Adds setting to configurable profiling result path (jazzband#135)

Using a custom setting referenced in this library. You can specify where to store the binary profile. This allows developers to save the profiled data to a folder location outside of their project if necessary.

silk/collector.py

@@ -1,5 +1,6 @@
 from threading import local
 
+import os.path
 import cProfile
 import pstats
 import logging
@@ -89,9 +90,13 @@ def silk_queries(self):
         return self._get_objects('silk_queries')
 
     def configure(self, request=None):
+        silky_config = SilkyConfig()
+
         self.request = request
+        self.profiler_result_path = silky_config.SILKY_PYTHON_PROFILER_RESULT_PATH
         self._configure()
-        if SilkyConfig().SILKY_PYTHON_PROFILER:
+
+        if silky_config.SILKY_PYTHON_PROFILER:
             self.pythonprofiler = cProfile.Profile()
             self.pythonprofiler.enable()
 
@@ -144,7 +149,12 @@ def finalise(self):
             self.request.pyprofile = profile_text
 
             if SilkyConfig().SILKY_PYTHON_PROFILER_BINARY:
-                file_name = "{}.prof".format(self.request.id)
+                file_name = "{}.prof".format(
+                    os.path.join(
+                        self.profiler_result_path,
+                        str(self.request.id)
+                    )
+                )
                 with open(file_name, 'w+b') as f:
                     ps.dump_stats(f.name)
                     self.request.prof_file.save(f.name, File(f))

silk/config.py

@@ -26,6 +26,7 @@ class SilkyConfig(six.with_metaclass(Singleton, object)):
         'SILKY_INTERCEPT_PERCENT': 100,
         'SILKY_INTERCEPT_FUNC': None,
         'SILKY_PYTHON_PROFILER': False,
+        'SILKY_PYTHON_PROFILER_RESULT_PATH': ''
     }
 
     def _setup(self):