Start adding operator precedence

Author: rocky

mathics_scanner/data/named-characters.yml

@@ -21,6 +21,9 @@
 #   operator-name: If present, this symbol is a Mathics operator with
 #                  whose class name is the given name. For example Divide.
 #
+#   precedence: If present, this symbol is a Mathics operator with the
+#               specific numeric precidence value.
+#
 #   unicode-equivalent: A unicode equivalent for the named-character, if it
 #                       exists.
 #
@@ -1243,6 +1246,7 @@ CompoundExpression:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: CompoundExpression
+  precedence: 10
 Condition:
   ascii: "/;"
   has-unicode-inverse: false
@@ -1453,12 +1457,14 @@ Decrement:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Decrement
+  precedence: 660
 # See also RawQuestion
 Definition:
   ascii: "?"
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Definition
+  precedence: 670
 
 # \[Degree] is letter-like, not an operator, which is
 # mutually exclusive.
@@ -1631,6 +1637,7 @@ Divide:
   has-unicode-inverse: true
   is-letter-like: false
   operator-name: Divide
+  precedence: 470
   unicode-equivalent: "\xF7"
   unicode-equivalent-name: DIVISION SIGN
   wl-unicode: "\xF7"
@@ -2694,6 +2701,7 @@ Factorial:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Factorial
+  precedence: 610
   unicode-equivalent: "!"
   unicode-equivalent-name: EXCLAMATION MARK
   wl-unicode: "!"
@@ -3711,6 +3719,7 @@ Get:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Get
+  precedence: 720
 Gimel:
   amslatex: "\\gimel"
   esc-alias: gi
@@ -4366,6 +4375,7 @@ Increment:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Increment
+  precedence: 660
 IndentingNewLine:
   esc-alias: nl
   has-unicode-inverse: false
@@ -4392,6 +4402,7 @@ Information:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Information
+  precedence: 0
 Integral:
   amslatex: "\\int"
   esc-alias: int
@@ -4978,6 +4989,7 @@ Minus:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Minus
+  precedence: 480
   unicode-equivalent: "-"
   unicode-equivalent-name: HYPHEN-MINUS
   wl-unicode: "-"
@@ -5751,6 +5763,7 @@ Plus:
   has-unicode-inverse: true
   is-letter-like: false
   operator-name: Plus
+  precedence: 310
   unicode-equivalent: "+"
   unicode-equivalent-name: PLUS SIGN
   wl-unicode: "+"
@@ -5780,6 +5793,7 @@ Power:
   has-unicode-inverse: true
   is-letter-like: false
   operator-name: Power
+  precedence: 590
   unicode-equivalent: "^"
   unicode-equivalent-name: CIRCUMFLEX ACCENT
   wl-unicode: "^"
@@ -5873,11 +5887,13 @@ Put:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Put
+  precedence: 30
 PutAppend:
   ascii: ">>>"
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: PutAppend
+  precedence: 30
 QuarterNote:
   has-unicode-inverse: false
   is-letter-like: false
@@ -6501,6 +6517,7 @@ SameQ:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: SameQ
+  precedence: 290
 Sampi:
   esc-alias: sa
   has-unicode-inverse: true
@@ -6984,6 +7001,7 @@ Set:
   has-unicode-inverse: true
   is-letter-like: false
   operator-name: Set
+  precedence: 40
   unicode-equivalent: "="
   unicode-equivalent-name: EQUALS SIGN
   wl-unicode: "="
@@ -7434,6 +7452,7 @@ Times:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Times
+  precedence: 400
   unicode-equivalent: "\xD7"
   unicode-equivalent-name: MULTIPLICATION SIGN
   wl-unicode: "\xD7"
@@ -7574,6 +7593,7 @@ Unset:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: Unset
+  precedence: 670
 UpArrow:
   amslatex: "\\uparrow"
   has-unicode-inverse: false
@@ -7675,6 +7695,7 @@ UpSet:
   has-unicode-inverse: false
   is-letter-like: false
   operator-name: UpSet
+  precedence: 40
 UpSetDelayed:
   ascii: "^:="
   has-unicode-inverse: false

test/test_general_yaml_sanity.py

@@ -1,8 +1,9 @@
 # -*- coding: utf-8 -*-
 
-from mathics_scanner.load import load_mathics_character_yaml
 import unicodedata
 
