(core) Add BulkAddOrUpdateRecord action for efficiency

Summary:
This diff adds a new `BulkAddOrUpdateRecord` user action which is what is sounds like:

- A bulk version of the existing `AddOrUpdateRecord` action.
- Much more efficient for operating on many records than applying many individual actions.
- Column values are specified as maps from `colId` to arrays of values as usual.
- Produces bulk versions of `AddRecord` and `UpdateRecord` actions instead of many individual actions.

Examples of users wanting to use something like `AddOrUpdateRecord` with large numbers of records:

- https://grist.slack.com/archives/C0234CPPXPA/p1651789710290879
- https://grist.slack.com/archives/C0234CPPXPA/p1660743493480119
- https://grist.slack.com/archives/C0234CPPXPA/p1660333148491559
- https://grist.slack.com/archives/C0234CPPXPA/p1663069291726159

I tested what made many `AddOrUpdateRecord` actions slow in the first place. It was almost entirely due to producing many individual `AddRecord` user actions. About half of that time was for processing the resulting `AddRecord` doc actions. Lookups and updates were not a problem. With these changes, the slowness is gone.

The Python user action implementation is more complex but there are no surprises. The JS API now groups `records` based on the keys of `require` and `fields` so that `BulkAddOrUpdateRecord` can be applied to each group.

Test Plan: Update and extend Python and DocApi tests.

Reviewers: jarek, paulfitz

Reviewed By: jarek, paulfitz

Subscribers: jarek

Differential Revision: https://phab.getgrist.com/D3642
This commit is contained in:
Alex Hall 2022-09-28 15:13:07 +02:00
parent df65219729
commit 1864b7ba5d
6 changed files with 261 additions and 40 deletions

View File

@ -355,6 +355,10 @@ export function parseUserAction(ua: UserAction, docData: DocData): UserAction {
ua = _parseUserActionColValues(ua, docData, false, 2);
ua = _parseUserActionColValues(ua, docData, false, 3);
return ua;
case 'BulkAddOrUpdateRecord':
ua = _parseUserActionColValues(ua, docData, true, 2);
ua = _parseUserActionColValues(ua, docData, true, 3);
return ua;
default:
return ua;
}

View File

@ -5,6 +5,7 @@ import { arrayRepeat } from './gutil';
import flatMap = require('lodash/flatMap');
import isEqual = require('lodash/isEqual');
import pick = require('lodash/pick');
import groupBy = require('lodash/groupBy');
/**
* An implementation of the TableOperations interface, given a platform
@ -59,8 +60,21 @@ export class TableOperationsImpl implements TableOperations {
allow_empty_require: upsertOptions?.allowEmptyRequire
};
const recordOptions: OpOptions = pick(upsertOptions, 'parseStrings');
const actions = records.map(rec =>
["AddOrUpdateRecord", tableId, rec.require, rec.fields || {}, options]);
// Group records based on having the same keys in `require` and `fields`.
// A single bulk action will be applied to each group.
// We don't want one bulk action for all records that might have different shapes,
// because that would require filling arrays with null values.
const recGroups = groupBy(records, rec => {
const requireKeys = Object.keys(rec.require).sort().join(',');
const fieldsKeys = Object.keys(rec.fields || {}).sort().join(',');
return `${requireKeys}:${fieldsKeys}`;
});
const actions = Object.values(recGroups).map(group => {
const require = convertToBulkColValues(group.map(r => ({fields: r.require})));
const fields = convertToBulkColValues(group.map(r => ({fields: r.fields || {}})));
return ["BulkAddOrUpdateRecord", tableId, require, fields, options];
});
await this._applyUserActions(tableId, [...fieldNames(records)],
actions, recordOptions);
return [];

View File

@ -77,6 +77,10 @@ function isAclTable(tableId: string): boolean {
return ['_grist_ACLRules', '_grist_ACLResources'].includes(tableId);
}
function isAddOrUpdateRecordAction(a: UserAction): boolean {
return ['AddOrUpdateRecord', 'BulkAddOrUpdateRecord'].includes(String(a[0]));
}
// A list of key metadata tables that need special handling. Other metadata tables may
// refer to material in some of these tables but don't need special handling.
// TODO: there are other metadata tables that would need access control, or redesign -
@ -128,8 +132,9 @@ const OTHER_RECOGNIZED_ACTIONS = new Set([
'BulkRemoveRecord',
'ReplaceTableData',
// A data action handled specially because of read needs.
// Data actions handled specially because of read needs.
'AddOrUpdateRecord',
'BulkAddOrUpdateRecord',
// Groups of actions.
'ApplyDocActions',
@ -979,21 +984,22 @@ export class GranularAccess implements GranularAccessForBundle {
// way to do that within the data engine as currently
// formulated. Could perhaps be done for on-demand tables though.
private async _checkAddOrUpdateAccess(docSession: OptDocSession, actions: UserAction[]) {
if (!scanActionsRecursively(actions, (a) => a[0] === 'AddOrUpdateRecord')) {
if (!scanActionsRecursively(actions, isAddOrUpdateRecordAction)) {
// Don't need to apply this particular check.
return;
}
// Fail if being combined with anything fancy.
if (scanActionsRecursively(actions, (a) => {
const name = a[0];
return !['ApplyUndoActions', 'ApplyDocActions', 'AddOrUpdateRecord'].includes(String(name)) &&
return !['ApplyUndoActions', 'ApplyDocActions'].includes(String(name)) &&
!isAddOrUpdateRecordAction(a) &&
!(isDataAction(a) && !getTableId(a).startsWith('_grist_'));
})) {
throw new Error('Can only combine AddOrUpdate with simple data changes');
}
// Check for read access, and that we're not touching metadata.
await applyToActionsRecursively(actions, async (a) => {
if (a[0] !== 'AddOrUpdateRecord') { return; }
if (!isAddOrUpdateRecordAction(a)) { return; }
const tableId = validTableIdString(a[1]);
if (tableId.startsWith('_grist_')) {
throw new Error(`AddOrUpdate cannot yet be used on metadata tables`);

