:merge_arrow_right_hook: Merge pull request #188 from dgzlopes/fix-180-multiple-file-support

Add multiple file support

Author: KevinHock

README.md

@@ -73,7 +73,7 @@ $ detect-secrets scan --update .secrets.baseline
 
 ### Command Line
 
-`detect-secrets` is designed to be used as a git pre-commit hook, but you can also invoke `detect-secrets scan [path]` directly (`path` defaults to `.` if not specified).
+`detect-secrets` is designed to be used as a git pre-commit hook, but you can also invoke `detect-secrets scan [path]` directly being `path` the file(s) and/or directory(ies) to scan (`path` defaults to `.` if not specified).
 
 It should be noted that by default, `detect-secrets scan` only operates on files that are tracked by git. So if you intend to scan files outside of a git repository, you will need to pass the `--all-files` flag.
 

detect_secrets/core/baseline.py

@@ -5,14 +5,17 @@
 import re
 import subprocess
 
+from detect_secrets.core.log import get_logger
 from detect_secrets.core.secrets_collection import SecretsCollection
 
+log = get_logger(format_string='%(message)s')
+
 
 def initialize(
+    path,
     plugins,
     exclude_files_regex=None,
     exclude_lines_regex=None,
-    path='.',
     scan_all_files=False,
 ):
     """Scans the entire codebase for secrets, and returns a
@@ -23,7 +26,7 @@ def initialize(
 
     :type exclude_files_regex: str|None
     :type exclude_lines_regex: str|None
-    :type path: str
+    :type path: list
     :type scan_all_files: bool
 
     :rtype: SecretsCollection
@@ -34,13 +37,17 @@ def initialize(
         exclude_lines=exclude_lines_regex,
     )
 
-    if os.path.isfile(path):
-        # This option allows for much easier adhoc usage.
-        files_to_scan = [path]
-    elif scan_all_files:
-        files_to_scan = _get_files_recursively(path)
-    else:
-        files_to_scan = _get_git_tracked_files(path)
+    files_to_scan = list()
+    for element in path:
+        if os.path.isdir(element):
+            if scan_all_files:
+                files_to_scan.extend(_get_files_recursively(element))
+            else:
+                files_to_scan.extend(_get_git_tracked_files(element))
+        elif os.path.isfile(element):
+            files_to_scan.append(element)
+        else:
+            log.error("detect-secrets: " + element + ": No such file or directory")
 
     if not files_to_scan:
         return output

detect_secrets/core/usage.py

@@ -120,7 +120,7 @@ def add_arguments(self):
     def _add_initialize_baseline_argument(self):
         self.parser.add_argument(
             'path',
-            nargs='?',
+            nargs='*',
             default='.',
             help=(
                 'Scans the entire codebase and outputs a snapshot of '

tests/core/baseline_test.py

@@ -31,24 +31,26 @@ def setup(self):
 
     def get_results(
         self,
-        path='./test_data/files',
+        path=['./test_data/files'],
         exclude_files_regex=None,
         scan_all_files=False,
     ):
         return baseline.initialize(
+            path,
             self.plugins,
-            path=path,
             exclude_files_regex=exclude_files_regex,
             scan_all_files=scan_all_files,
         ).json()
 
     @pytest.mark.parametrize(
         'path',
         [
-            './test_data/files',
+            [
+                './test_data/files',
 
-            # Test relative paths
-            'test_data/../test_data/files/tmp/..',
+                # Test relative paths
+                'test_data/../test_data/files/tmp/..',
+            ],
         ],
     )
     def test_basic_usage(self, path):
@@ -58,6 +60,31 @@ def test_basic_usage(self, path):
         assert len(results['test_data/files/file_with_secrets.py']) == 1
         assert len(results['test_data/files/tmp/file_with_secrets.py']) == 2
 
+    def test_with_multiple_files(self):
+        results = self.get_results(path=[
+            'test_data/files/file_with_secrets.py',
+            'test_data/files/tmp/file_with_secrets.py',
+        ])
+
+        assert len(results['test_data/files/file_with_secrets.py']) == 1
+        assert len(results['test_data/files/tmp/file_with_secrets.py']) == 2
+        assert 'test_data/files/file_with_secrets.py' in results
+        assert 'test_data/files/tmp/file_with_secrets.py' in results
+
+    def test_with_multiple_non_existent_files(self):
+        results = self.get_results(path=['non-existent-file.A', 'non-existent-file.B'])
+
+        # No expected results, because files don't exist
+        assert not results
+
+    def test_with_folders_and_files(self):
+        results = self.get_results(path=['test_data/', 'non-existent-file.B'])
+
+        assert 'test_data/files/file_with_secrets.py' in results
+        assert 'test_data/files/tmp/file_with_secrets.py' in results
+        assert 'test_data/files/file_with_no_secrets.py' not in results
+        assert 'non-existent-file.B' not in results
+
     def test_exclude_regex(self):
         results = self.get_results(exclude_files_regex='tmp*')
 
@@ -82,7 +109,7 @@ def test_no_files_in_git_repo(self):
                 ),
             ),
         ):
-            results = self.get_results(path='will_be_mocked')
+            results = self.get_results(path=['will_be_mocked'])
 
         assert not results
 
@@ -94,12 +121,12 @@ def test_single_non_tracked_git_file_should_work(self):
             'Super hidden value "BEEF0123456789a"',
             'detect_secrets.core.secrets_collection.codecs.open',
         ):
-            results = self.get_results('will_be_mocked')
+            results = self.get_results(path=['will_be_mocked'])
 
         assert len(results['will_be_mocked']) == 1
 
     def test_scan_all_files(self):
-        results = self.get_results(path='test_data/files', scan_all_files=True)
+        results = self.get_results(path=['test_data/files'], scan_all_files=True)
         assert len(results.keys()) == 2
 
 

tests/main_test.py

@@ -41,7 +41,7 @@ def test_scan_with_rootdir(self, mock_baseline_initialize):
             plugins=Any(tuple),
             exclude_files_regex=None,
             exclude_lines_regex=None,
-            path='test_data',
+            path=['test_data'],
             scan_all_files=False,
         )