rx:exit after full match; fix TX population after unused group

src/operators/rx.cc

@@ -38,7 +38,6 @@ bool Rx::init(const std::string &arg, std::string *error) {
 
 bool Rx::evaluate(Transaction *transaction, RuleWithActions *rule,
     const std::string& input, std::shared_ptr<RuleMessage> ruleMessage) {
-    std::list<SMatch> matches;
     Regex *re;
 
     if (m_param.empty() && !m_string->m_containsMacro) {
@@ -52,29 +51,29 @@ bool Rx::evaluate(Transaction *transaction, RuleWithActions *rule,
         re = m_re;
     }
 
-    matches = re->searchAll(input);
+    std::vector<Utils::SMatchCapture> captures;
+    re->searchOneMatch(input, captures);
+
     if (rule && rule->hasCaptureAction() && transaction) {
-        int i = 0;
-        matches.reverse();
-        for (const SMatch& a : matches) {
+        for (const Utils::SMatchCapture& capture : captures) {
+            const std::string capture_substring(input.substr(capture.m_offset,capture.m_length));
             transaction->m_collections.m_tx_collection->storeOrUpdateFirst(
-                std::to_string(i), a.str());
+                std::to_string(capture.m_group), capture_substring);
             ms_dbg_a(transaction, 7, "Added regex subexpression TX." +
-                std::to_string(i) + ": " + a.str());
-            transaction->m_matched.push_back(a.str());
-            i++;
+                std::to_string(capture.m_group) + ": " + capture_substring);
+            transaction->m_matched.push_back(capture_substring);
         }
     }
 
-    for (const auto & i : matches) {
-        logOffset(ruleMessage, i.offset(), i.str().size());
+    for (const auto & capture : captures) {
+        logOffset(ruleMessage, capture.m_offset, capture.m_length);
     }
 
     if (m_string->m_containsMacro) {
         delete re;
     }
 
-    if (matches.size() > 0) {
+    if (captures.size() > 0) {
         return true;
     }
 

src/utils/regex.cc

@@ -16,10 +16,6 @@
 #include "src/utils/regex.h"
 
 #include <pcre.h>
-#include <sys/socket.h>
-#include <sys/types.h>
-#include <netinet/in.h>
-#include <arpa/inet.h>
 #include <string>
 #include <list>
 
@@ -99,6 +95,26 @@ std::list<SMatch> Regex::searchAll(const std::string& s) const {
     return retList;
 }
 
+bool Regex::searchOneMatch(const std::string& s, std::vector<SMatchCapture>& captures) const {
+    const char *subject = s.c_str();
+    int ovector[OVECCOUNT];
+
+    int rc = pcre_exec(m_pc, m_pce, subject, s.size(), 0, 0, ovector, OVECCOUNT);
+
+    for (int i = 0; i < rc; i++) {
+        size_t start = ovector[2*i];
+        size_t end = ovector[2*i+1];
+        size_t len = end - start;
+        if (end > s.size()) {
+            continue;
+        }
+        SMatchCapture capture(i, start, len);
+        captures.push_back(capture);
+    }
+
+    return (rc > 0);
+}
+
 int Regex::search(const std::string& s, SMatch *match) const {
     int ovector[OVECCOUNT];
     int ret = pcre_exec(m_pc, m_pce, s.c_str(),

src/utils/regex.h

@@ -19,6 +19,7 @@
 #include <fstream>
 #include <string>
 #include <list>
+#include <vector>
 
 #ifndef SRC_UTILS_REGEX_H_
 #define SRC_UTILS_REGEX_H_
@@ -47,6 +48,16 @@ class SMatch {
     size_t m_offset;
 };
 
+struct SMatchCapture {
+    SMatchCapture(size_t group, size_t offset, size_t length) :
+	m_group(group),
+	m_offset(offset),
+	m_length(length) { }
+
+    size_t m_group; // E.g. 0 = full match; 6 = capture group 6
+    size_t m_offset; // offset of match within the analyzed string
+    size_t m_length;
+};
 
 class Regex {
  public:
@@ -58,6 +69,7 @@ class Regex {
     Regex& operator=(const Regex&) = delete;
 
     std::list<SMatch> searchAll(const std::string& s) const;
+    bool searchOneMatch(const std::string& s, std::vector<SMatchCapture>& captures) const;
     int search(const std::string &s, SMatch *match) const;
     int search(const std::string &s) const;