filter-attribute-mods.diff

From: Matt McCutchen <matt@mattmccutchen.net>

Implement the "m", "o", "g" include modifiers to tweak the permissions,
owner, or group of matching files.

To use this patch, run these commands for a successful build:

    patch -p1 <patches/filter-attribute-mods.diff
    ./configure                         (optional if already run)
    make

based-on: 6c8ca91c731b7bf2b081694bda85b7dadc2b7aff
diff --git a/exclude.c b/exclude.c
--- a/exclude.c
+++ b/exclude.c
@@ -51,12 +51,15 @@ filter_rule_list cvs_filter_list = { .debug_type = " [global CVS]" };
 filter_rule_list daemon_filter_list = { .debug_type = " [daemon]" };
 filter_rule_list implied_filter_list = { .debug_type = " [implied]" };
 
+filter_rule *last_hit_filter_rule;
+
 int saw_xattr_filter = 0;
 int trust_sender_args = 0;
 int trust_sender_filter = 0;
 
-/* Need room enough for ":MODS " prefix plus some room to grow. */
-#define MAX_RULE_PREFIX (16)
+/* Need room enough for ":MODS " prefix, which can now include
+ * chmod/user/group values. */
+#define MAX_RULE_PREFIX (256)
 
 #define SLASH_WILD3_SUFFIX "/***"
 
@@ -139,8 +142,27 @@ static void teardown_mergelist(filter_rule *ex)
 		mergelist_cnt--;
 }
 
+static struct filter_chmod_struct *ref_filter_chmod(struct filter_chmod_struct *chmod)
+{
+	chmod->ref_cnt++;
+	assert(chmod->ref_cnt != 0); /* Catch overflow. */
+	return chmod;
+}
+
+static void unref_filter_chmod(struct filter_chmod_struct *chmod)
+{
+	chmod->ref_cnt--;
+	if (chmod->ref_cnt == 0) {
+		free(chmod->modestr);
+		free_chmod_mode(chmod->modes);
+		free(chmod);
+	}
+}
+
 static void free_filter(filter_rule *ex)
 {
+	if (ex->rflags & FILTRULE_CHMOD)
+		unref_filter_chmod(ex->chmod);
 	if (ex->rflags & FILTRULE_PERDIR_MERGE)
 		teardown_mergelist(ex);
 	free(ex->pattern);
@@ -1006,7 +1028,9 @@ static void report_filter_result(enum logcode code, char const *name,
 
 /* This function is used to check if a file should be included/excluded
  * from the list of files based on its name and type etc.  The value of
- * filter_level is set to either SERVER_FILTERS or ALL_FILTERS. */
+ * filter_level is set to either SERVER_FILTERS or ALL_FILTERS.
+ * "last_hit_filter_rule" will be set to the operative filter, or NULL if none. */
+
 int name_is_excluded(const char *fname, int name_flags, int filter_level)
 {
 	if (daemon_filter_list.head && check_filter(&daemon_filter_list, FLOG, fname, name_flags) < 0) {
@@ -1015,6 +1039,9 @@ int name_is_excluded(const char *fname, int name_flags, int filter_level)
 		return 1;
 	}
 
+	/* Don't leave a daemon include in last_hit_filter_rule. */
+	last_hit_filter_rule = NULL;
+
 	if (filter_level != ALL_FILTERS)
 		return 0;
 
@@ -1034,7 +1061,8 @@ int check_server_filter(filter_rule_list *listp, enum logcode code, const char *
 }
 
 /* Return -1 if file "name" is defined to be excluded by the specified
- * exclude list, 1 if it is included, and 0 if it was not matched. */
+ * exclude list, 1 if it is included, and 0 if it was not matched.
+ * Sets last_hit_filter_rule to the filter that was hit, or NULL if none. */
 int check_filter(filter_rule_list *listp, enum logcode code,
 		 const char *name, int name_flags)
 {
@@ -1057,10 +1085,12 @@ int check_filter(filter_rule_list *listp, enum logcode code,
 		}
 		if (rule_matches(name, ent, name_flags)) {
 			report_filter_result(code, name, ent, name_flags, listp->debug_type);
+			last_hit_filter_rule = ent;
 			return ent->rflags & FILTRULE_INCLUDE ? 1 : -1;
 		}
 	}
 
