Add test as examples for macho/winpe symbol matching

Signed-off-by: Ayan Sinha Mahapatra <ayansmahapatra@gmail.com>

Author: AyanSinhaMahapatra

scanpipe/pipes/symbolmap.py

@@ -41,8 +41,8 @@
 SMALL_FILE_SYMBOLS_THRESHOLD = 20
 MATCHING_RATIO_RUST_SMALL_FILE = 0.4
 MATCHING_RATIO_ELF = 0.05
-MATCHING_RATIO_MACHO = 0.5
-MATCHING_RATIO_WINPE = 0.5
+MATCHING_RATIO_MACHO = 0.15
+MATCHING_RATIO_WINPE = 0.15
 MATCHING_RATIO_JAVASCRIPT = 0.7
 SMALL_FILE_SYMBOLS_THRESHOLD_JAVASCRIPT = 30
 MATCHING_RATIO_JAVASCRIPT_SMALL_FILE = 0.5

scanpipe/tests/data/d2d-macho/from-lumen.zip

@@ -1599,6 +1599,68 @@ def test_scanpipe_pipes_d2d_map_elf_symbols(self):
             ).count(),
         )
 
+    @skipIf(sys.platform == "darwin", "Test is failing on macOS")
+    def test_scanpipe_pipes_d2d_map_macho_symbols(self):
+        input_dir = self.project1.input_path
+        input_resources = [
+            self.data / "d2d-macho/from-lumen.zip",
+            self.data / "d2d-macho/to-lumen.zip",
+        ]
+        copy_inputs(input_resources, input_dir)
+        self.from_files, self.to_files = d2d.get_inputs(self.project1)
+        inputs_with_codebase_path_destination = [
+            (self.from_files, self.project1.codebase_path / d2d.FROM),
+            (self.to_files, self.project1.codebase_path / d2d.TO),
+        ]
+        for input_files, codebase_path in inputs_with_codebase_path_destination:
+            for input_file_path in input_files:
+                scancode.extract_archive(input_file_path, codebase_path)
+
+        scancode.extract_archives(
+            self.project1.codebase_path,
+            recurse=True,
+        )
+        pipes.collect_and_create_codebase_resources(self.project1)
+        buffer = io.StringIO()
+        d2d.map_macho_binaries_with_symbols(project=self.project1, logger=buffer.write)
+        self.assertEqual(
+            9,
+            CodebaseRelation.objects.filter(
+                project=self.project1, map_type="macho_symbols"
+            ).count(),
+        )
+
+    @skipIf(sys.platform == "darwin", "Test is failing on macOS")
+    def test_scanpipe_pipes_d2d_map_winpe_symbols(self):
+        input_dir = self.project1.input_path
+        input_resources = [
+            self.data / "d2d-winpe/to-translucent.zip",
+            self.data / "d2d-winpe/from-translucent.zip",
+        ]
+        copy_inputs(input_resources, input_dir)
+        self.from_files, self.to_files = d2d.get_inputs(self.project1)
+        inputs_with_codebase_path_destination = [
+            (self.from_files, self.project1.codebase_path / d2d.FROM),
+            (self.to_files, self.project1.codebase_path / d2d.TO),
+        ]
+        for input_files, codebase_path in inputs_with_codebase_path_destination:
+            for input_file_path in input_files:
+                scancode.extract_archive(input_file_path, codebase_path)
+
+        scancode.extract_archives(
+            self.project1.codebase_path,
+            recurse=True,
+        )
+        pipes.collect_and_create_codebase_resources(self.project1)
+        buffer = io.StringIO()
+        d2d.map_winpe_binaries_with_symbols(project=self.project1, logger=buffer.write)
+        self.assertEqual(
+            4,
+            CodebaseRelation.objects.filter(
+                project=self.project1, map_type="winpe_symbols"
+            ).count(),
+        )
+
     @mock.patch("scanpipe.pipes.purldb.match_resources")
     def test_scanpipe_pipes_d2d_match_purldb_resource_no_package_data(
         self, mock_match_resource

scanpipe/tests/data/d2d-macho/to-lumen.zip

@@ -84,7 +84,6 @@ install_requires =
     elf-inspector==0.0.3
     go-inspector==0.5.0
     rust-inspector==0.1.0
-    source-inspector==0.6.1
     binary-inspector==0.1.2
     python-inspector==0.14.0
     source-inspector==0.6.1; sys_platform != "darwin" and platform_machine != "arm64"