Add schema object dolt_status oracle coverage

[timsehn]

This adds a dedicated Dolt oracle suite for dolt_status schema-object behavior: index create/drop/replace, unique and composite indexes, column/schema changes, CHECK/UNIQUE/FK table definitions, mixed table/schema-object states, and filtered scans.

The new oracle exposed and this PR fixes:

• index-only schema changes were missing from dolt_status
• staging a named table could overwrite a different named staged entry when SQLite reused rootpages under FK DDL
• path reset of a staged dropped indexed table restored the table without its index entries
• fallback catalog serialization could not recover schema rows for unnamed index entries

Tests:

• make -C build -j$(sysctl -n hw.ncpu 2>/dev/null || nproc) DOLTLITE_PROLLY_CHECK=1 doltlite
• bash test/vc_oracle_status_schema_objects_test.sh build/doltlite dolt
• bash test/vc_oracle_status_test.sh build/doltlite dolt
• bash test/vc_oracle_add_test.sh build/doltlite dolt
• bash test/vc_oracle_reset_test.sh build/doltlite dolt
• bash test/vc_oracle_checkout_test.sh build/doltlite dolt
• bash test/vc_oracle_schemas_test.sh build/doltlite dolt
• bash test/vc_oracle_schema_diff_test.sh build/doltlite dolt
• bash test/vc_oracle_commit_test.sh build/doltlite dolt
• bash test/vc_oracle_diff_test.sh build/doltlite dolt

Fixes #1655

[github-actions]

Version-Control Performance Ceilings

Runs: median of 3 executions per benchmark, excluding fixture setup. The
working-set fixtures use 800 committed tables with 125 rows each
(100000 rows total); data-dirty cases update 1 row in 40
tables, and schema-dirty cases update 1 row plus add a column in 20 tables.
Branch tests use 300 branches, checkout uses a clean 400-table
branch switch over 100000 rows, and merge
tests use 100000-row tables with 2000 changed or conflicting
rows per side.

Benchmark	Median ms	Ceiling ms	Result
status_clean_many_tables	62	200	PASS
status_dirty_many_tables	63	200	PASS
diff_regular_working_one_table	58	150	PASS
diff_regular_working_many_tables	62	200	PASS
diff_stat_working_many_tables	62	200	PASS
diff_schema_working_many_tables	63	200	PASS
branch_list_many_branches	17	100	PASS
branch_create_delete	30	100	PASS
checkout_branch_clean	104	200	PASS
merge_data_no_conflicts	36	150	PASS
merge_schema_no_conflicts	18	100	PASS
merge_data_conflicts	98	250	PASS
merge_data_conflicts_with_resolve	98	250	PASS

[github-actions]

Sysbench-Style Benchmark: Doltlite vs SQLite

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	24,266	27,044	1.11
oltp_range_select	10,671	11,928	1.12
oltp_sum_range	9,342	11,349	1.21
oltp_order_range	2,623	2,910	1.11
oltp_distinct_range	3,716	4,057	1.09
oltp_index_scan	3,926	4,880	1.24
select_random_points	10,554	11,051	1.05
select_random_ranges	3,028	4,014	1.33
covering_index_scan	4,304	4,013	0.93
groupby_scan	31,818	34,488	1.08
index_join	5,908	7,885	1.33
index_join_scan	3,401	4,834	1.42
types_table_scan	1,108,528	1,236,178	1.12
table_scan	1,270,087	1,355,118	1.07
oltp_read_only	100,797	115,745	1.15
Average			1.16

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	181,666	240,752	1.33
oltp_insert	15,751	28,391	1.80
oltp_update_index	51,147	87,465	1.71
oltp_update_non_index	34,955	58,496	1.67
oltp_delete_insert	45,155	69,298	1.53
oltp_write_only	22,254	44,505	2.00
types_delete_insert	25,355	39,702	1.57
oltp_read_write	67,038	102,372	1.53
Average			1.64

