Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Obs AI assistant] Replace with double quotes #187327

Merged
merged 8 commits into from
Jul 5, 2024
Merged

[Obs AI assistant] Replace with double quotes #187327

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
3c2fe19
[Obs AI assistant] Replace with double quotes
stratoula Jul 2, 2024
39068df
Update x-pack/plugins/observability_solution/observability_ai_assista…
stratoula Jul 4, 2024
ffe4d57
Addresses multiple escaped indices
stratoula Jul 4, 2024
9998024
Merge branch 'main' into integrate-double-quotes
stratoula Jul 4, 2024
6cc8894
small nit
stratoula Jul 4, 2024
ee6f302
Make Soren happy
stratoula Jul 4, 2024
10d3818
Escape backticks
stratoula Jul 4, 2024
c8c27c6
Change the implementation
stratoula Jul 4, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
236 changes: 127 additions & 109 deletions ...bservability_ai_assistant_app/server/functions/query/correct_common_esql_mistakes.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,147 +11,165 @@ describe('correctCommonEsqlMistakes', () => {
return input.replaceAll(/[\t|\s]*\n[\t|\s]*/gms, '\n');
}

function expectQuery(input: string, expectedOutput: string) {
function expectQuery({ input, expectedOutput }: { input: string; expectedOutput: string }) {
expect(normalize(correctCommonEsqlMistakes(input).output)).toEqual(normalize(expectedOutput));
}

it('replaces aliasing via the AS keyword with the = operator', () => {
expectQuery(`FROM logs-* | STATS COUNT() AS count`, 'FROM logs-*\n| STATS count = COUNT()');

expectQuery(`FROM logs-* | STATS COUNT() as count`, 'FROM logs-*\n| STATS count = COUNT()');

expectQuery(
`FROM logs-* | STATS AVG(transaction.duration.histogram) AS avg_request_latency, PERCENTILE(transaction.duration.histogram, 95) AS p95`,
`FROM logs-*
| STATS avg_request_latency = AVG(transaction.duration.histogram), p95 = PERCENTILE(transaction.duration.histogram, 95)`
);

expectQuery(
`FROM traces-apm*
expectQuery({
input: `FROM logs-* | STATS COUNT() AS count`,
expectedOutput: 'FROM logs-*\n| STATS count = COUNT()',
});

expectQuery({
input: `FROM logs-* | STATS COUNT() as count`,
expectedOutput: 'FROM logs-*\n| STATS count = COUNT()',
});

expectQuery({
input: `FROM logs-* | STATS AVG(transaction.duration.histogram) AS avg_request_latency, PERCENTILE(transaction.duration.histogram, 95) AS p95`,
expectedOutput: `FROM logs-*
| STATS avg_request_latency = AVG(transaction.duration.histogram), p95 = PERCENTILE(transaction.duration.histogram, 95)`,
});

expectQuery({
input: `FROM traces-apm*
| WHERE @timestamp >= NOW() - 24 hours
| STATS AVG(transaction.duration.us) AS avg_duration, SUM(success) AS total_successes, COUNT(*) AS total_requests BY service.name`,
`FROM traces-apm*
expectedOutput: `FROM traces-apm*
| WHERE @timestamp >= NOW() - 24 hours
| STATS avg_duration = AVG(transaction.duration.us), total_successes = SUM(success), total_requests = COUNT(*) BY service.name`
);
| STATS avg_duration = AVG(transaction.duration.us), total_successes = SUM(success), total_requests = COUNT(*) BY service.name`,
});
});

it(`replaces " or ' escaping in FROM statements with backticks`, () => {
expectQuery(`FROM "logs-*" | LIMIT 10`, 'FROM logs-*\n| LIMIT 10');
expectQuery(`FROM 'logs-*' | LIMIT 10`, 'FROM logs-*\n| LIMIT 10');
expectQuery(`FROM logs-* | LIMIT 10`, 'FROM logs-*\n| LIMIT 10');
it("replaces ` or ' escaping in FROM statements with double quotes", () => {
expectQuery({ input: `FROM "logs-*" | LIMIT 10`, expectedOutput: 'FROM "logs-*"\n| LIMIT 10' });
expectQuery({ input: `FROM 'logs-*' | LIMIT 10`, expectedOutput: 'FROM "logs-*"\n| LIMIT 10' });
expectQuery({ input: 'FROM `logs-*` | LIMIT 10', expectedOutput: 'FROM "logs-*"\n| LIMIT 10' });
expectQuery({
input: `FROM 'logs-2024-07-01','logs-2024-07-02' | LIMIT 10`,
expectedOutput: 'FROM "logs-2024-07-01","logs-2024-07-02"\n| LIMIT 10',
});
expectQuery({
input: 'FROM `logs-2024-07-01`,`logs-2024-07-02` | LIMIT 10',
expectedOutput: 'FROM "logs-2024-07-01","logs-2024-07-02"\n| LIMIT 10',
});
expectQuery({ input: `FROM logs-* | LIMIT 10`, expectedOutput: 'FROM logs-*\n| LIMIT 10' });
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can we add one here for backticks as well? i see the LLM do this all the time now 😄

