Merge branch 'mh/check-ref-format-3'

* mh/check-ref-format-3: (23 commits)
  add_ref(): verify that the refname is formatted correctly
  resolve_ref(): expand documentation
  resolve_ref(): also treat a too-long SHA1 as invalid
  resolve_ref(): emit warnings for improperly-formatted references
  resolve_ref(): verify that the input refname has the right format
  remote: avoid passing NULL to read_ref()
  remote: use xstrdup() instead of strdup()
  resolve_ref(): do not follow incorrectly-formatted symbolic refs
  resolve_ref(): extract a function get_packed_ref()
  resolve_ref(): turn buffer into a proper string as soon as possible
  resolve_ref(): only follow a symlink that contains a valid, normalized refname
  resolve_ref(): use prefixcmp()
  resolve_ref(): explicitly fail if a symlink is not readable
  Change check_refname_format() to reject unnormalized refnames
  Inline function refname_format_print()
  Make collapse_slashes() allocate memory for its result
  Do not allow ".lock" at the end of any refname component
  Refactor check_refname_format()
  Change check_ref_format() to take a flags argument
  Change bad_ref_char() to return a boolean value
  ...

Documentation/git-check-ref-format.txt

@@ -8,8 +8,9 @@ git-check-ref-format - Ensures that a reference name is well formed
 SYNOPSIS
 --------
 [verse]
-'git check-ref-format' <refname>
-'git check-ref-format' --print <refname>
+'git check-ref-format' [--normalize]
+       [--[no-]allow-onelevel] [--refspec-pattern]
+       <refname>
 'git check-ref-format' --branch <branchname-shorthand>
 
 DESCRIPTION
@@ -28,22 +29,28 @@ git imposes the following rules on how references are named:
 
 . They can include slash `/` for hierarchical (directory)
   grouping, but no slash-separated component can begin with a
-  dot `.`.
+  dot `.` or end with the sequence `.lock`.
 
 . They must contain at least one `/`. This enforces the presence of a
   category like `heads/`, `tags/` etc. but the actual names are not
-  restricted.
+  restricted.  If the `--allow-onelevel` option is used, this rule
+  is waived.
 
 . They cannot have two consecutive dots `..` anywhere.
 
 . They cannot have ASCII control characters (i.e. bytes whose
   values are lower than \040, or \177 `DEL`), space, tilde `~`,
-  caret `{caret}`, colon `:`, question-mark `?`, asterisk `*`,
-  or open bracket `[` anywhere.
+  caret `{caret}`, or colon `:` anywhere.
 
-. They cannot end with a slash `/` nor a dot `.`.
+. They cannot have question-mark `?`, asterisk `{asterisk}`, or open
+  bracket `[` anywhere.  See the `--refspec-pattern` option below for
+  an exception to this rule.
 
-. They cannot end with the sequence `.lock`.
+. They cannot begin or end with a slash `/` or contain multiple
+  consecutive slashes (see the `--normalize` option below for an
+  exception to this rule)
+
+. They cannot end with a dot `.`.
 
 . They cannot contain a sequence `@{`.
 
@@ -68,16 +75,36 @@ reference name expressions (see linkgit:gitrevisions[7]):
 
 . at-open-brace `@{` is used as a notation to access a reflog entry.
 
-With the `--print` option, if 'refname' is acceptable, it prints the
-canonicalized name of a hypothetical reference with that name.  That is,
-it prints 'refname' with any extra `/` characters removed.
-
 With the `--branch` option, it expands the ``previous branch syntax''
 `@{-n}`.  For example, `@{-1}` is a way to refer the last branch you
 were on.  This option should be used by porcelains to accept this
 syntax anywhere a branch name is expected, so they can act as if you
 typed the branch name.
 
+OPTIONS
+-------
+--allow-onelevel::
+--no-allow-onelevel::
+	Controls whether one-level refnames are accepted (i.e.,
+	refnames that do not contain multiple `/`-separated
+	components).  The default is `--no-allow-onelevel`.
+
+--refspec-pattern::
+	Interpret <refname> as a reference name pattern for a refspec
+	(as used with remote repositories).  If this option is
+	enabled, <refname> is allowed to contain a single `{asterisk}`
+	in place of a one full pathname component (e.g.,
+	`foo/{asterisk}/bar` but not `foo/bar{asterisk}`).
+
+--normalize::
+	Normalize 'refname' by removing any leading slash (`/`)
+	characters and collapsing runs of adjacent slashes between
+	name components into a single slash.  Iff the normalized
+	refname is valid then print it to standard output and exit
+	with a status of 0.  (`--print` is a deprecated way to spell
+	`--normalize`.)
+
+
 EXAMPLES
 --------
 
