[PATCH 0/2] read-tree: improve untracked file support

[PATCH 0/2] read-tree: improve untracked file support

[Phillip Wood]

From: Phillip Wood <phillip.wood@dunelm.org.uk>

These two patches teach read-tree how to avoid overwriting untracked
files when doing '--reset -u' and also how to respect all of git's
standard excludes files. I'd like to see the porcelain commands stop
overwriting untracked files, this is a first step on the way. I'm not
sure if we want to add options to the porcelain commands to protect
untracked files or just change their behavior and add an option to
override that. I'm leaning towards the latter but I'd be interested to
hear what others think.

Phillip Wood (2):
  read-tree --reset: add --protect-untracked
  read-tree: add --exclude-standard

 Documentation/git-read-tree.txt | 19 ++++++++--
 builtin/am.c                    |  8 +++--
 builtin/checkout.c              |  2 +-
 builtin/read-tree.c             | 61 +++++++++++++++++++++++++++++--
 builtin/rebase.c                |  2 +-
 builtin/reset.c                 |  2 +-
 builtin/stash.c                 |  7 ++--
 t/lib-read-tree.sh              | 11 ++++++
 t/t1005-read-tree-reset.sh      | 63 +++++++++++++++++++++++++++++++--
 t/t1013-read-tree-submodule.sh  |  3 +-
 unpack-trees.c                  |  3 +-
 unpack-trees.h                  | 10 ++++--
 12 files changed, 170 insertions(+), 21 deletions(-)

-- 
2.21.0

[PATCH 1/2] read-tree --reset: add --protect-untracked

[Phillip Wood]

From: Phillip Wood <phillip.wood@dunelm.org.uk>

Currently there is no way to get git to discard changes to the
worktree without overwriting untracked files. `reset --hard`,
`checkout --force`, `checkout <rev> :/` and `read-tree --reset -u`
will all overwrite untracked files. unpack_trees() has known how to
avoid overwriting untracked files since commit fcc387db9b ("read-tree
-m -u: do not overwrite or remove untracked working tree files.",
2006-05-17) in response to concerns about lost files [1] but those
protections do not apply to --reset. This patch adds an
--protect-untracked option for read-tree/unpack_trees() to apply the
same protections with --reset. It does not change the behavior of any
of the porcelain commands but paves the way for adding options or
changing the default for those in future.

Note the actual change in unpack-trees.c is quite simple, the majority
of this patch is converting existing callers to use the new
unpack_trees_reset_type enum.

[1] https://public-inbox.org/git/8aa486160605161500m1dd8428cj@mail.gmail.com/

Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
---

Notes:
    adding --protect-untracked to the invocation of
    test_submodule_forced_switch() in t1013 fixes the known breakage of
    tests 57 & 58 but breaks test 64 which relies on forcing. I'm not sure
    that the expected results of 57 & 58 are correct if we're forcing.

 Documentation/git-read-tree.txt | 10 +++++++--
 builtin/am.c                    |  8 ++++---
 builtin/checkout.c              |  2 +-
 builtin/read-tree.c             | 12 +++++++++++
 builtin/rebase.c                |  2 +-
 builtin/reset.c                 |  2 +-
 builtin/stash.c                 |  7 ++++---
 t/lib-read-tree.sh              | 11 ++++++++++
 t/t1005-read-tree-reset.sh      | 37 +++++++++++++++++++++++++++++++--
 unpack-trees.c                  |  3 ++-
 unpack-trees.h                  | 10 +++++++--
 11 files changed, 88 insertions(+), 16 deletions(-)

diff --git a/Documentation/git-read-tree.txt b/Documentation/git-read-tree.txt
index d271842608..67864c6bbc 100644
--- a/Documentation/git-read-tree.txt
+++ b/Documentation/git-read-tree.txt
@@ -40,12 +40,18 @@ OPTIONS
 --reset::
 	Same as -m, except that unmerged entries are discarded instead
 	of failing. When used with `-u`, updates leading to loss of
-	working tree changes will not abort the operation.
+	working tree changes will not abort the operation and
+	untracked files will be overwritten. Use `--protect-untracked`
+	to avoid overwriting untracked files.
 
 -u::
 	After a successful merge, update the files in the work
 	tree with the result of the merge.
 
+--protect-untracked::
+	Prevent `--reset` `-u` from overwriting untracked files. Requires
+	`--reset` and `-u` (`-m` never overwrites untracked files).
+
 -i::
 	Usually a merge requires the index file as well as the
 	files in the working tree to be up to date with the
@@ -89,7 +95,7 @@ OPTIONS
 	existed in the original index file.
 
 --exclude-per-directory=<gitignore>::
-	When running the command with `-u` and `-m` options, the
+	When using `-u` with `-m` or `--reset` `--protect-untracked` options, the
 	merge result may need to overwrite paths that are not
 	tracked in the current branch.  The command usually
 	refuses to proceed with the merge to avoid losing such a
diff --git a/builtin/am.c b/builtin/am.c
index 912d9821b1..a92394b682 100644
--- a/builtin/am.c
+++ b/builtin/am.c
@@ -1854,7 +1854,8 @@ static void am_resolve(struct am_state *state)
  * true, any unmerged entries will be discarded. Returns 0 on success, -1 on
  * failure.
  */
-static int fast_forward_to(struct tree *head, struct tree *remote, int reset)
+static int fast_forward_to(struct tree *head, struct tree *remote,
+			   enum unpack_trees_reset_type reset)
 {
 	struct lock_file lock_file = LOCK_INIT;
 	struct unpack_trees_options opts;
@@ -1942,7 +1943,8 @@ static int clean_index(const struct object_id *head, const struct object_id *rem
 
 	read_cache_unmerged();
 
-	if (fast_forward_to(head_tree, head_tree, 1))
+	if (fast_forward_to(head_tree, head_tree,
+			    UNPACK_RESET_OVERWRITE_UNTRACKED))
 		return -1;
 
 	if (write_cache_as_tree(&index, 0, NULL))
@@ -1952,7 +1954,7 @@ static int clean_index(const struct object_id *head, const struct object_id *rem
 	if (!index_tree)
 		return error(_("Could not parse object '%s'."), oid_to_hex(&index));
 
-	if (fast_forward_to(index_tree, remote_tree, 0))
+	if (fast_forward_to(index_tree, remote_tree, UNPACK_NO_RESET))
 		return -1;
 
 	if (merge_tree(remote_tree))
diff --git a/builtin/checkout.c b/builtin/checkout.c
index ffa776c6e1..e9e70018f9 100644
--- a/builtin/checkout.c
+++ b/builtin/checkout.c
@@ -506,7 +506,7 @@ static int reset_tree(struct tree *tree, const struct checkout_opts *o,
 	opts.head_idx = -1;
 	opts.update = worktree;
 	opts.skip_unmerged = !worktree;
-	opts.reset = 1;
+	opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
 	opts.merge = 1;
 	opts.fn = oneway_merge;
 	opts.verbose_update = o->show_progress;
diff --git a/builtin/read-tree.c b/builtin/read-tree.c
index 5c9c082595..23735adde9 100644
--- a/builtin/read-tree.c
+++ b/builtin/read-tree.c
@@ -114,6 +114,7 @@ static int git_read_tree_config(const char *var, const char *value, void *cb)
 int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
 {
 	int i, stage = 0;
+	int protect_untracked = -1;
 	struct object_id oid;
 	struct tree_desc t[MAX_UNPACK_TREES];
 	struct unpack_trees_options opts;
@@ -140,6 +141,8 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
 		  PARSE_OPT_NONEG },
 		OPT_BOOL('u', NULL, &opts.update,
 			 N_("update working tree with merge result")),
+		OPT_BOOL(0, "protect-untracked", &protect_untracked,
+			 N_("do not overwrite untracked files")),
 		{ OPTION_CALLBACK, 0, "exclude-per-directory", &opts,
 		  N_("gitignore"),
 		  N_("allow explicitly ignored files to be overwritten"),
@@ -209,8 +212,17 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
 	if ((opts.update || opts.index_only) && !opts.merge)
 		die("%s is meaningless without -m, --reset, or --prefix",
 		    opts.update ? "-u" : "-i");
+	if (protect_untracked >= 0) {
+		if (!opts.reset || !opts.update)
+			die("-%s-protect-untracked requires --reset and -u",
+			    protect_untracked ? "" : "-no");
+		opts.reset = UNPACK_RESET_PROTECT_UNTRACKED;
+	}
 	if ((opts.dir && !opts.update))
 		die("--exclude-per-directory is meaningless unless -u");
+	if (opts.dir && opts.reset == UNPACK_RESET_OVERWRITE_UNTRACKED)
+		warning("--exclude-per-directory without --preserve-untracked "
+			"has no effect");
 	if (opts.merge && !opts.index_only)
 		setup_work_tree();
 
diff --git a/builtin/rebase.c b/builtin/rebase.c
index 2e41ad5644..feb30a71f5 100644
--- a/builtin/rebase.c
+++ b/builtin/rebase.c
@@ -398,7 +398,7 @@ static int reset_head(struct object_id *oid, const char *action,
 	unpack_tree_opts.update = 1;
 	unpack_tree_opts.merge = 1;
 	if (!detach_head)
-		unpack_tree_opts.reset = 1;
+		unpack_tree_opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
 
 	if (repo_read_index_unmerged(the_repository) < 0) {
 		ret = error(_("could not read index"));
diff --git a/builtin/reset.c b/builtin/reset.c
index 26ef9a7bd0..a39dd92fb2 100644
--- a/builtin/reset.c
+++ b/builtin/reset.c
@@ -70,7 +70,7 @@ static int reset_index(const struct object_id *oid, int reset_type, int quiet)
 		opts.update = 1;
 		/* fallthrough */
 	default:
-		opts.reset = 1;
+		opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
 	}
 
 	read_cache_unmerged();
diff --git a/builtin/stash.c b/builtin/stash.c
index 2a8e6d09b4..175d1b62d3 100644
--- a/builtin/stash.c
+++ b/builtin/stash.c
@@ -227,7 +227,8 @@ static int clear_stash(int argc, const char **argv, const char *prefix)
 	return do_clear_stash();
 }
 
