git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: Git clone and case sensitivity

From
Duy Nguyen <pclouds@gmail.com>
Date
Jul 28, 2018, 04:36 UTC
Message-ID
<20180728043559.GA29185@duynguyen.home>
In-Reply-To
<20180727205909.GC376343@genre.crustytoothpaste.net>
On Fri, Jul 27, 2018 at 08:59:09PM +0000, brian m. carlson wrote:
Show 15 quoted lines
> On Fri, Jul 27, 2018 at 11:59:33AM +0200, Paweł Paruzel wrote:
> > Hi,
> > 
> > Lately, I have been wondering why my test files in repo are modified
> > after I clone it. It turned out to be two files: boolStyle_t_f and
> > boolStyle_T_F.
> > The system that pushed those files was case sensitive while my mac
> > after High Sierra update had APFS which is by default
> > case-insensitive. I highly suggest that git clone threw an exception
> > when files are case sensitive and being cloned to a case insensitive
> > system. This has caused problems with overriding files for test cases
> > without any warning.
> 
> If we did what you proposed, it would be impossible to clone such a
> repository on a case-insensitive system.

I agree throwing a real exception would be bad. But how about detecting the problem and trying our best to keep the repo in somewhat usable state like this?

This patch uses sparse checkout to hide all those paths that we fail to checkout, so you can still have a clean worktree to do things, as long as you don't touch those paths.

-- 8< --
diff --git a/builtin/clone.c b/builtin/clone.c
index 1d939af9d8..a6b5e2c948 100644
--- a/builtin/clone.c
+++ b/builtin/clone.c
@@ -711,6 +711,30 @@ static void update_head(const struct ref *our, const struct ref *remote,
 	}
 }
 
+static int enable_sparse_checkout_on_icase_fs(struct index_state *istate)
+{
+	int i;
+	int skip_count = 0;
+	FILE *fp = fopen(git_path("info/sparse"), "a+");
+
+	for (i = 0; i < istate->cache_nr; i++) {
+		struct cache_entry *ce = istate->cache[i];
+		if (!ce_skip_worktree(ce))
+			continue;
+		if (!skip_count) {
+			git_config_set_multivar_gently("core.sparseCheckout",
+						       "true",
+						       CONFIG_REGEX_NONE, 0);
+			fprintf(fp, "# List of paths hidden by 'git clone'\n");
+		}
+		fprintf(fp, "/%s\n", ce->name);
+		skip_count++;
+	}
+	fclose(fp);
+
+	return skip_count;
+}
+
 static int checkout(int submodule_progress)
 {
 	struct object_id oid;
@@ -751,6 +775,7 @@ static int checkout(int submodule_progress)
 	opts.verbose_update = (option_verbosity >= 0);
 	opts.src_index = &the_index;
 	opts.dst_index = &the_index;
+	opts.clone_checkout = 1;
 
 	tree = parse_tree_indirect(&oid);
 	parse_tree(tree);
@@ -761,6 +786,12 @@ static int checkout(int submodule_progress)
 	if (write_locked_index(&the_index, &lock_file, COMMIT_LOCK))
 		die(_("unable to write new index file"));
 
+	if (enable_sparse_checkout_on_icase_fs(&the_index))
+		warning("Paths that differ only in case are detected "
+			"and will not work correctly on this case-insensitive "
+			"filesystem. Sparse checkout has been enabled to hide "
+			"these paths.");
+
 	err |= run_hook_le(NULL, "post-checkout", sha1_to_hex(null_sha1),
 			   oid_to_hex(&oid), "1", NULL);
 
diff --git a/cache.h b/cache.h
index 8b447652a7..9ecf7ad952 100644
--- a/cache.h
+++ b/cache.h
@@ -1455,6 +1455,7 @@ struct checkout {
 	unsigned force:1,
 		 quiet:1,
 		 not_new:1,
+		 set_skipworktree_on_updated:1,
 		 refresh_cache:1;
 };
 #define CHECKOUT_INIT { NULL, "" }
diff --git a/entry.c b/entry.c
index b5d1d3cf23..ba21db63e7 100644
--- a/entry.c
+++ b/entry.c
@@ -447,6 +447,11 @@ int checkout_entry(struct cache_entry *ce,
 
