Actually determine if we are dealing with a shallow clone instead of … (

#298)

* Actually determine if we are dealing with a shallow clone instead of assuming

* Switch tests to check the new shallow assumptions

* Mock out is_shallow_clone for all our tests

tartufo/scanner.py

@@ -767,9 +767,8 @@ def chunks(self) -> Generator[types.Chunk, None, None]:
                 branches = [self.git_options.branch]
             else:
                 # Everything
-                if not self._repo.listall_branches():
-                    # If no local branches are found, assume that this is a
-                    # shallow clone and examine the repo head as a single
+                if util.is_shallow_clone(self._repo):
+                    # If this is a shallow clone, examine the repo head as a single
                     # commit to scan all files at once
                     branches = ["HEAD"]
                 else:

tartufo/util.py

@@ -242,3 +242,17 @@ def process_issues(
         write_outputs(scan.issues, output_dir)
         if options.output_format != types.OutputFormat.Json.value:
             click.echo(f"Results have been saved in {output_dir}")
+
+
+def is_shallow_clone(repo: pygit2.Repository) -> bool:
+    """Determine whether a repository is a shallow clone
+
+    :param repo: The repository to check for "shallowness"
+
+    This is used to work around https://github.com/libgit2/libgit2/issues/3058
+    Basically, any time a git repository is a "shallow" clone (it was cloned
+    with `--max-depth N`), git will create a file at `.git/shallow`. So we
+    simply need to test whether that file exists to know whether we are
+    interacting with a shallow repository.
+    """
+    return (pathlib.Path(repo.path) / "shallow").exists()

tests/test_git_repo_scanner.py

@@ -151,12 +151,17 @@ class ChunkGeneratorTests(ScannerTestCase):
     def setUp(self) -> None:
         self.diff_patcher = mock.patch("tartufo.scanner.GitScanner._iter_diff_index")
         self.repo_patcher = mock.patch("pygit2.Repository")
+        self.shallow_patcher = mock.patch("tartufo.scanner.util.is_shallow_clone")
 
         self.mock_iter_diff = self.diff_patcher.start()
         self.mock_repo = self.repo_patcher.start()
+        self.mock_shallow = self.shallow_patcher.start()
+
+        self.mock_shallow.return_value = False
 
         self.addCleanup(self.diff_patcher.stop)
         self.addCleanup(self.repo_patcher.stop)
+        self.addCleanup(self.shallow_patcher.stop)
         return super().setUp()
 
     def test_single_branch_is_loaded_if_specified(self):
@@ -306,8 +311,8 @@ def test_error_is_raised_when_specified_branch_is_not_found(self):
             for _ in test_scanner.chunks:
                 pass
 
-    def test_head_is_scanned_when_no_local_branches_are_found(self):
-        self.mock_repo.return_value.listall_branches.return_value = []
+    def test_head_is_scanned_when_shallow_clone_is_found(self):
+        self.mock_shallow.return_value = True
         self.mock_iter_diff.return_value = []
         self.mock_repo.return_value.head.target = "commit-hash"
         mock_head = mock.MagicMock(spec=pygit2.Commit)