-static int reset_tree(struct object_id *i_tree, int update, int reset)
+static int reset_tree(struct object_id *i_tree, int update,
+		      enum unpack_trees_reset_type reset)
 {
 	int nr_trees = 1;
 	struct unpack_trees_options opts;
@@ -461,7 +462,7 @@ static int do_apply_stash(const char *prefix, struct stash_info *info,
 	}
 
 	if (has_index) {
-		if (reset_tree(&index_tree, 0, 0))
+		if (reset_tree(&index_tree, 0, UNPACK_NO_RESET))
 			return -1;
 	} else {
 		struct strbuf out = STRBUF_INIT;
@@ -471,7 +472,7 @@ static int do_apply_stash(const char *prefix, struct stash_info *info,
 			return -1;
 		}
 
-		if (reset_tree(&c_tree, 0, 1)) {
+		if (reset_tree(&c_tree, 0, UNPACK_RESET_OVERWRITE_UNTRACKED)) {
 			strbuf_release(&out);
 			return -1;
 		}
diff --git a/t/lib-read-tree.sh b/t/lib-read-tree.sh
index b95f485606..22f516d91f 100644
--- a/t/lib-read-tree.sh
+++ b/t/lib-read-tree.sh
@@ -39,3 +39,14 @@ read_tree_u_must_fail () {
 	test_cmp pre-dry-run-wt post-dry-run-wt &&
 	test_must_fail git read-tree "$@"
 }
+
+read_tree_u_must_fail_save_err () {
+	git ls-files -s >pre-dry-run &&
+	git diff-files -p >pre-dry-run-wt &&
+	test_must_fail git read-tree -n "$@" &&
+	git ls-files -s >post-dry-run &&
+	git diff-files -p >post-dry-run-wt &&
+	test_cmp pre-dry-run post-dry-run &&
+	test_cmp pre-dry-run-wt post-dry-run-wt &&
+	test_must_fail git read-tree "$@" 2>actual-err
+}
diff --git a/t/t1005-read-tree-reset.sh b/t/t1005-read-tree-reset.sh
index 83b09e1310..6c9dd6805b 100755
--- a/t/t1005-read-tree-reset.sh
+++ b/t/t1005-read-tree-reset.sh
@@ -19,15 +19,48 @@ test_expect_success 'setup' '
 	git add df &&
 	echo content >new &&
 	git add new &&
-	git commit -m two
+	git commit -m two &&
+	git ls-files >expect-two
 '
 
-test_expect_success 'reset should work' '
+test_expect_success '--protect-untracked option sanity checks' '
+	read_tree_u_must_fail --reset --protect-untracked HEAD &&
+	read_tree_u_must_fail --reset --no-protect-untracked HEAD &&
+	read_tree_u_must_fail -m -u --protect-untracked HEAD &&
+	read_tree_u_must_fail -m -u --no-protect-untracked
+'
+
+test_expect_success 'reset should reset worktree' '
+	echo changed >df &&
 	read_tree_u_must_succeed -u --reset HEAD^ &&
 	git ls-files >actual &&
 	test_cmp expect actual
 '
 
+test_expect_success 'reset --protect-untracked protects untracked file' '
+	echo changed >new &&
+	read_tree_u_must_fail_save_err -u --reset --protect-untracked HEAD &&
+	echo "error: Untracked working tree file '\'new\'' would be overwritten by merge." >expected-err &&
+	test_cmp expected-err actual-err
+'
+
+test_expect_success 'reset --protect-untracked protects untracked directory' '
+	rm new &&
+	mkdir new &&
+	echo untracked >new/untracked &&
+	read_tree_u_must_fail_save_err -u --reset --protect-untracked HEAD &&
+	echo "error: Updating '\'new\'' would lose untracked files in it" >expected-err &&
+	test_cmp expected-err actual-err
+'
+
+test_expect_success 'reset --protect-untracked resets' '
+	rm -rf new &&
+	echo changed >df/file &&
+	read_tree_u_must_succeed -u --reset --protect-untracked HEAD &&
+	git ls-files >actual-two &&
+	test_cmp expect-two actual-two
+'
+
 test_expect_success 'reset should remove remnants from a failed merge' '
 	read_tree_u_must_succeed --reset -u HEAD &&
 	git ls-files -s >expect &&
diff --git a/unpack-trees.c b/unpack-trees.c
index 50189909b8..b1722730fe 100644
--- a/unpack-trees.c
+++ b/unpack-trees.c
@@ -1917,7 +1917,8 @@ static int verify_absent_1(const struct cache_entry *ce,
 	int len;
 	struct stat st;
 
-	if (o->index_only || o->reset || !o->update)
+	if (o->index_only || o->reset == UNPACK_RESET_OVERWRITE_UNTRACKED ||
+	    !o->update)
 		return 0;
 
 	len = check_leading_path(ce->name, ce_namelen(ce));
diff --git a/unpack-trees.h b/unpack-trees.h
index d344d7d296..732b262c4d 100644
--- a/unpack-trees.h
+++ b/unpack-trees.h
@@ -41,9 +41,15 @@ void setup_unpack_trees_porcelain(struct unpack_trees_options *opts,
  */
 void clear_unpack_trees_porcelain(struct unpack_trees_options *opts);
 
+enum unpack_trees_reset_type {
+	UNPACK_NO_RESET = 0,
+	UNPACK_RESET_OVERWRITE_UNTRACKED,
+	UNPACK_RESET_PROTECT_UNTRACKED
+};
+
 struct unpack_trees_options {
-	unsigned int reset,
-		     merge,
+	enum unpack_trees_reset_type reset;
+	unsigned int merge,
 		     update,
 		     clone,
 		     index_only,
-- 
2.21.0

[PATCH 2/2] read-tree: add --exclude-standard

[Phillip Wood]

From: Phillip Wood <phillip.wood@dunelm.org.uk>

Currently there is no way to get read-tree to respect
.git/info/exclude or core.excludesFile so scripts using `read-tree -u`
have subtly different behavior to porcelain commands like checkout
even when they use --exclude-per-directory. This new option is copied
from ls-tree's --exclude-standard option to setup the standard
excludes. The new option is also used to fix a known submodule test
failure.

Note that KNOWN_FAILURE_SUBMODULE_OVERWRITE_IGNORED_UNTRACKED is still
used by t7112-reset-submodule.sh as it is not removed (apparently
reset does not call setup_standard_excludes()).

Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
---
 Documentation/git-read-tree.txt |  9 +++++-
 builtin/read-tree.c             | 55 ++++++++++++++++++++++++++++++---
 t/t1005-read-tree-reset.sh      | 36 ++++++++++++++++++---
 t/t1013-read-tree-submodule.sh  |  3 +-
 4 files changed, 90 insertions(+), 13 deletions(-)

diff --git a/Documentation/git-read-tree.txt b/Documentation/git-read-tree.txt
index 67864c6bbc..a2b8b73a99 100644
--- a/Documentation/git-read-tree.txt
+++ b/Documentation/git-read-tree.txt
@@ -107,7 +107,14 @@ OPTIONS
 	running `make clean` to remove the generated file.  This
 	option tells the command to read per-directory exclude
 	file (usually '.gitignore') and allows such an untracked
-	but explicitly ignored file to be overwritten.
+	but explicitly ignored file to be overwritten. Incompatible
+	with `--exclude-standard`.
+
+--exclude-standard::
+	When updating the worktree use the standard Git exclusions:
+	.git/info/exclude, .gitignore in each directory, and the user's global
+	exclusion file when deciding if it is safe to overwrite a file.
+	Incompatible with `--exclude-per-directory`.
 