});

it('replaces = as equal operator with ==', () => {
expectQuery(
`FROM logs-*\n| WHERE service.name = "foo"`,
`FROM logs-*\n| WHERE service.name == "foo"`
);

expectQuery(
`FROM logs-*\n| WHERE service.name = "foo" AND service.environment = "bar"`,
`FROM logs-*\n| WHERE service.name == "foo" AND service.environment == "bar"`
);

expectQuery(
`FROM logs-*\n| WHERE (service.name = "foo" AND service.environment = "bar") OR agent.name = "baz"`,
`FROM logs-*\n| WHERE (service.name == "foo" AND service.environment == "bar") OR agent.name == "baz"`
);

expectQuery(
`FROM logs-*\n| WHERE \`what=ever\` = "foo=bar"`,
`FROM logs-*\n| WHERE \`what=ever\` == "foo=bar"`
);
expectQuery({
input: `FROM logs-*\n| WHERE service.name = "foo"`,
expectedOutput: `FROM logs-*\n| WHERE service.name == "foo"`,
});

expectQuery({
input: `FROM logs-*\n| WHERE service.name = "foo" AND service.environment = "bar"`,
expectedOutput: `FROM logs-*\n| WHERE service.name == "foo" AND service.environment == "bar"`,
});

expectQuery({
input: `FROM logs-*\n| WHERE (service.name = "foo" AND service.environment = "bar") OR agent.name = "baz"`,
expectedOutput: `FROM logs-*\n| WHERE (service.name == "foo" AND service.environment == "bar") OR agent.name == "baz"`,
});

expectQuery({
input: `FROM logs-*\n| WHERE \`what=ever\` = "foo=bar"`,
expectedOutput: `FROM logs-*\n| WHERE \`what=ever\` == "foo=bar"`,
});
});

it('replaces single-quote escaped strings with double-quote escaped strings', () => {
expectQuery(
`FROM nyc_taxis
expectQuery({
input: `FROM nyc_taxis
| WHERE DATE_EXTRACT('hour', dropoff_datetime) >= 6 AND DATE_EXTRACT('hour', dropoff_datetime) < 10
| LIMIT 10`,
`FROM nyc_taxis
expectedOutput: `FROM nyc_taxis
| WHERE DATE_EXTRACT("hour", dropoff_datetime) >= 6 AND DATE_EXTRACT("hour", dropoff_datetime) < 10
| LIMIT 10`
);
expectQuery(
`FROM nyc_taxis
| LIMIT 10`,
});
expectQuery({
input: `FROM nyc_taxis
| WHERE DATE_EXTRACT('hour', "hh:mm a, 'of' d MMMM yyyy") >= 6 AND DATE_EXTRACT('hour', dropoff_datetime) < 10
| LIMIT 10`,
`FROM nyc_taxis
expectedOutput: `FROM nyc_taxis
| WHERE DATE_EXTRACT("hour", "hh:mm a, 'of' d MMMM yyyy") >= 6 AND DATE_EXTRACT("hour", dropoff_datetime) < 10
| LIMIT 10`
);
| LIMIT 10`,
});
});

it(`verifies if the SORT key is in KEEP, and if it's not, it will include it`, () => {
expectQuery(
'FROM logs-* \n| KEEP date \n| SORT @timestamp DESC',
'FROM logs-*\n| KEEP date, @timestamp\n| SORT @timestamp DESC'
);

expectQuery(
`FROM logs-* | KEEP date, whatever | EVAL my_truncated_date_field = DATE_TRUNC(1 year, date) | SORT @timestamp, my_truncated_date_field DESC`,
'FROM logs-*\n| KEEP date, whatever, @timestamp\n| EVAL my_truncated_date_field = DATE_TRUNC(1 year, date)\n| SORT @timestamp, my_truncated_date_field DESC'
);