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	106,613	55,842	0.52
oltp_range_select	19,695	14,739	0.75
oltp_sum_range	17,999	14,407	0.80
oltp_order_range	3,585	3,268	0.91
oltp_distinct_range	4,570	4,397	0.96
oltp_index_scan	12,081	8,070	0.67
select_random_points	18,334	14,470	0.79
select_random_ranges	11,463	6,928	0.60
covering_index_scan	12,910	7,179	0.56
groupby_scan	32,915	35,004	1.06
index_join	10,301	9,975	0.97
index_join_scan	4,243	5,313	1.25
types_table_scan	1,108,271	1,230,817	1.11
table_scan	1,268,674	1,353,919	1.07
oltp_read_only	219,936	156,886	0.71
Average			0.85

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	195,729	261,955	1.34
oltp_insert	21,742	35,746	1.64
oltp_update_index	76,996	114,729	1.49
oltp_update_non_index	58,289	80,173	1.38
oltp_delete_insert	66,997	94,259	1.41
oltp_write_only	42,438	64,895	1.53
types_delete_insert	40,365	52,777	1.31
oltp_read_write	86,222	121,248	1.41
Average			1.44

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	51,505	56,186	1.09
oltp_range_select	13,703	14,785	1.08
oltp_sum_range	12,415	14,236	1.15
oltp_order_range	3,082	3,253	1.06
oltp_distinct_range	4,050	4,376	1.08
oltp_index_scan	6,803	8,136	1.20
select_random_points	13,502	14,425	1.07
select_random_ranges	5,853	6,963	1.19
covering_index_scan	7,256	7,135	0.98
groupby_scan	32,268	34,790	1.08
index_join	7,582	10,016	1.32
index_join_scan	3,850	5,234	1.36
types_table_scan	1,104,468	1,226,408	1.11
table_scan	1,265,592	1,362,578	1.08
oltp_read_only	142,825	157,771	1.10
Average			1.13

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	14,873	59,888	4.03
oltp_insert_ac	17,269	78,020	4.52
oltp_update_index_ac	19,138	89,222	4.66
oltp_update_non_index_ac	15,357	69,236	4.51
oltp_delete_insert_ac	17,196	81,253	4.73
oltp_write_only_ac	17,072	80,723	4.73
types_delete_insert_ac	15,052	69,629	4.63
oltp_read_write_ac	22,540	90,295	4.01
Average			4.48

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark (TEXT PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 32-char hex TEXT PRIMARY KEY (UUID-shaped).
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	29,854	38,336	1.28
oltp_range_select	12,181	14,695	1.21
oltp_sum_range	11,173	14,128	1.26
oltp_order_range	2,837	3,183	1.12
oltp_distinct_range	3,968	4,145	1.04
oltp_index_scan	4,393	6,313	1.44
select_random_points	18,274	21,170	1.16
select_random_ranges	4,097	5,248	1.28
covering_index_scan	4,603	4,817	1.05
groupby_scan	31,690	33,851	1.07
index_join	6,734	9,325	1.38
index_join_scan	4,464	5,546	1.24
types_table_scan	1,183,813	1,250,631	1.06
table_scan	1,218,485	1,343,878	1.10
oltp_read_only	114,908	135,443	1.18
Average			1.19

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	231,868	354,697	1.53
oltp_insert	21,293	38,923	1.83
oltp_update_index	69,451	131,107	1.89
oltp_update_non_index	47,292	85,194	1.80
oltp_delete_insert	50,322	102,888	2.04
oltp_write_only	28,119	60,490	2.15
types_delete_insert	32,362	54,272	1.68
oltp_read_write	86,171	138,662	1.61
Average			1.82

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	99,473	61,885	0.62
oltp_range_select	20,446	17,458	0.85
oltp_sum_range	19,180	16,644	0.87
oltp_order_range	3,801	3,520	0.93
oltp_distinct_range	4,850	4,536	0.94
oltp_index_scan	11,886	8,909	0.75
select_random_points	25,309	24,127	0.95
select_random_ranges	10,795	7,700	0.71
covering_index_scan	11,394	7,209	0.63
groupby_scan	32,272	34,108	1.06
index_join	11,722	11,312	0.97
index_join_scan	5,494	5,923	1.08
types_table_scan	1,059,247	1,281,192	1.21
table_scan	1,424,313	1,346,958	0.95
oltp_read_only	228,494	177,253	0.78
Average			0.89