+	last_hit_filter_rule = NULL;
 	return 0;
 }
 
@@ -1077,9 +1107,45 @@ static const uchar *rule_strcmp(const uchar *str, const char *rule, int rule_len
 	return NULL;
 }
 
+static char *grab_paren_value(const uchar **s_ptr)
+{
+	const uchar *start, *end;
+	int val_sz;
+	char *val;
+
+	if ((*s_ptr)[1] != '(')
+		return NULL;
+	start = (*s_ptr) + 2;
+
+	for (end = start; *end != ')'; end++)
+		if (!*end || *end == ' ' || *end == '_')
+			return NULL;
+
+	val_sz = end - start + 1;
+	val = new_array(char, val_sz);
+	strlcpy(val, (const char *)start, val_sz);
+	*s_ptr = end; /* remember ++s in parse_rule_tok */
+	return val;
+}
+
+static struct filter_chmod_struct *make_chmod_struct(char *modestr)
+{
+	struct filter_chmod_struct *chmod;
+	struct chmod_mode_struct *modes = NULL;
+
+	if (!parse_chmod(modestr, &modes))
+		return NULL;
+
+	chmod = new(struct filter_chmod_struct);
+	chmod->ref_cnt = 1;
+	chmod->modestr = modestr;
+	chmod->modes = modes;
+	return chmod;
+}
+
 #define FILTRULES_FROM_CONTAINER (FILTRULE_ABS_PATH | FILTRULE_INCLUDE \
 				| FILTRULE_DIRECTORY | FILTRULE_NEGATE \
-				| FILTRULE_PERISHABLE)
+				| FILTRULE_PERISHABLE | FILTRULES_ATTRS)
 
 /* Gets the next include/exclude rule from *rulestr_ptr and advances
  * *rulestr_ptr to point beyond it.  Stores the pattern's start (within
@@ -1094,6 +1160,7 @@ static filter_rule *parse_rule_tok(const char **rulestr_ptr,
 				   const char **pat_ptr, unsigned int *pat_len_ptr)
 {
 	const uchar *s = (const uchar *)*rulestr_ptr;
+	char *val;
 	filter_rule *rule;
 	unsigned int len;
 
@@ -1112,6 +1179,12 @@ static filter_rule *parse_rule_tok(const char **rulestr_ptr,
 	/* Inherit from the template.  Don't inherit FILTRULES_SIDES; we check
 	 * that later. */
 	rule->rflags = template->rflags & FILTRULES_FROM_CONTAINER;
+	if (template->rflags & FILTRULE_CHMOD)
+		rule->chmod = ref_filter_chmod(template->chmod);
+	if (template->rflags & FILTRULE_FORCE_OWNER)
+		rule->force_uid = template->force_uid;
+	if (template->rflags & FILTRULE_FORCE_GROUP)
+		rule->force_gid = template->force_gid;
 
