[GSoC Patch 0/2] add unicode support to git repo structure

[GSoC Patch 0/2] add unicode support to git repo structure

[K Jayatheerth]

Currently, `git repo structure` always renders its table output using plain
ASCII characters (`|`, `-`, `*`), regardless of the user's locale. This
series introduces Unicode box-drawing characters (`│`, `├`, `─`, `┼`, `┤`)
and bullet points (`•`) when a UTF-8 locale is detected, providing a cleaner
and more visually distinct hierarchical output on modern terminals while
gracefully falling back to the existing ASCII formatting otherwise.

Summary of changes:

- Patch 1 (gettext: fall back to env-derived charset when unset):
  Fixes an issue where `is_utf8_locale()` would leave `charset` unset (NULL)
  when `git_setup_gettext()` returns early upon missing the locale directory
  (such as in uninstalled development builds). Because `is_encoding_utf8(NULL)`
  defaults to 1, `is_utf8_locale()` would mistakenly report a UTF-8 locale
  even under non-UTF-8 environments like `LC_ALL=C`. This patch enables the
  environment-derived fallback (`LC_ALL`, `LC_CTYPE`, `LANG`) whenever `charset`
  is unset, regardless of `NO_GETTEXT`.

- Patch 2 (repo: add Unicode support for `repo structure` output):
  Updates `builtin/repo.c` (`stats_table_setup_structure()` and
  `stats_table_print_structure()`) to query `is_utf8_locale()` and emit
  Unicode box-drawing borders and bullets when running under a UTF-8 locale,
  falling back to ASCII when not in UTF-8. In `t/t1901-repo-structure.sh`,
  tests follow the established `t/lib-git-svn.sh` convention for discovering
  `GIT_TEST_UTF8_LOCALE` to define a `UTF8_LOCALE` test prerequisite, existing
  tests are pinned to `LC_ALL=C` to reliably test the ASCII fallback path,
  and new tests are added to verify the UTF-8 output rendering (including
  when running with a missing locale directory).

K Jayatheerth (2):
  gettext: fall back to env-derived charset when unset
  repo: add Unicode support for `repo structure` output

 builtin/repo.c            | 110 ++++++++++++++++-------------
 gettext.c                 |  32 +++++----
 t/t1901-repo-structure.sh | 141 +++++++++++++++++++++++++++++++++++++-
 3 files changed, 221 insertions(+), 62 deletions(-)

-- 
2.55.GIT

[GSoC Patch 1/2] gettext: fall back to env-derived charset when unset

[K Jayatheerth]

`is_utf8_locale()` relies on the static `charset` variable, which is
normally initialized by `init_gettext_charset()`. That initialization
only happens when `git_setup_gettext()` successfully locates the locale
directory.

When running directly from the source tree without `make install`, or in
other environments where the locale directory is unavailable,
`git_setup_gettext()` returns early, leaving `charset` unset (NULL).
Because `is_encoding_utf8(NULL)` defaults to 1, `is_utf8_locale()` would
mistakenly report a UTF-8 locale even in non-UTF-8 environments (e.g.
under `LC_ALL=C`).

The fallback that derives the charset from `LC_ALL`, `LC_CTYPE`, or
`LANG` was previously compiled only under `NO_GETTEXT`. That left
gettext-enabled builds without a fallback when `charset` remains
uninitialized.

Make the fallback conditional on `charset` being unset rather than on
`NO_GETTEXT`. This ensures `is_utf8_locale()` accurately inspects the
environment-derived charset regardless of whether gettext support is
enabled.