 --index-output=<file>::
 	Instead of writing the results out to `$GIT_INDEX_FILE`,
diff --git a/builtin/read-tree.c b/builtin/read-tree.c
index 23735adde9..5df493c4a7 100644
--- a/builtin/read-tree.c
+++ b/builtin/read-tree.c
@@ -50,6 +50,40 @@ static int index_output_cb(const struct option *opt, const char *arg,
 	return 0;
 }
 
+enum exclude_type {
+	EXCLUDE_NONE,
+	EXCLUDE_PER_DIRECTORY,
+	EXCLUDE_STANDARD
+} exclude_opt = EXCLUDE_NONE;
+
+static int exclude_error(enum exclude_type exclude)
+{
+	if (exclude == exclude_opt)
+		return error("more than one --exclude-per-directory given");
+	else
+		return error("cannot combine --exclude-per-directory and "
+			     "--exclude-standard");
+}
+
+static int option_parse_exclude_standard(const struct option *opt,
+					 const char *arg, int unset)
+{
+	struct unpack_trees_options *opts;
+
+	BUG_ON_OPT_NEG(unset);
+	BUG_ON_OPT_ARG(arg);
+
+	if (exclude_opt == EXCLUDE_PER_DIRECTORY)
+		return exclude_error(EXCLUDE_STANDARD);
+
+	opts = (struct unpack_trees_options *)opt->value;
+	opts->dir = xcalloc(1, sizeof(*opts->dir));
+	setup_standard_excludes(opts->dir);
+	exclude_opt = EXCLUDE_STANDARD;
+
+	return 0;
+}
+
 static int exclude_per_directory_cb(const struct option *opt, const char *arg,
 				    int unset)
 {
@@ -61,12 +95,13 @@ static int exclude_per_directory_cb(const struct option *opt, const char *arg,
 	opts = (struct unpack_trees_options *)opt->value;
 
 	if (opts->dir)
-		die("more than one --exclude-per-directory given.");
+		return exclude_error(EXCLUDE_PER_DIRECTORY);
 
 	dir = xcalloc(1, sizeof(*opts->dir));
 	dir->flags |= DIR_SHOW_IGNORED;
 	dir->exclude_per_dir = arg;
 	opts->dir = dir;
+	exclude_opt = EXCLUDE_PER_DIRECTORY;
 	/* We do not need to nor want to do read-directory
 	 * here; we are merely interested in reusing the
 	 * per directory ignore stack mechanism.
@@ -147,6 +182,10 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
 		  N_("gitignore"),
 		  N_("allow explicitly ignored files to be overwritten"),
 		  PARSE_OPT_NONEG, exclude_per_directory_cb },
+		{ OPTION_CALLBACK, 0, "exclude-standard", &opts, NULL,
+			N_("add the standard git exclusions"),
+			PARSE_OPT_NOARG | PARSE_OPT_NONEG,
+			option_parse_exclude_standard },
 		OPT_BOOL('i', NULL, &opts.index_only,
 			 N_("don't check the working tree after merging")),
 		OPT__DRY_RUN(&opts.dry_run, N_("don't update the index or the work tree")),
@@ -219,10 +258,16 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
 		opts.reset = UNPACK_RESET_PROTECT_UNTRACKED;
 	}
 	if ((opts.dir && !opts.update))
-		die("--exclude-per-directory is meaningless unless -u");
-	if (opts.dir && opts.reset == UNPACK_RESET_OVERWRITE_UNTRACKED)
-		warning("--exclude-per-directory without --preserve-untracked "
-			"has no effect");
+		die("%s requires -u", exclude_opt == EXCLUDE_STANDARD ?
+			"--exclude-standard" :" --exclude-per-directory");
+	if (opts.dir && opts.reset == UNPACK_RESET_OVERWRITE_UNTRACKED) {
+		if (exclude_opt == EXCLUDE_STANDARD)
+			die("--reset with --exclude-standard requires "
+			    "--protect-untracked");
+		else
+			warning("--exclude-per-directory without "
+				"--preserve-untracked has no effect");
+	}
 	if (opts.merge && !opts.index_only)
 		setup_work_tree();
 
diff --git a/t/t1005-read-tree-reset.sh b/t/t1005-read-tree-reset.sh
index 6c9dd6805b..2e2a6a0c69 100755
--- a/t/t1005-read-tree-reset.sh
+++ b/t/t1005-read-tree-reset.sh
@@ -30,6 +30,20 @@ test_expect_success '--protect-untracked option sanity checks' '
 	read_tree_u_must_fail -m -u --no-protect-untracked
 '
 
+test_expect_success 'exclude option sanity checks' '
+	read_tree_u_must_fail --reset -u --exclude-standard HEAD &&
+	read_tree_u_must_fail --reset --protect-untracked --exclude-standard &&
+	read_tree_u_must_fail --reset -u --protect-untracked \
+			      --exclude-standard \
+			      --exclude-per-directory=.gitignore HEAD &&
+	read_tree_u_must_fail --reset -u --protect-untracked \
+			      --exclude-per-directory=gitignore \
+			      --exclude-per-directory=.gitignore HEAD &&
+	read_tree_u_must_fail --reset --exclude-per-directory=.gitignore HEAD &&
+	read_tree_u_must_succeed --reset -u --exclude-per-directory=.gitignore \
+				 HEAD
+'
+
 test_expect_success 'reset should reset worktree' '
 	echo changed >df &&
 	read_tree_u_must_succeed -u --reset HEAD^ &&
@@ -53,12 +67,24 @@ test_expect_success 'reset --protect-untracked protects untracked directory' '
 	test_cmp expected-err actual-err
 '
 
-test_expect_success 'reset --protect-untracked resets' '
-	rm -rf new &&
+test_expect_success 'reset --protect-untracked --exclude-standard overwrites ignored path' '
+	test_when_finished "rm .git/info/exclude" &&
+	echo missing >.git/info/exclude &&
+	read_tree_u_must_fail -u --reset --protect-untracked \
+			      --exclude-standard HEAD &&
+	echo new >.git/info/exclude &&
 	echo changed >df/file &&
-	read_tree_u_must_succeed -u --reset --protect-untracked HEAD &&
-	git ls-files >actual-two &&
-	test_cmp expect-two actual-two
+	read_tree_u_must_succeed -u --reset --protect-untracked \
+				 --exclude-standard HEAD &&
+	git ls-files >actual &&
+	test_cmp expect-two actual
+'
+
+test_expect_success 'reset --protect-untracked resets' '
+	echo changed >df &&
+	read_tree_u_must_succeed -u --reset --protect-untracked HEAD^ &&
+	git ls-files >actual &&
+	test_cmp expect actual
 '
 
 test_expect_success 'reset should remove remnants from a failed merge' '
diff --git a/t/t1013-read-tree-submodule.sh b/t/t1013-read-tree-submodule.sh
index 91a6fafcb4..728280d40d 100755
--- a/t/t1013-read-tree-submodule.sh
+++ b/t/t1013-read-tree-submodule.sh
@@ -6,9 +6,8 @@ test_description='read-tree can handle submodules'
 . "$TEST_DIRECTORY"/lib-submodule-update.sh
 
 KNOWN_FAILURE_DIRECTORY_SUBMODULE_CONFLICTS=1
-KNOWN_FAILURE_SUBMODULE_OVERWRITE_IGNORED_UNTRACKED=1
 
-test_submodule_switch_recursing_with_args "read-tree -u -m"
+test_submodule_switch_recursing_with_args "read-tree -u -m --exclude-standard"
 
 test_submodule_forced_switch_recursing_with_args "read-tree -u --reset"
 
-- 
2.21.0

Re: [PATCH 1/2] read-tree --reset: add --protect-untracked

[Duy Nguyen]

On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
> diff --git a/unpack-trees.h b/unpack-trees.h
> index d344d7d296..732b262c4d 100644
> --- a/unpack-trees.h
> +++ b/unpack-trees.h
> @@ -41,9 +41,15 @@ void setup_unpack_trees_porcelain(struct unpack_trees_options *opts,
>   */
>  void clear_unpack_trees_porcelain(struct unpack_trees_options *opts);
>
> +enum unpack_trees_reset_type {
> +       UNPACK_NO_RESET = 0,
> +       UNPACK_RESET_OVERWRITE_UNTRACKED,
> +       UNPACK_RESET_PROTECT_UNTRACKED
> +};
> +
>  struct unpack_trees_options {
> -       unsigned int reset,
> -                    merge,
> +       enum unpack_trees_reset_type reset;

Can we add protected_untracked that can be used in combination with
the current "reset"?

This opens up (or raises the question about) the opportunity to
protect untracked changes on non-reset updates.

> +       unsigned int merge,
>                      update,
>                      clone,
>                      index_only,
> --
> 2.21.0
>


-- 
Duy

Re: [PATCH 0/2] read-tree: improve untracked file support

[Duy Nguyen]

On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
>
> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>
> These two patches teach read-tree how to avoid overwriting untracked
> files when doing '--reset -u' and also how to respect all of git's
> standard excludes files. I'd like to see the porcelain commands stop
> overwriting untracked files, this is a first step on the way. I'm not
> sure if we want to add options to the porcelain commands to protect
> untracked files or just change their behavior and add an option to
> override that. I'm leaning towards the latter but I'd be interested to
> hear what others think.