 	/* Figure out what kind of a filter rule "s" is pointing at.  Note
 	 * that if FILTRULE_NO_PREFIXES is set, the rule is either an include
@@ -1258,11 +1331,63 @@ static filter_rule *parse_rule_tok(const char **rulestr_ptr,
 					goto invalid;
 				rule->rflags |= FILTRULE_EXCLUDE_SELF;
 				break;
+			case 'g': {
+				gid_t gid;
+
+				if (!(val = grab_paren_value(&s)))
+					goto invalid;
+				if (group_to_gid(val, &gid, True)) {
+					rule->rflags |= FILTRULE_FORCE_GROUP;
+					rule->force_gid = gid;
+				} else {
+					rprintf(FERROR,
+						"unknown group '%s' in filter rule: %s\n",
+						val, *rulestr_ptr);
+					exit_cleanup(RERR_SYNTAX);
+				}
+				free(val);
+				break;
+			}
+			case 'm': {
+				struct filter_chmod_struct *chmod;
+
+				if (!(val = grab_paren_value(&s)))
+					goto invalid;
+				if ((chmod = make_chmod_struct(val))) {
+					if (rule->rflags & FILTRULE_CHMOD)
+						unref_filter_chmod(rule->chmod);
+					rule->rflags |= FILTRULE_CHMOD;
+					rule->chmod = chmod;
+				} else {
+					rprintf(FERROR,
+						"unparseable chmod string '%s' in filter rule: %s\n",
+						val, *rulestr_ptr);
+					exit_cleanup(RERR_SYNTAX);
+				}
+				break;
+			}
 			case 'n':
 				if (!(rule->rflags & FILTRULE_MERGE_FILE))
 					goto invalid;
 				rule->rflags |= FILTRULE_NO_INHERIT;
 				break;
+			case 'o': {
+				uid_t uid;
+
+				if (!(val = grab_paren_value(&s)))
+					goto invalid;
+				if (user_to_uid(val, &uid, True)) {
+					rule->rflags |= FILTRULE_FORCE_OWNER;
+					rule->force_uid = uid;
+				} else {
+					rprintf(FERROR,
+						"unknown user '%s' in filter rule: %s\n",
+						val, *rulestr_ptr);
+					exit_cleanup(RERR_SYNTAX);
+				}
+				free(val);
+				break;
+			}
 			case 'p':
 				rule->rflags |= FILTRULE_PERISHABLE;
 				break;
@@ -1576,6 +1701,23 @@ char *get_rule_prefix(filter_rule *rule, const char *pat, int for_xfer,
 		else if (am_sender)
 			return NULL;
 	}
+	if (rule->rflags & FILTRULES_ATTRS) {
+		if (!for_xfer || protocol_version >= 31) {
+			if (rule->rflags & FILTRULE_CHMOD)
+				if (!snappendf(&op, (buf + sizeof buf) - op,
+					"m(%s)", rule->chmod->modestr))
+					return NULL;
+			if (rule->rflags & FILTRULE_FORCE_OWNER)
+				if (!snappendf(&op, (buf + sizeof buf) - op,
+					"o(%u)", (unsigned)rule->force_uid))
+					return NULL;
+			if (rule->rflags & FILTRULE_FORCE_GROUP)
+				if (!snappendf(&op, (buf + sizeof buf) - op,
+					"g(%u)", (unsigned)rule->force_gid))
+					return NULL;
+		} else if (!am_sender)
+			return NULL;
+	}
 	if (op - buf > legal_len)
 		return NULL;
 	if (legal_len)
diff --git a/flist.c b/flist.c
--- a/flist.c
+++ b/flist.c
@@ -86,6 +86,7 @@ extern char curr_dir[MAXPATHLEN];
 extern struct chmod_mode_struct *chmod_modes;
 
 extern filter_rule_list filter_list, implied_filter_list, daemon_filter_list;
+extern filter_rule *last_hit_filter_rule;
 
 #ifdef ICONV_OPTION
 extern int filesfrom_convert;
@@ -1259,7 +1260,7 @@ struct file_struct *make_file(const char *fname, struct file_list *flist,
 	} else if (readlink_stat(thisname, &st, linkname) != 0) {
 		int save_errno = errno;
 		/* See if file is excluded before reporting an error. */
-		if (filter_level != NO_FILTERS
+		if (filter_level != NO_FILTERS && filter_level != ALL_FILTERS_NO_EXCLUDE
 		 && (is_excluded(thisname, 0, filter_level)
 		  || is_excluded(thisname, 1, filter_level))) {
 			if (ignore_perishable && save_errno != ENOENT)
@@ -1304,6 +1305,12 @@ struct file_struct *make_file(const char *fname, struct file_list *flist,
 
 	if (filter_level == NO_FILTERS)
 		goto skip_filters;
+	if (filter_level == ALL_FILTERS_NO_EXCLUDE) {
+		/* Call only for the side effect of setting last_hit_filter_rule to
+		 * any operative include filter, which might affect attributes. */
+		is_excluded(thisname, S_ISDIR(st.st_mode) != 0, ALL_FILTERS);
+		goto skip_filters;
+	}
 
 	if (S_ISDIR(st.st_mode)) {
 		if (!xfer_dirs) {
@@ -1536,12 +1543,23 @@ static struct file_struct *send_file_name(int f, struct file_list *flist,
 					  int flags, int filter_level)
 {
 	struct file_struct *file;
+	BOOL can_tweak_mode;
 
 	file = make_file(fname, flist, stp, flags, filter_level);
 	if (!file)
 		return NULL;
 