Mentored-by: Justin Tobler <jltobler@gmail.com>
Mentored-by: Lucas Seiki Oshiro <lucasseikioshiro@gmail.com>
Signed-off-by: K Jayatheerth <jayatheerthkulkarni2005@gmail.com>
---
 gettext.c | 32 ++++++++++++++++++--------------
 1 file changed, 18 insertions(+), 14 deletions(-)

diff --git a/gettext.c b/gettext.c
index 8d08a61f84..5376a0de0f 100644
--- a/gettext.c
+++ b/gettext.c
@@ -141,19 +141,23 @@ int gettext_width(const char *s)
 
 int is_utf8_locale(void)
 {
-#ifdef NO_GETTEXT
-	if (!charset) {
-		const char *env = getenv("LC_ALL");
-		if (!env || !*env)
-			env = getenv("LC_CTYPE");
-		if (!env || !*env)
-			env = getenv("LANG");
-		if (!env)
-			env = "";
-		if (strchr(env, '.'))
-			env = strchr(env, '.') + 1;
-		charset = xstrdup(env);
+	const char *c = charset;
+
+	if (!c) {
+		static char fallback_charset[64];
+		if (!*fallback_charset) {
+			const char *env = getenv("LC_ALL");
+			if (!env || !*env)
+				env = getenv("LC_CTYPE");
+			if (!env || !*env)
+				env = getenv("LANG");
+			if (!env)
+				env = "";
+			if (strchr(env, '.'))
+				env = strchr(env, '.') + 1;
+			strlcpy(fallback_charset, env, sizeof(fallback_charset));
+		}
+		c = fallback_charset;
 	}
-#endif
-	return is_encoding_utf8(charset);
+	return is_encoding_utf8(c);
 }
-- 
2.55.GIT

[GSoC Patch 2/2] repo: add Unicode support for `repo structure` output

[K Jayatheerth]

ASCII output ignores locale support for UTF-8. Use box-drawing
characters and bullets when the locale supports UTF-8, since modern
terminals render tables cleanly this way, and fall back to ASCII
otherwise.

Tests now discover an available UTF-8 locale to set a `UTF8_LOCALE`
prerequisite, and existing ASCII table tests are explicitly pinned
to `LC_ALL=C` so their behavior remains deterministic regardless of the
runner's environment.

Mentored-by: Justin Tobler <jltobler@gmail.com>
Mentored-by: Lucas Seiki Oshiro <lucasseikioshiro@gmail.com>
Signed-off-by: K Jayatheerth <jayatheerthkulkarni2005@gmail.com>
---
 builtin/repo.c            | 110 ++++++++++++++++-------------
 t/t1901-repo-structure.sh | 141 +++++++++++++++++++++++++++++++++++++-
 2 files changed, 203 insertions(+), 48 deletions(-)

diff --git a/builtin/repo.c b/builtin/repo.c
index 84e012f83f..2c1cca6f2e 100644
--- a/builtin/repo.c
+++ b/builtin/repo.c
@@ -498,82 +498,86 @@ static void stats_table_setup_structure(struct stats_table *table,
 	size_t object_count_total;
 	size_t disk_object_total;
 	size_t ref_total;
+	const int utf8 = is_utf8_locale();
+	const char *bullet_l0 = utf8 ? "•" : "*";
+	const char *bullet_l1 = utf8 ? "  •" : "  *";
+	const char *bullet_l2 = utf8 ? "    •" : "    *";
 
 	ref_total = get_total_reference_count(refs);
-	stats_table_addf(table, "* %s", _("References"));
-	stats_table_count_addf(table, ref_total, "  * %s", _("Count"));
-	stats_table_count_addf(table, refs->branches, "    * %s", _("Branches"));
-	stats_table_count_addf(table, refs->tags, "    * %s", _("Tags"));
-	stats_table_count_addf(table, refs->remotes, "    * %s", _("Remotes"));
-	stats_table_count_addf(table, refs->others, "    * %s", _("Others"));
+	stats_table_addf(table, "%s %s", bullet_l0, _("References"));
+	stats_table_count_addf(table, ref_total, "%s %s", bullet_l1, _("Count"));
+	stats_table_count_addf(table, refs->branches, "%s %s", bullet_l2, _("Branches"));
+	stats_table_count_addf(table, refs->tags, "%s %s", bullet_l2, _("Tags"));
+	stats_table_count_addf(table, refs->remotes, "%s %s", bullet_l2, _("Remotes"));
+	stats_table_count_addf(table, refs->others, "%s %s", bullet_l2, _("Others"));
 
