Fix PowerApprox support after upstream Power/PowerApprox split

The upstream CVXPY split Power into exact (Power) and approximate
(PowerApprox) subclasses. This broke NLP solving because:

1. dnlp2smooth used `power` (function) as dict key but lookup uses
   `type(expr)` which returns the class (Power or PowerApprox)
2. power_canon.py used `p_rational` which doesn't exist on PowerApprox
3. diff_engine converters used "power" (lowercase) but class name is "Power"

Fixes:
- dnlp2smooth/__init__.py: Import Power and PowerApprox classes,
  register both with power_canon
- power_canon.py: Use `p_used` instead of `p_rational`
- converters.py: Use "Power" and "PowerApprox" as keys

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: Transurgeon

cvxpy/reductions/dnlp2smooth/canonicalizers/__init__.py

@@ -24,7 +24,7 @@
 from cvxpy.atoms.elementwise.kl_div import kl_div
 from cvxpy.atoms.elementwise.minimum import minimum
 from cvxpy.atoms.elementwise.maximum import maximum
-from cvxpy.atoms.elementwise.power import power
+from cvxpy.atoms.elementwise.power import Power, PowerApprox
 from cvxpy.atoms.elementwise.trig import cos, sin, tan
 from cvxpy.atoms.elementwise.hyperbolic import sinh, asinh, tanh, atanh
 from cvxpy.atoms.elementwise.huber import huber
@@ -77,7 +77,8 @@
     tanh: tanh_canon,
     atanh: atanh_canon,
     quad_over_lin: quad_over_lin_canon,
-    power: power_canon,
+    Power: power_canon,
+    PowerApprox: power_canon,
     Pnorm : pnorm_canon,
     DivExpression: div_canon,
     entr: entr_canon,

cvxpy/reductions/dnlp2smooth/canonicalizers/power_canon.py

@@ -24,7 +24,7 @@
 
 def power_canon(expr, args):
     x = args[0]
-    p = expr.p_rational 
+    p = expr.p_used
     shape = expr.shape
     ones = Constant(np.ones(shape))
     if p == 0:

cvxpy/reductions/solvers/nlp_solvers/diff_engine/converters.py

@@ -304,7 +304,8 @@ def _convert_diag_vec(expr, children):
     # Matrix multiplication
     "MulExpression": _convert_matmul,
     # Elementwise univariate with parameter
-    "power": lambda expr, children: _diffengine.make_power(children[0], float(expr.p.value)),
+    "Power": lambda expr, children: _diffengine.make_power(children[0], float(expr.p.value)),
+    "PowerApprox": lambda expr, children: _diffengine.make_power(children[0], float(expr.p.value)),
     # Trigonometric
     "sin": lambda _expr, children: _diffengine.make_sin(children[0]),
     "cos": lambda _expr, children: _diffengine.make_cos(children[0]),