-	if (chmod_modes && !S_ISLNK(file->mode) && file->mode)
+	can_tweak_mode = !S_ISLNK(file->mode) && file->mode;
+	if ((filter_level == ALL_FILTERS || filter_level == ALL_FILTERS_NO_EXCLUDE)
+		&& last_hit_filter_rule) {
+		if ((last_hit_filter_rule->rflags & FILTRULE_CHMOD) && can_tweak_mode)
+			file->mode = tweak_mode(file->mode, last_hit_filter_rule->chmod->modes);
+		if ((last_hit_filter_rule->rflags & FILTRULE_FORCE_OWNER) && uid_ndx)
+			F_OWNER(file) = last_hit_filter_rule->force_uid;
+		if ((last_hit_filter_rule->rflags & FILTRULE_FORCE_GROUP) && gid_ndx)
+			F_GROUP(file) = last_hit_filter_rule->force_gid;
+	}
+	if (chmod_modes && can_tweak_mode)
 		file->mode = tweak_mode(file->mode, chmod_modes);
 
 	if (f >= 0) {
@@ -2443,7 +2461,7 @@ struct file_list *send_file_list(int f, int argc, char *argv[])
 			struct file_struct *file;
 			file = send_file_name(f, flist, fbuf, &st,
 					      FLAG_TOP_DIR | FLAG_CONTENT_DIR | flags,
-					      NO_FILTERS);
+					      ALL_FILTERS_NO_EXCLUDE);
 			if (!file)
 				continue;
 			if (inc_recurse) {
@@ -2457,7 +2475,7 @@ struct file_list *send_file_list(int f, int argc, char *argv[])
 			} else
 				send_if_directory(f, flist, file, fbuf, len, flags);
 		} else
-			send_file_name(f, flist, fbuf, &st, flags, NO_FILTERS);
+			send_file_name(f, flist, fbuf, &st, flags, ALL_FILTERS_NO_EXCLUDE);
 	}
 
 	if (reenable_multiplex >= 0)
diff --git a/rsync.1.md b/rsync.1.md
--- a/rsync.1.md
+++ b/rsync.1.md
@@ -1513,7 +1513,9 @@ expand it.
     >     --chmod=D2775,F664
 
     It is also legal to specify multiple `--chmod` options, as each additional
-    option is just appended to the list of changes to make.
+    option is just appended to the list of changes to make.  To change
+    permissions of files matching a pattern, use an include filter with the `m`
+    modifier, which takes effect before any `--chmod` options.
 
     See the [`--perms`](#opt) and [`--executability`](#opt) options for how the
     resulting permission value can be applied to the files in the transfer.
@@ -3033,6 +3035,10 @@ expand it.
     An older rsync client may need to use [`-s`](#opt) to avoid a complaint
     about wildcard characters, but a modern rsync handles this automatically.
 
+    To change ownership of files matching a pattern, use an include filter with
+    a `o` or `g` modifier, which take effect before uid/gid mapping and
+    therefore *can* be mixed with [`--usermap` & `--groupmap`](#opt--usermap).
+
 0.  `--timeout=SECONDS`
 
     This option allows you to set a maximum I/O timeout in seconds.  If no data
@@ -4187,6 +4193,15 @@ The following modifiers are accepted after an include (+) or exclude (-) rule:
   like "CVS" and "`*.o`" are marked as perishable, and will not prevent a
   directory that was removed on the source from being deleted on the
   destination.
+- An `m(CHMOD)` on an include rule tweaks the permissions of matching source
+  files in the same way as [`--chmod`](#opt).  This happens before any tweaks
+  requested via [`--chmod`](#opt).
+- An `o(USER)` on an include rule pretends that matching source files are owned
+  by `USER` (a name or numeric uid).  This happens before any uid mapping by
+  name or [`--usermap`](#opt).
+- A `g(GROUP)` on an include rule pretends that matching source files are owned
+  by `GROUP` (a name or numeric gid).  This happens before any gid mapping by
+  name or [`--groupmap`](#opt--usermap).
 - An `x` indicates that a rule affects xattr names in xattr copy/delete
   operations (and is thus ignored when matching file/dir names).  If no
   xattr-matching rules are specified, a default xattr filtering rule is used
@@ -4244,6 +4259,12 @@ The following modifiers are accepted after a merge or dir-merge rule:
   rules in the file must not specify sides (via a modifier or a rule prefix
   such as `hide`).
 
