macOS: ls-files path fails if path of workdir is NFD

Under macOS, `git ls-files path` does not work (gives an error)
if the absolute 'path' contains characters in NFD (decomposed).
This happens when core.precomposeunicode is true, which is the
most common case. The bug report says:

$ cd somewhere          # some safe place, /tmp or ~/tmp etc.
$ mkdir $'u\xcc\x88'    # ü in NFD
$ cd ü                  # or cd $'u\xcc\x88' or cd $'\xc3\xbc'
$ git init
$ git ls-files $'/somewhere/u\xcc\x88'   # NFD
  fatal: /somewhere/ü: '/somewhere/ü' is outside repository at '/somewhere/ü'
$ git ls-files $'/somewhere/\xc3\xbc'    # NFC
(the same error as above)

In the 'fatal:' error message, there are three ü;
the 1st and 2nd are in NFC, the 3rd is in NFD.

Add a test case that follows the bug report, with the simplification
that the 'ü' is replaced by an 'ä', which is already used as NFD and
NFC in t0050.

Precompose the result of getcwd(), if needed, just like all other
paths we use internally.  That way, paths comparisons are all done
in NFC and we would correctly notice that the early part of the
path given as an absolute path matches the current directory.

One possible implementation would be to re-define getcwd() similar
to opendir(), readdir() and closedir(), but since there is already a
strbuf wrapper around getcwd(), and only this wrapper is used inside
the whole codebase, equip strbuf_getcwd() with a call to the newly
created function precompose_strbuf_if_needed().

Note that precompose_strbuf_if_needed() is a function under macOS,
and is a "no-op" on all other systems.

Reported-by: Jun T <takimoto-j@kba.biglobe.ne.jp>
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

5 files changed, 24 insertions, 0 deletions

diff --git a/compat/precompose_utf8.c b/compat/precompose_utf8.c
index 0bd5c24250..5a7c90c90d 100644
--- a/compat/precompose_utf8.c
+++ b/compat/precompose_utf8.c
@@ -94,6 +94,16 @@ const char *precompose_string_if_needed(const char *in)
 	return in;
 }
 
+void precompose_strbuf_if_needed(struct strbuf *sb)
+{
+	char *buf_prec = (char *)precompose_string_if_needed(sb->buf);
+	if (buf_prec != sb->buf) {
+		size_t buf_prec_len = strlen(buf_prec);
+		free(strbuf_detach(sb, NULL));
+		strbuf_attach(sb, buf_prec, buf_prec_len, buf_prec_len + 1);
+	}
+}
+
 const char *precompose_argv_prefix(int argc, const char **argv, const char *prefix)
 {
 	int i = 0;
diff --git a/compat/precompose_utf8.h b/compat/precompose_utf8.h
index fea06cf28a..7c3cfcadb0 100644
--- a/compat/precompose_utf8.h
+++ b/compat/precompose_utf8.h
@@ -30,6 +30,7 @@ typedef struct {
 
 const char *precompose_argv_prefix(int argc, const char **argv, const char *prefix);
 const char *precompose_string_if_needed(const char *in);
+void precompose_strbuf_if_needed(struct strbuf *sb);
 void probe_utf8_pathname_composition(void);
 
 PREC_DIR *precompose_utf8_opendir(const char *dirname);
diff --git a/git-compat-util.h b/git-compat-util.h
index 3e7a59b5ff..8b63108f16 100644
--- a/git-compat-util.h
+++ b/git-compat-util.h
@@ -331,6 +331,7 @@ static inline const char *precompose_string_if_needed(const char *in)
 	return in;
 }
 
+#define precompose_strbuf_if_needed(a)
 #define probe_utf8_pathname_composition()
 #endif
 
diff --git a/strbuf.c b/strbuf.c
index 4c9ac6dc5e..b05581d8e7 100644
--- a/strbuf.c
+++ b/strbuf.c
@@ -569,6 +569,7 @@ int strbuf_getcwd(struct strbuf *sb)
 		strbuf_grow(sb, guessed_len);
 		if (getcwd(sb->buf, sb->alloc)) {
 			strbuf_setlen(sb, strlen(sb->buf));
+			precompose_strbuf_if_needed(sb);
 			return 0;
 		}
 
diff --git a/t/t0050-filesystem.sh b/t/t0050-filesystem.sh
index 325eb1c3cd..925734c819 100755
--- a/t/t0050-filesystem.sh
+++ b/t/t0050-filesystem.sh
@@ -156,4 +156,15 @@ test_expect_success CASE_INSENSITIVE_FS 'checkout with no pathspec and a case in
 	)
 '
 
+test_expect_success 'git ls-files under NFD' '
+	(
+		mkdir -p "somewhere/$aumlcdiar" &&
+		mypwd=$PWD &&
+		cd "somewhere/$aumlcdiar" &&
+		git init &&
+		git --literal-pathspecs ls-files "$mypwd/somewhere/$aumlcdiar" 2>err &&
+		>expected &&
+		test_cmp expected err
+	)
+'
 test_done