For new commands like git-restore, it's definitely a good thing to not
overwrite untracked files. For existing commands I guess we have to go
over them one by one. For "git reset --hard", it should really just
overwrite whatever needed to get back to the known good state. "git
checkout -f" , not so sure (seems weird that we need force-level-two
option to override the protection provided by -f, if we change default
behavior)
-- 
Duy

Re: [PATCH 1/2] read-tree --reset: add --protect-untracked

[Duy Nguyen]

On Wed, May 1, 2019 at 5:18 PM Duy Nguyen <pclouds@gmail.com> wrote:
>
> On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
> > diff --git a/unpack-trees.h b/unpack-trees.h
> > index d344d7d296..732b262c4d 100644
> > --- a/unpack-trees.h
> > +++ b/unpack-trees.h
> > @@ -41,9 +41,15 @@ void setup_unpack_trees_porcelain(struct unpack_trees_options *opts,
> >   */
> >  void clear_unpack_trees_porcelain(struct unpack_trees_options *opts);
> >
> > +enum unpack_trees_reset_type {
> > +       UNPACK_NO_RESET = 0,
> > +       UNPACK_RESET_OVERWRITE_UNTRACKED,
> > +       UNPACK_RESET_PROTECT_UNTRACKED
> > +};
> > +
> >  struct unpack_trees_options {
> > -       unsigned int reset,
> > -                    merge,
> > +       enum unpack_trees_reset_type reset;
>
> Can we add protected_untracked that can be used in combination with
> the current "reset"?
>
> This opens up (or raises the question about) the opportunity to
> protect untracked changes on non-reset updates.

Bahh. "protect_untracked" is already on for non-reset updates. Sorry
for the noise.
-- 
Duy

Re: [PATCH 0/2] read-tree: improve untracked file support

[Phillip Wood]

On 01/05/2019 11:31, Duy Nguyen wrote:
> On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
>>
>> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>>
>> These two patches teach read-tree how to avoid overwriting untracked
>> files when doing '--reset -u' and also how to respect all of git's
>> standard excludes files. I'd like to see the porcelain commands stop
>> overwriting untracked files, this is a first step on the way. I'm not
>> sure if we want to add options to the porcelain commands to protect
>> untracked files or just change their behavior and add an option to
>> override that. I'm leaning towards the latter but I'd be interested to
>> hear what others think.
> 
> For new commands like git-restore, it's definitely a good thing to not
> overwrite untracked files.

I agree, unfortunately this series does not help with git-restore, only 
git-switch. For restore on an index without conflicts I think it could 
just use the pathspec in struct unpack_trees_options and set opts.rest = 
UNPACK_RESET_PROTECT_UNTRACKED but that does not help if we want to 
handle conflicted paths differently to non-conflicted paths.

> For existing commands I guess we have to go
> over them one by one. For "git reset --hard", it should really just
> overwrite whatever needed to get back to the known good state. "git
> checkout -f" , not so sure (seems weird that we need force-level-two
> option to override the protection provided by -f, if we change default
> behavior)

I think it's fine for "checkout -f" to overwrite untracked files (and if 
"switch --discard-changes" does not then there is no pressing need to 
add such a mode to checkout), --force is a good name for an option that 
nukes everything that gets in it's way. For "reset --hard" I'm not so 
sure, if I have changes to an untracked file I don't wont them 
overwritten by default. There is no porcelain equivalent to "read-tree 
--reset --protect-untracked -u" and I was hoping "reset --hard" may 
become that porcelain equivalent with a new --force or 
--overwrite-untracked option.

For the various "foo --abort" some (most?) are using "reset --merge" 
which I think declines to overwrite untracked files but rebase uses 
"reset --hard" which I'd like to change to protect untracked files in 
the same way that rebase does for the initial checkout and when picking 
commits. I haven't thought about stash.

Best Wishes

Phillip

Re: [PATCH 1/2] read-tree --reset: add --protect-untracked

[Duy Nguyen]

On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
>
> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>
> Currently there is no way to get git to discard changes to the
> worktree without overwriting untracked files. `reset --hard`,
> `checkout --force`, `checkout <rev> :/` and `read-tree --reset -u`

"checkout <rev> :/" does not use unpack-trees, I think, so it does not
belong to this group.

> will all overwrite untracked files. unpack_trees() has known how to
> avoid overwriting untracked files since commit fcc387db9b ("read-tree
> -m -u: do not overwrite or remove untracked working tree files.",
> 2006-05-17) in response to concerns about lost files [1] but those
> protections do not apply to --reset. This patch adds an
> --protect-untracked option for read-tree/unpack_trees() to apply th
> same protections with --reset.

Your enum makes me wonder if we should have --reset-tracked instead of
--protect-untracked (say what you want to reset seems tiny bit easier
to understand than "reset except untracked"). Which leads to another
question, do we ever need --reset-untracked (i.e. overwriting
untracked files but never ever touch tracked ones).

I have one use case that --reset-untracked might make sense. Suppose
you do "git reset HEAD^" where HEAD has some new files. The new files
will be left alone in worktree of course, but switching back to HEAD
after that, unpack-trees will cry murder because it would otherwise
overwrite untracked files (that have the exact same content).

--reset-untracked might be useful for that. I'm carrying a patch to
detect identical file content though which addresses this very
specific case. Maybe it's the only case that --reset-untracked may be
useful.

> It does not change the behavior of any
> of the porcelain commands but paves the way for adding options or
> changing the default for those in future.
>
> Note the actual change in unpack-trees.c is quite simple, the majority
> of this patch is converting existing callers to use the new
> unpack_trees_reset_type enum.

This could be split in a separate patch, then the actual change would
become small and more to the point.

