Re: [PATCH v1 3/3] convert: ce_compare_data() checks for a sha1 of a path

9 messages, 2 authors, 2016-06-16 · open the first message on its own page

Re: [PATCH v1 3/3] convert: ce_compare_data() checks for a sha1 of a path

From: Junio C Hamano <hidden>
Date: 2016-06-16 02:19:24

tboegi@web.de writes:
-static int has_cr_in_index(const char *path)
+static int has_cr_in_index(const char *path, const unsigned char *sha1)
 {
 	unsigned long sz;
 	void *data;
 	int has_cr;
-
-	data = read_blob_data_from_cache(path, &sz);
-	if (!data)
+	enum object_type type;
+	if (!sha1)
+		sha1 = get_sha1_from_cache(path);
+	if (!sha1)
+		return 0;
+	data = read_sha1_file(sha1, &type, &sz);
+	if (!data || type != OBJ_BLOB) {
+		free(data);
 		return 0;
+	}
+
 	has_cr = memchr(data, '\r', sz) != NULL;
 	free(data);
 	return has_cr;
 }
Does this really need 2/3?  Wouldn't this be equivalent to

	if (!sha1) {
        	data = read_blob_data_from_cache(path, &sz);
	} else {
        	data = read_sha1_file(sha1, &type, &sz);
	}
	if (!data || type != OBJ_BLOB) {
        	free(data);
                return 0;
	}

        has_cr = ...

[PATCH v3 1/1] convert: ce_compare_data() checks for a sha1 of a path

From: <hidden>
Date: 2016-06-16 02:19:24

From: Torsten Bögershausen <redacted>

To compare a file in working tree with the index, convert_to_git() is used,
the the result is hashed and the hash value compared with ce->sha1.

Deep down would_convert_crlf_at_commit() is invoked, to check if CRLF
are converted or not: When a CRLF had been in the index before, CRLF in
the working tree are not converted.

While in a merge, a file name in the working tree has different blobs
in the index with different hash values.
Forwarding ce->sha1 from ce_compare_data() into crlf_to_git() makes sure
the would_convert_crlf_at_commit() looks at the appropriate blob.

Signed-off-by: Torsten Bögershausen <redacted>
---
 cache.h      |  1 +
 convert.c    | 35 +++++++++++++++++++++--------------
 convert.h    | 23 +++++++++++++++++++----
 read-cache.c |  4 +++-
 sha1_file.c  | 17 +++++++++++++----
 5 files changed, 57 insertions(+), 23 deletions(-)
diff --git a/cache.h b/cache.h
index 160f8e3..5b25462 100644
--- a/cache.h
+++ b/cache.h
@@ -604,6 +604,7 @@ extern int ie_modified(const struct index_state *, const struct cache_entry *, s
 
 #define HASH_WRITE_OBJECT 1
 #define HASH_FORMAT_CHECK 2
+#define HASH_CE_HAS_SHA1  4
 extern int index_fd(unsigned char *sha1, int fd, struct stat *st, enum object_type type, const char *path, unsigned flags);
 extern int index_path(unsigned char *sha1, const char *path, struct stat *st, unsigned flags);
 
diff --git a/convert.c b/convert.c
index f524b8d..dc86899 100644
--- a/convert.c
+++ b/convert.c
@@ -217,21 +217,25 @@ static void check_safe_crlf(const char *path, enum crlf_action crlf_action,
 	}
 }
 
-static int has_cr_in_index(const char *path)
+static int has_cr_in_index(const char *path, const unsigned char *sha1)
 {
 	unsigned long sz;
 	void *data;
-	int has_cr;
-
-	data = read_blob_data_from_cache(path, &sz);
-	if (!data)
-		return 0;
-	has_cr = memchr(data, '\r', sz) != NULL;
+	int has_cr = 0;
+	enum object_type type = OBJ_BLOB;
+	if (sha1)
+		data = read_sha1_file(sha1, &type, &sz);
+	else
+		data = read_blob_data_from_cache(path, &sz);
+
+	if (data)
+		has_cr = memchr(data, '\r', sz) != NULL;
 	free(data);
 	return has_cr;
 }
 
-static int crlf_to_git(const char *path, const char *src, size_t len,
+static int crlf_to_git(const char *path, const unsigned char *sha1,
+		       const char *src, size_t len,
 		       struct strbuf *buf,
 		       enum crlf_action crlf_action, enum safe_crlf checksafe)
 {
@@ -260,7 +264,7 @@ static int crlf_to_git(const char *path, const char *src, size_t len,
 			 * If the file in the index has any CR in it, do not convert.
 			 * This is the new safer autocrlf handling.
 			 */
-			if (has_cr_in_index(path))
+			if (has_cr_in_index(path, sha1))
 				return 0;
 		}
 	}
@@ -852,8 +856,9 @@ const char *get_convert_attr_ascii(const char *path)
 	return "";
 }
 
-int convert_to_git(const char *path, const char *src, size_t len,
-                   struct strbuf *dst, enum safe_crlf checksafe)
+int convert_to_git_ce_sha1(const char *path, const unsigned char *sha1,
+			   const char *src, size_t len,
+			   struct strbuf *dst, enum safe_crlf checksafe)
 {
 	int ret = 0;
 	const char *filter = NULL;
@@ -874,7 +879,7 @@ int convert_to_git(const char *path, const char *src, size_t len,
 		src = dst->buf;
 		len = dst->len;
 	}
-	ret |= crlf_to_git(path, src, len, dst, ca.crlf_action, checksafe);
+	ret |= crlf_to_git(path, sha1, src, len, dst, ca.crlf_action, checksafe);
 	if (ret && dst) {
 		src = dst->buf;
 		len = dst->len;
@@ -882,7 +887,9 @@ int convert_to_git(const char *path, const char *src, size_t len,
 	return ret | ident_to_git(path, src, len, dst, ca.ident);
 }
 
-void convert_to_git_filter_fd(const char *path, int fd, struct strbuf *dst,
+void convert_to_git_filter_fd(const char *path,
+			      const unsigned char *sha1,
+			      int fd, struct strbuf *dst,
 			      enum safe_crlf checksafe)
 {
 	struct conv_attrs ca;
@@ -894,7 +901,7 @@ void convert_to_git_filter_fd(const char *path, int fd, struct strbuf *dst,
 	if (!apply_filter(path, NULL, 0, fd, dst, ca.drv->clean))
 		die("%s: clean filter '%s' failed", path, ca.drv->name);
 
-	crlf_to_git(path, dst->buf, dst->len, dst, ca.crlf_action, checksafe);
+	crlf_to_git(path, sha1, dst->buf, dst->len, dst, ca.crlf_action, checksafe);
 	ident_to_git(path, dst->buf, dst->len, dst, ca.ident);
 }
 
diff --git a/convert.h b/convert.h
index ccf436b..1634d37 100644
--- a/convert.h
+++ b/convert.h
@@ -37,8 +37,16 @@ extern const char *get_wt_convert_stats_ascii(const char *path);
 extern const char *get_convert_attr_ascii(const char *path);
 