 		if (!changed)
 			return 0;
+		if (state->set_skipworktree_on_updated) {
+			ce->ce_flags |= CE_SKIP_WORKTREE;
+			state->istate->cache_changed |= CE_ENTRY_CHANGED;
+			return 0;
+		}
 		if (!state->force) {
 			if (!state->quiet)
 				fprintf(stderr,
diff --git a/unpack-trees.c b/unpack-trees.c
index 66741130ae..a8a24e0b13 100644
--- a/unpack-trees.c
+++ b/unpack-trees.c
@@ -358,6 +358,7 @@ static int check_updates(struct unpack_trees_options *o)
 	state.quiet = 1;
 	state.refresh_cache = 1;
 	state.istate = index;
+	state.set_skipworktree_on_updated = o->clone_checkout;
 
 	progress = get_progress(o);
 
diff --git a/unpack-trees.h b/unpack-trees.h
index c2b434c606..8ebe2e2ec5 100644
--- a/unpack-trees.h
+++ b/unpack-trees.h
@@ -49,6 +49,7 @@ struct unpack_trees_options {
 		     aggressive,
 		     skip_unmerged,
 		     initial_checkout,
+		     clone_checkout,
 		     diff_index_cached,
 		     debug_unpack,
 		     skip_sparse_checkout,
-- 8< --

