Merge pull request #763 from danielnsilva/drop-unused-requests

Add an option to remove unused requests from cassette

docs/advanced.rst

@@ -427,3 +427,16 @@ If you want to save the cassette only when the test succeeds, set the Cassette
 
     # Since there was an exception, the cassette file hasn't been created.
     assert not os.path.exists('fixtures/vcr_cassettes/synopsis.yaml')
+
+Drop unused requests
+--------------------
+
+Even if any HTTP request is changed or removed from tests, previously recorded
+interactions remain in the cassette file.  If set the ``drop_unused_requests``
+option to ``True``, VCR will not save old HTTP interactions if they are not used.
+
+.. code:: python
+
+    my_vcr = VCR(drop_unused_requests=True)
+    with my_vcr.use_cassette('fixtures/vcr_cassettes/synopsis.yaml'):
+        ...  # your HTTP interactions here

tests/integration/test_config.py

@@ -5,6 +5,7 @@
 import pytest
 
 import vcr
+from vcr.cassette import Cassette
 
 
 @pytest.mark.online
@@ -85,3 +86,21 @@ def some_test():
             assert b"Not in content" in urlopen(httpbin.url).read()
 
     assert not os.path.exists(str(tmpdir.join("dontsave2.yml")))
+
+
+def test_set_drop_unused_requests(tmpdir, httpbin):
+    my_vcr = vcr.VCR(drop_unused_requests=True)
+    file = str(tmpdir.join("test.yaml"))
+
+    with my_vcr.use_cassette(file):
+        urlopen(httpbin.url)
+        urlopen(httpbin.url + "/get")
+
+    cassette = Cassette.load(path=file)
+    assert len(cassette) == 2
+
+    with my_vcr.use_cassette(file):
+        urlopen(httpbin.url)
+
+    cassette = Cassette.load(path=file)
+    assert len(cassette) == 1

tests/unit/test_cassettes.py

@@ -11,6 +11,7 @@
 from vcr.cassette import Cassette
 from vcr.errors import UnhandledHTTPRequestError
 from vcr.patch import force_reset
+from vcr.request import Request
 from vcr.stubs import VCRHTTPSConnection
 
 
@@ -410,3 +411,25 @@ def test_find_requests_with_most_matches_many_similar_requests(mock_get_matchers
         (1, ["method", "path"], [("query", "failed : query")]),
         (3, ["method", "path"], [("query", "failed : query")]),
     ]
+
+
+def test_used_interactions(tmpdir):
+    interactions = [
+        {"request": {"body": "", "uri": "foo1", "method": "GET", "headers": {}}, "response": "bar1"},
+        {"request": {"body": "", "uri": "foo2", "method": "GET", "headers": {}}, "response": "bar2"},
+        {"request": {"body": "", "uri": "foo3", "method": "GET", "headers": {}}, "response": "bar3"},
+    ]
+    file = tmpdir.join("test_cassette.yml")
+    file.write(yaml.dump({"interactions": [interactions[0], interactions[1]]}))
+
+    cassette = Cassette.load(path=str(file))
+    request = Request._from_dict(interactions[1]["request"])
+    cassette.play_response(request)
+    assert len(cassette._played_interactions) < len(cassette._old_interactions)
+
+    request = Request._from_dict(interactions[2]["request"])
+    cassette.append(request, interactions[2]["response"])
+    assert len(cassette._new_interactions()) == 1
+
+    used_interactions = cassette._played_interactions + cassette._new_interactions()
+    assert len(used_interactions) == 2

vcr/cassette.py

@@ -177,6 +177,7 @@ def __init__(
         custom_patches=(),
         inject=False,
         allow_playback_repeats=False,
+        drop_unused_requests=False,
     ):
         self._persister = persister or FilesystemPersister
         self._path = path
@@ -189,13 +190,18 @@ def __init__(
         self.record_mode = record_mode
         self.custom_patches = custom_patches
         self.allow_playback_repeats = allow_playback_repeats
+        self.drop_unused_requests = drop_unused_requests
 
         # self.data is the list of (req, resp) tuples
         self.data = []
         self.play_counts = collections.Counter()
         self.dirty = False
         self.rewound = False
 
+        # Subsets of self.data to store old and played interactions
+        self._old_interactions = []
+        self._played_interactions = []
+
     @property
     def play_count(self):
         return sum(self.play_counts.values())
@@ -257,6 +263,7 @@ def play_response(self, request):
         for index, response in self._responses(request):
             if self.play_counts[index] == 0 or self.allow_playback_repeats:
                 self.play_counts[index] += 1
+                self._played_interactions.append((request, response))
                 return response
         # The cassette doesn't contain the request asked for.
         raise UnhandledHTTPRequestError(
@@ -317,19 +324,44 @@ def find_requests_with_most_matches(self, request):
 
         return final_best_matches
 
+    def _new_interactions(self):
+        """List of new HTTP interactions (request/response tuples)"""
+        new_interactions = []
+        for request, response in self.data:
+            if all(
+                not requests_match(request, old_request, self._match_on)
+                for old_request, _ in self._old_interactions
+            ):
+                new_interactions.append((request, response))
+        return new_interactions
+
     def _as_dict(self):
         return {"requests": self.requests, "responses": self.responses}
 
+    def _build_used_interactions_dict(self):
+        interactions = self._played_interactions + self._new_interactions()
+        cassete_dict = {
+            "requests": [request for request, _ in interactions],
+            "responses": [response for _, response in interactions],
+        }
+        return cassete_dict
+
     def _save(self, force=False):
+        if self.drop_unused_requests and len(self._played_interactions) < len(self._old_interactions):
+            cassete_dict = self._build_used_interactions_dict()
+            force = True
+        else:
+            cassete_dict = self._as_dict()
         if force or self.dirty:
-            self._persister.save_cassette(self._path, self._as_dict(), serializer=self._serializer)
+            self._persister.save_cassette(self._path, cassete_dict, serializer=self._serializer)
             self.dirty = False
 
     def _load(self):
         try:
             requests, responses = self._persister.load_cassette(self._path, serializer=self._serializer)
             for request, response in zip(requests, responses):
                 self.append(request, response)
+                self._old_interactions.append((request, response))
             self.dirty = False
             self.rewound = True
         except (CassetteDecodeError, CassetteNotFoundError):

vcr/config.py

@@ -48,6 +48,7 @@ def __init__(
         func_path_generator=None,
         decode_compressed_response=False,
         record_on_exception=True,
+        drop_unused_requests=False,
     ):
         self.serializer = serializer
         self.match_on = match_on
@@ -81,6 +82,7 @@ def __init__(
         self.decode_compressed_response = decode_compressed_response
         self.record_on_exception = record_on_exception
         self._custom_patches = tuple(custom_patches)
+        self.drop_unused_requests = drop_unused_requests
 
     def _get_serializer(self, serializer_name):
         try:
@@ -151,6 +153,7 @@ def add_cassette_library_dir(path):
             "func_path_generator": func_path_generator,
             "allow_playback_repeats": kwargs.get("allow_playback_repeats", False),
             "record_on_exception": record_on_exception,
+            "drop_unused_requests": kwargs.get("drop_unused_requests", self.drop_unused_requests),
         }
         path = kwargs.get("path")
         if path: