commit: reject non-characters

Unicode clause D14 defines all characters U+nFFFE and U+nFFFF (where 0 <= n <= 10h) as well as the range U+FDD0..U+FDEF as non-characters, reserved for internal use only. Disallow these characters in commit messages as they are normally not recommended for interchange. Signed-off-by: Peter Krefting <peter@softwolves.pp.se> Signed-off-by: Junio C Hamano <gitster@pobox.com>

Peter Krefting committed Jul 9, 2013 at 12:16 UTC 81050ac604e27395509b5bffe80692a14d6ad3cd
2 files changed +23 -2
commit.c
+5 -2
@@ -1305,8 +1305,11 @@ static int find_invalid_utf8(const char *buf, int len)
1305 /* Surrogates are only for UTF-16 and cannot be encoded in UTF-8. */
1306 if ((codepoint & 0x1ff800) == 0xd800)
1307 return bad_offset;
1308 - /* U+FFFE and U+FFFF are guaranteed non-characters. */
1309 - if ((codepoint & 0x1ffffe) == 0xfffe)
1308 + /* U+xxFFFE and U+xxFFFF are guaranteed non-characters. */
1309 + if ((codepoint & 0xffffe) == 0xfffe)
1310 + return bad_offset;
1311 + /* So are anything in the range U+FDD0..U+FDEF. */
1312 + if (codepoint >= 0xfdd0 && codepoint <= 0xfdef)
1313 return bad_offset;
1314 }
1315 return -1;
t/t3900-i18n-commit.sh
+18
@@ -58,6 +58,24 @@ test_expect_success 'UTF-8 overlong sequences rejected' '
58 grep "did not conform" "$HOME"/stderr
59 '
60
61 +test_expect_success 'UTF-8 non-characters refused' '
62 + test_when_finished "rm -f $HOME/stderr $HOME/invalid" &&
63 + echo "UTF-8 non-character 1" >F &&
64 + printf "Commit message\n\nNon-character:\364\217\277\276\n" \
65 + >"$HOME/invalid" &&
66 + git commit -a -F "$HOME/invalid" 2>"$HOME"/stderr &&
67 + grep "did not conform" "$HOME"/stderr
68 +'
69 +
70 +test_expect_success 'UTF-8 non-characters refused' '
71 + test_when_finished "rm -f $HOME/stderr $HOME/invalid" &&
72 + echo "UTF-8 non-character 2." >F &&
73 + printf "Commit message\n\nNon-character:\357\267\220\n" \
74 + >"$HOME/invalid" &&
75 + git commit -a -F "$HOME/invalid" 2>"$HOME"/stderr &&
76 + grep "did not conform" "$HOME"/stderr
77 +'
78 +
79 for H in ISO8859-1 eucJP ISO-2022-JP
80 do
81 test_expect_success "$H setup" '