@@ -90,7 +117,7 @@ $ git check-ref-format --branch @{-1}
 * Determine the reference name to use for a new branch:
 +
 ------------
-$ ref=$(git check-ref-format --print "refs/heads/$newbranch") ||
+$ ref=$(git check-ref-format --normalize "refs/heads/$newbranch") ||
 die "we do not like '$newbranch' as a branch name."
 ------------
 

builtin/check-ref-format.c

@@ -8,29 +8,32 @@
 #include "strbuf.h"
 
 static const char builtin_check_ref_format_usage[] =
-"git check-ref-format [--print] <refname>\n"
+"git check-ref-format [--normalize] [options] <refname>\n"
 "   or: git check-ref-format --branch <branchname-shorthand>";
 
 /*
- * Remove leading slashes and replace each run of adjacent slashes in
- * src with a single slash, and write the result to dst.
+ * Return a copy of refname but with leading slashes removed and runs
+ * of adjacent slashes replaced with single slashes.
  *
  * This function is similar to normalize_path_copy(), but stripped down
  * to meet check_ref_format's simpler needs.
  */
-static void collapse_slashes(char *dst, const char *src)
+static char *collapse_slashes(const char *refname)
 {
+	char *ret = xmalloc(strlen(refname) + 1);
 	char ch;
 	char prev = '/';
+	char *cp = ret;
 
-	while ((ch = *src++) != '\0') {
+	while ((ch = *refname++) != '\0') {
 		if (prev == '/' && ch == prev)
 			continue;
 
-		*dst++ = ch;
+		*cp++ = ch;
 		prev = ch;
 	}
-	*dst = '\0';
+	*cp = '\0';
+	return ret;
 }
 
 static int check_ref_format_branch(const char *arg)
@@ -45,27 +48,41 @@ static int check_ref_format_branch(const char *arg)
 	return 0;
 }
 
-static int check_ref_format_print(const char *arg)
-{
-	char *refname = xmalloc(strlen(arg) + 1);
-
-	if (check_ref_format(arg))
-		return 1;
-	collapse_slashes(refname, arg);
-	printf("%s\n", refname);
-	return 0;
-}
-
 int cmd_check_ref_format(int argc, const char **argv, const char *prefix)
 {
+	int i;
+	int normalize = 0;
+	int flags = 0;
+	const char *refname;
+
 	if (argc == 2 && !strcmp(argv[1], "-h"))
 		usage(builtin_check_ref_format_usage);
 
 	if (argc == 3 && !strcmp(argv[1], "--branch"))
 		return check_ref_format_branch(argv[2]);
-	if (argc == 3 && !strcmp(argv[1], "--print"))
-		return check_ref_format_print(argv[2]);
-	if (argc != 2)
+
+	for (i = 1; i < argc && argv[i][0] == '-'; i++) {
+		if (!strcmp(argv[i], "--normalize") || !strcmp(argv[i], "--print"))
+			normalize = 1;
+		else if (!strcmp(argv[i], "--allow-onelevel"))
+			flags |= REFNAME_ALLOW_ONELEVEL;
+		else if (!strcmp(argv[i], "--no-allow-onelevel"))
+			flags &= ~REFNAME_ALLOW_ONELEVEL;
+		else if (!strcmp(argv[i], "--refspec-pattern"))
+			flags |= REFNAME_REFSPEC_PATTERN;
+		else
+			usage(builtin_check_ref_format_usage);
+	}
+	if (! (i == argc - 1))
 		usage(builtin_check_ref_format_usage);
-	return !!check_ref_format(argv[1]);
+
+	refname = argv[i];
+	if (normalize)
+		refname = collapse_slashes(refname);
+	if (check_refname_format(refname, flags))
+		return 1;
+	if (normalize)
+		printf("%s\n", refname);
+
+	return 0;
 }

builtin/checkout.c

@@ -871,7 +871,7 @@ static int parse_branchname_arg(int argc, const char **argv,
 	new->name = arg;
 	setup_branch_path(new);
 
-	if (check_ref_format(new->path) == CHECK_REF_FORMAT_OK &&
+	if (!check_refname_format(new->path, 0) &&
 	    resolve_ref(new->path, branch_rev, 1, NULL))
 		hashcpy(rev, branch_rev);
 	else

builtin/fetch-pack.c

@@ -546,7 +546,7 @@ static void filter_refs(struct ref **refs, int nr_match, char **match)
 	for (ref = *refs; ref; ref = next) {
 		next = ref->next;
 		if (!memcmp(ref->name, "refs/", 5) &&
-		    check_ref_format(ref->name + 5))
+		    check_refname_format(ref->name + 5, 0))
 			; /* trash */
 		else if (args.fetch_all &&
 			 (!args.depth || prefixcmp(ref->name, "refs/tags/") )) {

builtin/receive-pack.c

@@ -396,7 +396,7 @@ static const char *update(struct command *cmd)
 	struct ref_lock *lock;
 