> While this might be fine for a closed system such as inside a
> company, this would make many open source repositories unusable,
> even when the files differing in case are nonfunctional (like README
> and readme).
> 
> This is actually one of a few ways people can make repositories that
> will show as modified on Windows or macOS systems, due to limitations in
> those OSes.  If you want to be sure that your repository is unmodified
> after clone, you can ensure that the output of git status --porcelain is
> empty, such as by checking for a zero exit from
> "test $(git status --porcelain | wc -l) -eq 0".
> -- 
> brian m. carlson: Houston, Texas, US
> OpenPGP: https://keybase.io/bk2204
Previous: brian m. carlsonNext: Duy Nguyen
Message 3 of 96 in “Git clone and case sensitivity”
  1. Paweł ParuzelJul 27, 2018
  2. brian m. carlsonJul 27, 2018
  3. Duy NguyenJul 28, 2018
  4. Duy NguyenJul 28, 2018
  5. Jeff KingJul 28, 2018
  6. Duy NguyenJul 28, 2018
  7. Simon RuderichJul 28, 2018
  8. Jeff KingJul 28, 2018
  9. brian m. carlsonJul 28, 2018
  10. Duy NguyenJul 29, 2018
  11. Jeff KingJul 29, 2018
  12. clone: report duplicate entries on case-insensitive filesystemsNguyễn Thái Ngọc Duy, Jul 30, 2018
  13. Torsten BögershausenJul 31, 2018
  14. Duy NguyenAug 1, 2018
  15. Elijah NewrenJul 31, 2018
  16. Junio C HamanoJul 31, 2018
  17. Jeff KingJul 31, 2018
  18. Junio C HamanoJul 31, 2018
  19. Jeff KingJul 31, 2018
  20. Junio C HamanoJul 31, 2018
  21. Junio C HamanoAug 1, 2018
  22. Duy NguyenAug 2, 2018
  23. Junio C HamanoAug 2, 2018
  24. Jeff KingAug 2, 2018
  25. Junio C HamanoAug 2, 2018
  26. Jeff KingAug 2, 2018
  27. Jeff HostetlerAug 3, 2018
  28. Junio C HamanoAug 3, 2018
  29. Jeff KingAug 3, 2018
  30. Jeff HostetlerAug 5, 2018
  31. Torsten BögershausenAug 3, 2018
  32. Duy NguyenAug 1, 2018
  33. Junio C HamanoJul 31, 2018
  34. Duy NguyenAug 1, 2018
  35. clone: report duplicate entries on case-insensitive filesystemsNguyễn Thái Ngọc Duy, Aug 7, 2018
  36. Junio C HamanoAug 7, 2018
  37. Jeff HostetlerAug 8, 2018
  38. Jeff KingAug 8, 2018
  39. Junio C HamanoAug 9, 2018
  40. Jeff KingAug 9, 2018
  41. Jeff HostetlerAug 9, 2018
  42. Jeff KingAug 9, 2018
  43. Elijah NewrenAug 9, 2018
  44. Jeff KingAug 9, 2018
  45. Elijah NewrenAug 9, 2018
  46. Jeff KingAug 9, 2018
  47. Elijah NewrenAug 9, 2018
  48. Junio C HamanoAug 9, 2018
  49. 0/1 clone: warn on colidding entries on checkoutNguyễn Thái Ngọc Duy, Aug 10, 2018
  50. 1/1 clone: report duplicate entries on case-insensitive filesystemsNguyễn Thái Ngọc Duy, Aug 10, 2018
  51. Junio C HamanoAug 10, 2018
  52. SZEDER GáborAug 11, 2018
  53. Duy NguyenAug 11, 2018
  54. Junio C HamanoAug 13, 2018
  55. Duy NguyenAug 13, 2018
  56. Junio C HamanoAug 10, 2018
  57. clone: report duplicate entries on case-insensitive filesystemsNguyễn Thái Ngọc Duy, Aug 12, 2018
  58. Jeff HostetlerAug 13, 2018
  59. Junio C HamanoAug 13, 2018
  60. Torsten BögershausenAug 15, 2018
  61. Duy NguyenAug 15, 2018
  62. config.txt: clarify core.checkStat = minimalNguyễn Thái Ngọc Duy, Aug 16, 2018
  63. Junio C HamanoAug 16, 2018
  64. Duy NguyenAug 16, 2018
  65. Junio C HamanoAug 16, 2018
  66. Junio C HamanoAug 17, 2018
  67. Duy NguyenAug 17, 2018
  68. Junio C HamanoAug 15, 2018
  69. Torsten BögershausenAug 16, 2018
  70. Duy NguyenAug 16, 2018
  71. Junio C HamanoAug 16, 2018
  72. clone: report duplicate entries on case-insensitive filesystemsNguyễn Thái Ngọc Duy, Aug 17, 2018
  73. Junio C HamanoAug 17, 2018
  74. Duy NguyenAug 17, 2018
  75. Torsten BögershausenAug 17, 2018
  76. clone: report duplicate entries on case-insensitive filesystemsCarlo Marcelo Arenas Belón, Nov 19, 2018
  77. Torsten BögershausenNov 19, 2018
  78. Carlo ArenasNov 19, 2018
  79. Duy NguyenNov 19, 2018
  80. Duy NguyenNov 19, 2018
  81. Duy NguyenNov 19, 2018
  82. Duy NguyenNov 19, 2018
  83. Ramsay JonesNov 19, 2018
  84. Ramsay JonesNov 19, 2018
  85. Carlo ArenasNov 20, 2018
  86. Junio C HamanoNov 20, 2018
  87. clone: fix colliding file detection on APFSNguyễn Thái Ngọc Duy, Nov 20, 2018
  88. Ramsay JonesNov 20, 2018
  89. Carlo ArenasNov 20, 2018
  90. Duy NguyenNov 20, 2018
  91. 1/1 t5601-99: Enable colliding file detection for MINGWtboegi@web.de, Nov 22, 2018
  92. 1/1 t5601-99: Enable colliding file detection for MINGWCarlo Marcelo Arenas Belón, Nov 22, 2018
  93. Johannes SchindelinNov 23, 2018
  94. Ramsay JonesNov 19, 2018
  95. Carlo ArenasNov 19, 2018
  96. Jeff HostetlerJul 31, 2018

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.