expectQuery(
`FROM logs-*\n| STATS COUNT(*) BY BUCKET(@timestamp, 1m)\n| SORT \`BUCKET(@timestamp, 1m)\` DESC`,
`FROM logs-*\n| STATS COUNT(*) BY BUCKET(@timestamp, 1m)\n| SORT \`BUCKET(@timestamp, 1m)\` DESC`
);

expectQuery(
`FROM logs-* | KEEP date, whatever | RENAME whatever AS forever | SORT forever DESC`,
`FROM logs-*\n| KEEP date, whatever\n| RENAME whatever AS forever\n| SORT forever DESC`
);

expectQuery(
'FROM employees\n| KEEP first_name, last_name\n| RENAME first_name AS fn, last_name AS ln',
'FROM employees\n| KEEP first_name, last_name\n| RENAME first_name AS fn, last_name AS ln'
);
expectQuery({
input: 'FROM logs-* \n| KEEP date \n| SORT @timestamp DESC',
expectedOutput: 'FROM logs-*\n| KEEP date, @timestamp\n| SORT @timestamp DESC',
});

expectQuery({
input: `FROM logs-* | KEEP date, whatever | EVAL my_truncated_date_field = DATE_TRUNC(1 year, date) | SORT @timestamp, my_truncated_date_field DESC`,
expectedOutput:
'FROM logs-*\n| KEEP date, whatever, @timestamp\n| EVAL my_truncated_date_field = DATE_TRUNC(1 year, date)\n| SORT @timestamp, my_truncated_date_field DESC',
});

expectQuery({
input: `FROM logs-*\n| STATS COUNT(*) BY BUCKET(@timestamp, 1m)\n| SORT \`BUCKET(@timestamp, 1m)\` DESC`,
expectedOutput: `FROM logs-*\n| STATS COUNT(*) BY BUCKET(@timestamp, 1m)\n| SORT \`BUCKET(@timestamp, 1m)\` DESC`,
});

expectQuery({
input: `FROM logs-* | KEEP date, whatever | RENAME whatever AS forever | SORT forever DESC`,
expectedOutput: `FROM logs-*\n| KEEP date, whatever\n| RENAME whatever AS forever\n| SORT forever DESC`,
});

expectQuery({
input:
'FROM employees\n| KEEP first_name, last_name\n| RENAME first_name AS fn, last_name AS ln',
expectedOutput:
'FROM employees\n| KEEP first_name, last_name\n| RENAME first_name AS fn, last_name AS ln',
});
});

it(`escapes the column name if SORT uses an expression`, () => {
expectQuery(
'FROM logs-* \n| STATS COUNT(*) by service.name\n| SORT COUNT(*) DESC',
'FROM logs-*\n| STATS COUNT(*) BY service.name\n| SORT `COUNT(*)` DESC'
);

expectQuery(
'FROM logs-* \n| STATS COUNT(*) by service.name\n| SORT COUNT(*) DESC, @timestamp ASC',
'FROM logs-*\n| STATS COUNT(*) BY service.name\n| SORT `COUNT(*)` DESC, @timestamp ASC'
);

expectQuery(
`FROM employees\n| KEEP first_name, last_name, height\n| SORT first_name ASC NULLS FIRST`,
`FROM employees\n| KEEP first_name, last_name, height\n| SORT first_name ASC NULLS FIRST`
);

expectQuery(
`FROM employees
expectQuery({
input: 'FROM logs-* \n| STATS COUNT(*) by service.name\n| SORT COUNT(*) DESC',
expectedOutput: 'FROM logs-*\n| STATS COUNT(*) BY service.name\n| SORT `COUNT(*)` DESC',
});

expectQuery({
input: 'FROM logs-* \n| STATS COUNT(*) by service.name\n| SORT COUNT(*) DESC, @timestamp ASC',
expectedOutput:
'FROM logs-*\n| STATS COUNT(*) BY service.name\n| SORT `COUNT(*)` DESC, @timestamp ASC',
});

expectQuery({
input: `FROM employees\n| KEEP first_name, last_name, height\n| SORT first_name ASC NULLS FIRST`,
expectedOutput: `FROM employees\n| KEEP first_name, last_name, height\n| SORT first_name ASC NULLS FIRST`,
});

expectQuery({
input: `FROM employees
| STATS my_count = COUNT() BY LEFT(last_name, 1)
| SORT \`LEFT(last_name, 1)\``,
`FROM employees
expectedOutput: `FROM employees
| STATS my_count = COUNT() BY LEFT(last_name, 1)
| SORT \`LEFT(last_name, 1)\``
);
| SORT \`LEFT(last_name, 1)\``,
});
});