 /* returns 1 if *dst was used */
-extern int convert_to_git(const char *path, const char *src, size_t len,
-			  struct strbuf *dst, enum safe_crlf checksafe);
+extern int convert_to_git_ce_sha1(const char *path, const unsigned char *sha1,
+				  const char *src, size_t len,
+				  struct strbuf *dst, enum safe_crlf checksafe);
+
+static inline int convert_to_git(const char *path, const char *src, size_t len,
+				 struct strbuf *dst, enum safe_crlf checksafe)
+{
+	return convert_to_git_ce_sha1(path, NULL, src, len, dst, checksafe);
+}
+
 extern int convert_to_working_tree(const char *path, const char *src,
 				   size_t len, struct strbuf *dst);
 extern int renormalize_buffer(const char *path, const char *src, size_t len,
@@ -47,9 +55,16 @@ static inline int would_convert_to_git(const char *path)
 {
 	return convert_to_git(path, NULL, 0, NULL, 0);
 }
+static inline int would_convert_to_git_ce_sha1(const char *path,
+					       const unsigned char *sha1)
+{
+	return convert_to_git_ce_sha1(path, sha1, NULL, 0, NULL, 0);
+}
+
 /* Precondition: would_convert_to_git_filter_fd(path) == true */
-extern void convert_to_git_filter_fd(const char *path, int fd,
-				     struct strbuf *dst,
+extern void convert_to_git_filter_fd(const char *path,
+				     const unsigned char *sha1,
+				     int fd, struct strbuf *dst,
 				     enum safe_crlf checksafe);
 extern int would_convert_to_git_filter_fd(const char *path);
 
diff --git a/read-cache.c b/read-cache.c
index d9fb78b..4b0e3c3 100644
--- a/read-cache.c
+++ b/read-cache.c
@@ -163,7 +163,9 @@ static int ce_compare_data(const struct cache_entry *ce, struct stat *st)
 
 	if (fd >= 0) {
 		unsigned char sha1[20];
-		if (!index_fd(sha1, fd, st, OBJ_BLOB, ce->name, 0))
+		unsigned flags = HASH_CE_HAS_SHA1;
+		memcpy(sha1, ce->sha1, sizeof(sha1));
+		if (!index_fd(sha1, fd, st, OBJ_BLOB, ce->name, flags))
 			match = hashcmp(sha1, ce->sha1);
 		/* index_fd() closed the file descriptor already */
 	}
diff --git a/sha1_file.c b/sha1_file.c
index d0f2aa0..dd013d5 100644
--- a/sha1_file.c
+++ b/sha1_file.c
@@ -3275,6 +3275,7 @@ static int index_mem(unsigned char *sha1, void *buf, size_t size,
 {
 	int ret, re_allocated = 0;
 	int write_object = flags & HASH_WRITE_OBJECT;
+	const int valid_sha1 = flags & HASH_CE_HAS_SHA1;
 
 	if (!type)
 		type = OBJ_BLOB;
@@ -3284,8 +3285,11 @@ static int index_mem(unsigned char *sha1, void *buf, size_t size,
 	 */
 	if ((type == OBJ_BLOB) && path) {
 		struct strbuf nbuf = STRBUF_INIT;
-		if (convert_to_git(path, buf, size, &nbuf,
-				   write_object ? safe_crlf : SAFE_CRLF_FALSE)) {
+		if (convert_to_git_ce_sha1(path,
+					   valid_sha1 ? sha1 : NULL,
+					   buf, size, &nbuf,
+					   write_object ? safe_crlf : SAFE_CRLF_FALSE)){
+
 			buf = strbuf_detach(&nbuf, &size);
 			re_allocated = 1;
 		}
@@ -3313,12 +3317,15 @@ static int index_stream_convert_blob(unsigned char *sha1, int fd,
 {
 	int ret;
 	const int write_object = flags & HASH_WRITE_OBJECT;
+	const int valid_sha1 = flags & HASH_CE_HAS_SHA1;
 	struct strbuf sbuf = STRBUF_INIT;
 
 	assert(path);
 	assert(would_convert_to_git_filter_fd(path));
 
-	convert_to_git_filter_fd(path, fd, &sbuf,
+	convert_to_git_filter_fd(path,
+				 valid_sha1 ? sha1 : NULL,
+				 fd, &sbuf,
 				 write_object ? safe_crlf : SAFE_CRLF_FALSE);
 
 	if (write_object)
@@ -3396,6 +3403,8 @@ int index_fd(unsigned char *sha1, int fd, struct stat *st,
 	     enum object_type type, const char *path, unsigned flags)
 {
 	int ret;
+	const unsigned char *sha1_ce;
+	sha1_ce = flags & HASH_CE_HAS_SHA1 ? sha1 : NULL;
 
 	/*
 	 * Call xsize_t() only when needed to avoid potentially unnecessary
@@ -3406,7 +3415,7 @@ int index_fd(unsigned char *sha1, int fd, struct stat *st,
 	else if (!S_ISREG(st->st_mode))
 		ret = index_pipe(sha1, fd, type, path, flags);
 	else if (st->st_size <= big_file_threshold || type != OBJ_BLOB ||
-		 (path && would_convert_to_git(path)))
+		 (path && would_convert_to_git_ce_sha1(path,sha1_ce)))
 		ret = index_core(sha1, fd, xsize_t(st->st_size), type, path,
 				 flags);
 	else
-- 
2.0.0.rc1.6318.g0c2c796

[PATCH v3 0/1] CRLF-Handling: bug fix around ce_compare_data()

From: <hidden>
Date: 2016-06-16 02:19:24

From: Torsten Bögershausen <redacted>

Changes since v2:

- Only 1 patch, the t6038 needs to go as a separate "series"

has_cr_in_index() uses either
	data = read_sha1_file(sha1, &type, &sz);
or
	data = read_blob_data_from_cache(path, &sz);

(I like v2 better, since there is a single code to get a sha object
into memory, which later will be replaced by a streaming object)
But in any case, here is v3

Torsten Bögershausen (1):
  convert: ce_compare_data() checks for a sha1 of a path

 cache.h                    |  1 +
 convert.c                  | 35 +++++++++++++++++++++--------------
 convert.h                  | 23 +++++++++++++++++++----
 read-cache.c               |  4 +++-
 sha1_file.c                | 17 +++++++++++++----

-- 
2.0.0.rc1.6318.g0c2c796

[PATCH v1 1/1] t6038: different eol for "Merge addition of text=auto"

From: <hidden>
Date: 2016-06-16 02:19:39

From: Torsten Bögershausen <redacted>

t6038 uses different code, depending if NATIVE_CRLF is set ot not.

Enhance the test coverage for cross-platform testing and run
"Merge addition of text=auto" with both lf and crlf as core.eol.

It is important to be run this test with crlf under Linux,
the day when the "unified auto handling will be introduced.

Signed-off-by: Torsten Bögershausen <redacted>
---
 t/t6038-merge-text-auto.sh | 28 +++++++++++++++++++++++-----
 1 file changed, 23 insertions(+), 5 deletions(-)
diff --git a/t/t6038-merge-text-auto.sh b/t/t6038-merge-text-auto.sh
index 33b77ee..8846f5d 100755
--- a/t/t6038-merge-text-auto.sh
+++ b/t/t6038-merge-text-auto.sh
@@ -91,16 +91,13 @@ test_expect_success 'Merge after setting text=auto' '
 	compare_files expected file
 '
 
-test_expect_success 'Merge addition of text=auto' '
+test_expect_success 'Merge addition of text=auto eol=LF' '
+	git config core.eol lf &&
 	cat <<-\EOF >expected &&
 	first line
 	same line
 	EOF
 
-	if test_have_prereq NATIVE_CRLF; then
-		append_cr <expected >expected.temp &&
-		mv expected.temp expected
-	fi &&
 	git config merge.renormalize true &&
 	git rm -fr . &&
 	rm -f .gitattributes &&
@@ -109,7 +106,28 @@ test_expect_success 'Merge addition of text=auto' '
 	compare_files  expected file
 '
 
+test_expect_success 'Merge addition of text=auto eol=CRLF' '
+	git config core.eol crlf &&
+	cat <<-\EOF >expected &&
+	first line
+	same line
+	EOF
+
+	append_cr <expected >expected.temp &&
+	mv expected.temp expected &&
+	git config merge.renormalize true &&
+	git rm -fr . &&
+	rm -f .gitattributes &&
+	git reset --hard b &&
+	echo >&2 "After git reset --hard b" &&
+	git ls-files -s --eol >&2 &&
+	git merge a &&
+	compare_files  expected file
+'
+
+
 test_expect_success 'Detect CRLF/LF conflict after setting text=auto' '
+	git config core.eol native &&
 	echo "<<<<<<<" >expected &&
 	if test_have_prereq NATIVE_CRLF; then
 		echo first line | append_cr >>expected &&
-- 
2.0.0.rc1.6318.g0c2c796

[PATCH v1 0/1] t6038-merge-text-auto.sh

From: <hidden>
Date: 2016-06-16 02:19:39

From: Torsten Bögershausen <redacted>

Split of the old 10/10 series.
This is the update of t6038, which is needed to
motivate the patch
   `convert: ce_compare_data() checks for a sha1 of a path`
on top of
   `convert: unify the "auto" handling of CRLF`

When files with different eols are merged with
merge.renormalize = true,
it is important to look at the right blob to determine if
the crlf came from the blob or are a result of a coversion.
This is a little bit of a hen-and-egg problem:
The problem comes up after the "unified auto handling".
In theory, it should have been since before:
get_sha1_from_index() says:

 * We might be in the middle of a merge, in which
 * case we would read stage #2 (ours).