 	/* only refs/... are allowed */
-	if (prefixcmp(name, "refs/") || check_ref_format(name + 5)) {
+	if (prefixcmp(name, "refs/") || check_refname_format(name + 5, 0)) {
 		rp_error("refusing to create funny ref '%s' remotely", name);
 		return "funny refname";
 	}

builtin/replace.c

@@ -94,7 +94,7 @@ static int replace_object(const char *object_ref, const char *replace_ref,
 		     "refs/replace/%s",
 		     sha1_to_hex(object)) > sizeof(ref) - 1)
 		die("replace ref name too long: %.*s...", 50, ref);
-	if (check_ref_format(ref))
+	if (check_refname_format(ref, 0))
 		die("'%s' is not a valid ref name.", ref);
 
 	if (!resolve_ref(ref, prev, 1, NULL))

builtin/show-ref.c

@@ -145,7 +145,7 @@ static int exclude_existing(const char *match)
 			if (strncmp(ref, match, matchlen))
 				continue;
 		}
-		if (check_ref_format(ref)) {
+		if (check_refname_format(ref, 0)) {
 			warning("ref '%s' ignored", ref);
 			continue;
 		}

builtin/tag.c

@@ -407,12 +407,12 @@ static int parse_msg_arg(const struct option *opt, const char *arg, int unset)
 static int strbuf_check_tag_ref(struct strbuf *sb, const char *name)
 {
 	if (name[0] == '-')
-		return CHECK_REF_FORMAT_ERROR;
+		return -1;
 
 	strbuf_reset(sb);
 	strbuf_addf(sb, "refs/tags/%s", name);
 
-	return check_ref_format(sb->buf);
+	return check_refname_format(sb->buf, 0);
 }
 
 int cmd_tag(int argc, const char **argv, const char *prefix)

cache.h

@@ -819,10 +819,51 @@ static inline int get_sha1_with_context(const char *str, unsigned char *sha1, st
 {
 	return get_sha1_with_context_1(str, sha1, orc, 0, NULL);
 }
+
+/*
+ * Try to read a SHA1 in hexadecimal format from the 40 characters
+ * starting at hex.  Write the 20-byte result to sha1 in binary form.
+ * Return 0 on success.  Reading stops if a NUL is encountered in the
+ * input, so it is safe to pass this function an arbitrary
+ * null-terminated string.
+ */
 extern int get_sha1_hex(const char *hex, unsigned char *sha1);
+
 extern char *sha1_to_hex(const unsigned char *sha1);	/* static buffer result! */
 extern int read_ref(const char *filename, unsigned char *sha1);
-extern const char *resolve_ref(const char *path, unsigned char *sha1, int, int *);
+
+/*
+ * Resolve a reference, recursively following symbolic refererences.
+ *
+ * Store the referred-to object's name in sha1 and return the name of
+ * the non-symbolic reference that ultimately pointed at it.  The
+ * return value, if not NULL, is a pointer into either a static buffer
+ * or the input ref.
+ *
+ * If the reference cannot be resolved to an object, the behavior
+ * depends on the "reading" argument:
+ *
+ * - If reading is set, return NULL.
+ *
+ * - If reading is not set, clear sha1 and return the name of the last
+ *   reference name in the chain, which will either be a non-symbolic
+ *   reference or an undefined reference.  If this is a prelude to
+ *   "writing" to the ref, the return value is the name of the ref
+ *   that will actually be created or changed.
+ *
+ * If flag is non-NULL, set the value that it points to the
+ * combination of REF_ISPACKED (if the reference was found among the
+ * packed references) and REF_ISSYMREF (if the initial reference was a
+ * symbolic reference).
+ *
+ * If ref is not a properly-formatted, normalized reference, return
+ * NULL.  If more than MAXDEPTH recursive symbolic lookups are needed,
+ * give up and return NULL.
+ *
+ * errno is sometimes set on errors, but not always.
+ */
+extern const char *resolve_ref(const char *ref, unsigned char *sha1, int reading, int *flag);
+
 extern int dwim_ref(const char *str, int len, unsigned char *sha1, char **ref);
 extern int dwim_log(const char *str, int len, unsigned char *sha1, char **ref);
 extern int interpret_branch_name(const char *str, struct strbuf *);

connect.c

@@ -22,7 +22,7 @@ static int check_ref(const char *name, int len, unsigned int flags)
 	len -= 5;
 
