Fix several accidentally quadratic functions

Signed-off-by: Jon Johnson <jon.johnson@chainguard.dev>

Author: jonjohnsonjr

src/spdx_tools/spdx/model/document.py

@@ -81,3 +81,6 @@ def __init__(
         relationships = [] if relationships is None else relationships
         extracted_licensing_info = [] if extracted_licensing_info is None else extracted_licensing_info
         check_types_and_set_values(self, locals())
+
+    def __hash__(self):
+        return id(self)

src/spdx_tools/spdx/model/relationship.py

@@ -73,3 +73,6 @@ def __init__(
         comment: Optional[str] = None,
     ):
         check_types_and_set_values(self, locals())
+
+    def __hash__(self):
+        return hash("{} -> {} ({})".format(self.spdx_element_id, str(self.related_spdx_element_id), str(self.relationship_type)))

src/spdx_tools/spdx/parser/jsonlikedict/relationship_parser.py

@@ -1,7 +1,7 @@
 # SPDX-FileCopyrightText: 2022 spdx contributors
 #
 # SPDX-License-Identifier: Apache-2.0
-from beartype.typing import Dict, List, Optional
+from beartype.typing import Dict, List, Optional, Set
 
 from spdx_tools.common.typing.constructor_type_errors import ConstructorTypeErrors
 from spdx_tools.spdx.model import Relationship, RelationshipType
@@ -35,9 +35,9 @@ def parse_all_relationships(self, input_doc_dict: Dict) -> List[Relationship]:
         document_describes: List[str] = delete_duplicates_from_list(input_doc_dict.get("documentDescribes", []))
         doc_spdx_id: Optional[str] = input_doc_dict.get("SPDXID")
 
-        existing_relationships_without_comments: List[Relationship] = self.get_all_relationships_without_comments(
+        existing_relationships_without_comments: Set[Relationship] = set(self.get_all_relationships_without_comments(
             relationships
-        )
+        ))
         relationships.extend(
             parse_field_or_log_error(
                 self.logger,
@@ -52,9 +52,6 @@ def parse_all_relationships(self, input_doc_dict: Dict) -> List[Relationship]:
         )
 
         package_dicts: List[Dict] = input_doc_dict.get("packages", [])
-        existing_relationships_without_comments: List[Relationship] = self.get_all_relationships_without_comments(
-            relationships
-        )
 
         relationships.extend(
             parse_field_or_log_error(
@@ -110,7 +107,7 @@ def parse_relationship_type(relationship_type_str: str) -> RelationshipType:
         return relationship_type
 
     def parse_document_describes(
-        self, doc_spdx_id: str, described_spdx_ids: List[str], existing_relationships: List[Relationship]
+        self, doc_spdx_id: str, described_spdx_ids: List[str], existing_relationships: Set[Relationship]
     ) -> List[Relationship]:
         logger = Logger()
         describes_relationships = []
@@ -131,10 +128,11 @@ def parse_document_describes(
         return describes_relationships
 
     def parse_has_files(
-        self, package_dicts: List[Dict], existing_relationships: List[Relationship]
+        self, package_dicts: List[Dict], existing_relationships: Set[Relationship]
     ) -> List[Relationship]:
         # assume existing relationships are stripped of comments
         logger = Logger()
+
         contains_relationships = []
         for package in package_dicts:
             package_spdx_id: Optional[str] = package.get("SPDXID")
@@ -160,7 +158,7 @@ def parse_has_files(
         return contains_relationships
 
     def check_if_relationship_exists(
-        self, relationship: Relationship, existing_relationships: List[Relationship]
+        self, relationship: Relationship, existing_relationships: Set[Relationship]
     ) -> bool:
         # assume existing relationships are stripped of comments
         if relationship in existing_relationships:

src/spdx_tools/spdx/validation/spdx_id_validators.py

@@ -4,7 +4,9 @@
 
 import re
 
-from beartype.typing import List
+from beartype.typing import List, Set
+
+from functools import cache
 
 from spdx_tools.spdx.document_utils import get_contained_spdx_element_ids
 from spdx_tools.spdx.model import Document, File
@@ -22,11 +24,16 @@ def is_spdx_id_present_in_files(spdx_id: str, files: List[File]) -> bool:
     return spdx_id in [file.spdx_id for file in files]
 
 
+@cache
 def is_spdx_id_present_in_document(spdx_id: str, document: Document) -> bool:
-    all_spdx_ids_in_document: List[str] = get_list_of_all_spdx_ids(document)
+    all_spdx_ids_in_document: Set[str] = get_set_of_all_spdx_ids(document)
 
     return spdx_id in all_spdx_ids_in_document
 
+@cache
+def get_set_of_all_spdx_ids(document: Document) -> Set[str]:
+    return set(get_list_of_all_spdx_ids(document))
+
 
 def get_list_of_all_spdx_ids(document: Document) -> List[str]:
     all_spdx_ids_in_document: List[str] = [document.creation_info.spdx_id]