Use different regexes in KeywordDetector to improve accuracy

detect_secrets/core/audit.py

@@ -9,6 +9,7 @@
 
 from ..plugins.core import initialize
 from ..plugins.high_entropy_strings import HighEntropyStringsPlugin
+from ..plugins.keyword import KeywordDetector
 from .baseline import format_baseline_for_output
 from .baseline import merge_results
 from .bidirectional_iterator import BidirectionalIterator
@@ -518,7 +519,13 @@ def _get_secret_with_context(
     )
 
 
-def _highlight_secret(secret_line, secret_lineno, secret, filename, plugin_settings):
+def _highlight_secret(
+    secret_line,
+    secret_lineno,
+    secret,
+    filename,
+    plugin_settings,
+):
     """
     :type secret_line: str
     :param secret_line: the line on which the secret is found
@@ -544,7 +551,11 @@ def _highlight_secret(secret_line, secret_lineno, secret, filename, plugin_setti
         plugin_settings,
     )
 
-    for raw_secret in _raw_secret_generator(plugin, secret_line):
+    for raw_secret in _raw_secret_generator(
+        plugin,
+        secret_line,
+        is_php_file=filename.endswith('.php'),
+    ):
         secret_obj = PotentialSecret(
             plugin.secret_type,
             filename,
@@ -572,10 +583,14 @@ def _highlight_secret(secret_line, secret_lineno, secret, filename, plugin_setti
     )
 
 
-def _raw_secret_generator(plugin, secret_line):
+def _raw_secret_generator(plugin, secret_line, is_php_file):
     """Generates raw secrets by re-scanning the line, with the specified plugin"""
-    for raw_secret in plugin.secret_generator(secret_line):
-        yield raw_secret
+    if isinstance(plugin, KeywordDetector):
+        for raw_secret in plugin.secret_generator(secret_line, is_php_file):
+            yield raw_secret
+    else:
+        for raw_secret in plugin.secret_generator(secret_line):
+            yield raw_secret
 
     if issubclass(plugin.__class__, HighEntropyStringsPlugin):
         with plugin.non_quoted_string_regex(strict=False):

test_data/short_files/first_line.py → test_data/short_files/first_line.php

@@ -1,4 +1,4 @@
-secret = 'BEEF0123456789a'
+seecret = 'BEEF0123456789a'
 skipped_sequential_false_positive = '0123456789a'
 print('second line')
 var = 'third line'

tests/core/audit_test.py

@@ -335,7 +335,7 @@ def test_compare(self, mock_printer):
         # These files come after, because filenames are sorted first
         assert headers[2] == textwrap.dedent("""
             Secret:      3 of 4
-            Filename:    test_data/short_files/first_line.py
+            Filename:    test_data/short_files/first_line.php
             Secret Type: Hex High Entropy String
             Status:      >> REMOVED <<
         """)[1:]
@@ -406,7 +406,7 @@ def old_baseline(self):
                 ],
 
                 # This entire file will be "removed"
-                'test_data/short_files/first_line.py': [
+                'test_data/short_files/first_line.php': [
                     {
                         'hashed_secret': '0de9a11b3f37872868ca49ecd726c955e25b6e21',
                         'line_number': 1,

tests/main_test.py

@@ -194,9 +194,9 @@ def test_old_baseline_ignored_with_update_flag(
         'filename, expected_output',
         [
             (
-                'test_data/short_files/first_line.py',
+                'test_data/short_files/first_line.php',
                 textwrap.dedent("""
-                    1:secret = 'BEEF0123456789a'
+                    1:seecret = 'BEEF0123456789a'
                     2:skipped_sequential_false_positive = '0123456789a'
                     3:print('second line')
                     4:var = 'third line'

tests/pre_commit_hook_test.py

@@ -145,6 +145,9 @@ def test_that_baseline_gets_updated(
                     'hex_limit': 3,
                     'name': 'HexHighEntropyString',
                 },
+                {
+                    'name': 'KeywordDetector',
+                },
                 {
                     'name': 'PrivateKeyDetector',
                 },