Skip to content

Commit

Permalink
sql: omit dropped columns from SHOW STATISTICS output
Browse files Browse the repository at this point in the history 
Fixes #76573

Previously, statistics on dropped columns were displayed in `SHOW
STATISTICS [USING JSON]` output.

This was inadequate because statement bundle scripts with
`ALTER TABLE ... INJECT STATISTICS` statements, generated via
`SHOW STATISTICS USING JSON` output, required manual editing
in order to run successfully if the table in question had a dropped
column.

To address this, this patch omits printing of statistics involving
dropped columns in `SHOW STATISTICS` output, including multicolumn
statistics.

Release note (bug fix): This fixes SHOW STATISTICS output so statistics
involving dropped columns are not displayed.
  • Loading branch information
Mark Sirek committed Jun 1, 2022
1 parent 97d0c6e commit 70c3b24
Show file tree
Hide file tree
Showing 2 changed files with 127 additions and 18 deletions.
88 changes: 88 additions & 0 deletions pkg/sql/logictest/testdata/logic_test/distsql_stats
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1542,3 +1542,91 @@ SELECT jsonb_pretty(COALESCE(json_agg(stat), '[]'))
"row_count": 1
}
]

# Regression test for #76573
statement ok
CREATE TABLE t1 (a INT, b INT, c INT)

statement ok
ANALYZE t1

statement ok
CREATE STATISTICS t1_ab ON a,b FROM t1

statement ok
CREATE STATISTICS t1_ac ON a,c FROM t1

statement ok
CREATE STATISTICS t1_bc ON b,c FROM t1

statement ok
ALTER TABLE t1 drop column c

statement ok
show statistics for table t1

query TTIII colnames
SELECT
statistics_name,
column_names,
row_count,
distinct_count,
null_count
FROM
[SHOW STATISTICS FOR TABLE t1]
ORDER BY statistics_name, column_names::STRING
----
statistics_name column_names row_count distinct_count null_count
NULL {a} 0 0 0
NULL {b} 0 0 0
NULL {rowid} 0 0 0
t1_ab {a,b} 0 0 0

query T
SELECT jsonb_pretty(COALESCE(json_agg(stat), '[]'))
FROM (
SELECT json_array_elements(statistics) - 'created_at' - 'histo_col_type' - 'histo_version' AS stat
FROM [SHOW STATISTICS USING JSON FOR TABLE t1]
)
----
[
{
"avg_size": 0,
"columns": [
"rowid"
],
"distinct_count": 0,
"null_count": 0,
"row_count": 0
},
{
"avg_size": 0,
"columns": [
"a"
],
"distinct_count": 0,
"null_count": 0,
"row_count": 0
},
{
"avg_size": 0,
"columns": [
"b"
],
"distinct_count": 0,
"null_count": 0,
"row_count": 0
},
{
"avg_size": 0,
"columns": [
"a",
"b"
],
"distinct_count": 0,
"name": "t1_ab",
"null_count": 0,
"row_count": 0
}
]

57 changes: 39 additions & 18 deletions pkg/sql/show_stats.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@ import (
"github.com/cockroachdb/cockroach/pkg/sql/catalog/colinfo"
"github.com/cockroachdb/cockroach/pkg/sql/catalog/descpb"
"github.com/cockroachdb/cockroach/pkg/sql/sem/tree"
"github.com/cockroachdb/cockroach/pkg/sql/sqlerrors"
"github.com/cockroachdb/cockroach/pkg/sql/stats"
"github.com/cockroachdb/cockroach/pkg/sql/types"
"github.com/cockroachdb/cockroach/pkg/util/errorutil"
Expand Down Expand Up @@ -152,27 +153,37 @@ func (p *planner) ShowTableStats(ctx context.Context, n *tree.ShowTableStats) (p