> [1] https://public-inbox.org/git/8aa486160605161500m1dd8428cj@mail.gmail.com/
>
> Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
> ---
>
> Notes:
>     adding --protect-untracked to the invocation of
>     test_submodule_forced_switch() in t1013 fixes the known breakage of
>     tests 57 & 58 but breaks test 64 which relies on forcing. I'm not sure
>     that the expected results of 57 & 58 are correct if we're forcing.
>
>  Documentation/git-read-tree.txt | 10 +++++++--
>  builtin/am.c                    |  8 ++++---
>  builtin/checkout.c              |  2 +-
>  builtin/read-tree.c             | 12 +++++++++++
>  builtin/rebase.c                |  2 +-
>  builtin/reset.c                 |  2 +-
>  builtin/stash.c                 |  7 ++++---
>  t/lib-read-tree.sh              | 11 ++++++++++
>  t/t1005-read-tree-reset.sh      | 37 +++++++++++++++++++++++++++++++--
>  unpack-trees.c                  |  3 ++-
>  unpack-trees.h                  | 10 +++++++--
>  11 files changed, 88 insertions(+), 16 deletions(-)
>
> diff --git a/Documentation/git-read-tree.txt b/Documentation/git-read-tree.txt
> index d271842608..67864c6bbc 100644
> --- a/Documentation/git-read-tree.txt
> +++ b/Documentation/git-read-tree.txt
> @@ -40,12 +40,18 @@ OPTIONS
>  --reset::
>         Same as -m, except that unmerged entries are discarded instead
>         of failing. When used with `-u`, updates leading to loss of
> -       working tree changes will not abort the operation.
> +       working tree changes will not abort the operation and
> +       untracked files will be overwritten. Use `--protect-untracked`
> +       to avoid overwriting untracked files.
>
>  -u::
>         After a successful merge, update the files in the work
>         tree with the result of the merge.
>
> +--protect-untracked::
> +       Prevent `--reset` `-u` from overwriting untracked files. Requires
> +       `--reset` and `-u` (`-m` never overwrites untracked files).
> +
>  -i::
>         Usually a merge requires the index file as well as the
>         files in the working tree to be up to date with the
> @@ -89,7 +95,7 @@ OPTIONS
>         existed in the original index file.
>
>  --exclude-per-directory=<gitignore>::
> -       When running the command with `-u` and `-m` options, the
> +       When using `-u` with `-m` or `--reset` `--protect-untracked` options, the
>         merge result may need to overwrite paths that are not
>         tracked in the current branch.  The command usually
>         refuses to proceed with the merge to avoid losing such a
> diff --git a/builtin/am.c b/builtin/am.c
> index 912d9821b1..a92394b682 100644
> --- a/builtin/am.c
> +++ b/builtin/am.c
> @@ -1854,7 +1854,8 @@ static void am_resolve(struct am_state *state)
>   * true, any unmerged entries will be discarded. Returns 0 on success, -1 on
>   * failure.
>   */
> -static int fast_forward_to(struct tree *head, struct tree *remote, int reset)
> +static int fast_forward_to(struct tree *head, struct tree *remote,
> +                          enum unpack_trees_reset_type reset)
>  {
>         struct lock_file lock_file = LOCK_INIT;
>         struct unpack_trees_options opts;
> @@ -1942,7 +1943,8 @@ static int clean_index(const struct object_id *head, const struct object_id *rem
>
>         read_cache_unmerged();
>
> -       if (fast_forward_to(head_tree, head_tree, 1))
> +       if (fast_forward_to(head_tree, head_tree,
> +                           UNPACK_RESET_OVERWRITE_UNTRACKED))
>                 return -1;
>
>         if (write_cache_as_tree(&index, 0, NULL))
> @@ -1952,7 +1954,7 @@ static int clean_index(const struct object_id *head, const struct object_id *rem
>         if (!index_tree)
>                 return error(_("Could not parse object '%s'."), oid_to_hex(&index));
>
> -       if (fast_forward_to(index_tree, remote_tree, 0))
> +       if (fast_forward_to(index_tree, remote_tree, UNPACK_NO_RESET))
>                 return -1;
>
>         if (merge_tree(remote_tree))
> diff --git a/builtin/checkout.c b/builtin/checkout.c
> index ffa776c6e1..e9e70018f9 100644
> --- a/builtin/checkout.c
> +++ b/builtin/checkout.c
> @@ -506,7 +506,7 @@ static int reset_tree(struct tree *tree, const struct checkout_opts *o,
>         opts.head_idx = -1;
>         opts.update = worktree;
>         opts.skip_unmerged = !worktree;
> -       opts.reset = 1;
> +       opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
>         opts.merge = 1;
>         opts.fn = oneway_merge;
>         opts.verbose_update = o->show_progress;
> diff --git a/builtin/read-tree.c b/builtin/read-tree.c
> index 5c9c082595..23735adde9 100644
> --- a/builtin/read-tree.c
> +++ b/builtin/read-tree.c
> @@ -114,6 +114,7 @@ static int git_read_tree_config(const char *var, const char *value, void *cb)
>  int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
>  {
>         int i, stage = 0;
> +       int protect_untracked = -1;
>         struct object_id oid;
>         struct tree_desc t[MAX_UNPACK_TREES];
>         struct unpack_trees_options opts;
> @@ -140,6 +141,8 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
>                   PARSE_OPT_NONEG },
>                 OPT_BOOL('u', NULL, &opts.update,
>                          N_("update working tree with merge result")),
> +               OPT_BOOL(0, "protect-untracked", &protect_untracked,
> +                        N_("do not overwrite untracked files")),
>                 { OPTION_CALLBACK, 0, "exclude-per-directory", &opts,
>                   N_("gitignore"),
>                   N_("allow explicitly ignored files to be overwritten"),
> @@ -209,8 +212,17 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
>         if ((opts.update || opts.index_only) && !opts.merge)
>                 die("%s is meaningless without -m, --reset, or --prefix",
>                     opts.update ? "-u" : "-i");
> +       if (protect_untracked >= 0) {
> +               if (!opts.reset || !opts.update)
> +                       die("-%s-protect-untracked requires --reset and -u",
> +                           protect_untracked ? "" : "-no");
> +               opts.reset = UNPACK_RESET_PROTECT_UNTRACKED;
> +       }
>         if ((opts.dir && !opts.update))
>                 die("--exclude-per-directory is meaningless unless -u");
> +       if (opts.dir && opts.reset == UNPACK_RESET_OVERWRITE_UNTRACKED)
> +               warning("--exclude-per-directory without --preserve-untracked "
> +                       "has no effect");
>         if (opts.merge && !opts.index_only)
>                 setup_work_tree();
>
> diff --git a/builtin/rebase.c b/builtin/rebase.c
> index 2e41ad5644..feb30a71f5 100644
> --- a/builtin/rebase.c
> +++ b/builtin/rebase.c
> @@ -398,7 +398,7 @@ static int reset_head(struct object_id *oid, const char *action,
>         unpack_tree_opts.update = 1;
>         unpack_tree_opts.merge = 1;
>         if (!detach_head)
> -               unpack_tree_opts.reset = 1;
> +               unpack_tree_opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
>
>         if (repo_read_index_unmerged(the_repository) < 0) {
>                 ret = error(_("could not read index"));
> diff --git a/builtin/reset.c b/builtin/reset.c
> index 26ef9a7bd0..a39dd92fb2 100644
> --- a/builtin/reset.c
> +++ b/builtin/reset.c
> @@ -70,7 +70,7 @@ static int reset_index(const struct object_id *oid, int reset_type, int quiet)
>                 opts.update = 1;
>                 /* fallthrough */
>         default:
> -               opts.reset = 1;
> +               opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
>         }
>
>         read_cache_unmerged();
> diff --git a/builtin/stash.c b/builtin/stash.c
> index 2a8e6d09b4..175d1b62d3 100644
> --- a/builtin/stash.c
> +++ b/builtin/stash.c
> @@ -227,7 +227,8 @@ static int clear_stash(int argc, const char **argv, const char *prefix)
>         return do_clear_stash();
>  }
>
> -static int reset_tree(struct object_id *i_tree, int update, int reset)
> +static int reset_tree(struct object_id *i_tree, int update,
> +                     enum unpack_trees_reset_type reset)
>  {
>         int nr_trees = 1;
>         struct unpack_trees_options opts;
> @@ -461,7 +462,7 @@ static int do_apply_stash(const char *prefix, struct stash_info *info,
>         }
>
>         if (has_index) {
> -               if (reset_tree(&index_tree, 0, 0))
> +               if (reset_tree(&index_tree, 0, UNPACK_NO_RESET))
>                         return -1;
>         } else {
>                 struct strbuf out = STRBUF_INIT;
> @@ -471,7 +472,7 @@ static int do_apply_stash(const char *prefix, struct stash_info *info,
>                         return -1;
>                 }
>
> -               if (reset_tree(&c_tree, 0, 1)) {
> +               if (reset_tree(&c_tree, 0, UNPACK_RESET_OVERWRITE_UNTRACKED)) {
>                         strbuf_release(&out);
>                         return -1;
>                 }
> diff --git a/t/lib-read-tree.sh b/t/lib-read-tree.sh
> index b95f485606..22f516d91f 100644
> --- a/t/lib-read-tree.sh
> +++ b/t/lib-read-tree.sh
> @@ -39,3 +39,14 @@ read_tree_u_must_fail () {
>         test_cmp pre-dry-run-wt post-dry-run-wt &&
>         test_must_fail git read-tree "$@"
>  }
> +
> +read_tree_u_must_fail_save_err () {
> +       git ls-files -s >pre-dry-run &&
> +       git diff-files -p >pre-dry-run-wt &&
> +       test_must_fail git read-tree -n "$@" &&
> +       git ls-files -s >post-dry-run &&
> +       git diff-files -p >post-dry-run-wt &&
> +       test_cmp pre-dry-run post-dry-run &&
> +       test_cmp pre-dry-run-wt post-dry-run-wt &&
> +       test_must_fail git read-tree "$@" 2>actual-err
> +}
> diff --git a/t/t1005-read-tree-reset.sh b/t/t1005-read-tree-reset.sh
> index 83b09e1310..6c9dd6805b 100755
> --- a/t/t1005-read-tree-reset.sh
> +++ b/t/t1005-read-tree-reset.sh
> @@ -19,15 +19,48 @@ test_expect_success 'setup' '
>         git add df &&
>         echo content >new &&
>         git add new &&
> -       git commit -m two
> +       git commit -m two &&
> +       git ls-files >expect-two
>  '
>
> -test_expect_success 'reset should work' '
> +test_expect_success '--protect-untracked option sanity checks' '
> +       read_tree_u_must_fail --reset --protect-untracked HEAD &&
> +       read_tree_u_must_fail --reset --no-protect-untracked HEAD &&
> +       read_tree_u_must_fail -m -u --protect-untracked HEAD &&
> +       read_tree_u_must_fail -m -u --no-protect-untracked
> +'
> +
> +test_expect_success 'reset should reset worktree' '
> +       echo changed >df &&
>         read_tree_u_must_succeed -u --reset HEAD^ &&
>         git ls-files >actual &&
>         test_cmp expect actual
>  '
>
> +test_expect_success 'reset --protect-untracked protects untracked file' '
> +       echo changed >new &&
> +       read_tree_u_must_fail_save_err -u --reset --protect-untracked HEAD &&
> +       echo "error: Untracked working tree file '\'new\'' would be overwritten by merge." >expected-err &&
> +       test_cmp expected-err actual-err
> +'
> +
> +test_expect_success 'reset --protect-untracked protects untracked directory' '
> +       rm new &&
> +       mkdir new &&
> +       echo untracked >new/untracked &&
> +       read_tree_u_must_fail_save_err -u --reset --protect-untracked HEAD &&
> +       echo "error: Updating '\'new\'' would lose untracked files in it" >expected-err &&
> +       test_cmp expected-err actual-err
> +'
> +
> +test_expect_success 'reset --protect-untracked resets' '
> +       rm -rf new &&
> +       echo changed >df/file &&
> +       read_tree_u_must_succeed -u --reset --protect-untracked HEAD &&
> +       git ls-files >actual-two &&
> +       test_cmp expect-two actual-two
> +'
> +
>  test_expect_success 'reset should remove remnants from a failed merge' '
>         read_tree_u_must_succeed --reset -u HEAD &&
>         git ls-files -s >expect &&
> diff --git a/unpack-trees.c b/unpack-trees.c
> index 50189909b8..b1722730fe 100644
> --- a/unpack-trees.c
> +++ b/unpack-trees.c
> @@ -1917,7 +1917,8 @@ static int verify_absent_1(const struct cache_entry *ce,
>         int len;
>         struct stat st;
>
> -       if (o->index_only || o->reset || !o->update)
> +       if (o->index_only || o->reset == UNPACK_RESET_OVERWRITE_UNTRACKED ||
> +           !o->update)
>                 return 0;
>
>         len = check_leading_path(ce->name, ce_namelen(ce));
> diff --git a/unpack-trees.h b/unpack-trees.h
> index d344d7d296..732b262c4d 100644
> --- a/unpack-trees.h
> +++ b/unpack-trees.h
> @@ -41,9 +41,15 @@ void setup_unpack_trees_porcelain(struct unpack_trees_options *opts,
>   */
>  void clear_unpack_trees_porcelain(struct unpack_trees_options *opts);
>
> +enum unpack_trees_reset_type {
> +       UNPACK_NO_RESET = 0,
> +       UNPACK_RESET_OVERWRITE_UNTRACKED,
> +       UNPACK_RESET_PROTECT_UNTRACKED
> +};
> +
>  struct unpack_trees_options {
> -       unsigned int reset,
> -                    merge,
> +       enum unpack_trees_reset_type reset;
> +       unsigned int merge,
>                      update,
>                      clone,
>                      index_only,
> --
> 2.21.0
>


