From: Vsevolod Stakhov <vsevolod@rspamd.com>
Date: Wed, 8 Jul 2026 09:09:41 +0000 (+0100)
Subject: [Minor] maps: move glob pattern anchoring into rspamd_str_regexp_escape
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=38ffc00b25b37ac28afa10c6677c692ccbcc72ba;p=thirdparty%2Frspamd.git

[Minor] maps: move glob pattern anchoring into rspamd_str_regexp_escape

Add RSPAMD_REGEXP_ESCAPE_ANCHOR flag that wraps the escaped pattern
into ^(?:...)$ within the same allocation instead of a second
g_strdup_printf in the glob map insertion path.
---

diff --git a/src/libserver/maps/map_helpers.c b/src/libserver/maps/map_helpers.c
index e0112feb90..aafde1d7e1 100644
--- a/src/libserver/maps/map_helpers.c
+++ b/src/libserver/maps/map_helpers.c
@@ -685,14 +685,11 @@ void rspamd_map_helper_insert_re(gpointer st, gconstpointer key, gconstpointer v
 
 	/* Check regexp stuff */
 	if (re_map->map_flags & RSPAMD_REGEXP_MAP_FLAG_GLOB) {
-		char *anchored;
-
-		escaped = rspamd_str_regexp_escape(key, strlen(key), &escaped_len,
-										   RSPAMD_REGEXP_ESCAPE_GLOB | RSPAMD_REGEXP_ESCAPE_UTF);
 		/* Glob semantics: the pattern must match the subject as a whole */
-		anchored = g_strdup_printf("^(?:%s)$", escaped);
-		re = rspamd_regexp_new(anchored, NULL, &err);
-		g_free(anchored);
+		escaped = rspamd_str_regexp_escape(key, strlen(key), &escaped_len,
+										   RSPAMD_REGEXP_ESCAPE_GLOB | RSPAMD_REGEXP_ESCAPE_UTF |
+											   RSPAMD_REGEXP_ESCAPE_ANCHOR);
+		re = rspamd_regexp_new(escaped, NULL, &err);
 		g_free(escaped);
 	}
 	else {
diff --git a/src/libutil/str_util.c b/src/libutil/str_util.c
index fc78cd22de..c884af20eb 100644
--- a/src/libutil/str_util.c
+++ b/src/libutil/str_util.c
@@ -3515,6 +3515,11 @@ rspamd_str_regexp_escape(const char *pattern, gsize slen,
 		}
 	}
 
+	if (flags & RSPAMD_REGEXP_ESCAPE_ANCHOR) {
+		/* Room for ^(?: and )$ */
+		len += 6;
+	}
+
 	if (flags & RSPAMD_REGEXP_ESCAPE_UTF) {
 		if (rspamd_fast_utf8_validate(pattern, slen) != 0) {
 			tmp_utf = rspamd_str_make_utf_valid(pattern, slen, NULL, NULL);
@@ -3547,6 +3552,11 @@ rspamd_str_regexp_escape(const char *pattern, gsize slen,
 	d = res;
 	dend = d + len;
 
+	if (flags & RSPAMD_REGEXP_ESCAPE_ANCHOR) {
+		memcpy(d, "^(?:", 4);
+		d += 4;
+	}
+
 	while (p < end) {
 		g_assert(d < dend);
 		t = *p++;
@@ -3646,6 +3656,11 @@ rspamd_str_regexp_escape(const char *pattern, gsize slen,
 		*d++ = t;
 	}
 
+	if (flags & RSPAMD_REGEXP_ESCAPE_ANCHOR) {
+		*d++ = ')';
+		*d++ = '$';
+	}
+
 	*d = '\0';
 
 	if (dst_len) {
diff --git a/src/libutil/str_util.h b/src/libutil/str_util.h
index 8de4b361ed..c28f6d6e10 100644
--- a/src/libutil/str_util.h
+++ b/src/libutil/str_util.h
@@ -505,6 +505,8 @@ enum rspamd_regexp_escape_flags {
 	RSPAMD_REGEXP_ESCAPE_UTF = 1u << 0,
 	RSPAMD_REGEXP_ESCAPE_GLOB = 1u << 1,
 	RSPAMD_REGEXP_ESCAPE_RE = 1u << 2,
+	/* Wrap the result into ^(?:...)$ so it matches the subject as a whole */
+	RSPAMD_REGEXP_ESCAPE_ANCHOR = 1u << 3,
 };
 
 /**