Parser and other bug fixes

codesectools/datasets/BenchmarkJava/dataset.py

@@ -12,6 +12,7 @@
 
 from codesectools.datasets.core.dataset import File, PrebuiltFileDataset
 from codesectools.shared.cwe import CWE, CWEs
+from codesectools.utils import CPU_COUNT
 
 
 class TestCode(File):
@@ -65,7 +66,7 @@ class BenchmarkJava(PrebuiltFileDataset):
     license = "GPL-2.0"
     license_url = "https://github.com/OWASP-Benchmark/BenchmarkJava/blob/master/LICENSE"
 
-    build_command = "mvn clean compile"
+    build_command = f"mvn clean compile -T {CPU_COUNT // 2}"
     prebuilt_expected = (Path("target/classes/org/owasp/benchmark/testcode"), "*.class")
     artifacts_arg = "."
 

codesectools/sasts/all/parser.py

@@ -151,10 +151,11 @@ def stats_by_scores(self) -> dict:
                     ]
                 )
 
-                for line in defect.lines:
-                    if not defect_locations.get(line):
-                        defect_locations[line] = []
-                    defect_locations[line].append(defect)
+                if defect.lines:
+                    for line in defect.lines:
+                        if not defect_locations.get(line):
+                            defect_locations[line] = []
+                        defect_locations[line].append(defect)
 
             same_location = 0
             same_location_same_cwe = 0

codesectools/sasts/all/report.py

@@ -130,7 +130,7 @@ def generate_single_defect(self, defect_file: dict) -> str:
                     else "None"
                 )
                 rows.append(
-                    (float("inf"), "None", defect.sast, cwe_link, defect.message)
+                    (float("inf"), "None", defect.sast_name, cwe_link, defect.message)
                 )
 
         for row in sorted(rows, key=lambda r: r[0]):
@@ -176,7 +176,12 @@ def generate_single_defect(self, defect_file: dict) -> str:
 
         html_content = file_page.export_html(code_format=self.TEMPLATE)
         html_content = html_content.replace('href="HACK', 'id="')
-        html_content = html_content.replace("[name]", defect_file["source_path"])
+        html_content = html_content.replace(
+            "[name]",
+            str(
+                Path(defect_file["source_path"]).relative_to(self.result.source_path)  # ty:ignore[no-matching-overload]
+            ),
+        )
         html_content = html_content.replace("[tippy_calls]", tippy_calls)
 
         return html_content

codesectools/sasts/core/parser/__init__.py

@@ -52,8 +52,10 @@ def __init__(
             lines: A list of line numbers where the defect is located.
 
         """
+        if not filepath.is_file():
+            raise FileNotFoundError(filepath.resolve())
         self.filepath = filepath
-        self.filepath_str = str(filepath)
+        self.filepath_str = str(filepath.resolve())
         self.filename = filepath.name
         self.sast_name = sast_name
         self.checker = checker
@@ -70,6 +72,7 @@ def __repr__(self) -> str:
 
         """
         return f"""{self.__class__.__name__}(
+    sast: \t{self.sast_name}       
     filepath: \t{self.filepath}
     checker: \t{self.checker}
     level: \t{self.level}

codesectools/sasts/core/parser/format/SARIF/parser.py

@@ -54,6 +54,9 @@ def __init__(
         for result in self.results:
             filepath, lines = self.get_location(result)
 
+            if not filepath:
+                continue
+
             if rule_id := result.rule_id:
                 rule = self.rules[rule_id]
 
@@ -114,15 +117,16 @@ def get_rule_properties(self, rule_id: str) -> PropertyBag | None:
                 return properties
         return None
 