-- 
Duy

Re: [PATCH 0/2] read-tree: improve untracked file support

[Duy Nguyen]

On Wed, May 1, 2019 at 9:58 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
>
>
>
> On 01/05/2019 11:31, Duy Nguyen wrote:
> > On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
> >>
> >> From: Phillip Wood <phillip.wood@dunelm.org.uk>
> >>
> >> These two patches teach read-tree how to avoid overwriting untracked
> >> files when doing '--reset -u' and also how to respect all of git's
> >> standard excludes files. I'd like to see the porcelain commands stop
> >> overwriting untracked files, this is a first step on the way. I'm not
> >> sure if we want to add options to the porcelain commands to protect
> >> untracked files or just change their behavior and add an option to
> >> override that. I'm leaning towards the latter but I'd be interested to
> >> hear what others think.
> >
> > For new commands like git-restore, it's definitely a good thing to not
> > overwrite untracked files.
>
> I agree, unfortunately this series does not help with git-restore, only
> git-switch. For restore on an index without conflicts I think it could
> just use the pathspec in struct unpack_trees_options and set opts.rest =
> UNPACK_RESET_PROTECT_UNTRACKED but that does not help if we want to
> handle conflicted paths differently to non-conflicted paths.

Right. I got confused. You did mention "git checkout <rev> :/" in 1/2,
which is the same as "git restore --source <rev> --staged --worktree
:/" and  can also potentially overwrite untracked files, even though
it does not use unpack-trees and cannot be fixed with this. Never
mind. Let's leave git-restore out of the discussion for now.

> > For existing commands I guess we have to go
> > over them one by one. For "git reset --hard", it should really just
> > overwrite whatever needed to get back to the known good state. "git
> > checkout -f" , not so sure (seems weird that we need force-level-two
> > option to override the protection provided by -f, if we change default
> > behavior)
>
> I think it's fine for "checkout -f" to overwrite untracked files (and if
> "switch --discard-changes" does not then there is no pressing need to
> add such a mode to checkout), --force is a good name for an option that
> nukes everything that gets in it's way. For "reset --hard" I'm not so
> sure, if I have changes to an untracked file I don't wont them
> overwritten by default. There is no porcelain equivalent to "read-tree
> --reset --protect-untracked -u" and I was hoping "reset --hard" may
> become that porcelain equivalent with a new --force or
> --overwrite-untracked option.

My (biased, obviously) view is that "git reset --hard" is very
dangerous and I'm not trying to change that, especially when its
behavior has been like this since forever and I'm sure it's used in
scripts.

Instead "git restore" should be used when you need "git reset --hard
HEAD", the most often use case. And since it's new, changing default
behavior is not a problem. Which brings us back to git-restore :)

But either way, git-restore or git-reset, I still don't see why
untracked files are more valuable in this case than tracked ones to
change the default. I can see that sometimes you may want to restore
just tracked files, or untracked files, almost like filtering with
pathspec.

> For the various "foo --abort" some (most?) are using "reset --merge"
> which I think declines to overwrite untracked files but rebase uses
> "reset --hard" which I'd like to change to protect untracked files in
> the same way that rebase does for the initial checkout and when picking
> commits. I haven't thought about stash.

Yeah it looks like cherry-pick and revert use "reset --merge" too
(reset_for_rollback function). That's all of them. Probably a stupid
question, why can't rebase just use "rebase --merge" like everybody
else?

>
> Best Wishes
>
> Phillip
>


-- 
Duy

Re: [PATCH 1/2] read-tree --reset: add --protect-untracked

[Phillip Wood]

On 02/05/2019 11:38, Duy Nguyen wrote:
> On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
>>
>> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>>
>> Currently there is no way to get git to discard changes to the
>> worktree without overwriting untracked files. `reset --hard`,
>> `checkout --force`, `checkout <rev> :/` and `read-tree --reset -u`
> 
> "checkout <rev> :/" does not use unpack-trees, I think, so it does not
> belong to this group.

You're right that it does not use unpack-trees, but I'm making the point 
that there is no way to discard changes to tracked files without 
overwriting untracked files whatever one uses.

>> will all overwrite untracked files. unpack_trees() has known how to
>> avoid overwriting untracked files since commit fcc387db9b ("read-tree
>> -m -u: do not overwrite or remove untracked working tree files.",
>> 2006-05-17) in response to concerns about lost files [1] but those
>> protections do not apply to --reset. This patch adds an
>> --protect-untracked option for read-tree/unpack_trees() to apply th
>> same protections with --reset.
> 
> Your enum makes me wonder if we should have --reset-tracked instead of
> --protect-untracked (say what you want to reset seems tiny bit easier
> to understand than "reset except untracked"). Which leads to another
> question, do we ever need --reset-untracked (i.e. overwriting
> untracked files but never ever touch tracked ones).
> 
> I have one use case that --reset-untracked might make sense. Suppose
> you do "git reset HEAD^" where HEAD has some new files. The new files
> will be left alone in worktree of course, but switching back to HEAD
> after that, unpack-trees will cry murder because it would otherwise
> overwrite untracked files (that have the exact same content).
> 
> --reset-untracked might be useful for that. I'm carrying a patch to
> detect identical file content though which addresses this very
> specific case. Maybe it's the only case that --reset-untracked may be
> useful.

That patch sounds useful (I played around with "hg update" the other day 
and I think it does the same). I think that's probably a common case, it 
also works if one does "git checkout <other-branch> file" then "git 
reset HEAD" followed "git checkout <other-branch>". --reset-untracked is 
just like --reset which we can't get rid of without breaking backwards 
compatibility. Having --reset-tracked is maybe better than having a 
--protect-tracked option that only works with --reset -u.

>> It does not change the behavior of any
>> of the porcelain commands but paves the way for adding options or
>> changing the default for those in future.
>>
>> Note the actual change in unpack-trees.c is quite simple, the majority
>> of this patch is converting existing callers to use the new
>> unpack_trees_reset_type enum.
> 
> This could be split in a separate patch, then the actual change would
> become small and more to the point.

Yes that would make it clearer.