 	object_count_total = get_total_object_values(&objects->type_counts);
 	stats_table_addf(table, "");
-	stats_table_addf(table, "* %s", _("Reachable objects"));
-	stats_table_count_addf(table, object_count_total, "  * %s", _("Count"));
+	stats_table_addf(table, "%s %s", bullet_l0, _("Reachable objects"));
+	stats_table_count_addf(table, object_count_total, "%s %s", bullet_l1, _("Count"));
 	stats_table_count_addf(table, objects->type_counts.commits,
-			       "    * %s", _("Commits"));
+			       "%s %s", bullet_l2, _("Commits"));
 	stats_table_count_addf(table, objects->type_counts.trees,
-			       "    * %s", _("Trees"));
+			       "%s %s", bullet_l2, _("Trees"));
 	stats_table_count_addf(table, objects->type_counts.blobs,
-			       "    * %s", _("Blobs"));
+			       "%s %s", bullet_l2, _("Blobs"));
 	stats_table_count_addf(table, objects->type_counts.tags,
-			       "    * %s", _("Tags"));
+			       "%s %s", bullet_l2, _("Tags"));
 
 	inflated_object_total = get_total_object_values(&objects->inflated_sizes);
 	stats_table_size_addf(table, inflated_object_total,
-			      "  * %s", _("Inflated size"));
+			      "%s %s", bullet_l1, _("Inflated size"));
 	stats_table_size_addf(table, objects->inflated_sizes.commits,
-			      "    * %s", _("Commits"));
+			      "%s %s", bullet_l2, _("Commits"));
 	stats_table_size_addf(table, objects->inflated_sizes.trees,
-			      "    * %s", _("Trees"));
+			      "%s %s", bullet_l2, _("Trees"));
 	stats_table_size_addf(table, objects->inflated_sizes.blobs,
-			      "    * %s", _("Blobs"));
+			      "%s %s", bullet_l2, _("Blobs"));
 	stats_table_size_addf(table, objects->inflated_sizes.tags,
-			      "    * %s", _("Tags"));
+			      "%s %s", bullet_l2, _("Tags"));
 
 	disk_object_total = get_total_object_values(&objects->disk_sizes);
 	stats_table_size_addf(table, disk_object_total,
-			      "  * %s", _("Disk size"));
+			      "%s %s", bullet_l1, _("Disk size"));
 	stats_table_size_addf(table, objects->disk_sizes.commits,
-			      "    * %s", _("Commits"));
+			      "%s %s", bullet_l2, _("Commits"));
 	stats_table_size_addf(table, objects->disk_sizes.trees,
-			      "    * %s", _("Trees"));
+			      "%s %s", bullet_l2, _("Trees"));
 	stats_table_size_addf(table, objects->disk_sizes.blobs,
-			      "    * %s", _("Blobs"));
+			      "%s %s", bullet_l2, _("Blobs"));
 	stats_table_size_addf(table, objects->disk_sizes.tags,
-			      "    * %s", _("Tags"));
+			      "%s %s", bullet_l2, _("Tags"));
 
