Allow to generate more informative profile file name (#638)

* Add SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME settings to generate more informative profile file name

* Improve the algorithm that generate a file name from request path

README.md

@@ -182,6 +182,11 @@ SILKY_PYTHON_PROFILER_RESULT_PATH = '/path/to/profiles/'
 
 A download button will become available with a binary `.prof` file for every request. This file can be used for further analysis using [snakeviz](https://github.com/jiffyclub/snakeviz) or other cProfile tools
 
+To retrieve which endpoint generates a specific profile file it is possible to add a stub of the request path in the file name with the following:
+
+```python
+SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME = True
+```
 
 Silk can also be used to profile specific blocks of code/functions. It provides a decorator and a context
 manager for this purpose.

project/tests/test_collector.py

@@ -2,6 +2,7 @@
 from tests.util import DictStorage
 
 from silk.collector import DataCollector
+from silk.config import SilkyConfig
 
 from .factories import RequestMinFactory
 
@@ -45,3 +46,39 @@ def test_finalise(self):
                 content = f.read()
                 self.assertTrue(content)
                 self.assertGreater(len(content), 0)
+
+    def test_profile_file_name_with_disabled_extended_file_name(self):
+        SilkyConfig().SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME = False
+        request_path = 'normal/uri/'
+        resulting_prefix = self._get_prof_file_name(request_path)
+        self.assertEqual(resulting_prefix, '')
+
+    def test_profile_file_name_with_enabled_extended_file_name(self):
+
+        SilkyConfig().SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME = True
+        request_path = 'normal/uri/'
+        resulting_prefix = self._get_prof_file_name(request_path)
+        self.assertEqual(resulting_prefix, 'normal_uri_')
+
+    def test_profile_file_name_with_path_traversal_and_special_char(self):
+        SilkyConfig().SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME = True
+        request_path = 'spÉciàl/.././大/uri/@É/'
+        resulting_prefix = self._get_prof_file_name(request_path)
+        self.assertEqual(resulting_prefix, 'special_uri_e_')
+
+    def test_profile_file_name_with_long_path(self):
+        SilkyConfig().SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME = True
+        request_path = 'long/path/' + 'a' * 100
+        resulting_prefix = self._get_prof_file_name(request_path)
+        # the path is limited to 50 char plus the last `_`
+        self.assertEqual(len(resulting_prefix), 51)
+
+    @classmethod
+    def _get_prof_file_name(cls, request_path: str) -> str:
+        request = RequestMinFactory()
+        request.path = request_path
+        DataCollector().configure(request)
+        DataCollector().finalise()
+        file_path = DataCollector().request.prof_file.name
+        filename = file_path.rsplit('/')[-1]
+        return filename.replace(f"{request.id}.prof", "")

silk/collector.py

@@ -2,6 +2,8 @@
 import logging
 import marshal
 import pstats
+import re
+import unicodedata
 from io import StringIO
 from threading import local
 
@@ -143,7 +145,8 @@ def finalise(self):
             self.request.pyprofile = profile_text
 
             if SilkyConfig().SILKY_PYTHON_PROFILER_BINARY:
-                file_name = self.request.prof_file.storage.get_available_name(f"{str(self.request.id)}.prof")
+                proposed_file_name = self._get_proposed_file_name()
+                file_name = self.request.prof_file.storage.get_available_name(proposed_file_name)
                 with self.request.prof_file.storage.open(file_name, 'w+b') as f:
                     marshal.dump(ps.stats, f)
                 self.request.prof_file = f.name
@@ -189,3 +192,29 @@ def finalise(self):
 
     def register_silk_query(self, *args):
         self.register_objects(TYP_SILK_QUERIES, *args)
+
+    def _get_proposed_file_name(self) -> str:
+        """Retrieve the profile file name to be proposed to the storage"""
+
+        if SilkyConfig().SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME:
+            slugified_path = slugify_path(self.request.path)
+            return f"{slugified_path}_{str(self.request.id)}.prof"
+        return f"{str(self.request.id)}.prof"
+
+
+def slugify_path(request_path: str) -> str:
+    """
+    Convert any characters not included in [a-zA-Z0-9_]) with a single underscore.
+    Convert to lowercase. Also strip any leading and trailing char that are not in the
+    accepted list
+
+    Inspired from django slugify
+    """
+    request_path = str(request_path)
+    request_path = (
+        unicodedata.normalize("NFKD", request_path)
+        .encode("ascii", "ignore")
+        .decode("ascii")
+    )
+    request_path = request_path.lower()[:50]
+    return re.sub(r'\W+', '_', request_path).strip('_')

silk/config.py

@@ -28,6 +28,7 @@ class SilkyConfig(metaclass=Singleton):
         'SILKY_PYTHON_PROFILER': False,
         'SILKY_PYTHON_PROFILER_FUNC': None,
         'SILKY_STORAGE_CLASS': 'silk.storage.ProfilerResultStorage',
+        'SILKY_PYTHON_PROFILER_EXTENDED_FILE_NAME': False,
         'SILKY_MIDDLEWARE_CLASS': 'silk.middleware.SilkyMiddleware',
         'SILKY_JSON_ENSURE_ASCII': True,
         'SILKY_ANALYZE_QUERIES': False,