Fix conversion of non-utf8 sequences to AnyValue (#1253)

String values which are not valid Unicode sequences SHOULD be converted to AnyValue's bytes_value with the bytes representing the string in the original order and format of the source string.
open-telemetry · Mar 13, 2024 · 1753fbe · 1753fbe
1 parent 01c46a2
commit 1753fbe
Show file tree

Hide file tree

Showing 3 changed files with 63 additions and 1 deletion.
diff --git a/src/Contrib/Otlp/AttributesConverter.php b/src/Contrib/Otlp/AttributesConverter.php
@@ -46,12 +46,23 @@ public static function convertAnyValue($value): AnyValue
             $result->setDoubleValue($value);
         }
         if (is_string($value)) {
-            $result->setStringValue($value);
+            if (self::isUtf8($value)) {
+                $result->setStringValue($value);
+            } else {
+                $result->setBytesValue($value);
+            }
         }
 
         return $result;
     }
 
+    private static function isUtf8(string $value): bool
+    {
+        return \extension_loaded('mbstring')
+            ? \mb_check_encoding($value, 'UTF-8')
+            : (bool) \preg_match('//u', $value);
+    }
+
     /**
      * Test whether an array is simple (non-KeyValue)
      */

diff --git a/tests/Unit/Contrib/Otlp/AttributesConverterTest.php b/tests/Unit/Contrib/Otlp/AttributesConverterTest.php
@@ -36,6 +36,13 @@ public static function basicTypesProvider(): array
         ];
     }
 
+    public function test_convert_bytes(): void
+    {
+        $anyValue = AttributesConverter::convertAnyValue("\xe2");
+        $this->assertTrue($anyValue->hasBytesValue());
+        $this->assertSame("\xe2", $anyValue->getBytesValue());
+    }
+
     /**
      * @dataProvider arrayProvider
      */

diff --git a/tests/Unit/Contrib/Otlp/SpanExporterTest.php b/tests/Unit/Contrib/Otlp/SpanExporterTest.php
@@ -104,4 +104,48 @@ public function test_json_span_and_trace_id_hex_format(): void
             }
             TRACE, stream_get_contents($stream));
     }
+
+    public function test_json_invalid_utf8_sequence_is_encoded_as_bytes_value(): void
+    {
+        $stream = fopen('php://memory', 'a+b');
+        $transport = new StreamTransport($stream, 'application/json');
+        $exporter = new SpanExporter($transport);
+
+        $exporter->export([
+            (new SpanData())
+                ->setContext(SpanContext::create('0af7651916cd43dd8448eb211c80319c', 'b7ad6b7169203331'))
+                ->addAttribute('invalid-utf8', "\xe2"),
+        ]);
+
+        fseek($stream, 0);
+        $this->assertJsonStringEqualsJsonString(<<<TRACE
+            {
+                "resourceSpans": [
+                    {
+                        "resource": {},
+                        "scopeSpans": [
+                            {
+                                "scope": {},
+                                "spans": [
+                                    {
+                                        "traceId": "0af7651916cd43dd8448eb211c80319c",
+                                        "spanId": "b7ad6b7169203331",
+                                        "name": "test-span-data",
+                                        "kind": 1,
+                                        "startTimeUnixNano": "1505855794194009601",
+                                        "endTimeUnixNano": "1505855799465726528",
+                                        "status": {},
+                                        "attributes": [{
+                                            "key": "invalid-utf8",
+                                            "value": {"bytesValue": "4g=="}
+                                        }]
+                                    }
+                                ]
+                            }
+                        ]
+                    }
+                ]
+            }
+            TRACE, stream_get_contents($stream));
+    }
 }