fix 1260

CHANGES.rst

@@ -11,6 +11,8 @@ Enhancements
 * ``ToString`` accepts an optional *form* parameter.
 * ``ToExpression`` handles multi-line string input
 * ``FileNames`` returns a sorted list (#1250).
+* ``ReplaceRepeated`` and ``FixedPoint`` now supports the ``MaxIteration`` option (#1260).
+* ``FixedPoint`` now supports the ``SameTest`` option.
 
 Bug fixes
 +++++++++

mathics/builtin/control.py

@@ -506,7 +506,7 @@ class FixedPoint(Builtin):
     <dt>'FixedPoint[$f$, $expr$]'
         <dd>starting with $expr$, iteratively applies $f$ until the result no longer changes.
     <dt>'FixedPoint[$f$, $expr$, $n$]'
-        <dd>performs at most $n$ iterations.
+        <dd>performs at most $n$ iterations. The same that using $MaxIterations->n$
     </dl>
 
     >> FixedPoint[Cos, 1.0]
@@ -524,24 +524,46 @@ class FixedPoint(Builtin):
      = 0.739085
     """
 
-    def apply(self, f, expr, n, evaluation):
-        "FixedPoint[f_, expr_, n_:DirectedInfinity[1]]"
+    options = {
+        "MaxIterations": "Infinity",
+        "SameTest": "Automatic",
+    }
 
+    def apply(self, f, expr, n, evaluation, options):
+        "FixedPoint[f_, expr_, n_:DirectedInfinity[1], OptionsPattern[FixedPoint]]"
         if n == Expression("DirectedInfinity", 1):
             count = None
         else:
             count = n.get_int_value()
             if count is None or count < 0:
                 evaluation.message("FixedPoint", "intnn")
                 return
+
+        if count is None:
+            count = self.get_option(options, "MaxIterations", evaluation)
+            if count.is_numeric():
+                count = count.get_int_value()
+            else:
+                count = None
+
         result = expr
         index = 0
+        sametest = self.get_option(options, "SameTest", evaluation)
+        if sametest == Symbol("Automatic"):
+            sametest = None
+
         while count is None or index < count:
             evaluation.check_stopped()
             new_result = Expression(f, result).evaluate(evaluation)
-            if new_result == result:
-                result = new_result
-                break
+            if sametest:
+                same = Expression(sametest, result, new_result).evaluate(evaluation)
+                same = same.is_true()
+                if same:
+                    break
+            else:
+                if new_result == result:
+                    result = new_result
+                    break
             result = new_result
             index += 1
 

mathics/builtin/patterns.py

@@ -305,6 +305,11 @@ class ReplaceRepeated(BinaryOperator):
     >> a+b+c //. c->d
      = a + b + d
 
+    >> f = ReplaceRepeated[c->d];
+    >> f[a+b+c]
+     = a + b + d
+    >> Clear[f];
+
     Simplification of logarithms:
     >> logrules = {Log[x_ * y_] :> Log[x] + Log[y], Log[x_ ^ y_] :> y * Log[x]};
     >> Log[a * (b * c) ^ d ^ e * f] //. logrules
@@ -324,8 +329,16 @@ class ReplaceRepeated(BinaryOperator):
         "rmix": "Elements of `1` are a mixture of lists and nonlists.",
     }
 
-    def apply_list(self, expr, rules, evaluation):
-        "ReplaceRepeated[expr_, rules_]"
+    options = {
+        "MaxIterations": "65535",
+    }
+
+    rules = {
+        "ReplaceRepeated[rules_][expr_]": "ReplaceRepeated[expr, rules]",
+    }
+
+    def apply_list(self, expr, rules, evaluation, options):
+        "ReplaceRepeated[expr_, rules_, OptionsPattern[ReplaceRepeated]]"
         try:
             rules, ret = create_rules(rules, expr, "ReplaceRepeated", evaluation)
         except PatternError:
@@ -335,8 +348,17 @@ def apply_list(self, expr, rules, evaluation):
         if ret:
             return rules
 
+        maxit = self.get_option(options, "MaxIterations", evaluation)
+        if maxit.is_numeric():
+            maxit = maxit.get_int_value()
+        else:
+            maxit = -1
+
         while True:
             evaluation.check_stopped()
+            if maxit == 0:
+                break
+            maxit -= 1
             result, applied = expr.apply_rules(rules, evaluation)
             if applied:
                 result = result.evaluate(evaluation)