Best Wishes

Phillip

> 
>> [1] https://public-inbox.org/git/8aa486160605161500m1dd8428cj@mail.gmail.com/
>>
>> Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
>> ---
>>
>> Notes:
>>      adding --protect-untracked to the invocation of
>>      test_submodule_forced_switch() in t1013 fixes the known breakage of
>>      tests 57 & 58 but breaks test 64 which relies on forcing. I'm not sure
>>      that the expected results of 57 & 58 are correct if we're forcing.
>>
>>   Documentation/git-read-tree.txt | 10 +++++++--
>>   builtin/am.c                    |  8 ++++---
>>   builtin/checkout.c              |  2 +-
>>   builtin/read-tree.c             | 12 +++++++++++
>>   builtin/rebase.c                |  2 +-
>>   builtin/reset.c                 |  2 +-
>>   builtin/stash.c                 |  7 ++++---
>>   t/lib-read-tree.sh              | 11 ++++++++++
>>   t/t1005-read-tree-reset.sh      | 37 +++++++++++++++++++++++++++++++--
>>   unpack-trees.c                  |  3 ++-
>>   unpack-trees.h                  | 10 +++++++--
>>   11 files changed, 88 insertions(+), 16 deletions(-)
>>
>> diff --git a/Documentation/git-read-tree.txt b/Documentation/git-read-tree.txt
>> index d271842608..67864c6bbc 100644
>> --- a/Documentation/git-read-tree.txt
>> +++ b/Documentation/git-read-tree.txt
>> @@ -40,12 +40,18 @@ OPTIONS
>>   --reset::
>>          Same as -m, except that unmerged entries are discarded instead
>>          of failing. When used with `-u`, updates leading to loss of
>> -       working tree changes will not abort the operation.
>> +       working tree changes will not abort the operation and
>> +       untracked files will be overwritten. Use `--protect-untracked`
>> +       to avoid overwriting untracked files.
>>
>>   -u::
>>          After a successful merge, update the files in the work
>>          tree with the result of the merge.
>>
>> +--protect-untracked::
>> +       Prevent `--reset` `-u` from overwriting untracked files. Requires
>> +       `--reset` and `-u` (`-m` never overwrites untracked files).
>> +
>>   -i::
>>          Usually a merge requires the index file as well as the
>>          files in the working tree to be up to date with the
>> @@ -89,7 +95,7 @@ OPTIONS
>>          existed in the original index file.
>>
>>   --exclude-per-directory=<gitignore>::
>> -       When running the command with `-u` and `-m` options, the
>> +       When using `-u` with `-m` or `--reset` `--protect-untracked` options, the
>>          merge result may need to overwrite paths that are not
>>          tracked in the current branch.  The command usually
>>          refuses to proceed with the merge to avoid losing such a
>> diff --git a/builtin/am.c b/builtin/am.c
>> index 912d9821b1..a92394b682 100644
>> --- a/builtin/am.c
>> +++ b/builtin/am.c
>> @@ -1854,7 +1854,8 @@ static void am_resolve(struct am_state *state)
>>    * true, any unmerged entries will be discarded. Returns 0 on success, -1 on
>>    * failure.
>>    */
>> -static int fast_forward_to(struct tree *head, struct tree *remote, int reset)
>> +static int fast_forward_to(struct tree *head, struct tree *remote,
>> +                          enum unpack_trees_reset_type reset)
>>   {
>>          struct lock_file lock_file = LOCK_INIT;
>>          struct unpack_trees_options opts;
>> @@ -1942,7 +1943,8 @@ static int clean_index(const struct object_id *head, const struct object_id *rem
>>
>>          read_cache_unmerged();
>>
>> -       if (fast_forward_to(head_tree, head_tree, 1))
>> +       if (fast_forward_to(head_tree, head_tree,
>> +                           UNPACK_RESET_OVERWRITE_UNTRACKED))
>>                  return -1;
>>
>>          if (write_cache_as_tree(&index, 0, NULL))
>> @@ -1952,7 +1954,7 @@ static int clean_index(const struct object_id *head, const struct object_id *rem
>>          if (!index_tree)
>>                  return error(_("Could not parse object '%s'."), oid_to_hex(&index));
>>
>> -       if (fast_forward_to(index_tree, remote_tree, 0))
>> +       if (fast_forward_to(index_tree, remote_tree, UNPACK_NO_RESET))
>>                  return -1;
>>
>>          if (merge_tree(remote_tree))
>> diff --git a/builtin/checkout.c b/builtin/checkout.c
>> index ffa776c6e1..e9e70018f9 100644
>> --- a/builtin/checkout.c
>> +++ b/builtin/checkout.c
>> @@ -506,7 +506,7 @@ static int reset_tree(struct tree *tree, const struct checkout_opts *o,
>>          opts.head_idx = -1;
>>          opts.update = worktree;
>>          opts.skip_unmerged = !worktree;
>> -       opts.reset = 1;
>> +       opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
>>          opts.merge = 1;
>>          opts.fn = oneway_merge;
>>          opts.verbose_update = o->show_progress;
>> diff --git a/builtin/read-tree.c b/builtin/read-tree.c
>> index 5c9c082595..23735adde9 100644
>> --- a/builtin/read-tree.c
>> +++ b/builtin/read-tree.c
>> @@ -114,6 +114,7 @@ static int git_read_tree_config(const char *var, const char *value, void *cb)
>>   int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
>>   {
>>          int i, stage = 0;
>> +       int protect_untracked = -1;
>>          struct object_id oid;
>>          struct tree_desc t[MAX_UNPACK_TREES];
>>          struct unpack_trees_options opts;
>> @@ -140,6 +141,8 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
>>                    PARSE_OPT_NONEG },
>>                  OPT_BOOL('u', NULL, &opts.update,
>>                           N_("update working tree with merge result")),
>> +               OPT_BOOL(0, "protect-untracked", &protect_untracked,
>> +                        N_("do not overwrite untracked files")),
>>                  { OPTION_CALLBACK, 0, "exclude-per-directory", &opts,
>>                    N_("gitignore"),
>>                    N_("allow explicitly ignored files to be overwritten"),
>> @@ -209,8 +212,17 @@ int cmd_read_tree(int argc, const char **argv, const char *unused_prefix)
>>          if ((opts.update || opts.index_only) && !opts.merge)
>>                  die("%s is meaningless without -m, --reset, or --prefix",
>>                      opts.update ? "-u" : "-i");
>> +       if (protect_untracked >= 0) {
>> +               if (!opts.reset || !opts.update)
>> +                       die("-%s-protect-untracked requires --reset and -u",
>> +                           protect_untracked ? "" : "-no");
>> +               opts.reset = UNPACK_RESET_PROTECT_UNTRACKED;
>> +       }
>>          if ((opts.dir && !opts.update))
>>                  die("--exclude-per-directory is meaningless unless -u");
>> +       if (opts.dir && opts.reset == UNPACK_RESET_OVERWRITE_UNTRACKED)
>> +               warning("--exclude-per-directory without --preserve-untracked "
>> +                       "has no effect");
>>          if (opts.merge && !opts.index_only)
>>                  setup_work_tree();
>>
>> diff --git a/builtin/rebase.c b/builtin/rebase.c
>> index 2e41ad5644..feb30a71f5 100644
>> --- a/builtin/rebase.c
>> +++ b/builtin/rebase.c
>> @@ -398,7 +398,7 @@ static int reset_head(struct object_id *oid, const char *action,
>>          unpack_tree_opts.update = 1;
>>          unpack_tree_opts.merge = 1;
>>          if (!detach_head)
>> -               unpack_tree_opts.reset = 1;
>> +               unpack_tree_opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
>>
>>          if (repo_read_index_unmerged(the_repository) < 0) {
>>                  ret = error(_("could not read index"));
>> diff --git a/builtin/reset.c b/builtin/reset.c
>> index 26ef9a7bd0..a39dd92fb2 100644
>> --- a/builtin/reset.c
>> +++ b/builtin/reset.c
>> @@ -70,7 +70,7 @@ static int reset_index(const struct object_id *oid, int reset_type, int quiet)
>>                  opts.update = 1;
>>                  /* fallthrough */
>>          default:
>> -               opts.reset = 1;
>> +               opts.reset = UNPACK_RESET_OVERWRITE_UNTRACKED;
>>          }
>>
>>          read_cache_unmerged();
>> diff --git a/builtin/stash.c b/builtin/stash.c
>> index 2a8e6d09b4..175d1b62d3 100644
>> --- a/builtin/stash.c
>> +++ b/builtin/stash.c
>> @@ -227,7 +227,8 @@ static int clear_stash(int argc, const char **argv, const char *prefix)
>>          return do_clear_stash();
>>   }
>>
>> -static int reset_tree(struct object_id *i_tree, int update, int reset)
>> +static int reset_tree(struct object_id *i_tree, int update,
>> +                     enum unpack_trees_reset_type reset)
>>   {
>>          int nr_trees = 1;
>>          struct unpack_trees_options opts;
>> @@ -461,7 +462,7 @@ static int do_apply_stash(const char *prefix, struct stash_info *info,
>>          }
>>
>>          if (has_index) {
>> -               if (reset_tree(&index_tree, 0, 0))
>> +               if (reset_tree(&index_tree, 0, UNPACK_NO_RESET))
>>                          return -1;
>>          } else {
>>                  struct strbuf out = STRBUF_INIT;
>> @@ -471,7 +472,7 @@ static int do_apply_stash(const char *prefix, struct stash_info *info,
>>                          return -1;
>>                  }
>>
>> -               if (reset_tree(&c_tree, 0, 1)) {
>> +               if (reset_tree(&c_tree, 0, UNPACK_RESET_OVERWRITE_UNTRACKED)) {
>>                          strbuf_release(&out);
>>                          return -1;
>>                  }
>> diff --git a/t/lib-read-tree.sh b/t/lib-read-tree.sh
>> index b95f485606..22f516d91f 100644
>> --- a/t/lib-read-tree.sh
>> +++ b/t/lib-read-tree.sh
>> @@ -39,3 +39,14 @@ read_tree_u_must_fail () {
>>          test_cmp pre-dry-run-wt post-dry-run-wt &&
>>          test_must_fail git read-tree "$@"
>>   }
>> +
>> +read_tree_u_must_fail_save_err () {
>> +       git ls-files -s >pre-dry-run &&
>> +       git diff-files -p >pre-dry-run-wt &&
>> +       test_must_fail git read-tree -n "$@" &&
>> +       git ls-files -s >post-dry-run &&
>> +       git diff-files -p >post-dry-run-wt &&
>> +       test_cmp pre-dry-run post-dry-run &&
>> +       test_cmp pre-dry-run-wt post-dry-run-wt &&
>> +       test_must_fail git read-tree "$@" 2>actual-err
>> +}
>> diff --git a/t/t1005-read-tree-reset.sh b/t/t1005-read-tree-reset.sh
>> index 83b09e1310..6c9dd6805b 100755
>> --- a/t/t1005-read-tree-reset.sh
>> +++ b/t/t1005-read-tree-reset.sh
>> @@ -19,15 +19,48 @@ test_expect_success 'setup' '
>>          git add df &&
>>          echo content >new &&
>>          git add new &&
>> -       git commit -m two
>> +       git commit -m two &&
>> +       git ls-files >expect-two
>>   '
>>
>> -test_expect_success 'reset should work' '
>> +test_expect_success '--protect-untracked option sanity checks' '
>> +       read_tree_u_must_fail --reset --protect-untracked HEAD &&
>> +       read_tree_u_must_fail --reset --no-protect-untracked HEAD &&
>> +       read_tree_u_must_fail -m -u --protect-untracked HEAD &&
>> +       read_tree_u_must_fail -m -u --no-protect-untracked
>> +'
>> +
>> +test_expect_success 'reset should reset worktree' '
>> +       echo changed >df &&
>>          read_tree_u_must_succeed -u --reset HEAD^ &&
>>          git ls-files >actual &&
>>          test_cmp expect actual
>>   '
>>
>> +test_expect_success 'reset --protect-untracked protects untracked file' '
>> +       echo changed >new &&
>> +       read_tree_u_must_fail_save_err -u --reset --protect-untracked HEAD &&
>> +       echo "error: Untracked working tree file '\'new\'' would be overwritten by merge." >expected-err &&
>> +       test_cmp expected-err actual-err
>> +'
>> +
>> +test_expect_success 'reset --protect-untracked protects untracked directory' '
>> +       rm new &&
>> +       mkdir new &&
>> +       echo untracked >new/untracked &&
>> +       read_tree_u_must_fail_save_err -u --reset --protect-untracked HEAD &&
>> +       echo "error: Updating '\'new\'' would lose untracked files in it" >expected-err &&
>> +       test_cmp expected-err actual-err
>> +'
>> +
>> +test_expect_success 'reset --protect-untracked resets' '
>> +       rm -rf new &&
>> +       echo changed >df/file &&
>> +       read_tree_u_must_succeed -u --reset --protect-untracked HEAD &&
>> +       git ls-files >actual-two &&
>> +       test_cmp expect-two actual-two
>> +'
>> +
>>   test_expect_success 'reset should remove remnants from a failed merge' '
>>          read_tree_u_must_succeed --reset -u HEAD &&
>>          git ls-files -s >expect &&
>> diff --git a/unpack-trees.c b/unpack-trees.c
>> index 50189909b8..b1722730fe 100644
>> --- a/unpack-trees.c
>> +++ b/unpack-trees.c
>> @@ -1917,7 +1917,8 @@ static int verify_absent_1(const struct cache_entry *ce,
>>          int len;
>>          struct stat st;
>>
>> -       if (o->index_only || o->reset || !o->update)
>> +       if (o->index_only || o->reset == UNPACK_RESET_OVERWRITE_UNTRACKED ||
>> +           !o->update)
>>                  return 0;
>>
>>          len = check_leading_path(ce->name, ce_namelen(ce));
>> diff --git a/unpack-trees.h b/unpack-trees.h
>> index d344d7d296..732b262c4d 100644
>> --- a/unpack-trees.h
>> +++ b/unpack-trees.h
>> @@ -41,9 +41,15 @@ void setup_unpack_trees_porcelain(struct unpack_trees_options *opts,
>>    */
>>   void clear_unpack_trees_porcelain(struct unpack_trees_options *opts);
>>
>> +enum unpack_trees_reset_type {
>> +       UNPACK_NO_RESET = 0,
>> +       UNPACK_RESET_OVERWRITE_UNTRACKED,
>> +       UNPACK_RESET_PROTECT_UNTRACKED
>> +};
>> +
>>   struct unpack_trees_options {
>> -       unsigned int reset,
>> -                    merge,
>> +       enum unpack_trees_reset_type reset;
>> +       unsigned int merge,
>>                       update,
>>                       clone,
>>                       index_only,
>> --
>> 2.21.0
>>
> 
> 

