Provide dedicated capability to isolate logger name

[djaglowski]

This change follows from (specification) #2276

This library should provide an simple mechanism for isolating a LoggerName value, which will be used later in the data pipeline (specifically here), to populate the InstrumentationScope.

The following is suggested:

1. Add LoggerName to entry.Entry. (This does not imply that LoggerName is a top level field in the data model. It is simply a means to convey this information until the log entry is batched by Instrumentation Scope.)
2. Add a logger_name block, which follows the same pattern as timestamp and severity. Specifically, this means that every parser may optionally assign a string value to LoggerName.