feat: Use unicodedata.name on non-alphanumeric attr names (#993)

.pre-commit-config.yaml

@@ -8,7 +8,7 @@ repos:
       - id: end-of-file-fixer
       - id: debug-statements
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.3.3
+    rev: v0.3.4
     hooks:
       - id: ruff
         args: [ --fix, --show-fixes]

tests/codegen/handlers/test_create_compound_fields.py

@@ -237,7 +237,8 @@ def test_build_attr_choice(self):
 
         actual = self.processor.build_attr_choice(attr)
 
-        self.assertEqual(attr.local_name, actual.name)
+        self.assertEqual(attr.name, actual.name)
+        self.assertEqual(attr.local_name, actual.local_name)
         self.assertEqual(attr.namespace, actual.namespace)
         self.assertIsNone(actual.default)
         self.assertEqual(attr.tag, actual.tag)

tests/codegen/models/test_attr.py

@@ -6,6 +6,15 @@
 
 
 class AttrTests(FactoryTestCase):
+    def test__post__init__(self):
+        attr = AttrFactory.create(name="$")
+        self.assertEqual("$", attr.local_name)
+        self.assertEqual("DOLLAR SIGN", attr.name)
+
+        attr = AttrFactory.create(local_name="$", name="dollar")
+        self.assertEqual("$", attr.local_name)
+        self.assertEqual("dollar", attr.name)
+
     def test__eq__(self):
         attr = AttrFactory.element()
         clone = attr.clone()

tests/codegen/test_utils.py

@@ -324,8 +324,6 @@ def test_rename_duplicate_attributes(self):
             AttrFactory.create(name="g[A]", tag=Tag.ENUMERATION),
             AttrFactory.create(name="g_a", tag=Tag.ENUMERATION),
             AttrFactory.create(name="g_a_1", tag=Tag.ENUMERATION),
-            AttrFactory.create(name="%", tag=Tag.ENUMERATION),
-            AttrFactory.create(name="$", tag=Tag.ENUMERATION),
         ]
         target = ClassFactory.create(attrs=attrs)
 
@@ -346,8 +344,6 @@ def test_rename_duplicate_attributes(self):
             "g[A]_2",
             "g_a_3",
             "g_a_1",
-            "%",
-            "$_1",
         ]
         self.assertEqual(expected, [x.name for x in attrs])
 

tests/fixtures/annotations/units.py

@@ -8,5 +8,5 @@
 class unit(Enum):
     M = "m"
     KG = "kg"
-    VALUE = "%"
+    PERCENT_SIGN = "%"
     NA = "NA"

tests/formats/dataclass/test_filters.py

@@ -517,6 +517,7 @@ def test_field_choices(self):
         attr = AttrFactory.create(
             choices=[
                 AttrFactory.element(
+                    name="$",
                     namespace="foo",
                     types=[type_float],
                     restrictions=Restrictions(max_exclusive="10"),
@@ -530,8 +531,8 @@ def test_field_choices(self):
 
         actual = self.filters.field_choices(attr, "foo", ["a", "b"])
         expected = (
-            {"name": "attr_B", "type": "Type[float]", "max_exclusive": 10.0},
-            {"name": "attr_C", "namespace": "bar", "type": "Type[str]"},
+            {"name": "$", "type": "Type[float]", "max_exclusive": 10.0},
+            {"name": "attr_B", "namespace": "bar", "type": "Type[str]"},
             {
                 "namespace": "##other",
                 "wildcard": True,

xsdata/codegen/handlers/create_compound_fields.py

@@ -252,7 +252,8 @@ def build_attr_choice(cls, attr: Attr) -> Attr:
         restrictions.sequence = None
 
         return Attr(
-            name=attr.local_name,
+            name=attr.name,
+            local_name=attr.local_name,
             namespace=attr.namespace,
             types=[x.clone() for x in attr.types],
             tag=attr.tag,

xsdata/codegen/models.py

@@ -1,6 +1,7 @@
 import copy
 import operator
 import sys
+import unicodedata
 from dataclasses import asdict, dataclass, field, fields, replace
 from enum import IntEnum
 from typing import Any, Dict, Iterator, List, Optional, Tuple, Type, TypeVar
@@ -271,6 +272,7 @@ class Attr(CodegenModel):
         tag: The xml tag that produced this attr
         name: The final attr name
         local_name: The original attr name
+        wrapper: The wrapper element name
         index: The index position of this attr in the class
         default: The default value
         fixed: Specifies if the default value is fixed
@@ -286,7 +288,7 @@ class Attr(CodegenModel):
 
     tag: str
     name: str = field(compare=False)
-    local_name: str = field(init=False)
+    local_name: str = field(default="")
     wrapper: Optional[str] = field(default=None)
     index: int = field(compare=False, default_factory=int)
     default: Optional[str] = field(default=None, compare=False)
@@ -301,8 +303,12 @@ class Attr(CodegenModel):
     substitution: Optional[str] = field(default=None, compare=False)
 
     def __post_init__(self):
-        """Set the original attr name on init."""
-        self.local_name = self.name
+        """Post init processing."""
+        if not self.local_name:
+            self.local_name = self.name
+
+        if text.alnum(self.name) == "":
+            self.name = "_".join(unicodedata.name(char) for char in self.name)
 
     @property
     def key(self) -> str:

xsdata/formats/dataclass/filters.py

@@ -456,7 +456,10 @@ def field_metadata(
         return self.filter_metadata(metadata)
 
     def field_choices(
-        self, attr: Attr, parent_namespace: Optional[str], parents: List[str]
+        self,
+        attr: Attr,
+        parent_namespace: Optional[str],
+        parents: List[str],
     ) -> Optional[Tuple]:
         """Return a tuple of field metadata if the attr has choices."""
         if not attr.choices:
@@ -471,7 +474,7 @@ def field_choices(
             )
 
             metadata = {
-                "name": choice.name,
+                "name": choice.local_name,
                 "wildcard": choice.is_wildcard,
                 "type": self.choice_type(choice, parents),
                 "namespace": namespace,