This seams wrong, as in the merge we sometimes need to
look at "theirs".
(But I haven't managed to construct a TC)

t6038-merge-text-auto.sh

Torsten Bögershausen (1):
  t6038: different eol for "Merge addition of text=auto"

 t/t6038-merge-text-auto.sh | 28 +++++++++++++++++++++++-----
 1 file changed, 23 insertions(+), 5 deletions(-)

-- 
2.0.0.rc1.6318.g0c2c796

[PATCH v2 0/3] unified auto CRLF handling, V2

From: <hidden>
Date: 2016-06-16 02:19:47

From: Torsten Bögershausen <redacted>

unified auto CRLF handling, V2
  1/3 is 7/10 of the old 10/10 series
  2/3 and 
  3/3 is a replacement for tb/convert-peek-in-index:
      Better commit message, added test case

All in all we are getting closer.
Most of the patches had been send & reviewed earlier,
but a critical review won't hurt.

Torsten Bögershausen (3):
  convert: unify the "auto" handling of CRLF
  read-cache: factor out get_sha1_from_index() helper
  Correct ce_compare_data() in a middle of a merge

 Documentation/config.txt        | 12 +++----
 Documentation/gitattributes.txt | 15 +++++----
 builtin/apply.c                 |  3 +-
 builtin/blame.c                 |  2 +-
 cache.h                         |  4 +++
 combine-diff.c                  |  3 +-
 convert.c                       | 65 +++++++++++++++++++++++-------------
 convert.h                       | 18 +++++++---
 diff.c                          |  3 +-
 dir.c                           |  2 +-
 read-cache.c                    | 33 +++++++++++-------
 sha1_file.c                     | 12 +++++--
 t/t0025-crlf-auto.sh            |  4 +--
 t/t0027-auto-crlf.sh            | 32 +++++++++---------
 t/t6038-merge-text-auto.sh      | 74 ++++++++++++++++++++++++-----------------
 15 files changed, 172 insertions(+), 110 deletions(-)

-- 
2.0.0.rc1.6318.g0c2c796

[PATCH v2 2/3] read-cache: factor out get_sha1_from_index() helper

From: <hidden>
Date: 2016-06-16 02:19:47

From: Torsten Bögershausen <redacted>

Factor out the retrieval of the sha1 for a given path in
read_blob_data_from_index() into the function get_sha1_from_index().

This will be used in the next commit, when convert.c can do the
analyze for "text=auto" without slurping the whole blob into memory
at once.

Add a wrapper definition get_sha1_from_cache().

Signed-off-by: Torsten Bögershausen <redacted>
---
 cache.h      |  3 +++
 read-cache.c | 29 ++++++++++++++++++-----------
 2 files changed, 21 insertions(+), 11 deletions(-)
diff --git a/cache.h b/cache.h
index 6049f86..322ee40 100644
--- a/cache.h
+++ b/cache.h
@@ -379,6 +379,7 @@ extern void free_name_hash(struct index_state *istate);
 #define unmerge_cache_entry_at(at) unmerge_index_entry_at(&the_index, at)
 #define unmerge_cache(pathspec) unmerge_index(&the_index, pathspec)
 #define read_blob_data_from_cache(path, sz) read_blob_data_from_index(&the_index, (path), (sz))
+#define get_sha1_from_cache(path)  get_sha1_from_index (&the_index, (path))
 #endif
 
 enum object_type {
@@ -1050,6 +1051,8 @@ static inline void *read_sha1_file(const unsigned char *sha1, enum object_type *
 	return read_sha1_file_extended(sha1, type, size, LOOKUP_REPLACE_OBJECT);
 }
 
+const unsigned char *get_sha1_from_index(struct index_state *istate, const char *path);
+
 /*
  * This internal function is only declared here for the benefit of
  * lookup_replace_object().  Please do not call it directly.
diff --git a/read-cache.c b/read-cache.c
index d9fb78b..a3ef967 100644
--- a/read-cache.c
+++ b/read-cache.c
@@ -2263,13 +2263,27 @@ int index_name_is_other(const struct index_state *istate, const char *name,
 
 void *read_blob_data_from_index(struct index_state *istate, const char *path, unsigned long *size)
 {
-	int pos, len;
+	const unsigned char *sha1;
 	unsigned long sz;
 	enum object_type type;
 	void *data;
 
-	len = strlen(path);
-	pos = index_name_pos(istate, path, len);
+	sha1 = get_sha1_from_index(istate, path);
+	if (!sha1)
+		return NULL;
+	data = read_sha1_file(sha1, &type, &sz);
+	if (!data || type != OBJ_BLOB) {
+		free(data);
+		return NULL;
+	}
+	if (size)
+		*size = sz;
+	return data;
+}
+
+const unsigned char *get_sha1_from_index(struct index_state *istate, const char *path)
+{
+	int pos = index_name_pos(istate, path, strlen(path));
 	if (pos < 0) {
 		/*
 		 * We might be in the middle of a merge, in which
@@ -2285,14 +2299,7 @@ void *read_blob_data_from_index(struct index_state *istate, const char *path, un
 	}
 	if (pos < 0)
 		return NULL;
-	data = read_sha1_file(istate->cache[pos]->sha1, &type, &sz);
-	if (!data || type != OBJ_BLOB) {
-		free(data);
-		return NULL;
-	}
-	if (size)
-		*size = sz;
-	return data;
+	return (istate->cache[pos]->sha1);
 }
 
 void stat_validity_clear(struct stat_validity *sv)
-- 
2.0.0.rc1.6318.g0c2c796

[PATCH v2 3/3] Correct ce_compare_data() in a middle of a merge

From: <hidden>
Date: 2016-06-16 02:19:47

From: Torsten Bögershausen <redacted>

The following didn't work as expected:
- In a middle of a merge
- merge.renormalize is true,
- .gitattributes = "* text=auto"
- core.eol = crlf

Merge a blob with CRLF "first line\r\nsame line\r\n" and a blob
with LF "first line\nsame line\n".

The expected result of the merge is "first line\nsame line\n".

The content in the working tree is "first line\r\nsame line\r\n",
and ce_compare_data() should find that the content is clean and return 0.

Deep down crlf_to_git() is invoked, to check if CRLF are converted or not.

The "new safer autocrlf handling" calls blob_has_cr().

Instead of using the sha1 of the blob, (CRLF in this example),
the function get_sha1_from_index() is invoked.
get_sha1_from_index() decides to return "ours" when in the middle of
the merge, which is LF.

As a result, the CRLF in the worktree are converted into LF before
the comparison.
The contents of LF and CRLF don't match any more.

The problem is that ce_compare_data() has ce->sha1, but the sha1 is lost
on it's way into blob_has_cr().

Forwarding ce->sha1 from ce_compare_data() into crlf_to_git() makes sure
that blob_has_cr() looks at the appropriate blob.

Add a new parameter index_blob_sha1 to convert_to_git(), and forward the
sha1 from ce_compare_data() into convert_to_git(). Other callers use NULL
for index_blob_sha1, and the sha1 is determined from path
using get_sha1_from_cache(path). This is the same handling as before.

In the same spirit, forward the sha1 into would_convert_to_git().

While at it, rename has_cr_in_index() into blob_has_cr()
and replace 0 with SAFE_CRLF_FALSE.

Add a TC in t6038 to have a test coverage under Linux.