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	254,037	383,318	1.51
oltp_insert	49,310	52,940	1.07
oltp_update_index	113,482	168,469	1.48
oltp_update_non_index	86,197	112,152	1.30
oltp_delete_insert	93,085	134,690	1.45
oltp_write_only	98,740	83,634	0.85
types_delete_insert	56,424	77,318	1.37
oltp_read_write	140,075	166,498	1.19
Average			1.28

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	54,270	62,542	1.15
oltp_range_select	16,142	17,333	1.07
oltp_sum_range	14,886	16,724	1.12
oltp_order_range	3,487	3,507	1.01
oltp_distinct_range	4,432	4,542	1.02
oltp_index_scan	7,231	9,071	1.25
select_random_points	20,838	24,379	1.17
select_random_ranges	6,649	7,741	1.16
covering_index_scan	7,666	7,390	0.96
groupby_scan	31,942	34,231	1.07
index_join	8,693	10,948	1.26
index_join_scan	5,026	5,976	1.19
types_table_scan	1,047,811	1,274,397	1.22
table_scan	1,352,704	1,346,302	1.00
oltp_read_only	154,766	172,148	1.11
Average			1.12

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	23,938	90,261	3.77
oltp_insert_ac	27,844	107,334	3.85
oltp_update_index_ac	30,235	121,470	4.02
oltp_update_non_index_ac	24,918	106,527	4.28
oltp_delete_insert_ac	27,149	121,400	4.47
oltp_write_only_ac	27,668	110,947	4.01
types_delete_insert_ac	23,473	103,328	4.40
oltp_read_write_ac	32,218	115,199	3.58
Average			4.05

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark (composite PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 2-column INTEGER PRIMARY KEY(a, b) WITHOUT ROWID.
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	32,827	40,244	1.23
oltp_range_select	18,138	21,878	1.21
oltp_sum_range	17,572	20,617	1.17
oltp_order_range	3,409	3,852	1.13
oltp_distinct_range	4,500	4,883	1.09
oltp_index_scan	4,505	6,126	1.36
select_random_points	27,758	31,752	1.14
select_random_ranges	7,544	8,963	1.19
covering_index_scan	4,162	4,198	1.01
groupby_scan	35,852	39,390	1.10
index_join	8,117	10,378	1.28
index_join_scan	4,184	5,473	1.31
types_table_scan	1,051,699	1,260,393	1.20
table_scan	1,188,836	1,354,327	1.14
oltp_read_only	143,827	168,640	1.17
Average			1.18

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	246,013	355,497	1.45
oltp_insert	19,073	36,648	1.92
oltp_update_index	68,536	116,261	1.70
oltp_update_non_index	49,964	85,476	1.71
oltp_delete_insert	49,861	96,338	1.93
oltp_write_only	27,222	57,969	2.13
types_delete_insert	32,991	55,493	1.68
oltp_read_write	101,053	158,870	1.57
Average			1.76

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	102,266	65,636	0.64
oltp_range_select	25,581	24,567	0.96
oltp_sum_range	25,595	23,691	0.93
oltp_order_range	4,312	4,297	1.00
oltp_distinct_range	5,557	5,324	0.96
oltp_index_scan	11,593	9,222	0.80
select_random_points	38,086	36,656	0.96
select_random_ranges	15,421	11,775	0.76
covering_index_scan	11,187	7,122	0.64
groupby_scan	36,743	39,315	1.07
index_join	12,578	12,722	1.01
index_join_scan	5,165	5,929	1.15
types_table_scan	1,043,908	1,269,797	1.22
table_scan	1,285,578	1,365,498	1.06
oltp_read_only	255,645	210,921	0.83
Average			0.93