it(`handles complicated queries correctly`, () => {
expectQuery(
`FROM "postgres-logs*"
expectQuery({
input: `FROM "postgres-logs*"
| GROK message "%{TIMESTAMP_ISO8601:timestamp} %{TZ} \[%{NUMBER:process_id}\]: \[%{NUMBER:log_line}\] user=%{USER:user},db=%{USER:database},app=\[%{DATA:application}\],client=%{IP:client_ip} LOG: duration: %{NUMBER:duration:float} ms statement: %{GREEDYDATA:statement}"
| EVAL "@timestamp" = TO_DATETIME(timestamp)
| WHERE statement LIKE 'SELECT%'
| STATS avg_duration = AVG(duration)`,
`FROM postgres-logs*
expectedOutput: `FROM "postgres-logs*"
| GROK message "%{TIMESTAMP_ISO8601:timestamp} %{TZ} \[%{NUMBER:process_id}\]: \[%{NUMBER:log_line}\] user=%{USER:user},db=%{USER:database},app=\[%{DATA:application}\],client=%{IP:client_ip} LOG: duration: %{NUMBER:duration:float} ms statement: %{GREEDYDATA:statement}"
| EVAL @timestamp = TO_DATETIME(timestamp)
| WHERE statement LIKE "SELECT%"
| STATS avg_duration = AVG(duration)`
);
| STATS avg_duration = AVG(duration)`,
});

expectQuery(
`FROM metrics-apm*
expectQuery({
input: `FROM metrics-apm*
| WHERE metricset.name == "service_destination" AND @timestamp > NOW() - 24 hours
| EVAL total_events = span.destination.service.response_time.count
| EVAL total_latency = span.destination.service.response_time.sum.us
Expand All @@ -161,27 +179,27 @@ describe('correctCommonEsqlMistakes', () => {
avg_latency_per_request = AVG(total_latency / total_events),
failure_rate = AVG(is_failure)
BY span.destination.service.resource`,
`FROM metrics-apm*
expectedOutput: `FROM metrics-apm*
| WHERE metricset.name == "service_destination" AND @timestamp > NOW() - 24 hours
| EVAL total_events = span.destination.service.response_time.count
| EVAL total_latency = span.destination.service.response_time.sum.us
| EVAL is_failure = CASE(event.outcome == "failure", 1, 0)
| STATS avg_throughput = AVG(total_events), avg_latency_per_request = AVG(total_latency / total_events), failure_rate = AVG(is_failure) BY span.destination.service.resource`
);
| STATS avg_throughput = AVG(total_events), avg_latency_per_request = AVG(total_latency / total_events), failure_rate = AVG(is_failure) BY span.destination.service.resource`,
});

expectQuery(
`FROM sample_data
expectQuery({
input: `FROM sample_data
| EVAL successful = CASE(
STARTS_WITH(message, "Connected to"), 1,
message == "Connection error", 0
)
| STATS success_rate = AVG(successful)`,
`FROM sample_data
expectedOutput: `FROM sample_data
| EVAL successful = CASE(
STARTS_WITH(message, "Connected to"), 1,
message == "Connection error", 0
)
| STATS success_rate = AVG(successful)`
);
| STATS success_rate = AVG(successful)`,
});
});
});
12 changes: 5 additions & 7 deletions ...ion/observability_ai_assistant_app/server/functions/query/correct_common_esql_mistakes.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -234,15 +234,13 @@ export function correctCommonEsqlMistakes(query: string): {

const formattedCommands: string[] = commands.map(({ name, command }, index) => {
let formattedCommand = command;

switch (name) {
case 'FROM':
formattedCommand = formattedCommand
.replaceAll(/FROM "(.*)"/g, 'FROM $1')
.replaceAll(/FROM '(.*)'/g, 'FROM $1')
.replaceAll(/FROM `(.*)`/g, 'FROM $1');
case 'FROM': {
formattedCommand = split(formattedCommand, ',')
.map((singlePattern) => singlePattern.replaceAll(/`/g, '"').replaceAll(/'/g, '"'))
.join(',');
break;

}
case 'WHERE':
formattedCommand = replaceSingleQuotesWithDoubleQuotes(formattedCommand);
formattedCommand = ensureEqualityOperators(formattedCommand);
Expand Down