+The attribute-affecting modifiers `m`, `o`, and `g` work only in client filters
+(not in daemon filters), and only the modifiers of the first matching rule are
+applied.  As an example, assuming [`--super`](#opt) is enabled, the rule
+"`+o(root),g(root),m(go=) *~`" would ensure that all "backup"
+files belong to root and are not accessible to anyone else.
+
 Per-directory rules are inherited in all subdirectories of the directory where
 the merge-file was found unless the 'n' modifier was used.  Each subdirectory's
 rules are prefixed to the inherited per-directory rules from its parents, which
diff --git a/rsync.h b/rsync.h
--- a/rsync.h
+++ b/rsync.h
@@ -181,6 +181,9 @@
 #define NO_FILTERS	0
 #define SERVER_FILTERS	1
 #define ALL_FILTERS	2
+/* Don't let the file be excluded, but check for a filter that might affect
+ * its attributes via FILTRULES_ATTRS. */
+#define ALL_FILTERS_NO_EXCLUDE	3
 
 #define XFLG_FATAL_ERRORS	(1<<0)
 #define XFLG_OLD_PREFIXES	(1<<1)
@@ -982,6 +985,8 @@ struct map_struct {
 	int status;		/* first errno from read errors		*/
 };
 
+struct chmod_mode_struct;
+
 #define NAME_IS_FILE		(0)    /* filter name as a file */
 #define NAME_IS_DIR		(1<<0) /* filter name as a dir */
 #define NAME_IS_XATTR		(1<<2) /* filter name as an xattr */
@@ -1007,8 +1012,18 @@ struct map_struct {
 #define FILTRULE_CLEAR_LIST	(1<<18)/* this item is the "!" token */
 #define FILTRULE_PERISHABLE	(1<<19)/* perishable if parent dir goes away */
 #define FILTRULE_XATTR		(1<<20)/* rule only applies to xattr names */
+#define FILTRULE_CHMOD		(1<<21)/* chmod-tweak matching files */
+#define FILTRULE_FORCE_OWNER	(1<<22)/* force owner of matching files */
+#define FILTRULE_FORCE_GROUP	(1<<23)/* force group of matching files */
 
 #define FILTRULES_SIDES (FILTRULE_SENDER_SIDE | FILTRULE_RECEIVER_SIDE)
+#define FILTRULES_ATTRS (FILTRULE_CHMOD | FILTRULE_FORCE_OWNER | FILTRULE_FORCE_GROUP)
+
+struct filter_chmod_struct {
+	unsigned int ref_cnt;
+	char *modestr;
+	struct chmod_mode_struct *modes;
+};
 
 typedef struct filter_struct {
 	struct filter_struct *next;
@@ -1018,6 +1033,11 @@ typedef struct filter_struct {
 		int slash_cnt;
 		struct filter_list_struct *mergelist;
 	} u;
+	/* TODO: Use an "extras" mechanism to avoid
+	 * allocating this memory when we don't need it. */
+	struct filter_chmod_struct *chmod;
+	uid_t force_uid;
+	gid_t force_gid;
 	uchar elide;
 } filter_rule;
 
diff --git a/util1.c b/util1.c
--- a/util1.c
+++ b/util1.c
@@ -918,6 +918,25 @@ size_t stringjoin(char *dest, size_t destsize, ...)
 	return ret;
 }
 
+/* Append formatted text at *dest_ptr up to a maximum of sz (like snprintf).
+ * On success, advance *dest_ptr and return True; on overflow, return False. */
+BOOL snappendf(char **dest_ptr, size_t sz, const char *format, ...)
+{
+	va_list ap;
+	size_t len;
+
+	va_start(ap, format);
+	len = vsnprintf(*dest_ptr, sz, format, ap);
+	va_end(ap);
+
+	if (len >= sz)
+		return False;
+	else {
+		*dest_ptr += len;
+		return True;
+	}
+}
+
 int count_dir_elements(const char *p)
 {
 	int cnt = 0, new_component = 1;