 	stats_table_addf(table, "");
-	stats_table_addf(table, "* %s", _("Largest objects"));
-	stats_table_addf(table, "  * %s", _("Commits"));
+	stats_table_addf(table, "%s %s", bullet_l0, _("Largest objects"));
+	stats_table_addf(table, "%s %s", bullet_l1, _("Commits"));
 	stats_table_object_size_addf(table,
 				     &objects->largest.commit_size.oid,
 				     objects->largest.commit_size.value,
-				     "    * %s", _("Maximum size"));
+				     "%s %s", bullet_l2, _("Maximum size"));
 	stats_table_object_count_addf(table,
 				      &objects->largest.parent_count.oid,
 				      objects->largest.parent_count.value,
-				      "    * %s", _("Maximum parents"));
-	stats_table_addf(table, "  * %s", _("Trees"));
+				      "%s %s", bullet_l2, _("Maximum parents"));
+	stats_table_addf(table, "%s %s", bullet_l1, _("Trees"));
 	stats_table_object_size_addf(table,
 				     &objects->largest.tree_size.oid,
 				     objects->largest.tree_size.value,
-				     "    * %s", _("Maximum size"));
+				     "%s %s", bullet_l2, _("Maximum size"));
 	stats_table_object_count_addf(table,
 				      &objects->largest.tree_entries.oid,
 				      objects->largest.tree_entries.value,
-				      "    * %s", _("Maximum entries"));
-	stats_table_addf(table, "  * %s", _("Blobs"));
+				      "%s %s", bullet_l2, _("Maximum entries"));
+	stats_table_addf(table, "%s %s", bullet_l1, _("Blobs"));
 	stats_table_object_size_addf(table,
 				     &objects->largest.blob_size.oid,
 				     objects->largest.blob_size.value,
-				     "    * %s", _("Maximum size"));
-	stats_table_addf(table, "  * %s", _("Tags"));
+				     "%s %s", bullet_l2, _("Maximum size"));
+	stats_table_addf(table, "%s %s", bullet_l1, _("Tags"));
 	stats_table_object_size_addf(table,
 				     &objects->largest.tag_size.oid,
 				     objects->largest.tag_size.value,
-				     "    * %s", _("Maximum size"));
+				     "%s %s", bullet_l2, _("Maximum size"));
 }
 
 #define INDEX_WIDTH 4
@@ -589,28 +593,42 @@ static void stats_table_print_structure(const struct stats_table *table)
 	int unit_col_width = table->unit_col_width;
 	struct string_list_item *item;
 	struct strbuf buf = STRBUF_INIT;
+	const int utf8 = is_utf8_locale();
+	const char *border_left = utf8 ? "│ " : "| ";
+	const char *border_mid = utf8 ? " │ " : " | ";
+	const char *border_right = utf8 ? " │" : " |";
 
 	if (title_name_width > name_col_width)
 		name_col_width = title_name_width;
 	if (title_value_width > value_col_width + unit_col_width + 1)
 		value_col_width = title_value_width - unit_col_width;
 
-	strbuf_addstr(&buf, "| ");
+	strbuf_addstr(&buf, border_left);
 	strbuf_utf8_align(&buf, ALIGN_LEFT, name_col_width + INDEX_WIDTH,
 			  name_col_title);
-	strbuf_addstr(&buf, " | ");
+	strbuf_addstr(&buf, border_mid);
 	strbuf_utf8_align(&buf, ALIGN_LEFT,
 			  value_col_width + unit_col_width + 1, value_col_title);
-	strbuf_addstr(&buf, " |");
+	strbuf_addstr(&buf, border_right);
 	printf("%s\n", buf.buf);
 
-	printf("| ");
-	for (int i = 0; i < name_col_width + INDEX_WIDTH; i++)
-		putchar('-');
-	printf(" | ");
-	for (int i = 0; i < value_col_width + unit_col_width + 1; i++)
-		putchar('-');
-	printf(" |\n");
+	if (utf8) {
+		printf("├─");
+		for (int i = 0; i < name_col_width + INDEX_WIDTH; i++)
+			printf("─");
+		printf("─┼─");
+		for (int i = 0; i < value_col_width + unit_col_width + 1; i++)
+			printf("─");
+		printf("─┤\n");
+	} else {
+		printf("| ");
+		for (int i = 0; i < name_col_width + INDEX_WIDTH; i++)
+			putchar('-');
+		printf(" | ");
+		for (int i = 0; i < value_col_width + unit_col_width + 1; i++)
+			putchar('-');
+		printf(" |\n");
+	}
 
 	for_each_string_list_item(item, &table->rows) {
 		struct stats_table_entry *entry = item->util;
@@ -624,7 +642,7 @@ static void stats_table_print_structure(const struct stats_table *table)
 		}
 
 		strbuf_reset(&buf);
-		strbuf_addstr(&buf, "| ");
+		strbuf_addstr(&buf, border_left);
 		strbuf_utf8_align(&buf, ALIGN_LEFT, name_col_width, item->string);
 
 		if (entry && entry->oid)
@@ -633,11 +651,11 @@ static void stats_table_print_structure(const struct stats_table *table)
 		else
 			strbuf_addchars(&buf, ' ', INDEX_WIDTH);
 