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	261,519	383,957	1.47
oltp_insert	26,306	46,041	1.75
oltp_update_index	98,770	146,684	1.49
oltp_update_non_index	78,464	105,408	1.34
oltp_delete_insert	77,467	121,640	1.57
oltp_write_only	50,378	78,720	1.56
types_delete_insert	50,846	69,331	1.36
oltp_read_write	129,650	179,058	1.38
Average			1.49

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	58,174	66,337	1.14
oltp_range_select	21,927	24,670	1.13
oltp_sum_range	20,660	23,953	1.16
oltp_order_range	4,002	4,323	1.08
oltp_distinct_range	5,002	5,322	1.06
oltp_index_scan	7,093	9,111	1.28
select_random_points	31,176	36,038	1.16
select_random_ranges	10,365	11,927	1.15
covering_index_scan	7,030	7,144	1.02
groupby_scan	36,464	39,531	1.08
index_join	9,767	12,786	1.31
index_join_scan	4,685	5,992	1.28
types_table_scan	1,075,881	1,274,094	1.18
table_scan	1,337,299	1,360,534	1.02
oltp_read_only	187,583	209,244	1.12
Average			1.14

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	22,453	84,664	3.77
oltp_insert_ac	25,320	105,256	4.16
oltp_update_index_ac	27,525	113,278	4.12
oltp_update_non_index_ac	23,518	92,565	3.94
oltp_delete_insert_ac	24,145	109,462	4.53
oltp_write_only_ac	25,481	107,095	4.20
types_delete_insert_ac	22,948	94,658	4.12
oltp_read_write_ac	32,225	109,776	3.41
Average			4.03

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[github-actions]

Sysbench-Style Benchmark (BLOB PK): Doltlite vs SQLite

Companion to the classic Sysbench-Style Benchmark. Every workload here
runs against tables with a 16-byte big-endian BLOB PRIMARY KEY.
Individual ratios gated at 2.5×; section averages gated at 2×. Autocommit writes use 6× / 5×.

In-Memory

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	30,195	37,111	1.23
oltp_range_select	12,250	14,532	1.19
oltp_sum_range	11,270	13,759	1.22
oltp_order_range	2,755	3,128	1.14
oltp_distinct_range	3,856	4,146	1.08
oltp_index_scan	4,234	6,089	1.44
select_random_points	17,239	20,850	1.21
select_random_ranges	3,804	5,246	1.38
covering_index_scan	4,341	4,396	1.01
groupby_scan	30,789	33,434	1.09
index_join	6,662	8,893	1.33
index_join_scan	3,832	5,526	1.44
types_table_scan	1,042,599	1,237,185	1.19
table_scan	1,176,528	1,326,606	1.13
oltp_read_only	114,658	135,271	1.18
Average			1.22

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	243,245	352,883	1.45
oltp_insert	19,758	38,703	1.96
oltp_update_index	66,730	126,416	1.89
oltp_update_non_index	47,961	83,012	1.73
oltp_delete_insert	48,353	100,416	2.08
oltp_write_only	28,257	60,195	2.13
types_delete_insert	32,321	52,933	1.64
oltp_read_write	81,578	136,811	1.68
Average			1.82

File-Backed

Reads

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	98,637	60,811	0.62
oltp_range_select	19,479	17,302	0.89
oltp_sum_range	18,650	16,444	0.88
oltp_order_range	3,663	3,529	0.96
oltp_distinct_range	4,727	4,642	0.98
oltp_index_scan	11,333	8,985	0.79
select_random_points	25,258	25,126	0.99
select_random_ranges	10,649	7,838	0.74
covering_index_scan	11,311	7,263	0.64
groupby_scan	31,536	34,101	1.08
index_join	10,700	11,591	1.08
index_join_scan	5,104	5,848	1.15
types_table_scan	1,041,897	1,246,620	1.20
table_scan	1,240,440	1,339,075	1.08
oltp_read_only	220,128	174,243	0.79
Average			0.93

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert	260,350	376,231	1.45
oltp_insert	31,019	51,460	1.66
oltp_update_index	99,319	161,635	1.63
oltp_update_non_index	75,280	107,990	1.43
oltp_delete_insert	79,758	129,651	1.63
oltp_write_only	55,559	82,627	1.49
types_delete_insert	50,845	72,791	1.43
oltp_read_write	112,791	161,847	1.43
Average			1.52

File-Backed (autocommit)

Each statement runs as its own transaction — exposes per-commit
fixed costs that the wrapped-in-BEGIN/COMMIT tests amortize away.
SQLite uses WAL mode with synchronous=FULL in this section so
the comparison uses SQLite's durable WAL autocommit path.

