From b5b5a1df3d5509153673fa3b148ed52806c457d8 Mon Sep 17 00:00:00 2001
From: George Robinson <george.robinson@grafana.com>
Date: Mon, 30 Oct 2023 13:56:54 +0000
Subject: [PATCH] Support UTF-8 label matchers: Do not allow unquoted escape
 sequences (#3571)

* Do not allow unquoted escape sequences

This commit updates the matchers parser to reject unquoted
openmetrics escape sequences. As an example, foo=bar\n
will no longer parse, and must instead be written as
foo="bar\n". This avoids an issue where the input is valid
in both the matchers and classic parsers, but results
in two different parsings.

---------

Signed-off-by: George Robinson <george.robinson@grafana.com>
---
 matchers/compat/parse_test.go          | 8 ++++----
 matchers/compliance/compliance_test.go | 3 +++
 matchers/parse/lexer.go                | 2 +-
 matchers/parse/parse_test.go           | 4 ++++
 4 files changed, 12 insertions(+), 5 deletions(-)

diff --git a/matchers/compat/parse_test.go b/matchers/compat/parse_test.go
index b3d1f15a..d31e275b 100644
--- a/matchers/compat/parse_test.go
+++ b/matchers/compat/parse_test.go
@@ -38,8 +38,8 @@ func TestFallbackMatcherParser(t *testing.T) {
 		expected: mustNewMatcher(t, labels.MatchEqual, "foo🙂", "bar"),
 	}, {
 		name:     "is accepted in old parser but not new",
-		input:    "foo=!bar",
-		expected: mustNewMatcher(t, labels.MatchEqual, "foo", "!bar"),
+		input:    "foo=!bar\\n",
+		expected: mustNewMatcher(t, labels.MatchEqual, "foo", "!bar\n"),
 	}, {
 		name:  "is accepted in neither",
 		input: "foo!bar",
@@ -81,10 +81,10 @@ func TestFallbackMatchersParser(t *testing.T) {
 		},
 	}, {
 		name:  "is accepted in old parser but not new",
-		input: "{foo=!bar,bar=$baz}",
+		input: "{foo=!bar,bar=$baz\\n}",
 		expected: labels.Matchers{
 			mustNewMatcher(t, labels.MatchEqual, "foo", "!bar"),
-			mustNewMatcher(t, labels.MatchEqual, "bar", "$baz"),
+			mustNewMatcher(t, labels.MatchEqual, "bar", "$baz\n"),
 		},
 	}, {
 		name:  "is accepted in neither",
diff --git a/matchers/compliance/compliance_test.go b/matchers/compliance/compliance_test.go
index 705c4c43..844f80a5 100644
--- a/matchers/compliance/compliance_test.go
+++ b/matchers/compliance/compliance_test.go
@@ -76,6 +76,7 @@ func TestCompliance(t *testing.T) {
 				m, _ := labels.NewMatcher(labels.MatchEqual, "foo", "\\t")
 				return append(ms, m)
 			}(),
+			skip: true,
 		},
 		{
 			input: `{foo=bar\t}`,
@@ -84,6 +85,7 @@ func TestCompliance(t *testing.T) {
 				m, _ := labels.NewMatcher(labels.MatchEqual, "foo", "bar\\t")
 				return append(ms, m)
 			}(),
+			skip: true,
 		},
 		{
 			input: `{foo=bar\}`,
@@ -92,6 +94,7 @@ func TestCompliance(t *testing.T) {
 				m, _ := labels.NewMatcher(labels.MatchEqual, "foo", "bar\\")
 				return append(ms, m)
 			}(),
+			skip: true,
 		},
 		{
 			input: `{foo=bar\\}`,
diff --git a/matchers/parse/lexer.go b/matchers/parse/lexer.go
index a25c9d7e..d6daa6a9 100644
--- a/matchers/parse/lexer.go
+++ b/matchers/parse/lexer.go
@@ -25,7 +25,7 @@ const (
 )
 
 func isReserved(r rune) bool {
-	return unicode.IsSpace(r) || strings.ContainsRune("{}!=~,\"'`", r)
+	return unicode.IsSpace(r) || strings.ContainsRune("{}!=~,\\\"'`", r)
 }
 
 // expectedError is returned when the next rune does not match what is expected.
diff --git a/matchers/parse/parse_test.go b/matchers/parse/parse_test.go
index 03bef2f1..9b99bde5 100644
--- a/matchers/parse/parse_test.go
+++ b/matchers/parse/parse_test.go
@@ -195,6 +195,10 @@ func TestMatchers(t *testing.T) {
 		name:  "invalid escape sequence",
 		input: "{foo=\"bar\\w\"}",
 		error: "5:12: \"bar\\w\": invalid input",
+	}, {
+		name:  "no unquoted escape sequences",
+		input: "{foo=bar\\n}",
+		error: "8:9: \\: invalid input: expected a comma or close brace",
 	}}
 
 	for _, test := range tests {