View File

@ -1033,8 +1033,7 @@ class TestUserActions(test_engine.EngineTestCase):
{"color": "green"},
{"on_many": "all"},
[
["UpdateRecord", "Table1", 1, {"color": "green"}],
["UpdateRecord", "Table1", 2, {"color": "green"}],
["BulkUpdateRecord", "Table1", [1, 2], {"color": ["green", "green"]}],
],
)
@ -1044,8 +1043,7 @@ class TestUserActions(test_engine.EngineTestCase):
{"color": "greener"},
{"on_many": "all", "allow_empty_require": True},
[
["UpdateRecord", "Table1", 1, {"color": "greener"}],
["UpdateRecord", "Table1", 2, {"color": "greener"}],
["BulkUpdateRecord", "Table1", [1, 2], {"color": ["greener", "greener"]}],
],
)
@ -1159,6 +1157,128 @@ class TestUserActions(test_engine.EngineTestCase):
[["UpdateRecord", "Table1", 102, {"date": 1900800}]],
)
# Empty both does nothing
check(
{},
{},
{"allow_empty_require": True},
[],
)
def test_bulk_add_or_update(self):
sample = testutil.parse_test_sample({
"SCHEMA": [
[1, "Table1", [
[1, "first_name", "Text", False, "", "first_name", ""],
[2, "last_name", "Text", False, "", "last_name", ""],
[4, "color", "Text", False, "", "color", ""],
]],
],
"DATA": {
"Table1": [
["id", "first_name", "last_name"],
[1, "John", "Doe"],
[2, "John", "Smith"],
],
}
})
self.load_sample(sample)
def check(require, values, options, stored):
self.assertPartialOutActions(
self.apply_user_action(["BulkAddOrUpdateRecord", "Table1", require, values, options]),
{"stored": stored},
)
check(
{
"first_name": [
"John",
"John",
"John",
"Bob",
],
"last_name": [
"Doe",
"Smith",
"Johnson",
"Johnson",
],
},
{
"color": [
"red",
"blue",
"green",
"yellow",
],
},
{},
[
["BulkAddRecord", "Table1", [3, 4], {
"color": ["green", "yellow"],
"first_name": ["John", "Bob"],
"last_name": ["Johnson", "Johnson"],
}],
["BulkUpdateRecord", "Table1", [1, 2], {"color": ["red", "blue"]}],
],
)
with self.assertRaises(ValueError) as cm:
check(
{"color": ["yellow"]},
{"color": ["red", "blue", "green"]},
{},
[],
)
self.assertEqual(
str(cm.exception),
'Value lists must all have the same length, '
'got {"col_values color": 3, "require color": 1}',
)
with self.assertRaises(ValueError) as cm:
check(
{
"first_name": [
"John",
"John",
],
"last_name": [
"Doe",
],
},
{},
{},
[],
)
self.assertEqual(
str(cm.exception),
'Value lists must all have the same length, '
'got {"require first_name": 2, "require last_name": 1}',
)
with self.assertRaises(ValueError) as cm:
check(
{
"first_name": [
"John",
"John",
],
"last_name": [
"Doe",
"Doe",
],
},
{},
{},
[],
)
self.assertEqual(
str(cm.exception),
"require values must be unique",
)
def test_reference_lookup(self):
sample = testutil.parse_test_sample({
"SCHEMA": [