Signed-off-by: Torsten Bögershausen <redacted>
---
 builtin/apply.c            |  3 ++-
 builtin/blame.c            |  2 +-
 cache.h                    |  1 +
 combine-diff.c             |  3 ++-
 convert.c                  | 43 ++++++++++++++++++++++++++------------
 convert.h                  | 15 ++++++++++----
 diff.c                     |  3 ++-
 dir.c                      |  2 +-
 read-cache.c               |  4 +++-
 sha1_file.c                | 12 ++++++++---
 t/t6038-merge-text-auto.sh | 51 +++++++++++++++++++++++++---------------------
 11 files changed, 90 insertions(+), 49 deletions(-)
diff --git a/builtin/apply.c b/builtin/apply.c
index 8e4da2e..0cf9a0a 100644
--- a/builtin/apply.c
+++ b/builtin/apply.c
@@ -2140,7 +2140,8 @@ static int read_old_data(struct stat *st, const char *path, struct strbuf *buf)
 	case S_IFREG:
 		if (strbuf_read_file(buf, path, st->st_size) != st->st_size)
 			return error(_("unable to open or read %s"), path);
-		convert_to_git(path, buf->buf, buf->len, buf, 0);
+		convert_to_git(path, buf->buf, buf->len, buf,
+			       SAFE_CRLF_FALSE, NULL);
 		return 0;
 	default:
 		return -1;
diff --git a/builtin/blame.c b/builtin/blame.c
index 21f42b0..1c523b6 100644
--- a/builtin/blame.c
+++ b/builtin/blame.c
@@ -2377,7 +2377,7 @@ static struct commit *fake_working_tree_commit(struct diff_options *opt,
 		if (strbuf_read(&buf, 0, 0) < 0)
 			die_errno("failed to read from stdin");
 	}
-	convert_to_git(path, buf.buf, buf.len, &buf, 0);
+	convert_to_git(path, buf.buf, buf.len, &buf, SAFE_CRLF_FALSE, NULL);
 	origin->file.ptr = buf.buf;
 	origin->file.size = buf.len;
 	pretend_sha1_file(buf.buf, buf.len, OBJ_BLOB, origin->blob_sha1);
diff --git a/cache.h b/cache.h
index 322ee40..86f48f9 100644
--- a/cache.h
+++ b/cache.h
@@ -605,6 +605,7 @@ extern int ie_modified(const struct index_state *, const struct cache_entry *, s
 
 #define HASH_WRITE_OBJECT 1
 #define HASH_FORMAT_CHECK 2
+#define HASH_USE_SHA_NOT_PATH 4
 extern int index_fd(unsigned char *sha1, int fd, struct stat *st, enum object_type type, const char *path, unsigned flags);
 extern int index_path(unsigned char *sha1, const char *path, struct stat *st, unsigned flags);
 
diff --git a/combine-diff.c b/combine-diff.c
index 8f2313d..8b535a7 100644
--- a/combine-diff.c
+++ b/combine-diff.c
@@ -1052,7 +1052,8 @@ static void show_patch_diff(struct combine_diff_path *elem, int num_parent,
 			if (is_file) {
 				struct strbuf buf = STRBUF_INIT;
 
-				if (convert_to_git(elem->path, result, len, &buf, safe_crlf)) {
+				if (convert_to_git(elem->path, result, len,
+						   &buf, safe_crlf, NULL)) {
 					free(result);
 					result = strbuf_detach(&buf, &len);
 					result_size = len;
diff --git a/convert.c b/convert.c
index 67d69b5..802ee7c 100644
--- a/convert.c
+++ b/convert.c
@@ -219,23 +219,28 @@ static void check_safe_crlf(const char *path, enum crlf_action crlf_action,
 	}
 }
 
-static int has_cr_in_index(const char *path)
+static int blob_has_cr(const unsigned char *index_blob_sha1)
 {
 	unsigned long sz;
 	void *data;
-	int has_cr;
-
-	data = read_blob_data_from_cache(path, &sz);
+	int has_cr = 0;
+	enum object_type type;
+	if (!index_blob_sha1)
+		return 0;
+	data = read_sha1_file(index_blob_sha1, &type, &sz);
 	if (!data)
 		return 0;
-	has_cr = memchr(data, '\r', sz) != NULL;
+	if (type == OBJ_BLOB)
+		has_cr = memchr(data, '\r', sz) != NULL;
+
 	free(data);
 	return has_cr;
 }
 
 static int crlf_to_git(const char *path, const char *src, size_t len,
 		       struct strbuf *buf,
-		       enum crlf_action crlf_action, enum safe_crlf checksafe)
+		       enum crlf_action crlf_action, enum safe_crlf checksafe,
+		       const unsigned char *index_blob_sha1)
 {
 	struct text_stat stats;
 	char *dst;
@@ -256,14 +261,23 @@ static int crlf_to_git(const char *path, const char *src, size_t len,
 	if (crlf_action == CRLF_AUTO || crlf_action == CRLF_AUTO_INPUT || crlf_action == CRLF_AUTO_CRLF) {
 		if (convert_is_binary(len, &stats))
 			return 0;
+
 		/*
 		 * If the file in the index has any CR in it, do not convert.
 		 * This is the new safer autocrlf handling.
 		 */
 		if (checksafe == SAFE_CRLF_RENORMALIZE)
 			checksafe = SAFE_CRLF_FALSE;
-		else if (has_cr_in_index(path))
-			return 0;
+		else {
+			/*
+			 * If the file in the index has any CR in it, do not convert.
+			 * This is the new safer autocrlf handling.
+			 */
+			if (!index_blob_sha1)
+				index_blob_sha1 = get_sha1_from_cache(path);
+			if (blob_has_cr(index_blob_sha1))
+				return 0;
+		}
 	}
 	check_safe_crlf(path, crlf_action, &stats, checksafe);
 
@@ -855,7 +869,8 @@ const char *get_convert_attr_ascii(const char *path)
 }
 
 int convert_to_git(const char *path, const char *src, size_t len,
-                   struct strbuf *dst, enum safe_crlf checksafe)
+		   struct strbuf *dst, enum safe_crlf checksafe,
+		   const unsigned char *index_blob_sha1)
 {
 	int ret = 0;
 	const char *filter = NULL;
@@ -876,7 +891,7 @@ int convert_to_git(const char *path, const char *src, size_t len,
 		src = dst->buf;
 		len = dst->len;
 	}
-	ret |= crlf_to_git(path, src, len, dst, ca.crlf_action, checksafe);
+	ret |= crlf_to_git(path, src, len, dst, ca.crlf_action, checksafe, index_blob_sha1);
 	if (ret && dst) {
 		src = dst->buf;
 		len = dst->len;
@@ -885,7 +900,8 @@ int convert_to_git(const char *path, const char *src, size_t len,
 }
 
 void convert_to_git_filter_fd(const char *path, int fd, struct strbuf *dst,
-			      enum safe_crlf checksafe)
+			      enum safe_crlf checksafe,
+			      const unsigned char *index_blob_sha1)
 {
 	struct conv_attrs ca;
 	convert_attrs(&ca, path);
@@ -896,7 +912,8 @@ void convert_to_git_filter_fd(const char *path, int fd, struct strbuf *dst,
 	if (!apply_filter(path, NULL, 0, fd, dst, ca.drv->clean))
 		die("%s: clean filter '%s' failed", path, ca.drv->name);
 
-	crlf_to_git(path, dst->buf, dst->len, dst, ca.crlf_action, checksafe);
+	crlf_to_git(path, dst->buf, dst->len, dst, ca.crlf_action,
+		    checksafe, index_blob_sha1);
 	ident_to_git(path, dst->buf, dst->len, dst, ca.ident);
 }
 
@@ -951,7 +968,7 @@ int renormalize_buffer(const char *path, const char *src, size_t len, struct str
 		src = dst->buf;
 		len = dst->len;
 	}
-	return ret | convert_to_git(path, src, len, dst, SAFE_CRLF_RENORMALIZE);
+	return ret | convert_to_git(path, src, len, dst, SAFE_CRLF_RENORMALIZE, NULL);
 }
 