 	/* REF_NORMAL means that we don't want the magic fake tag refs */
-	if ((flags & REF_NORMAL) && check_ref_format(name) < 0)
+	if ((flags & REF_NORMAL) && check_refname_format(name, 0))
 		return 0;
 
 	/* REF_HEADS means that we want regular branch heads */

environment.c

@@ -106,7 +106,7 @@ static char *expand_namespace(const char *raw_namespace)
 		if (strcmp((*c)->buf, "/") != 0)
 			strbuf_addf(&buf, "refs/namespaces/%s", (*c)->buf);
 	strbuf_list_free(components);
-	if (check_ref_format(buf.buf) != CHECK_REF_FORMAT_OK)
+	if (check_refname_format(buf.buf, 0))
 		die("bad git namespace path \"%s\"", raw_namespace);
 	strbuf_addch(&buf, '/');
 	return strbuf_detach(&buf, NULL);

fast-import.c

@@ -722,13 +722,8 @@ static struct branch *new_branch(const char *name)
 
 	if (b)
 		die("Invalid attempt to create duplicate branch: %s", name);
-	switch (check_ref_format(name)) {
-	case 0: break; /* its valid */
-	case CHECK_REF_FORMAT_ONELEVEL:
-		break; /* valid, but too few '/', allow anyway */
-	default:
+	if (check_refname_format(name, REFNAME_ALLOW_ONELEVEL))
 		die("Branch name doesn't conform to GIT standards: %s", name);
-	}
 
 	b = pool_calloc(1, sizeof(struct branch));
 	b->name = pool_strdup(name);

git_remote_helpers/git/git.py

@@ -54,7 +54,7 @@ def valid_git_ref (ref_name):
     # The following is a reimplementation of the git check-ref-format
     # command.  The rules were derived from the git check-ref-format(1)
     # manual page.  This code should be replaced by a call to
-    # check_ref_format() in the git library, when such is available.
+    # check_refname_format() in the git library, when such is available.
     if ref_name.endswith('/') or \
        ref_name.startswith('.') or \
        ref_name.count('/.') or \

hex.c

@@ -39,7 +39,15 @@ int get_sha1_hex(const char *hex, unsigned char *sha1)
 {
 	int i;
 	for (i = 0; i < 20; i++) {
-		unsigned int val = (hexval(hex[0]) << 4) | hexval(hex[1]);
+		unsigned int val;
+		/*
+		 * hex[1]=='\0' is caught when val is checked below,
+		 * but if hex[0] is NUL we have to avoid reading
+		 * past the end of the string:
+		 */
+		if (!hex[0])
+			return -1;
+		val = (hexval(hex[0]) << 4) | hexval(hex[1]);
 		if (val & ~0xff)
 			return -1;
 		*sha1++ = val;

notes-merge.c

@@ -570,7 +570,8 @@ int notes_merge(struct notes_merge_options *o,
 	/* Dereference o->local_ref into local_sha1 */
 	if (!resolve_ref(o->local_ref, local_sha1, 0, NULL))
 		die("Failed to resolve local notes ref '%s'", o->local_ref);
-	else if (!check_ref_format(o->local_ref) && is_null_sha1(local_sha1))
+	else if (!check_refname_format(o->local_ref, 0) &&
+		is_null_sha1(local_sha1))
 		local = NULL; /* local_sha1 == null_sha1 indicates unborn ref */
 	else if (!(local = lookup_commit_reference(local_sha1)))
 		die("Could not parse local commit %s (%s)",
@@ -583,7 +584,7 @@ int notes_merge(struct notes_merge_options *o,
 		 * Failed to get remote_sha1. If o->remote_ref looks like an
 		 * unborn ref, perform the merge using an empty notes tree.
 		 */
-		if (!check_ref_format(o->remote_ref)) {
+		if (!check_refname_format(o->remote_ref, 0)) {
 			hashclr(remote_sha1);
 			remote = NULL;
 		} else {

pack-refs.c

@@ -72,7 +72,7 @@ static void try_remove_empty_parents(char *name)
 	for (i = 0; i < 2; i++) { /* refs/{heads,tags,...}/ */
 		while (*p && *p != '/')
 			p++;
-		/* tolerate duplicate slashes; see check_ref_format() */
+		/* tolerate duplicate slashes; see check_refname_format() */
 		while (*p == '/')
 			p++;
 	}