Add error message byte order mark (#63)

* there was an issue with the bom test, this is fixed now

* fixed issue with line number

* Added an error for a file with a byte order mark

* increased version

---------

Co-authored-by: SanderKools <s.w.a.kools@minezk.nl>

README.md

@@ -77,6 +77,7 @@ a dict with three keys:
 | "no_csaf_file"        | "All CSAF fields must point to a provider-metadata.json file."                                                                                                         |
 | "pgp_data_error"      | "Signed message did not contain a correct ASCII-armored PGP block."                                                                                                    |
 | "pgp_error"           | "Decoding or parsing of the pgp message failed."                                                                                                                       |
+| "bom_in_file"         | "The Byte-Order Mark was found in the UTF-8 File. Security.txt must be encoded using UTF-8 in Net-Unicode form, the BOM signature must not appear at the beginning."   |
 
 
 ### Possible recommendations

requirements.txt

@@ -1,6 +1,6 @@
 requests==2.31.0
 python-dateutil==2.8.2
 langcodes==3.3.0
-pytest==7.4.0
+pytest==7.4.3
 requests-mock==1.11.0
 PGPy==0.6.0

sectxt/__init__.py

@@ -22,7 +22,7 @@
 import dateutil.parser
 import requests
 
-__version__ = "0.9.0"
+__version__ = "0.9.1"
 
 s = requests.Session()
 
@@ -103,8 +103,10 @@ def _add_error(
         explicit_line_no=None
     ) -> None:
         if explicit_line_no:
-            self._line_no = explicit_line_no
-        err_dict: ErrorDict = {"code": code, "message": message, "line": self._line_no}
+            error_line = explicit_line_no
+        else:
+            error_line = self._line_no
+        err_dict: ErrorDict = {"code": code, "message": message, "line": error_line}
         self._errors.append(err_dict)
 
     def _add_recommendation(
@@ -418,6 +420,12 @@ def _get_str(self, content: bytes) -> str:
         try:
             if content.startswith(codecs.BOM_UTF8):
                 content = content.replace(codecs.BOM_UTF8, b'')
+            self._add_error(
+                "bom_in_file",
+                "The Byte-Order Mark was found in the UTF-8 File. "
+                "Security.txt must be encoded using UTF-8 in Net-Unicode form, "
+                "the BOM signature must not appear at the beginning."
+            )
             return content.decode('utf-8')
         except UnicodeError:
             self._add_error("utf8", "Content must be utf-8 encoded.")

test/test_sectxt.py

@@ -292,12 +292,15 @@ def test_invalid_uri_scheme(requests_mock: Mocker):
 
 def test_byte_order_mark(requests_mock: Mocker):
     with Mocker() as m:
+        expires = f"Expires: {(date.today() + timedelta(days=10)).isoformat()}T18:37:07z\n"
         byte_content_with_bom = b'\xef\xbb\xbf\xef\xbb\xbfContact: mailto:me@example.com\n' \
-                                b'Expires: 2023-08-11T18:37:07z\n'
+                                + bytes(expires, "utf-8")
         m.get(
             "https://example.com/.well-known/security.txt",
             headers={"content-type": "text/plain"},
             content=byte_content_with_bom,
         )
         s = SecurityTXT("example.com")
-        assert(s.is_valid())
+        assert(not s.is_valid())
+        if not any(d["code"] == "bom_in_file" for d in s.errors):
+            pytest.fail("bom_in_file error code should be given")