 /*****************************************************************
diff --git a/convert.h b/convert.h
index 81b6cdf..8c34dd5 100644
--- a/convert.h
+++ b/convert.h
@@ -39,19 +39,26 @@ extern const char *get_convert_attr_ascii(const char *path);
 
 /* returns 1 if *dst was used */
 extern int convert_to_git(const char *path, const char *src, size_t len,
-			  struct strbuf *dst, enum safe_crlf checksafe);
+			  struct strbuf *dst, enum safe_crlf checksafe,
+			  const unsigned char *index_blob_sha1);
+
 extern int convert_to_working_tree(const char *path, const char *src,
 				   size_t len, struct strbuf *dst);
 extern int renormalize_buffer(const char *path, const char *src, size_t len,
 			      struct strbuf *dst);
-static inline int would_convert_to_git(const char *path)
+static inline int would_convert_to_git(const char *path,
+				       const unsigned char *index_blob_sha1)
 {
-	return convert_to_git(path, NULL, 0, NULL, 0);
+	return convert_to_git(path, NULL, 0, NULL, SAFE_CRLF_FALSE,
+			      index_blob_sha1);
 }
+
 /* Precondition: would_convert_to_git_filter_fd(path) == true */
 extern void convert_to_git_filter_fd(const char *path, int fd,
 				     struct strbuf *dst,
-				     enum safe_crlf checksafe);
+				     enum safe_crlf checksafe,
+				     const unsigned char *index_blob_sha1);
+
 extern int would_convert_to_git_filter_fd(const char *path);
 
 /*****************************************************************
diff --git a/diff.c b/diff.c
index d3734d3..a8308e0 100644
--- a/diff.c
+++ b/diff.c
@@ -2810,7 +2810,8 @@ int diff_populate_filespec(struct diff_filespec *s, unsigned int flags)
 		/*
 		 * Convert from working tree format to canonical git format
 		 */
-		if (convert_to_git(s->path, s->data, s->size, &buf, crlf_warn)) {
+		if (convert_to_git(s->path, s->data, s->size, &buf,
+				   crlf_warn, NULL)) {
 			size_t size = 0;
 			munmap(s->data, s->size);
 			s->should_munmap = 0;
diff --git a/dir.c b/dir.c
index 6172b34..66f1ffb 100644
--- a/dir.c
+++ b/dir.c
@@ -712,7 +712,7 @@ static int add_excludes(const char *fname, const char *base, int baselen,
 				 (pos = cache_name_pos(fname, strlen(fname))) >= 0 &&
 				 !ce_stage(active_cache[pos]) &&
 				 ce_uptodate(active_cache[pos]) &&
-				 !would_convert_to_git(fname))
+				 !would_convert_to_git(fname, NULL))
 				hashcpy(sha1_stat->sha1, active_cache[pos]->sha1);
 			else
 				hash_sha1_file(buf, size, "blob", sha1_stat->sha1);
diff --git a/read-cache.c b/read-cache.c
index a3ef967..c109b6d 100644
--- a/read-cache.c
+++ b/read-cache.c
@@ -163,7 +163,9 @@ static int ce_compare_data(const struct cache_entry *ce, struct stat *st)
 
 	if (fd >= 0) {
 		unsigned char sha1[20];
-		if (!index_fd(sha1, fd, st, OBJ_BLOB, ce->name, 0))
+		unsigned flags = HASH_USE_SHA_NOT_PATH;
+		memcpy(sha1, ce->sha1, sizeof(sha1));
+		if (!index_fd(sha1, fd, st, OBJ_BLOB, ce->name, flags))
 			match = hashcmp(sha1, ce->sha1);
 		/* index_fd() closed the file descriptor already */
 	}
diff --git a/sha1_file.c b/sha1_file.c
index d5e1121..10630f0 100644
--- a/sha1_file.c
+++ b/sha1_file.c
@@ -3273,6 +3273,7 @@ static int index_mem(unsigned char *sha1, void *buf, size_t size,
 {
 	int ret, re_allocated = 0;
 	int write_object = flags & HASH_WRITE_OBJECT;
+	const int valid_sha1 = flags & HASH_USE_SHA_NOT_PATH;
 
 	if (!type)
 		type = OBJ_BLOB;
@@ -3283,7 +3284,8 @@ static int index_mem(unsigned char *sha1, void *buf, size_t size,
 	if ((type == OBJ_BLOB) && path) {
 		struct strbuf nbuf = STRBUF_INIT;
 		if (convert_to_git(path, buf, size, &nbuf,
-				   write_object ? safe_crlf : SAFE_CRLF_FALSE)) {
+				   write_object ? safe_crlf : SAFE_CRLF_FALSE,
+				   valid_sha1 ? sha1 : NULL)) {
 			buf = strbuf_detach(&nbuf, &size);
 			re_allocated = 1;
 		}
@@ -3311,13 +3313,15 @@ static int index_stream_convert_blob(unsigned char *sha1, int fd,
 {
 	int ret;
 	const int write_object = flags & HASH_WRITE_OBJECT;
+	const int valid_sha1 = flags & HASH_USE_SHA_NOT_PATH;
 	struct strbuf sbuf = STRBUF_INIT;
 
 	assert(path);
 	assert(would_convert_to_git_filter_fd(path));
 
 	convert_to_git_filter_fd(path, fd, &sbuf,
-				 write_object ? safe_crlf : SAFE_CRLF_FALSE);
+				 write_object ? safe_crlf : SAFE_CRLF_FALSE,
+				 valid_sha1 ? sha1 : NULL);
 
 	if (write_object)
 		ret = write_sha1_file(sbuf.buf, sbuf.len, typename(OBJ_BLOB),
@@ -3394,6 +3398,8 @@ int index_fd(unsigned char *sha1, int fd, struct stat *st,
 	     enum object_type type, const char *path, unsigned flags)
 {
 	int ret;
+	const unsigned char *sha1_ce;
+	sha1_ce = flags & HASH_USE_SHA_NOT_PATH ? sha1 : NULL;
 
 	/*
 	 * Call xsize_t() only when needed to avoid potentially unnecessary
@@ -3404,7 +3410,7 @@ int index_fd(unsigned char *sha1, int fd, struct stat *st,
 	else if (!S_ISREG(st->st_mode))
 		ret = index_pipe(sha1, fd, type, path, flags);
 	else if (st->st_size <= big_file_threshold || type != OBJ_BLOB ||
-		 (path && would_convert_to_git(path)))
+		 (path && would_convert_to_git(path,sha1_ce)))
 		ret = index_core(sha1, fd, xsize_t(st->st_size), type, path,
 				 flags);
 	else
diff --git a/t/t6038-merge-text-auto.sh b/t/t6038-merge-text-auto.sh
index 33b77ee..5e8d5fa 100755
--- a/t/t6038-merge-text-auto.sh
+++ b/t/t6038-merge-text-auto.sh
@@ -91,16 +91,13 @@ test_expect_success 'Merge after setting text=auto' '
 	compare_files expected file
 '
 
-test_expect_success 'Merge addition of text=auto' '
+test_expect_success 'Merge addition of text=auto eol=LF' '
+	git config core.eol lf &&
 	cat <<-\EOF >expected &&
 	first line
 	same line
 	EOF
 
-	if test_have_prereq NATIVE_CRLF; then
-		append_cr <expected >expected.temp &&
-		mv expected.temp expected
-	fi &&
 	git config merge.renormalize true &&
 	git rm -fr . &&
 	rm -f .gitattributes &&
@@ -109,17 +106,31 @@ test_expect_success 'Merge addition of text=auto' '
 	compare_files  expected file
 '
 
