lib/logstorage: properly parse unquoted regexp filters ending with *

Use getCompoundToken() instead of getCompoundFuncArg() for obtaining regexp filter value,
since getCompoundFuncArg() skips trailing '*' chars.

This allows detecting invalid queries in the https://github.com/VictoriaMetrics/VictoriaMetrics/issues/8582 .

Author: valyala

docs/victorialogs/CHANGELOG.md

@@ -49,6 +49,7 @@ according to [these docs](https://docs.victoriametrics.com/victorialogs/quicksta
 * BUGFIX: [query API](https://docs.victoriametrics.com/victorialogs/querying/#querying-logs): properly set storage node authorization in cluster mode when [Basic Auth](https://docs.victoriametrics.com/victorialogs/cluster/#security) is enabled. See [#9080](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/9080).
 * BUGFIX: [Journald data ingestion](https://docs.victoriametrics.com/victorialogs/data-ingestion/journald/): properly read log timestamp from `__REALTIME_TIMESTAMP` field according to [the docs](https://docs.victoriametrics.com/victorialogs/data-ingestion/journald/#time-field). See [#9144](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/9144). The bug has been introduced in [v1.22.0-victorialogs](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.22.0-victorialogs).
 * BUGFIX: [data ingestion](https://docs.victoriametrics.com/victorialogs/data-ingestion/): support `-` as a timestamp value, as described in [RFC5424](https://datatracker.ietf.org/doc/html/rfc5424#section-6.2.3).
+* BUGFIX: [LogsQL regexp filter](https://docs.victoriametrics.com/victorialogs/logsql/#regexp-filter): properly parse unquoted filter ending with `*`, such as `foo:~bar.*`. It must be parsed as `foo:~"bar.*"`, while previously it was incorrectly parsed as `foo:~"bar."`.
 
 ## [v1.23.3](https://github.com/VictoriaMetrics/VictoriaMetrics/releases/tag/v1.23.3-victorialogs)
 

lib/logstorage/parser.go

@@ -2014,7 +2014,7 @@ func newFilterRegexp(fieldName, arg string) (filter, error) {
 
 	re, err := regexutil.NewRegex(arg)
 	if err != nil {
-		return nil, fmt.Errorf("invalid regexp %q: %w", arg, err)
+		return nil, fmt.Errorf("invalid regexp %q:%q: %w", getCanonicalColumnName(fieldName), arg, err)
 	}
 	fr := &filterRegexp{
 		fieldName: getCanonicalColumnName(fieldName),
@@ -2025,7 +2025,10 @@ func newFilterRegexp(fieldName, arg string) (filter, error) {
 
 func parseFilterTilda(lex *lexer, fieldName string) (filter, error) {
 	lex.nextToken()
-	arg := getCompoundFuncArg(lex)
+	arg, err := getCompoundToken(lex)
+	if err != nil {
+		return nil, fmt.Errorf("cannot read regexp for field %q: %w", getCanonicalColumnName(fieldName), err)
+	}
 	return newFilterRegexp(fieldName, arg)
 }
 

lib/logstorage/parser_test.go

@@ -64,6 +64,8 @@ func TestLexer(t *testing.T) {
 		[]string{"{", "foo", "=", "bar", ",", "a", "=~", "baz", ",", "b", "!=", "cd", ",", "d,}a", "!~", "abc", "}", "def"})
 	f(`_stream:{foo="bar",a=~"baz", b != 'cd',"d,}a"!~abc}`,
 		[]string{"_stream", ":", "{", "foo", "=", "bar", ",", "a", "=~", "baz", ",", "b", "!=", "cd", ",", "d,}a", "!~", "abc", "}"})
+
+	f(`foo:~*`, []string{"foo", ":", "~", "*"})
 }
 
 func TestQuery_AddTimeFilter(t *testing.T) {
@@ -1325,6 +1327,10 @@ func TestParseQuery_Success(t *testing.T) {
 	f(`foo bar:~".*"`, `foo`)
 	f(`foo bar:~""`, `foo`)
 	f(`foo bar:~".+"`, `foo bar:*`)
+	f(`x:~.*`, `*`)
+	f(`~.*`, `*`)
+	f(`x:~a*`, `x:~"a*"`)
+	f(`~a*`, `~"a*"`)
 
 	// seq filter
 	f(`seq()`, `seq()`)
@@ -2022,6 +2028,8 @@ func TestParseQuery_Failure(t *testing.T) {
 	f("foo:re(bar")
 	f("re(`ab(`)")
 	f(`re(a b)`)
+	f(`foo:~*`)
+	f(`~*`)
 
 	// invalid seq
 	f(`seq(`)