Update Latin-1 detection with allowed 0x1B symbol (#802)

* ESC code may be latin-1 too

* Update CredSweeper banner check in workflow

Author: babenek

.github/workflows/check.yml

@@ -92,7 +92,7 @@ jobs:
         run: |
           banner="$(python -m credsweeper --banner | head -1)"
           echo "banner = '${banner}'"
-          if [ "CredSweeper 1.14.2 crc32:a2e9bad4" != "${banner}" ]; then
+          if [ "CredSweeper 1.14.2 crc32:4b184311" != "${banner}" ]; then
             echo "Update the check for '${banner}'"
             exit 1
           fi

credsweeper/utils/util.py

@@ -153,6 +153,7 @@ def is_binary(data: Union[bytes, bytearray]) -> bool:
 
     NOT_LATIN1_PRINTABLE_SET = set(range(0, 256)) \
         .difference(set(x for x in string.printable.encode(ASCII))) \
+        .difference({0x1B}) \
         .difference(set(x for x in range(0xA0, 0x100)))
 
     @staticmethod

tests/utils/test_util.py

@@ -406,6 +406,8 @@ def test_is_latin1_p(self):
         self.assertTrue(Util.is_latin1((self.DEUTSCH_PANGRAM + '\0').encode(UTF_8)))
         # obsolete encoding may be recognized as Latin1
         self.assertTrue(Util.is_latin1(self.KOREAN_PANGRAM.encode("euc_kr")))
+        # 0x1B ESCAPE code in log
+        self.assertTrue(Util.is_latin1(b"PASSWORD=2IWJD88FH4Y;"))
 
     def test_is_ascii_entropy_validate_p(self):
         self.assertTrue(Util.is_ascii_entropy_validate(b''))