Merge pull request #2348 from martinhsv/v3/master

rx:exit after full match; fix TX population after unused group
This commit is contained in:
martinhsv 2020-07-06 07:54:44 -04:00 committed by GitHub
commit 0eb3c123f4
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
5 changed files with 184 additions and 16 deletions

View File

@ -1,6 +1,9 @@
v3.x.y - YYYY-MMM-DD (to be released)
-------------------------------------
- rx: exit after full match (remove /g emulation); ensure capture
groups occuring after unused groups still populate TX vars
[Issue #2336 - @martinhsv]
- Correct CHANGES file entry for #2234
- Add support to test framework for audit log content verification
and add regression tests for issues #2000, #2196

View File

@ -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;
}

View File

@ -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(),

View File

@ -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;

View File

@ -80,5 +80,143 @@
"SecRule REQUEST_HEADERS \"@rx ([A-z]+)\" \"id:1,log,pass,capture,id:14\"",
"SecRule TX:0 \"@rx ([A-z]+)\" \"id:15\""
]
},
{
"enabled":1,
"version_min":300000,
"title":"Testing Variables :: capture group match after unused group",
"client":{
"ip":"200.249.12.31",
"port":123
},
"server":{
"ip":"200.249.12.31",
"port":80
},
"request":{
"uri":"/?key=aadd",
"method":"GET"
},
"response":{
"headers":{
"Date":"Mon, 13 Jul 2015 20:02:41 GMT",
"Last-Modified":"Sun, 26 Oct 2014 22:33:37 GMT",
"Content-Type":"text/html"
},
"body":[
"no need."
]
},
"expected":{
"debug_log":"Added regex subexpression TX\\.3: dd[\\s\\S]*Target value: \"dd\" \\(Variable\\: TX\\:3[\\s\\S]*Rule returned 1"
},
"rules":[
"SecRuleEngine On",
"SecRule ARGS \"@rx (aa)(bb|cc)?(dd)\" \"id:1,log,pass,capture,id:16\"",
"SecRule TX:3 \"@streq dd\" \"id:19,phase:2,log,pass\""
]
},
{
"enabled":1,
"version_min":300000,
"title":"Testing Variables :: empty capture group match followed by nonempty capture group",
"client":{
"ip":"200.249.12.31",
"port":123
},
"server":{
"ip":"200.249.12.31",
"port":80
},
"request":{
"uri":"/?key=aadd",
"method":"GET"
},
"response":{
"headers":{
"Date":"Mon, 13 Jul 2015 20:02:41 GMT",
"Last-Modified":"Sun, 26 Oct 2014 22:33:37 GMT",
"Content-Type":"text/html"
},
"body":[
"no need."
]
},
"expected":{
"debug_log":"Added regex subexpression TX\\.3: dd[\\s\\S]*Target value: \"dd\" \\(Variable\\: TX\\:3[\\s\\S]*Rule returned 1"
},
"rules":[
"SecRuleEngine On",
"SecRule ARGS \"@rx (aa)(bb|cc|)(dd)\" \"id:18,phase:1,log,pass,capture\"",
"SecRule TX:3 \"@streq dd\" \"id:19,phase:2,log,pass\""
]
},
{
"enabled":1,
"version_min":300000,
"title":"Testing Variables :: repeating capture group -- alternates",
"client":{
"ip":"200.249.12.31",
"port":123
},
"server":{
"ip":"200.249.12.31",
"port":80
},
"request":{
"uri":"/?key=_abc123_",
"method":"GET"
},
"response":{
"headers":{
"Date":"Mon, 13 Jul 2015 20:02:41 GMT",
"Last-Modified":"Sun, 26 Oct 2014 22:33:37 GMT",
"Content-Type":"text/html"
},
"body":[
"no need."
]
},
"expected":{
"debug_log":"Added regex subexpression TX\\.2: abc[\\s\\S]*Added regex subexpression TX\\.3: 123"
},
"rules":[
"SecRuleEngine On",
"SecRule ARGS \"@rx _((?:(abc)|(123))+)_\" \"id:18,phase:1,log,pass,capture\""
]
},
{
"enabled":1,
"version_min":300000,
"title":"Testing Variables :: repeating capture group -- same (nested)",
"client":{
"ip":"200.249.12.31",
"port":123
},
"server":{
"ip":"200.249.12.31",
"port":80
},
"request":{
"uri":"/?key=a:5a:8a:9",
"method":"GET"
},
"response":{
"headers":{
"Date":"Mon, 13 Jul 2015 20:02:41 GMT",
"Last-Modified":"Sun, 26 Oct 2014 22:33:37 GMT",
"Content-Type":"text/html"
},
"body":[
"no need."
]
},
"expected":{
"debug_log":"Added regex subexpression TX\\.1: 5[\\s\\S]*Added regex subexpression TX\\.2: 8[\\s\\S]*Added regex subexpression TX\\.3: 9"
},
"rules":[
"SecRuleEngine On",
"SecRule ARGS \"@rx a:([0-9])(?:a:([0-9])(?:a:([0-9]))*)*\" \"id:18,phase:1,log,pass,capture\""
]
}
]