UTF7: enable detection of empty-document with byte-order-mark (#717)

* Tests: add detection coverage for UTF7 with BOM

This means that it would never provide detection of UTF7 at
runtime, because the other prefix would always take priority.

* 🐛 Allow UTF7 BOM to detect empty-content case

Checking for this special-case UTF7 BOM before the other cases
allows us to detect an empty document, which Python 3 encodes
into ASCII with a trailing minus symbol ('-').

Author: jayaddison

src/charset_normalizer/constant.py

@@ -9,11 +9,11 @@
 ENCODING_MARKS: dict[str, bytes | list[bytes]] = {
     "utf_8": BOM_UTF8,
     "utf_7": [
+        b"\x2b\x2f\x76\x38\x2d",
         b"\x2b\x2f\x76\x38",
         b"\x2b\x2f\x76\x39",
         b"\x2b\x2f\x76\x2b",
         b"\x2b\x2f\x76\x2f",
-        b"\x2b\x2f\x76\x38\x2d",
     ],
     "gb18030": b"\x84\x31\x95\x33",
     "utf_32": [BOM_UTF32_BE, BOM_UTF32_LE],

tests/test_base_detection.py

@@ -33,6 +33,7 @@ def test_bool_matches():
     [
         (b"\xfe\xff", "utf_16"),
         ("\uFEFF".encode("gb18030"), "gb18030"),
+        ("\uFEFF".encode("utf-7"), "utf_7"),
         (b"\xef\xbb\xbf", "utf_8"),
         ("".encode("utf_32"), "utf_32"),
     ],
@@ -90,6 +91,10 @@ def test_md_triggered_but_with_bom_or_sig(payload, expected_encoding):
             "我没有埋怨，磋砣的只是一些时间。".encode("utf_8_sig"),
             "utf_8",
         ),
+        (
+            ("\uFEFF" + "🐕").encode("utf-7"),
+            "utf_7",
+        ),
     ],
 )
 def test_content_with_bom_or_sig(payload, expected_encoding):