is_hfs_dotgit: loosen over-eager match of \u{..47}

Our is_hfs_dotgit function relies on the hackily-implemented
next_hfs_char to give us the next character that an HFS+
filename comparison would look at. It's hacky because it
doesn't implement the full case-folding table of HFS+; it
gives us just enough to see if the path matches ".git".

At the end of next_hfs_char, we use tolower() to convert our
32-bit code point to lowercase. Our tolower() implementation
only takes an 8-bit char, though; it throws away the upper
24 bits. This means we can't have any false negatives for
is_hfs_dotgit. We only care about matching 7-bit ASCII
characters in ".git", and we will correctly process 'G' or
'g'.

However, we _can_ have false positives. Because we throw
away the upper bits, code point \u{0147} (for example) will
look like 'G' and get downcased to 'g'. It's not known
whether a sequence of code points whose truncation ends up
as ".git" is meaningful in any language, but it does not
hurt to be more accurate here. We can just pass out the full
32-bit code point, and compare it manually to the upper and
lowercase characters we care about.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

diff --git a/t/t1450-fsck.sh b/t/t1450-fsck.sh
index 6edd99a81ef235808ffd36327813ccf1e8832e06..0279b2b1c00fab5f7ff25b760443a6675ffbb940 100755 (executable)
--- a/t/t1450-fsck.sh
+++ b/t/t1450-fsck.sh
@@ -273,4 +273,19 @@ dot-backslash-case .\\\\.GIT\\\\foobar
 dotgit-case-backslash .git\\\\foobar
 EOF
 
+test_expect_success 'fsck allows .Ňit' '
+       (
+               git init not-dotgit &&
+               cd not-dotgit &&
+               echo content >file &&
+               git add file &&
+               git commit -m base &&
+               blob=$(git rev-parse :file) &&
+               printf "100644 blob $blob\t.\\305\\207it" >tree &&
+               tree=$(git mktree <tree) &&
+               git fsck 2>err &&
+               test_line_count = 0 err
+       )
+'
+
 test_done

diff --git a/utf8.c b/utf8.c
index 2c6442cc112dc065fd49d0374af14c6966acb78b..9c9fa3a75781baa7507d9318c8a03ec54625679d 100644 (file)
--- a/utf8.c
+++ b/utf8.c
@@ -630,8 +630,8 @@ int mbs_chrlen(const char **text, size_t *remainder_p, const char *encoding)
 }
 
 /*
- * Pick the next char from the stream, folding as an HFS+ filename comparison
- * would. Note that this is _not_ complete by any means. It's just enough
+ * Pick the next char from the stream, ignoring codepoints an HFS+ would.
+ * Note that this is _not_ complete by any means. It's just enough
  * to make is_hfs_dotgit() work, and should not be used otherwise.
  */
 static ucs_char_t next_hfs_char(const char **in)
@@ -668,12 +668,7 @@ static ucs_char_t next_hfs_char(const char **in)
                        continue;
                }
 
-               /*
-                * there's a great deal of other case-folding that occurs,
-                * but this is enough to catch anything that will convert
-                * to ".git"
-                */
-               return tolower(out);
+               return out;
        }
 }
 
@@ -681,10 +676,23 @@ int is_hfs_dotgit(const char *path)
 {
        ucs_char_t c;
 
-       if (next_hfs_char(&path) != '.' ||
-           next_hfs_char(&path) != 'g' ||
-           next_hfs_char(&path) != 'i' ||
-           next_hfs_char(&path) != 't')
+       c = next_hfs_char(&path);
+       if (c != '.')
+               return 0;
+       c = next_hfs_char(&path);
+
+       /*
+        * there's a great deal of other case-folding that occurs
+        * in HFS+, but this is enough to catch anything that will
+        * convert to ".git"
+        */
+       if (c != 'g' && c != 'G')
+               return 0;
+       c = next_hfs_char(&path);
+       if (c != 'i' && c != 'I')
+               return 0;
+       c = next_hfs_char(&path);
+       if (c != 't' && c != 'T')
                return 0;
        c = next_hfs_char(&path);
        if (c && !is_dir_sep(c))