000+00:00) canary-fraction=1.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=1000]
+ ├── key: (1)
+ └── ordering: +1
+
+opt stats-as-of=(2018-01-01 10:00:00.00000+00:00) canary-fraction=0.0
+SELECT a FROM sel ORDER BY a
+----
+scan sel
+ ├── columns: a:1(int!null)
+ ├── stats: [rows=1000]
+ ├── key: (1)
+ └── ordering: +1