+test_expect_success 'Merge addition of text=auto eol=CRLF' '
+	git config core.eol crlf &&
+	cat <<-\EOF >expected &&
+	first line
+	same line
+	EOF
+
+	append_cr <expected >expected.temp &&
+	mv expected.temp expected &&
+	git config merge.renormalize true &&
+	git rm -fr . &&
+	rm -f .gitattributes &&
+	git reset --hard b &&
+	echo >&2 "After git reset --hard b" &&
+	git ls-files -s --eol >&2 &&
+	git merge a &&
+	compare_files  expected file
+'
+
 test_expect_success 'Detect CRLF/LF conflict after setting text=auto' '
+	git config core.eol native &&
 	echo "<<<<<<<" >expected &&
-	if test_have_prereq NATIVE_CRLF; then
-		echo first line | append_cr >>expected &&
-		echo same line | append_cr >>expected &&
-		echo ======= | append_cr >>expected
-	else
-		echo first line >>expected &&
-		echo same line >>expected &&
-		echo ======= >>expected
-	fi &&
+	echo first line >>expected &&
+	echo same line >>expected &&
+	echo ======= >>expected &&
 	echo first line | append_cr >>expected &&
 	echo same line | append_cr >>expected &&
 	echo ">>>>>>>" >>expected &&
@@ -135,15 +146,9 @@ test_expect_success 'Detect LF/CRLF conflict from addition of text=auto' '
 	echo "<<<<<<<" >expected &&
 	echo first line | append_cr >>expected &&
 	echo same line | append_cr >>expected &&
-	if test_have_prereq NATIVE_CRLF; then
-		echo ======= | append_cr >>expected &&
-		echo first line | append_cr >>expected &&
-		echo same line | append_cr >>expected
-	else
-		echo ======= >>expected &&
-		echo first line >>expected &&
-		echo same line >>expected
-	fi &&
+	echo ======= >>expected &&
+	echo first line >>expected &&
+	echo same line >>expected &&
 	echo ">>>>>>>" >>expected &&
 	git config merge.renormalize false &&
 	rm -f .gitattributes &&
-- 
2.0.0.rc1.6318.g0c2c796

[PATCH v2 1/3] convert: unify the "auto" handling of CRLF

From: <hidden>
Date: 2016-06-16 02:19:47

From: Torsten Bögershausen <redacted>

Before this change,
$ echo "* text=auto" >.gitattributes
$ echo "* eol=crlf" >>.gitattributes

would have the same effect as
$ echo "* text" >.gitattributes
$ git config core.eol crlf

Since the 'eol' attribute had higher priority than 'text=auto', this may
corrupt binary files and is not what most users expect to happen.

Make the 'eol' attribute to obey 'text=auto', and now
$ echo "* text=auto" >.gitattributes
$ echo "* eol=crlf" >>.gitattributes
behaves the same as
$ echo "* text=auto" >.gitattributes
$ git config core.eol crlf

In other words,
$ echo "* text=auto eol=crlf" >.gitattributes
has the same effect as
$ git config core.autocrlf true

and
$ echo "* text=auto eol=lf" >.gitattributes
has the same effect as
$ git config core.autocrlf input

Signed-off-by: Torsten Bögershausen <redacted>
---
 Documentation/config.txt        | 12 +++++-------
 Documentation/gitattributes.txt | 15 +++++++++------
 convert.c                       | 42 +++++++++++++++++++++--------------------
 convert.h                       |  3 ++-
 t/t0025-crlf-auto.sh            |  4 ++--
 t/t0027-auto-crlf.sh            | 32 +++++++++++++++----------------
 t/t6038-merge-text-auto.sh      | 23 ++++++++++++++--------
 7 files changed, 71 insertions(+), 60 deletions(-)
diff --git a/Documentation/config.txt b/Documentation/config.txt
index 53f00db..208b5de 100644
--- a/Documentation/config.txt
+++ b/Documentation/config.txt
@@ -405,13 +405,11 @@ file with mixed line endings would be reported by the `core.safecrlf`
 mechanism.
 
 core.autocrlf::
-	Setting this variable to "true" is almost the same as setting
-	the `text` attribute to "auto" on all files except that text
-	files are not guaranteed to be normalized: files that contain
-	`CRLF` in the repository will not be touched.  Use this
-	setting if you want to have `CRLF` line endings in your
-	working directory even though the repository does not have
-	normalized line endings.  This variable can be set to 'input',
+	Setting this variable to "true" is the same as setting
+	the `text` attribute to "auto" on all files and core.eol to "crlf".
+	Set to true if you want to have `CRLF` line endings in your
+	working directory and the repository has LF line endings.
+	This variable can be set to 'input',
 	in which case no output conversion is performed.
 
 core.symlinks::
diff --git a/Documentation/gitattributes.txt b/Documentation/gitattributes.txt
index e3b1de8..d7a124b 100644
--- a/Documentation/gitattributes.txt
+++ b/Documentation/gitattributes.txt
@@ -115,6 +115,7 @@ text file is normalized, its line endings are converted to LF in the
 repository.  To control what line ending style is used in the working
 directory, use the `eol` attribute for a single file and the
 `core.eol` configuration variable for all text files.
+Note that `core.autocrlf` overrides `core.eol`
 
 Set::
 
@@ -130,8 +131,9 @@ Unset::
 Set to string value "auto"::
 
 	When `text` is set to "auto", the path is marked for automatic
-	end-of-line normalization.  If Git decides that the content is
-	text, its line endings are normalized to LF on checkin.
+	end-of-line conversion.  If Git decides that the content is
+	text, its line endings are converted to LF on checkin.
+	When the file has been commited with CRLF, no conversion is done.
 
 Unspecified::
 
@@ -146,7 +148,7 @@ unspecified.
 ^^^^^
 
 This attribute sets a specific line-ending style to be used in the
-working directory.  It enables end-of-line normalization without any
+working directory.  It enables end-of-line conversion without any
 content checks, effectively setting the `text` attribute.
 
 Set to string value "crlf"::
@@ -186,9 +188,10 @@ the working directory, and prevent .jpg files from being normalized
 regardless of their content.
 
 ------------------------
+*               text=auto
 *.txt		text
-*.vcproj	eol=crlf
-*.sh		eol=lf
+*.vcproj	text eol=crlf
+*.sh		text eol=lf
 *.jpg		-text
 ------------------------
 
@@ -198,7 +201,7 @@ normalization in Git.
 
 If you simply want to have CRLF line endings in your working directory
 regardless of the repository you are working with, you can set the
-config variable "core.autocrlf" without changing any attributes.
+config variable "core.autocrlf" without using any attributes.
 
 ------------------------
 [core]
diff --git a/convert.c b/convert.c
index b1614bf..67d69b5 100644
--- a/convert.c
+++ b/convert.c
@@ -176,7 +176,9 @@ static enum eol output_eol(enum crlf_action crlf_action)
 		return EOL_LF;
 	case CRLF_UNDEFINED:
 	case CRLF_AUTO_CRLF:
+		return EOL_CRLF;
 	case CRLF_AUTO_INPUT:
+		return EOL_LF;
 	case CRLF_TEXT:
 	case CRLF_AUTO:
 		/* fall through */
@@ -254,17 +256,15 @@ static int crlf_to_git(const char *path, const char *src, size_t len,
 	if (crlf_action == CRLF_AUTO || crlf_action == CRLF_AUTO_INPUT || crlf_action == CRLF_AUTO_CRLF) {
 		if (convert_is_binary(len, &stats))
 			return 0;
-
-		if (crlf_action == CRLF_AUTO_INPUT || crlf_action == CRLF_AUTO_CRLF) {
-			/*
-			 * If the file in the index has any CR in it, do not convert.
-			 * This is the new safer autocrlf handling.
-			 */
-			if (has_cr_in_index(path))
-				return 0;
-		}
+		/*
+		 * If the file in the index has any CR in it, do not convert.
+		 * This is the new safer autocrlf handling.
+		 */
+		if (checksafe == SAFE_CRLF_RENORMALIZE)
+			checksafe = SAFE_CRLF_FALSE;
+		else if (has_cr_in_index(path))
+			return 0;
 	}
-
 	check_safe_crlf(path, crlf_action, &stats, checksafe);
 