-		strbuf_addstr(&buf, " | ");
+		strbuf_addstr(&buf, border_mid);
 		strbuf_utf8_align(&buf, ALIGN_RIGHT, value_col_width, value);
 		strbuf_addch(&buf, ' ');
 		strbuf_utf8_align(&buf, ALIGN_LEFT, unit_col_width, unit);
-		strbuf_addstr(&buf, " |");
+		strbuf_addstr(&buf, border_right);
 		printf("%s\n", buf.buf);
 	}
 
diff --git a/t/t1901-repo-structure.sh b/t/t1901-repo-structure.sh
index 02cc2b594a..926db09175 100755
--- a/t/t1901-repo-structure.sh
+++ b/t/t1901-repo-structure.sh
@@ -4,6 +4,29 @@ test_description='test git repo structure'
 
 . ./test-lib.sh
 
+# Detect if a UTF-8 locale is available on the test system.
+if test -z "$GIT_TEST_UTF8_LOCALE"
+then
+	case "${LC_ALL:-$LANG}" in
+	*.[Uu][Tt][Ff]8 | *.[Uu][Tt][Ff]-8)
+		GIT_TEST_UTF8_LOCALE="${LC_ALL:-$LANG}"
+		;;
+	*)
+		if type locale >/dev/null 2>&1
+		then
+			GIT_TEST_UTF8_LOCALE=$(locale -a 2>/dev/null | sed -n '/\.[uU][tT][fF]-*8$/{
+				p
+				q
+			}')
+		fi
+		;;
+	esac
+fi
+if test -n "$GIT_TEST_UTF8_LOCALE"
+then
+	test_set_prereq UTF8_LOCALE
+fi
+
 object_type_disk_usage() {
 	disk_usage_opt="--disk-usage"
 
@@ -66,7 +89,10 @@ test_expect_success 'empty repository' '
 		|     * Maximum size        |    0 B |
 		EOF
 
-		git repo structure >out 2>err &&
+		# Force a non-UTF8 locale so this test always exercises the
+		# ASCII fallback formatting, regardless of what locale the
+		# runner defaults to.
+		LC_ALL=C git repo structure >out 2>err &&
 
 		test_cmp expect out &&
 		test_line_count = 0 err
@@ -137,7 +163,10 @@ test_expect_success SHA1 'repository with references and objects' '
 		[6] 4dae4f5954f5e6feb3577cfb1b181daa3fd3afd2
 		EOF
 
-		git repo structure >out 2>err &&
+		# Force a non-UTF8 locale so this test always exercises the
+		# ASCII fallback formatting, regardless of what locale the
+		# runner defaults to.
+		LC_ALL=C git repo structure >out 2>err &&
 
 		test_cmp expect out &&
 		test_line_count = 0 err
@@ -230,4 +259,112 @@ test_expect_success 'git repo structure -h shows only repo structure usage' '
 	test_grep ! "git repo info" actual
 '
 