Re: [PATCH 0/2] read-tree: improve untracked file support

[Phillip Wood]

On 02/05/2019 11:53, Duy Nguyen wrote:
> On Wed, May 1, 2019 at 9:58 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
>> On 01/05/2019 11:31, Duy Nguyen wrote:
>>> On Wed, May 1, 2019 at 5:14 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
>>>>
>>>> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>>>>
>>>> These two patches teach read-tree how to avoid overwriting untracked
>>>> files when doing '--reset -u' and also how to respect all of git's
>>>> standard excludes files. I'd like to see the porcelain commands stop
>>>> overwriting untracked files, this is a first step on the way. I'm not
>>>> sure if we want to add options to the porcelain commands to protect
>>>> untracked files or just change their behavior and add an option to
>>>> override that. I'm leaning towards the latter but I'd be interested to
>>>> hear what others think.
>>>
>>> For new commands like git-restore, it's definitely a good thing to not
>>> overwrite untracked files.
>>
>> I agree, unfortunately this series does not help with git-restore, only
>> git-switch. For restore on an index without conflicts I think it could
>> just use the pathspec in struct unpack_trees_options and set opts.rest =
>> UNPACK_RESET_PROTECT_UNTRACKED but that does not help if we want to
>> handle conflicted paths differently to non-conflicted paths.
> 
> Right. I got confused. You did mention "git checkout <rev> :/" in 1/2,
> which is the same as "git restore --source <rev> --staged --worktree
> :/" and  can also potentially overwrite untracked files, even though
> it does not use unpack-trees and cannot be fixed with this. Never
> mind. Let's leave git-restore out of the discussion for now.
> 
>>> For existing commands I guess we have to go
>>> over them one by one. For "git reset --hard", it should really just
>>> overwrite whatever needed to get back to the known good state. "git
>>> checkout -f" , not so sure (seems weird that we need force-level-two
>>> option to override the protection provided by -f, if we change default
>>> behavior)
>>
>> I think it's fine for "checkout -f" to overwrite untracked files (and if
>> "switch --discard-changes" does not then there is no pressing need to
>> add such a mode to checkout), --force is a good name for an option that
>> nukes everything that gets in it's way. For "reset --hard" I'm not so
>> sure, if I have changes to an untracked file I don't wont them
>> overwritten by default. There is no porcelain equivalent to "read-tree
>> --reset --protect-untracked -u" and I was hoping "reset --hard" may
>> become that porcelain equivalent with a new --force or
>> --overwrite-untracked option.
> 
> My (biased, obviously) view is that "git reset --hard" is very
> dangerous and I'm not trying to change that, especially when its
> behavior has been like this since forever and I'm sure it's used in
> scripts.
> 
> Instead "git restore" should be used when you need "git reset --hard
> HEAD", the most often use case. And since it's new, changing default
> behavior is not a problem. Which brings us back to git-restore :)

Does restore clean up the branch state like reset? It's tricky because 
you only want to do that if there is no pathspec (or the pathspec is :/ 
or equivalent - I can't remember if restore always requires paths or not)

> But either way, git-restore or git-reset, I still don't see why
> untracked files are more valuable in this case than tracked ones to
> change the default. 

My issue is that is easy to see what changes you're going to lose in 
tracked files by running diff. For untracked files diff just says a new 
file will be created, it ignores the current contents as the path is in 
the index so it is easy to overwrite changes without realizing. There's 
also a philosophical point that git should not be stomping on paths that 
it is not tracking though that's a bit moot if a path is tracked in one 
revision but not another.

> I can see that sometimes you may want to restore
> just tracked files, or untracked files, almost like filtering with
> pathspec.
> 
>> For the various "foo --abort" some (most?) are using "reset --merge"
>> which I think declines to overwrite untracked files but rebase uses
>> "reset --hard" which I'd like to change to protect untracked files in
>> the same way that rebase does for the initial checkout and when picking
>> commits. I haven't thought about stash.
> 
> Yeah it looks like cherry-pick and revert use "reset --merge" too
> (reset_for_rollback function). That's all of them. Probably a stupid
> question, why can't rebase just use "rebase --merge" like everybody
> else?

I'm not sure - if --merge works for the others I can't see why it 
shouldn't work for rebase as well.

Best Wishes

Phillip
>> Best Wishes
>>
>> Phillip
>>
> 
> 

Re: [PATCH 0/2] read-tree: improve untracked file support

[Duy Nguyen]

On Tue, May 7, 2019 at 5:02 PM Phillip Wood <phillip.wood123@gmail.com> wrote:
> > My (biased, obviously) view is that "git reset --hard" is very
> > dangerous and I'm not trying to change that, especially when its
> > behavior has been like this since forever and I'm sure it's used in
> > scripts.
> >
> > Instead "git restore" should be used when you need "git reset --hard
> > HEAD", the most often use case. And since it's new, changing default
> > behavior is not a problem. Which brings us back to git-restore :)
>
> Does restore clean up the branch state like reset? It's tricky because
> you only want to do that if there is no pathspec (or the pathspec is :/
> or equivalent - I can't remember if restore always requires paths or not)

Nope. git-restore cares about files, not branches. Yes git-restore
always requires paths, just in case people type "git restore" and
expect to see help usage or something.

> > But either way, git-restore or git-reset, I still don't see why
> > untracked files are more valuable in this case than tracked ones to
> > change the default.
>
> My issue is that is easy to see what changes you're going to lose in
> tracked files by running diff. For untracked files diff just says a new
> file will be created, it ignores the current contents as the path is in
> the index so it is easy to overwrite changes without realizing. There's
> also a philosophical point that git should not be stomping on paths that
> it is not tracking though that's a bit moot if a path is tracked in one
> revision but not another.

Ah good point about diff. If only we had "git reset --dry-run" (that
shows the diff, including untracked files; or perhaps --diff would be
a better name for that imaginary option)
-- 
Duy