Enforce header entity length in grammar

__init__.py

@@ -77,6 +77,27 @@ def build():
             yield " " * 8 + "^" * len(d["line"].rstrip())
 
         return "\n".join(build())
+
+class HeaderFieldError(ValidationError):
+    def __init__(self, field, found_len, expected_len):
+        self.field = field
+        self.found_len = found_len
+        self.expected_len = expected_len
+
+    def asdict(self, with_message=True):
+        return {
+            "type": "invalid_header_field",
+            "field": self.field,
+            "expected_field_count": self.expected_len,
+            "actual_field_count": self.found_len,
+            **({"message": str(self)} if with_message else {}),
+        }
+
+    def __str__(self):
+        return (
+            f"Invalid number of parameters for HEADER field '{self.field}'. "
+            f"Expected {self.expected_len}, found {self.found_len}."
+        )
 
 
 grammar = r"""
@@ -184,6 +205,12 @@ def build():
 %ignore /[ \t\f\r\n]/+
 """
 
+HEADER_FIELDS = {
+    "file_description": namedtuple('file_description', ['description', 'implementation_level']),
+    "file_name": namedtuple('file_name', ['name', 'time_stamp', 'author', 'organization', 'preprocessor_version', 'originating_system', 'authorization']),
+    "file_schema":  namedtuple('file_schema', ['schema_identifiers']),
+}
+
 
 class Ref:
     def __init__(self, id):
@@ -304,6 +331,11 @@ def process_tree(filecontent, file_tree, with_progress, with_header=False):
 
     if with_header:
         header = dict(map(make_header_ent, header.children[0].children))
+        for field in HEADER_FIELDS.keys():
+            observed = header.get(field.upper(), [])
+            expected = HEADER_FIELDS.get(field)._fields
+            if len(header.get(field.upper(), [])) != len(expected):
+                raise HeaderFieldError(field.upper(), len(observed), len(expected))
 
     n = len(data.children)
     if n:
@@ -373,6 +405,11 @@ def replace_fn(match):
         header_tree = ast.children[0]  # HEADER section
 
         header = dict(map(make_header_ent, header_tree.children[0].children))
+        for field in HEADER_FIELDS.keys():
+            observed = header.get(field.upper(), [])
+            expected = HEADER_FIELDS.get(field)._fields
+            if len(header.get(field.upper(), [])) != len(expected):
+                raise HeaderFieldError(field.upper(), len(observed), len(expected))
         return header
 
 
@@ -473,13 +510,7 @@ def schema_version(self) -> tuple[int, int, int, int]:
 
     @property
     def header(self):
-        HEADER_FIELDS = {
-            "file_description": namedtuple('file_description', ['description', 'implementation_level']),
-            "file_name": namedtuple('file_name', ['name', 'time_stamp', 'author', 'organization', 'preprocessor_version', 'originating_system', 'authorization']),
-            "file_schema":  namedtuple('file_schema', ['schema_identifiers']),
-        }
         header = {}
-
         for field_name, namedtuple_class in HEADER_FIELDS.items():
             field_data = self.header_.get(field_name.upper(), [])
             header[field_name.lower()] = namedtuple_class(*field_data)

fixtures/too_many_header_entity_fields.ifc

@@ -0,0 +1,30 @@
+ISO-10303-21;
+HEADER;
+FILE_DESCRIPTION(('ViewDefinition [ReferenceView_V1.2]', 'ExchangeRequirement [Any]'),'2;1');
+FILE_NAME('Header.ifc','2025-02-13T15:58:45',('tricott'),('Trimble Inc.'),'TrimBimToIFC rel. 4.0.2','Example - Example - 2025.0','IFC4 model', '');
+FILE_SCHEMA(('IFC4'));
+ENDSEC;
+DATA;
+#1=IFCPERSON($,$,'',$,$,$,$,$);
+#2=IFCORGANIZATION($,'',$,$,$);
+#3=IFCPERSONANDORGANIZATION(#1,#2,$);
+#4=IFCAPPLICATION(#2,'v0.7.0-6c9e130ca','IfcOpenShell-v0.7.0-6c9e130ca','');
+#5=IFCOWNERHISTORY(#3,#4,$,.NOTDEFINED.,$,#3,#4,1700419055);
+#6=IFCDIRECTION((1.,0.,0.));
+#7=IFCDIRECTION((0.,0.,1.));
+#8=IFCCARTESIANPOINT((0.,0.,0.));
+#9=IFCAXIS2PLACEMENT3D(#8,#7,#6);
+#10=IFCDIRECTION((0.,1.));
+#11=IFCGEOMETRICREPRESENTATIONCONTEXT($,'Model',3,1.E-05,#9,#10);
+#12=IFCDIMENSIONALEXPONENTS(0,0,0,0,0,0,0);
+#13=IFCSIUNIT(*,.LENGTHUNIT.,$,.METRE.);
+#14=IFCSIUNIT(*,.AREAUNIT.,$,.SQUARE_METRE.);
+#15=IFCSIUNIT(*,.VOLUMEUNIT.,$,.CUBIC_METRE.);
+#16=IFCSIUNIT(*,.PLANEANGLEUNIT.,$,.RADIAN.);
+#17=IFCMEASUREWITHUNIT(IFCPLANEANGLEMEASURE(0.017453292519943295),#16);
+#18=IFCCONVERSIONBASEDUNIT(#12,.PLANEANGLEUNIT.,'DEGREE',#17);
+#19=IFCUNITASSIGNMENT((#13,#14,#15,#18));
+#20=IFCPROJECT('0iDmeiiLP3AOllitM2Favn',#5,'',$,$,$,$,(#11),#19);
+#21=IFCSITE('3rg2jGkIH10RFhrQsGZKRk',#5,$,$,$,$,$,$,$,$,$,$,$,$);
+ENDSEC;
+END-ISO-10303-21;

test_parser.py

@@ -126,3 +126,12 @@ def test_valid_headers(filename):
     # error in body; with_header should not raise an error
     with nullcontext():
         parse(filename=filename, with_tree=False, only_header=True, with_header=True)
+
+def test_too_many_header_entity_fields():
+    with pytest.raises(ValidationError):
+        parse(filename='fixtures/too_many_header_entity_fields.ifc', only_header=True)
+
+def test_too_many_header_entity_fields_whole_file():
+    with pytest.raises(ValidationError):
+        parse(filename='fixtures/too_many_header_entity_fields.ifc', with_header=True)
+