v := p.newContainerValuesNode(columns, 0)
if n.UsingJSON {
result := make([]stats.JSONStatistic, len(rows))
for i, r := range rows {
result[i].CreatedAt = tree.AsStringWithFlags(r[createdAtIdx], tree.FmtBareStrings)
result[i].RowCount = (uint64)(*r[rowCountIdx].(*tree.DInt))
result[i].DistinctCount = (uint64)(*r[distinctCountIdx].(*tree.DInt))
result[i].NullCount = (uint64)(*r[nullCountIdx].(*tree.DInt))
result := make([]stats.JSONStatistic, 0, len(rows))
for _, r := range rows {
var statsRow stats.JSONStatistic
colIDs := r[columnIDsIdx].(*tree.DArray).Array
statsRow.Columns = make([]string, len(colIDs))
ignoreStatsRowWithDroppedColumn := false
for j, d := range colIDs {
statsRow.Columns[j], err = statColumnString(desc, d)
if err != nil && sqlerrors.IsUndefinedColumnError(err) {
ignoreStatsRowWithDroppedColumn = true
break
}
}
if ignoreStatsRowWithDroppedColumn {
continue
}
statsRow.CreatedAt = tree.AsStringWithFlags(r[createdAtIdx], tree.FmtBareStrings)
statsRow.RowCount = (uint64)(*r[rowCountIdx].(*tree.DInt))
statsRow.DistinctCount = (uint64)(*r[distinctCountIdx].(*tree.DInt))
statsRow.NullCount = (uint64)(*r[nullCountIdx].(*tree.DInt))
if avgSizeColVerActive {
result[i].AvgSize = (uint64)(*r[avgSizeIdx].(*tree.DInt))
statsRow.AvgSize = (uint64)(*r[avgSizeIdx].(*tree.DInt))
}
if r[nameIdx] != tree.DNull {
result[i].Name = string(*r[nameIdx].(*tree.DString))
statsRow.Name = string(*r[nameIdx].(*tree.DString))
}
colIDs := r[columnIDsIdx].(*tree.DArray).Array
result[i].Columns = make([]string, len(colIDs))
for j, d := range colIDs {
result[i].Columns[j] = statColumnString(desc, d)
}
if err := result[i].DecodeAndSetHistogram(ctx, &p.semaCtx, r[histIdx]); err != nil {
if err := statsRow.DecodeAndSetHistogram(ctx, &p.semaCtx, r[histIdx]); err != nil {
v.Close(ctx)
return nil, err
}
result = append(result, statsRow)
}
encoded, err := encjson.Marshal(result)
if err != nil {
Expand Down Expand Up @@ -200,8 +211,18 @@ func (p *planner) ShowTableStats(ctx context.Context, n *tree.ShowTableStats) (p
colIDs := r[columnIDsIdx].(*tree.DArray).Array
colNames := tree.NewDArray(types.String)
colNames.Array = make(tree.Datums, len(colIDs))
ignoreStatsRowWithDroppedColumn := false
var colName string
for i, d := range colIDs {
colNames.Array[i] = tree.NewDString(statColumnString(desc, d))
colName, err = statColumnString(desc, d)
if err != nil && sqlerrors.IsUndefinedColumnError(err) {
ignoreStatsRowWithDroppedColumn = true
break
}
colNames.Array[i] = tree.NewDString(colName)
}
if ignoreStatsRowWithDroppedColumn {
continue
}

histogramID := tree.DNull
Expand Down Expand Up @@ -243,12 +264,12 @@ func (p *planner) ShowTableStats(ctx context.Context, n *tree.ShowTableStats) (p
}, nil
}

func statColumnString(desc catalog.TableDescriptor, colID tree.Datum) string {
func statColumnString(desc catalog.TableDescriptor, colID tree.Datum) (colName string, err error) {
id := descpb.ColumnID(*colID.(*tree.DInt))
colDesc, err := desc.FindColumnWithID(id)
if err != nil {
// This can happen if a column was removed.
return "<unknown>"
return "<unknown>", err
}
return colDesc.GetName()
return colDesc.GetName(), nil
}

0 comments on commit 70c3b24

Please sign in to comment.