 	/* Optimization: No CRLF? Nothing to convert, regardless. */
@@ -320,12 +320,10 @@ static int crlf_to_worktree(const char *path, const char *src, size_t len,
 		return 0;
 
 	if (crlf_action == CRLF_AUTO || crlf_action == CRLF_AUTO_INPUT || crlf_action == CRLF_AUTO_CRLF) {
-		if (crlf_action == CRLF_AUTO_INPUT || crlf_action == CRLF_AUTO_CRLF) {
-			/* If we have any CR or CRLF line endings, we do not touch it */
-			/* This is the new safer autocrlf-handling */
-			if (stats.lonecr || stats.crlf )
-				return 0;
-		}
+		/* If we have any CR or CRLF line endings, we do not touch it */
+		/* This is the new safer autocrlf-handling */
+		if (stats.lonecr || stats.crlf )
+			return 0;
 
 		if (convert_is_binary(len, &stats))
 			return 0;
@@ -786,7 +784,11 @@ static void convert_attrs(struct conv_attrs *ca, const char *path)
 		ca->drv = git_path_check_convert(ccheck + 2);
 		if (ca->crlf_action != CRLF_BINARY) {
 			enum eol eol_attr = git_path_check_eol(ccheck + 3);
-			if (eol_attr == EOL_LF)
+			if (ca->crlf_action == CRLF_AUTO && eol_attr == EOL_LF)
+				ca->crlf_action = CRLF_AUTO_INPUT;
+			else if (ca->crlf_action == CRLF_AUTO && eol_attr == EOL_CRLF)
+				ca->crlf_action = CRLF_AUTO_CRLF;
+			else if (eol_attr == EOL_LF)
 				ca->crlf_action = CRLF_TEXT_INPUT;
 			else if (eol_attr == EOL_CRLF)
 				ca->crlf_action = CRLF_TEXT_CRLF;
@@ -845,9 +847,9 @@ const char *get_convert_attr_ascii(const char *path)
 	case CRLF_AUTO:
 		return "text=auto";
 	case CRLF_AUTO_CRLF:
-		return "text=auto eol=crlf"; /* This is not supported yet */
+		return "text=auto eol=crlf";
 	case CRLF_AUTO_INPUT:
-		return "text=auto eol=lf"; /* This is not supported yet */
+		return "text=auto eol=lf";
 	}
 	return "";
 }
@@ -949,7 +951,7 @@ int renormalize_buffer(const char *path, const char *src, size_t len, struct str
 		src = dst->buf;
 		len = dst->len;
 	}
-	return ret | convert_to_git(path, src, len, dst, SAFE_CRLF_FALSE);
+	return ret | convert_to_git(path, src, len, dst, SAFE_CRLF_RENORMALIZE);
 }
 
 /*****************************************************************
diff --git a/convert.h b/convert.h
index ccf436b..81b6cdf 100644
--- a/convert.h
+++ b/convert.h
@@ -7,7 +7,8 @@
 enum safe_crlf {
 	SAFE_CRLF_FALSE = 0,
 	SAFE_CRLF_FAIL = 1,
-	SAFE_CRLF_WARN = 2
+	SAFE_CRLF_WARN = 2,
+	SAFE_CRLF_RENORMALIZE = 4
 };
 
 extern enum safe_crlf safe_crlf;
diff --git a/t/t0025-crlf-auto.sh b/t/t0025-crlf-auto.sh
index c164b46..d0bee08 100755
--- a/t/t0025-crlf-auto.sh
+++ b/t/t0025-crlf-auto.sh
@@ -114,7 +114,7 @@ test_expect_success 'autocrlf=true does not normalize CRLF files' '
 	test -z "$LFonlydiff" -a -z "$CRLFonlydiff" -a -z "$LFwithNULdiff"
 '
 
-test_expect_success 'text=auto, autocrlf=true _does_ normalize CRLF files' '
+test_expect_success 'text=auto, autocrlf=true does not normalize CRLF files' '
 
 	rm -f .gitattributes tmp LFonly CRLFonly LFwithNUL &&
 	git config core.autocrlf true &&
@@ -126,7 +126,7 @@ test_expect_success 'text=auto, autocrlf=true _does_ normalize CRLF files' '
 	LFonlydiff=$(git diff LFonly) &&
 	CRLFonlydiff=$(git diff CRLFonly) &&
 	LFwithNULdiff=$(git diff LFwithNUL) &&
-	test -z "$LFonlydiff" -a -n "$CRLFonlydiff" -a -z "$LFwithNULdiff"
+	test -z "$LFonlydiff" -a -z "$CRLFonlydiff" -a -z "$LFwithNULdiff"
 '
 
 test_expect_success 'text=auto, autocrlf=true does not normalize binary files' '
diff --git a/t/t0027-auto-crlf.sh b/t/t0027-auto-crlf.sh
index 9372589..8367d0b 100755
--- a/t/t0027-auto-crlf.sh
+++ b/t/t0027-auto-crlf.sh
@@ -175,8 +175,8 @@ attr_ascii () {
 	text,lf)   echo "text eol=lf" ;;
 	text,crlf) echo "text eol=crlf" ;;
 	auto,)     echo "text=auto" ;;
-	auto,lf)   echo "text eol=lf" ;;
-	auto,crlf) echo "text eol=crlf" ;;
+	auto,lf)   echo "text=auto eol=lf" ;;
+	auto,crlf) echo "text=auto eol=crlf" ;;
 	lf,)       echo "text eol=lf" ;;
 	crlf,)     echo "text eol=crlf" ;;
 	,) echo "" ;;
@@ -397,10 +397,9 @@ commit_chk_wrnNNO ""      ""      false   ""        ""        ""          ""
 commit_chk_wrnNNO ""      ""      true    LF_CRLF   ""        ""          ""          ""
 commit_chk_wrnNNO ""      ""      input   ""        ""        ""          ""          ""
 
-commit_chk_wrnNNO "auto"  ""      false   "$WILC"   "$WICL"   "$WAMIX"    ""          ""
-commit_chk_wrnNNO "auto"  ""      true    LF_CRLF   ""        LF_CRLF     ""          ""
-commit_chk_wrnNNO "auto"  ""      input   ""        CRLF_LF   CRLF_LF     ""          ""
-
+commit_chk_wrnNNO "auto"  ""      false   "$WILC"   ""        ""          ""          ""
+commit_chk_wrnNNO "auto"  ""      true    LF_CRLF   ""        ""          ""          ""
+commit_chk_wrnNNO "auto"  ""      input   ""        ""        ""          ""          ""
 for crlf in true false input
 do
 	commit_chk_wrnNNO -text ""      $crlf   ""        ""        ""          ""          ""
@@ -408,8 +407,8 @@ do
 	commit_chk_wrnNNO -text crlf    $crlf   ""        ""        ""          ""          ""
 	commit_chk_wrnNNO ""    lf      $crlf   ""       CRLF_LF    CRLF_LF      ""         CRLF_LF
 	commit_chk_wrnNNO ""    crlf    $crlf   LF_CRLF   ""        LF_CRLF     LF_CRLF     ""
-	commit_chk_wrnNNO auto  lf    	$crlf   ""       CRLF_LF    CRLF_LF     ""          CRLF_LF
-	commit_chk_wrnNNO auto  crlf  	$crlf   LF_CRLF   ""        LF_CRLF     LF_CRLF     ""
+	commit_chk_wrnNNO auto  lf    	$crlf   ""        ""        ""          ""          ""
+	commit_chk_wrnNNO auto  crlf  	$crlf   LF_CRLF   ""        ""          ""          ""
 	commit_chk_wrnNNO text  lf    	$crlf   ""       CRLF_LF    CRLF_LF     ""          CRLF_LF
 	commit_chk_wrnNNO text  crlf  	$crlf   LF_CRLF   ""        LF_CRLF     LF_CRLF     ""
 done
@@ -454,9 +453,9 @@ do
 	check_in_repo_NNO -text ""     $crlf   LF  CRLF  CRLF_mix_LF  LF_mix_CR  CRLF_nul
 	check_in_repo_NNO -text lf     $crlf   LF  CRLF  CRLF_mix_LF  LF_mix_CR  CRLF_nul
 	check_in_repo_NNO -text crlf   $crlf   LF  CRLF  CRLF_mix_LF  LF_mix_CR  CRLF_nul
-	check_in_repo_NNO auto  ""     $crlf   LF  LF    LF           LF_mix_CR  CRLF_nul
-	check_in_repo_NNO auto  lf     $crlf   LF  LF    LF           LF_mix_CR  LF_nul
-	check_in_repo_NNO auto  crlf   $crlf   LF  LF    LF           LF_mix_CR  LF_nul
+	check_in_repo_NNO auto  ""     $crlf   LF  CRLF  CRLF_mix_LF  LF_mix_CR  CRLF_nul
+	check_in_repo_NNO auto  lf     $crlf   LF  CRLF  CRLF_mix_LF  LF_mix_CR  CRLF_nul
+	check_in_repo_NNO auto  crlf   $crlf   LF  CRLF  CRLF_mix_LF  LF_mix_CR  CRLF_nul
 	check_in_repo_NNO text  ""     $crlf   LF  LF    LF           LF_mix_CR  LF_nul
 	check_in_repo_NNO text  lf     $crlf   LF  LF    LF           LF_mix_CR  LF_nul
 	check_in_repo_NNO text  crlf   $crlf   LF  LF    LF           LF_mix_CR  LF_nul
@@ -493,7 +492,8 @@ fi
 export CRLF_MIX_LF_CR MIX NL
 