+test_expect_success UTF8_LOCALE 'unicode output under UTF-8 locale with missing locale dir' '
+	test_when_finished "rm -rf repo" &&
+	git init repo &&
+	(
+		cd repo &&
+		cat >expect <<-\EOF &&
+		│ Repository structure      │ Value  │
+		├───────────────────────────┼────────┤
+		│ • References              │        │
+		│   • Count                 │    0   │
+		│     • Branches            │    0   │
+		│     • Tags                │    0   │
+		│     • Remotes             │    0   │
+		│     • Others              │    0   │
+		│                           │        │
+		│ • Reachable objects       │        │
+		│   • Count                 │    0   │
+		│     • Commits             │    0   │
+		│     • Trees               │    0   │
+		│     • Blobs               │    0   │
+		│     • Tags                │    0   │
+		│   • Inflated size         │    0 B │
+		│     • Commits             │    0 B │
+		│     • Trees               │    0 B │
+		│     • Blobs               │    0 B │
+		│     • Tags                │    0 B │
+		│   • Disk size             │    0 B │
+		│     • Commits             │    0 B │
+		│     • Trees               │    0 B │
+		│     • Blobs               │    0 B │
+		│     • Tags                │    0 B │
+		│                           │        │
+		│ • Largest objects         │        │
+		│   • Commits               │        │
+		│     • Maximum size        │    0 B │
+		│     • Maximum parents     │    0   │
+		│   • Trees                 │        │
+		│     • Maximum size        │    0 B │
+		│     • Maximum entries     │    0   │
+		│   • Blobs                 │        │
+		│     • Maximum size        │    0 B │
+		│   • Tags                  │        │
+		│     • Maximum size        │    0 B │
+		EOF
+		# Point GIT_TEXTDOMAINDIR at a nonexistent path so
+		# git_setup_gettext() takes its early-return path (its
+		# locale-directory check fails) and never populates the
+		# gettext-internal charset. This exercises the
+		# is_utf8_locale() fallback that derives the charset
+		# directly from LC_ALL/LC_CTYPE/LANG instead, regardless of
+		# whether gettext itself was able to initialize.
+		GIT_TEXTDOMAINDIR="$TRASH_DIRECTORY/nonexistent-locale-dir" \
+		LC_ALL="$GIT_TEST_UTF8_LOCALE" git repo structure >out 2>err &&
+		test_cmp expect out &&
+		test_line_count = 0 err
+	)
+'
+
+test_expect_success UTF8_LOCALE 'unicode output under UTF-8 locale' '
+	test_when_finished "rm -rf repo" &&
+	git init repo &&
+	(
+		cd repo &&
+		cat >expect <<-\EOF &&
+		│ Repository structure      │ Value  │
+		├───────────────────────────┼────────┤
+		│ • References              │        │
+		│   • Count                 │    0   │
+		│     • Branches            │    0   │
+		│     • Tags                │    0   │
+		│     • Remotes             │    0   │
+		│     • Others              │    0   │
+		│                           │        │
+		│ • Reachable objects       │        │
+		│   • Count                 │    0   │
+		│     • Commits             │    0   │
+		│     • Trees               │    0   │
+		│     • Blobs               │    0   │
+		│     • Tags                │    0   │
+		│   • Inflated size         │    0 B │
+		│     • Commits             │    0 B │
+		│     • Trees               │    0 B │
+		│     • Blobs               │    0 B │
+		│     • Tags                │    0 B │
+		│   • Disk size             │    0 B │
+		│     • Commits             │    0 B │
+		│     • Trees               │    0 B │
+		│     • Blobs               │    0 B │
+		│     • Tags                │    0 B │
+		│                           │        │
+		│ • Largest objects         │        │
+		│   • Commits               │        │
+		│     • Maximum size        │    0 B │
+		│     • Maximum parents     │    0   │
+		│   • Trees                 │        │
+		│     • Maximum size        │    0 B │
+		│     • Maximum entries     │    0   │
+		│   • Blobs                 │        │
+		│     • Maximum size        │    0 B │
+		│   • Tags                  │        │
+		│     • Maximum size        │    0 B │
+		EOF
+		LC_ALL="$GIT_TEST_UTF8_LOCALE" git repo structure >out 2>err &&
+		test_cmp expect out &&
+		test_line_count = 0 err
+	)
+'
+
 test_done
-- 
2.55.GIT

Re: [GSoC Patch 0/2] add unicode support to git repo structure

[Junio C Hamano]

K Jayatheerth <jayatheerthkulkarni2005@gmail.com> writes:

> Currently, `git repo structure` always renders its table output using plain
> ASCII characters (`|`, `-`, `*`), regardless of the user's locale. This
> series introduces Unicode box-drawing characters (`│`, `├`, `─`, `┼`, `┤`)
> and bullet points (`•`) when a UTF-8 locale is detected, providing a cleaner
> and more visually distinct hierarchical output on modern terminals while
> gracefully falling back to the existing ASCII formatting otherwise.

Generally speaking, Unicode box-drawing characters do not work as
ASCII art components as well as they should, because terminals often
do not agree on how wide they should be rendered.

So I am not very enthusiastic about this change.