From b9de5cf0fc11b4478f7d3969069a393e70031589 Mon Sep 17 00:00:00 2001
From: Zac Hatfield-Dodds <zac.hatfield.dodds@gmail.com>
Date: Sun, 4 Dec 2022 14:14:25 -0800
Subject: [PATCH] Improve database keys

---
 hypothesis-python/RELEASE.rst                 |  7 +++
 .../src/hypothesis/internal/reflection.py     | 60 +++++++++++++------
 2 files changed, 50 insertions(+), 17 deletions(-)
 create mode 100644 hypothesis-python/RELEASE.rst

diff --git a/hypothesis-python/RELEASE.rst b/hypothesis-python/RELEASE.rst
new file mode 100644
index 0000000000..3ed885d9ee
--- /dev/null
+++ b/hypothesis-python/RELEASE.rst
@@ -0,0 +1,7 @@
+RELEASE_TYPE: minor
+
+This release improves our treatment of database keys, which based on (among other things)
+the source code of your test function.  We now post-process this source to ignore
+decorators, comments, trailing whitespace, and blank lines - so that you can add
+:obj:`@example() <hypothesis.example>`\ s or make some small no-op edits to your code
+without preventing replay of any known failing or covering examples.
diff --git a/hypothesis-python/src/hypothesis/internal/reflection.py b/hypothesis-python/src/hypothesis/internal/reflection.py
index 1c1419c609..243dce6767 100644
--- a/hypothesis-python/src/hypothesis/internal/reflection.py
+++ b/hypothesis-python/src/hypothesis/internal/reflection.py
@@ -20,8 +20,9 @@
 import textwrap
 import types
 from functools import wraps
+from io import StringIO
 from keyword import iskeyword
-from tokenize import detect_encoding
+from tokenize import COMMENT, detect_encoding, generate_tokens, untokenize
 from types import ModuleType
 from typing import TYPE_CHECKING, Callable
 from unittest.mock import _patch as PatchType
@@ -48,6 +49,41 @@ def is_mock(obj):
     return hasattr(obj, "hypothesis_internal_is_this_a_mock_check")
 
 
+def _get_clean_source(function) -> bytes:
+    """Return the function's source code as bytes, without decorators or comments.
+
+    Because this is part of our database key, we reduce the cache invalidation
+    rate by ignoring decorators, comments, trailing whitespace, and empty lines.
+    We can't just use the (dumped) AST directly because it changes between Python
+    versions (e.g. ast.Constant)
+    """
+    try:
+        src = inspect.getsource(function)
+    except (OSError, TypeError):
+        # If we can't actually get the source code, try for the name as a fallback.
+        try:
+            return function.__name__.encode()
+        except AttributeError:
+            return b""
+    # Get the (one-indexed) line number of the function definition, and drop preceding
+    # lines - i.e. any decorators, so that adding `@example()`s keeps the same key.
+    try:
+        func_lineno = ast.parse(src).body[0].lineno - 1
+        src = "".join(src.splitlines(keepends=True)[func_lineno:])
+    except Exception:
+        pass
+    # Remove blank lines and use the tokenize module to strip out comments,
+    # so that those can be changed without changing the database key.
+    try:
+        src = untokenize(
+            t for t in generate_tokens(StringIO(src).readline) if t.type != COMMENT
+        )
+    except Exception:
+        pass
+    # Finally, remove any trailing whitespace and empty lines as a last cleanup.
+    return "\n".join(x.rstrip() for x in src.splitlines() if x.rstrip()).encode()
+
+
 def function_digest(function):
     """Returns a string that is stable across multiple invocations across
     multiple processes and is prone to changing significantly in response to
@@ -56,25 +92,15 @@ def function_digest(function):
     No guarantee of uniqueness though it usually will be.
     """
     hasher = hashlib.sha384()
+    hasher.update(_get_clean_source(function))
     try:
-        hasher.update(inspect.getsource(function).encode())
-    except (OSError, TypeError):
-        pass
-    try:
-        hasher.update(function.__name__.encode())
-    except AttributeError:
-        pass
-    try:
-        # We prefer to use the modern signature API, but left this for compatibility.
-        # While we don't promise stability of the database, there's no advantage to
-        # using signature here, so we might as well keep the existing keys for now.
-        spec = inspect.getfullargspec(function)
-        if inspect.ismethod(function):
-            del spec.args[0]
-        hasher.update(repr(spec).encode())
-    except TypeError:
+        # This is additional to the source code because it can include the effects
+        # of decorators, or of post-hoc assignment to the .__signature__ attribute.
+        hasher.update(repr(get_signature(function)).encode())
+    except Exception:
         pass
     try:
+        # We set this in order to distinguish e.g. @pytest.mark.parametrize cases.
         hasher.update(function._hypothesis_internal_add_digest)
     except AttributeError:
         pass