Add isolating log record processor

CHANGELOG.md

@@ -15,6 +15,11 @@ release.
 
 ### Logs
 
+- SDK SHOULD allow independent log processing pipelines
+  ([#4062](https://github.com/open-telemetry/opentelemetry-specification/pull/4062))
+- SDK MAY offer a way to copy `ReadWriteLogRecord`
+  ([#4062](https://github.com/open-telemetry/opentelemetry-specification/pull/4062))
+
 ### Events
 
 - Add specification for EventLogger and EventLoggerProvider

specification/logs/sdk.md

@@ -226,6 +226,9 @@ the following information added to the [LogRecord](data-model.md#log-and-event-r
 * [`SpanId`](./data-model.md#field-spanid)
 * [`TraceFlags`](./data-model.md#field-traceflags)
 
+The SDK MAY offer a way to make a copy of `ReadWriteLogRecord`
+so that the user can set up independent log processing pipelines.
+
 ## LogRecord Limits
 
 `LogRecord` attributes MUST adhere to
@@ -274,6 +277,11 @@ The SDK MUST allow users to implement and configure custom processors and
 decorate built-in processors for advanced scenarios such as enriching with
 attributes.
 
+The SDK SHOULD allow users to set up independent processing pipelines.
+For instance, the user should be able to make a log record modification in
+a processor that is not visible to and does not affect the behavior of other
+log record processors.
+
 The following diagram shows `LogRecordProcessor`'s relationship to other
 components in the SDK:
 
@@ -313,6 +321,10 @@ A `LogRecordProcessor` may freely modify `logRecord` for the duration of
 the `OnEmit` call. If `logRecord` is needed after `OnEmit` returns (i.e. for
 asynchronous processing) only reads are permitted.
 
+It is implementation specific whether `logRecord` modification are
+also applied to the subsequent registered processors.
+For instance, `logRecord` may be passed by value or by reference.
+
 #### ShutDown
 
 Shuts down the processor. Called when the SDK is shut down. This is an