View File

@ -918,14 +918,15 @@ class UserActions(object):
return values
@useraction
def AddOrUpdateRecord(self, table_id, require, col_values, options):
def BulkAddOrUpdateRecord(self, table_id, require, col_values, options):
"""
Add or Update ('upsert') a single record depending on `options`
and on whether a record matching `require` already exists.
Add or Update ('upsert') records depending on `options`
and on whether records matching `require` already exist.
`require` and `col_values` are dictionaries mapping column IDs to single cell values.
`require` and `col_values` are dictionaries mapping column IDs to lists of cell values.
All lists across both dictionaries must have the same length.
By default, if `table.lookupRecords(**require)` returns any records,
By default, for a single record, if `table.lookupRecords(**require)` returns any records,
update the first one with the values in `col_values`.
Otherwise create a new record with values `{**require, **col_values}`.
@ -940,39 +941,97 @@ class UserActions(object):
"""
table = self._engine.tables[table_id]
if not require and not options.get("allow_empty_require", False):
update = options.get("update", True)
add = options.get("add", True)
on_many = options.get("on_many", "first")
if on_many not in ("first", "none", "all"):
raise ValueError("on_many should be 'first', 'none', or 'all', not %r" % on_many)
allow_empty_require = options.get("allow_empty_require", False)
if not require and not allow_empty_require:
raise ValueError("require is empty but allow_empty_require isn't set")
# Decode `require` before looking up, but let AddRecord/UpdateRecord decode the final
# values when adding/updating
decoded_require = {k: decode_object(v) for k, v in six.iteritems(require)}
records = list(table.lookup_records(**decoded_require))
if not require and not col_values:
return # nothing to do
if records and options.get("update", True):
if len(records) > 1:
on_many = options.get("on_many", "first")
if on_many == "first":
records = records[:1]
elif on_many == "none":
return
elif on_many != "all":
raise ValueError("on_many should be 'first', 'none', or 'all', not %r" % on_many)
lengths = {}
lengths.update({'require ' + k:
len(v) for k, v in six.iteritems(require)})
lengths.update({'col_values ' + k:
len(v) for k, v in six.iteritems(col_values)})
unique_lengths = set(lengths.values())
if len(unique_lengths) != 1:
raise ValueError("Value lists must all have the same length, got %s" %
json.dumps(lengths, sort_keys=True))
[length] = unique_lengths
for record in records:
self.UpdateRecord(table_id, record.id, col_values)
decoded_require = actions.decode_bulk_values(require)
num_unique_keys = len(set(zip(*decoded_require.values())))
if require and num_unique_keys < length:
raise ValueError("require values must be unique")
if not records and options.get("add", True):
values = {
key: value
for key, value in six.iteritems(require)
if not (
# Column IDs in `require` that can be used to set values when creating new records,
# i.e. not formula columns that don't allow setting values.
# `col_values` is not checked for this because setting such a column there should raise an error
# This doesn't apply to `require` since it's also used to match existing records.
require_add_keys = {
key for key in require
if not (
table.get_column(key).is_formula() and
# Check that there actually is a formula and this isn't just an empty column
self._engine.docmodel.get_column_rec(table_id, key).formula
)
}
values.update(col_values)
self.AddRecord(table_id, values.pop("id", None), values)
)
}
col_keys = set(col_values.keys())
# Arguments for `BulkAddRecord` and `BulkUpdateRecord` below
add_record_ids = []
add_record_values = {k: [] for k in col_keys | require_add_keys - {'id'}}
update_record_ids = []
update_record_values = {k: [] for k in col_keys - {'id'}}
for i in range(length):
current_require = {key: vals[i] for key, vals in six.iteritems(decoded_require)}
records = list(table.lookup_records(**current_require))
if not records and add:
values = {key: require[key][i] for key in require_add_keys}
values.update({key: vals[i] for key, vals in six.iteritems(col_values)})
add_record_ids.append(values.pop("id", None))
for key, value in six.iteritems(values):
add_record_values[key].append(value)
if records and update:
if len(records) > 1:
if on_many == "first":
records = records[:1]
elif on_many == "none":
continue
for record in records:
update_record_ids.append(record.id)
for key, vals in six.iteritems(col_values):
update_record_values[key].append(vals[i])
if add_record_ids:
self.BulkAddRecord(table_id, add_record_ids, add_record_values)
if update_record_ids:
self.BulkUpdateRecord(table_id, update_record_ids, update_record_values)
@useraction
def AddOrUpdateRecord(self, table_id, require, col_values, options):
"""
Add or Update ('upsert') a record depending on `options`
and on whether a record matching `require` already exists.
`require` and `col_values` are dictionaries mapping column IDs to cell values.
See `BulkAddOrUpdateRecord` for more details.
"""
require = {k: [v] for k, v in six.iteritems(require)}
col_values = {k: [v] for k, v in six.iteritems(col_values)}
self.BulkAddOrUpdateRecord(table_id, require, col_values, options)
#----------------------------------------
# RemoveRecords & co.

View File

@ -1029,6 +1029,24 @@ function testDocApi() {
{id: [1, 2, 3, 4], A: [1, 33, 66, "$33"], B: [2, 1, 6, 0]},
);
// Test bulk case with a mixture of record shapes
await check([
{
require: {A: 1},
fields: {A: 111},
},
{
require: {A: 33},
fields: {A: 222, B: 444},
},
{
require: {id: 3},
fields: {A: 555, B: 666},
},
],
{id: [1, 2, 3, 4], A: [111, 222, 555, "$33"], B: [2, 444, 666, 0]},
);
// allow_empty_require option with empty `require` updates all records
await check([
{