Reads

Reads have no commit cost; these are the same SQL files as the
File-Backed Reads section, included here for symmetry and to
catch any per-statement overhead doltlite pays on the read path.

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_point_select	54,106	61,565	1.14
oltp_range_select	15,840	17,589	1.11
oltp_sum_range	14,480	16,905	1.17
oltp_order_range	3,265	3,633	1.11
oltp_distinct_range	4,461	4,763	1.07
oltp_index_scan	6,588	8,983	1.36
select_random_points	20,405	24,592	1.21
select_random_ranges	6,379	7,802	1.22
covering_index_scan	7,601	7,248	0.95
groupby_scan	32,258	34,163	1.06
index_join	8,974	11,448	1.28
index_join_scan	4,705	6,048	1.29
types_table_scan	1,048,628	1,253,314	1.20
table_scan	1,184,958	1,327,912	1.12
oltp_read_only	154,880	174,616	1.13
Average			1.16

Writes

Test	SQLite (us)	Doltlite (us)	Multiplier
oltp_bulk_insert_ac	22,545	79,951	3.55
oltp_insert_ac	24,630	102,619	4.17
oltp_update_index_ac	27,767	110,669	3.99
oltp_update_non_index_ac	22,292	91,069	4.09
oltp_delete_insert_ac	24,628	101,770	4.13
oltp_write_only_ac	24,348	98,837	4.06
types_delete_insert_ac	21,270	88,734	4.17
oltp_read_write_ac	29,325	105,729	3.61
Average			3.97

100000 rows, median of 5 invocations per test; autocommit writes use 9, workload-only timing via host monotonic clock when available.

Performance Ceiling Check (2.5x individual, 2x average; autocommit writes: 6x / 5x)

All tests within ceilings.

[timsehn]

Reviewed and verified on the branch (checked build):

• The dolt_status does not attribute index changes to their table #1655 repro is fixed: CREATE INDEX and DROP INDEX on a committed table both report t | 0 | modified in dolt_status.
• Green against the branch: the new schema-objects status oracle 18/18, reset oracle vs Dolt 43/43, and the index×VC matrix 38/38 — the reset index-restoration and fallback-row changes compose cleanly with the staging fixes they build on.
• No file overlap with Attribute index changes to their table in the dolt_diff summary #1656 (diff-summary index attribution) — the two use the same schema-row attribution approach in disjoint surfaces; consolidating the two comparators into one shared helper is a reasonable follow-up, not a blocker.

Two things to address before merge:

1. Stale base — needs a rebase onto current master. The branch forked before Keep unstaged views and triggers out of named-add commits #1652 (view/trigger staging) merged: doltlite_commit.sh fails its two view-staging gates here (view_stays_out_of_named_commit, view_in_am_commit) purely because the branch lacks doltliteBuildNamedStageMasterRoot. The pairing-hardening hunk sits in the same addStageNamedTables region Keep unstaged views and triggers out of named-add commits #1652 touched, so after rebasing please re-run doltlite_commit.sh (56 cases) and doltlite_index_vc_matrix.sh to confirm the merge composes.

2. Minor, non-blocking: the new unnamedRootMatch arm still allows a named working table to pair with an unnamed staged entry on a bare number match — that's the pre-existing exposure (a cross-domain number collision would overwrite an index entry), narrowed but not eliminated for the unnamed case. Worth a follow-up tightening with the same by-schema-row resolution used elsewhere; also statusCompareIndexSchemaObjects runs two full catalog schema loads plus an O(from×to) name scan on every status query — fine at typical schema sizes, just noting it against the scaling suite's future appetite.