-    def get_location(self, result: Result) -> tuple[Path, list[int] | None]:
+    def get_location(self, result: Result) -> tuple[Path | None, list[int]]:
         """Extract the file path and line numbers from a SARIF result."""
-        lines = None
+        filepath = None
+        lines = []
         if result.locations:
             if physical_location := result.locations[0].physical_location:
                 if root := physical_location.root:
                     if artifact_location := root.artifact_location:
                         if uri := artifact_location.uri:
-                            filepath = Path(uri)
+                            filepath = self.source_path / uri
 
                     if region := root.region:
                         lines = [

codesectools/sasts/tools/SpotBugs/parser.py

@@ -7,6 +7,7 @@
 
 from __future__ import annotations
 
+from collections import defaultdict
 from pathlib import Path
 from typing import TYPE_CHECKING, Any
 
@@ -29,30 +30,27 @@ class SpotBugsAnalysisResult(SARIFAnalysisResult):
     # /home/user/mypackage/src/main/java/org/mypackage/...
     def patch_dict(self, sarif_dict: dict) -> dict:
         """Patch the SARIF dictionary to resolve relative Java class paths."""
-        partial_parents = {}
+        file_index = defaultdict(list)
+        for file_path in self.source_path.rglob("*.java"):  # SpotBugs only support Java
+            file_index[file_path.name].append(file_path)
 
         def recursive_patch(data: Any) -> None:  # noqa: ANN401
             if isinstance(data, dict):
                 for key, value in data.items():
                     if key == "uri":
                         partial_filepath = Path(value)
-                        if partial_filepath.parent not in partial_parents:
-                            if next(
-                                self.source_path.rglob(str(partial_filepath)), None
+                        candidates = file_index.get(partial_filepath.name, [])
+
+                        found = None
+                        for candidate in candidates:
+                            if (
+                                candidate.parts[-len(partial_filepath.parts) :]
+                                == partial_filepath.parts
                             ):
-                                filepath = next(
-                                    self.source_path.rglob(str(partial_filepath))
-                                ).relative_to(self.source_path)
-                                partial_parents[partial_filepath.parent] = (
-                                    filepath.parent
-                                )
-                        else:
-                            filepath = (
-                                partial_parents[partial_filepath.parent]
-                                / partial_filepath.name
-                            )
+                                found = str(candidate.resolve())
+                                break
 
-                        data[key] = str(filepath)
+                        data[key] = found if found else None
                     else:
                         recursive_patch(value)
 

codesectools/sasts/tools/SpotBugs/sast.py

@@ -61,6 +61,8 @@ class SpotBugsSAST(PrebuiltSAST):
         [
             "spotbugs",
             "-textui",
+            "-maxHeap",
+            "4096",
             "-nested:true",
             "-progress",
             "-sarif=spotbugs.sarif",

codesectools/utils.py

@@ -259,4 +259,4 @@ def shorten_path(path: str, max_len: int = 20) -> str:
     return str(shortened_path)
 
 
-CPU_COUNT = os.cpu_count()
+CPU_COUNT = os.cpu_count() or 2

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "CodeSecTools"
-version = "0.15.1"
+version = "0.15.2"
 description = "A framework for code security that provides abstractions for static analysis tools and datasets to support their integration, testing, and evaluation."
 readme = "README.md"
 license = "AGPL-3.0-only"

tests/test_all_sasts.py

@@ -11,7 +11,7 @@
 from codesectools.sasts import SASTS_ALL
 from codesectools.sasts.all.cli import build_cli
 from codesectools.sasts.all.sast import AllSAST
-from codesectools.utils import run_command
+from codesectools.utils import CPU_COUNT, run_command
 
 all_sast = AllSAST()
 
@@ -46,7 +46,9 @@ def test_analyze(monkeypatch: pytest.MonkeyPatch) -> None:
     git.Repo.clone_from("https://github.com/appsecco/dvja.git", "/tmp/dvja")
     monkeypatch.chdir("/tmp/dvja")
 
-    retcode, stdout = run_command("mvn clean compile".split(" "), cwd=Path("/tmp/dvja"))
+    retcode, stdout = run_command(
+        f"mvn clean compile -T {CPU_COUNT // 2}".split(" "), cwd=Path("/tmp/dvja")
+    )
     assert retcode == 0
 
     result = runner.invoke(