Refactor TraceData to encapsulate internal traces' representation

telemetry/telemetry/timeline/trace_event_importer.py

Index: telemetry/telemetry/timeline/trace_event_importer.py
diff --git a/telemetry/telemetry/timeline/trace_event_importer.py b/telemetry/telemetry/timeline/trace_event_importer.py
index c909fb2a67fda5ce7ef0661f3f8450527b983119..e75b19ab8bc316fa4b6e040dbc6afd103355f404 100644
--- a/telemetry/telemetry/timeline/trace_event_importer.py
+++ b/telemetry/telemetry/timeline/trace_event_importer.py
@@ -9,7 +9,6 @@ https://code.google.com/p/trace-viewer/
 
 import collections
 import copy
-import json
 
 import telemetry.timeline.async_slice as tracing_async_slice
 import telemetry.timeline.flow_event as tracing_flow_event
@@ -33,25 +32,6 @@ class TraceEventTimelineImporter(importer.TimelineImporter):
     self._events = []
     self._metadata = []
     for trace in trace_data.GetTracesFor(trace_data_module.CHROME_TRACE_PART):
-      if isinstance(trace, trace_data_module.TraceFileHandle):
-        trace_handle = trace
-        file_path = trace_handle.file_path
-        try:
-          with open(file_path, 'r') as f:
-            trace = json.load(f)
-        finally:
-          # HACK: during the effort of converting chrome trace's representation
-          # from in-memory to disk, we need to either:
-          # 1) Audit all the clients of python trace importer to explictly
-          # cleaned up the trace handle after use.
-          # 2) Clean up the trace handle for them.
-          # We decide to clean up the trace handle for them here (option 2) as
-          # it's lots of work to fix all the call sites. Once all the trace
-          # converting code are converted to use the Javascript engine, this
-          # problem also goes away as we only have a single trace parser
-          # engine.
-          # https://bugs.chromium.org/p/chromium/issues/detail?id=682355
-          trace_handle.Clean()
       self._events.extend(trace['traceEvents'])
       self.CollectMetadataRecords(trace)