 # Same handling with and without ident
-for id in "" ident
+#for id in "" ident
+for id in ""
 do
 	for ceol in lf crlf native
 	do
@@ -509,7 +509,7 @@ do
 			checkout_files text  "$id" "crlf" "$crlf" "$ceol"  CRLF  CRLF  CRLF         CRLF_mix_CR  CRLF_nul
 			# currently the same as text, eol=XXX
 			checkout_files auto  "$id" "lf"   "$crlf" "$ceol"  LF    CRLF  CRLF_mix_LF  LF_mix_CR    LF_nul
-			checkout_files auto  "$id" "crlf" "$crlf" "$ceol"  CRLF  CRLF  CRLF         CRLF_mix_CR  CRLF_nul
+			checkout_files auto  "$id" "crlf" "$crlf" "$ceol"  CRLF  CRLF  CRLF_mix_LF  LF_mix_CR    LF_nul
 		done
 
 		# core.autocrlf false, different core.eol
@@ -517,7 +517,7 @@ do
 		# core.autocrlf true
 		checkout_files   ""    "$id" ""     true    "$ceol"  CRLF  CRLF  CRLF_mix_LF  LF_mix_CR    LF_nul
 		# text: core.autocrlf = true overrides core.eol
-		checkout_files   auto  "$id" ""     true    "$ceol"  CRLF  CRLF  CRLF         LF_mix_CR    LF_nul
+		checkout_files   auto  "$id" ""     true    "$ceol"  CRLF  CRLF  CRLF_mix_LF  LF_mix_CR    LF_nul
 		checkout_files   text  "$id" ""     true    "$ceol"  CRLF  CRLF  CRLF         CRLF_mix_CR  CRLF_nul
 		# text: core.autocrlf = input overrides core.eol
 		checkout_files   text  "$id" ""     input   "$ceol"  LF    CRLF  CRLF_mix_LF  LF_mix_CR    LF_nul
@@ -531,8 +531,8 @@ do
 	checkout_files     text  "$id" ""     false   ""       $NL   CRLF  $MIX_CRLF_LF $MIX_LF_CR   $LFNUL
 	checkout_files     text  "$id" ""     false   native   $NL   CRLF  $MIX_CRLF_LF $MIX_LF_CR   $LFNUL
 	# auto: core.autocrlf=false and core.eol unset(or native) uses native eol
-	checkout_files     auto  "$id" ""     false   ""       $NL   CRLF  $MIX_CRLF_LF LF_mix_CR    LF_nul
-	checkout_files     auto  "$id" ""     false   native   $NL   CRLF  $MIX_CRLF_LF LF_mix_CR    LF_nul
+	checkout_files     auto  "$id" ""     false   ""       $NL   CRLF  CRLF_mix_LF  LF_mix_CR    LF_nul
+	checkout_files     auto  "$id" ""     false   native   $NL   CRLF  CRLF_mix_LF  LF_mix_CR    LF_nul
 done
 
 # Should be the last test case: remove some files from the worktree
diff --git a/t/t6038-merge-text-auto.sh b/t/t6038-merge-text-auto.sh
index 85c10b0..33b77ee 100755
--- a/t/t6038-merge-text-auto.sh
+++ b/t/t6038-merge-text-auto.sh
@@ -16,6 +16,13 @@ test_description='CRLF merge conflict across text=auto change
 
 test_have_prereq SED_STRIPS_CR && SED_OPTIONS=-b
 
+compare_files () {
+	tr '\015\000' QN <"$1" >"$1".expect &&
+	tr '\015\000' QN <"$2" >"$2".actual &&
+	test_cmp "$1".expect "$2".actual &&
+	rm "$1".expect "$2".actual
+}
+
 test_expect_success setup '
 	git config core.autocrlf false &&
 
@@ -30,7 +37,7 @@ test_expect_success setup '
 	git branch side &&
 
 	echo "* text=auto" >.gitattributes &&
-	touch file &&
+	echo first line >file &&
 	git add .gitattributes file &&
 	test_tick &&
 	git commit -m "normalize file" &&
@@ -81,7 +88,7 @@ test_expect_success 'Merge after setting text=auto' '
 	rm -f .gitattributes &&
 	git reset --hard a &&
 	git merge b &&
-	test_cmp expected file
+	compare_files expected file
 '
 
 test_expect_success 'Merge addition of text=auto' '
@@ -99,7 +106,7 @@ test_expect_success 'Merge addition of text=auto' '
 	rm -f .gitattributes &&
 	git reset --hard b &&
 	git merge a &&
-	test_cmp expected file
+	compare_files  expected file
 '
 
 test_expect_success 'Detect CRLF/LF conflict after setting text=auto' '
@@ -121,7 +128,7 @@ test_expect_success 'Detect CRLF/LF conflict after setting text=auto' '
 	git reset --hard a &&
 	test_must_fail git merge b &&
 	fuzz_conflict file >file.fuzzy &&
-	test_cmp expected file.fuzzy
+	compare_files expected file.fuzzy
 '
 
 test_expect_success 'Detect LF/CRLF conflict from addition of text=auto' '
@@ -143,7 +150,7 @@ test_expect_success 'Detect LF/CRLF conflict from addition of text=auto' '
 	git reset --hard b &&
 	test_must_fail git merge a &&
 	fuzz_conflict file >file.fuzzy &&
-	test_cmp expected file.fuzzy
+	compare_files expected file.fuzzy
 '
 
 test_expect_failure 'checkout -m after setting text=auto' '
@@ -158,7 +165,7 @@ test_expect_failure 'checkout -m after setting text=auto' '
 	git reset --hard initial &&
 	git checkout a -- . &&
 	git checkout -m b &&
-	test_cmp expected file
+	compare_files expected file
 '
 
 test_expect_failure 'checkout -m addition of text=auto' '
@@ -173,7 +180,7 @@ test_expect_failure 'checkout -m addition of text=auto' '
 	git reset --hard initial &&
 	git checkout b -- . &&
 	git checkout -m a &&
-	test_cmp expected file
+	compare_files expected file
 '
 
 test_expect_failure 'cherry-pick patch from after text=auto was added' '
@@ -187,7 +194,7 @@ test_expect_failure 'cherry-pick patch from after text=auto was added' '
 	git reset --hard b &&
 	test_must_fail git cherry-pick a >err 2>&1 &&
 	grep "[Nn]othing added" err &&
-	test_cmp expected file
+	compare_files expected file
 '
 
 test_expect_success 'Test delete/normalize conflict' '
-- 
2.0.0.rc1.6318.g0c2c796
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help