gh-85267: Improvements to inspect.signature __text_signature__ handling (#98796)

This makes a couple related changes to inspect.signature's behaviour
when parsing a signature from `__text_signature__`.

First, `inspect.signature` is documented as only raising ValueError or
TypeError. However, in some cases, we could raise RuntimeError.  This PR
changes that, thereby fixing #83685.

(Note that the new ValueErrors in RewriteSymbolics are caught and then
reraised with a message)

Second, `inspect.signature` could randomly drop parameters that it
didn't understand (corresponding to `return None` in the `p` function).
This is the core issue in #85267. I think this is very surprising
behaviour and it seems better to fail outright.

Third, adding this new failure broke a couple tests. To fix them (and to
e.g. allow `inspect.signature(select.epoll.register)` as in #85267), I
add constant folding of a couple binary operations to RewriteSymbolics.

(There's some discussion of making signature expression evaluation
arbitrary powerful in #68155. I think that's out of scope. The
additional constant folding here is pretty straightforward, useful, and
not much of a slippery slope)

Fourth, while #85267 is incorrect about the cause of the issue, it turns
out if you had consecutive newlines in __text_signature__, you'd get
`tokenize.TokenError`.

Finally, the `if name is invalid:` code path was dead, since
`parse_name` never returned `invalid`.

Author: hauntsaninja

Lib/inspect.py

@@ -2122,7 +2122,7 @@ def _signature_strip_non_python_syntax(signature):
     self_parameter = None
     last_positional_only = None
 
-    lines = [l.encode('ascii') for l in signature.split('\n')]
+    lines = [l.encode('ascii') for l in signature.split('\n') if l]
     generator = iter(lines).__next__
     token_stream = tokenize.tokenize(generator)
 
@@ -2221,11 +2221,11 @@ def wrap_value(s):
             try:
                 value = eval(s, sys_module_dict)
             except NameError:
-                raise RuntimeError()
+                raise ValueError
 
         if isinstance(value, (str, int, float, bytes, bool, type(None))):
             return ast.Constant(value)
-        raise RuntimeError()
+        raise ValueError
 
     class RewriteSymbolics(ast.NodeTransformer):
         def visit_Attribute(self, node):
@@ -2235,7 +2235,7 @@ def visit_Attribute(self, node):
                 a.append(n.attr)
                 n = n.value
             if not isinstance(n, ast.Name):
-                raise RuntimeError()
+                raise ValueError
             a.append(n.id)
             value = ".".join(reversed(a))
             return wrap_value(value)
@@ -2245,14 +2245,29 @@ def visit_Name(self, node):
                 raise ValueError()
             return wrap_value(node.id)
 
+        def visit_BinOp(self, node):
+            # Support constant folding of a couple simple binary operations
+            # commonly used to define default values in text signatures
+            left = self.visit(node.left)
+            right = self.visit(node.right)
+            if not isinstance(left, ast.Constant) or not isinstance(right, ast.Constant):
+                raise ValueError
+            if isinstance(node.op, ast.Add):
+                return ast.Constant(left.value + right.value)
+            elif isinstance(node.op, ast.Sub):
+                return ast.Constant(left.value - right.value)
+            elif isinstance(node.op, ast.BitOr):
+                return ast.Constant(left.value | right.value)
+            raise ValueError
+
     def p(name_node, default_node, default=empty):
         name = parse_name(name_node)
         if default_node and default_node is not _empty:
             try:
                 default_node = RewriteSymbolics().visit(default_node)
                 default = ast.literal_eval(default_node)
             except ValueError:
-                return None
+                raise ValueError("{!r} builtin has invalid signature".format(obj)) from None
         parameters.append(Parameter(name, kind, default=default, annotation=empty))
 
     # non-keyword-only parameters

Lib/test/test_inspect.py

@@ -2525,7 +2525,7 @@ def p(name): return signature.parameters[name].default
         self.assertEqual(p('f'), False)
         self.assertEqual(p('local'), 3)
         self.assertEqual(p('sys'), sys.maxsize)
-        self.assertNotIn('exp', signature.parameters)
+        self.assertEqual(p('exp'), sys.maxsize - 1)
 
         test_callable(object)
 
@@ -4323,10 +4323,29 @@ def func(*args, **kwargs):
         sig = inspect.signature(func)
         self.assertIsNotNone(sig)
         self.assertEqual(str(sig), '(self, /, a, b=1, *args, c, d=2, **kwargs)')
+
         func.__text_signature__ = '($self, a, b=1, /, *args, c, d=2, **kwargs)'
         sig = inspect.signature(func)
         self.assertEqual(str(sig), '(self, a, b=1, /, *args, c, d=2, **kwargs)')
 
+        func.__text_signature__ = '(self, a=1+2, b=4-3, c=1 | 3 | 16)'
+        sig = inspect.signature(func)
+        self.assertEqual(str(sig), '(self, a=3, b=1, c=19)')
+
+        func.__text_signature__ = '(self, a=1,\nb=2,\n\n\n   c=3)'
+        sig = inspect.signature(func)
+        self.assertEqual(str(sig), '(self, a=1, b=2, c=3)')
+
+        func.__text_signature__ = '(self, x=does_not_exist)'
+        with self.assertRaises(ValueError):
+            inspect.signature(func)
+        func.__text_signature__ = '(self, x=sys, y=inspect)'
+        with self.assertRaises(ValueError):
+            inspect.signature(func)
+        func.__text_signature__ = '(self, 123)'
+        with self.assertRaises(ValueError):
+            inspect.signature(func)
+
     def test_base_class_have_text_signature(self):
         # see issue 43118
         from test.ann_module7 import BufferedReader

Misc/NEWS.d/next/Library/2022-10-28-07-24-34.gh-issue-85267.xUy_Wm.rst

@@ -0,0 +1,6 @@
+Several improvements to :func:`inspect.signature`'s handling of ``__text_signature``.
+- Fixes a case where :func:`inspect.signature` dropped parameters
+- Fixes a case where :func:`inspect.signature` raised :exc:`tokenize.TokenError`
+- Allows :func:`inspect.signature` to understand defaults involving binary operations of constants
+- :func:`inspect.signature` is documented as only raising :exc:`TypeError` or :exc:`ValueError`, but sometimes raised :exc:`RuntimeError`. These cases now raise :exc:`ValueError`
+- Removed a dead code path