+from mathics_scanner.load import load_mathics_character_yaml
+
 yaml_data = load_mathics_character_yaml()
 
 
@@ -21,7 +22,6 @@ def check_attr_is_invertible(attr: str):
             ), f"{attr_vs} all have the same {attr} field set to {attr_v}"
 
 
-
 def check_has_attr(attr: str):
     for k, v in yaml_data.items():
         assert attr in v, f"{k} has no {attr} attribute"
@@ -37,6 +37,7 @@ def test_yaml_field_names():
             "has-unicode-inverse",
             "is-letter-like",
             "operator-name",
+            "precedence",
             "unicode-equivalent",
             "unicode-equivalent-name",
             "wl-unicode",
@@ -56,11 +57,11 @@ def test_amslatex():
         if a.startswith("$"):
             assert a.endswith("$"), (
                 msg_prefix + "if something starts in math mode, it must end with '$'"
-                )
+            )
         if a.endswith("$"):
             assert a.startswith("$"), (
                 msg_prefix + "if something ends in math mode, it must start with '$'"
-                )
+            )
         if a in dup_amslatex:
             continue
         assert a not in amslatex_seen, msg_prefix + "value seen before"
@@ -103,38 +104,14 @@ def test_operators():
         operator_name_seen.add(operator_name)
 
 
-def test_wl_unicode_name():
+def test_precedence():
     for k, v in yaml_data.items():
-        if "wl-unicode" not in v:
-            continue
-        wl = v["wl-unicode"]
-
-        try:
-            expected_name = unicodedata.name(wl)
-        except (ValueError, TypeError):
-            continue
-
-        real_name = v.get("wl-unicode-name")
-
-        if real_name is None:
-            raise ValueError("Section {k}'s wl-unicode has a name but it isn't listed")
-
-        assert (
-            real_name == expected_name
-        ), f"Section {k} has wl-unicode-name set to {real_name} but it should be {expected_name}"
-
-
-def test_unicode_operators():
-    exclude_list = frozenset("Apply3Ats FunctionAmpersand".split(" "))
-    for k, v in yaml_data.items():
-        if k in exclude_list:
-            continue
-        if "operator-name" not in v:
-            continue
-        operator_name = v["operator-name"]
-        assert (
-            k == operator_name
-        ), f"Section name {k} should match operator-name {operator_name} when a section has an operator"
+        if "precedence" in v:
+            p = v.get("precedence", None)
+            msg_prefix = f"In {k} with precedence {p}, "
+            assert isinstance(p, int), msg_prefix + "must be an integer"
+            assert p >= 0, msg_prefix + "must be a positive integer"
+            assert "operator-name" in v, msg_prefix + "must have an operator name"
 
 
 def test_unicode_name():
@@ -186,6 +163,40 @@ def test_wl_unicode():
         ), f"{k} has neither wl-unicode nor unicode-equivalent attribute"
 
 
+def test_unicode_operators():
+    exclude_list = frozenset("Apply3Ats FunctionAmpersand".split(" "))
+    for k, v in yaml_data.items():
+        if k in exclude_list:
+            continue
+        if "operator-name" not in v:
+            continue
+        operator_name = v["operator-name"]
+        assert (
+            k == operator_name
+        ), f"Section name {k} should match operator-name {operator_name} when a section has an operator"
+
+
+def test_wl_unicode_name():
+    for k, v in yaml_data.items():
+        if "wl-unicode" not in v:
+            continue
+        wl = v["wl-unicode"]
+
+        try:
+            expected_name = unicodedata.name(wl)
+        except (ValueError, TypeError):
+            continue
+
+        real_name = v.get("wl-unicode-name")
+
+        if real_name is None:
+            raise ValueError("Section {k}'s wl-unicode has a name but it isn't listed")
+
+        assert (
+            real_name == expected_name
+        ), f"Section {k} has wl-unicode-name set to {real_name} but it should be {expected_name}"
+
+
 def test_general_yaml_sanity():
     # Check if required attributes are in place
     check_has_attr("is-letter-like")