Migrate from PCRE to PCRE2

Replace legacy PCRE (libpcre3) with PCRE2 (libpcre2-8). PCRE1 has been
end-of-life since 2021.

- src/rule.h: use pcre2_code and pcre2_match_data types
- src/rule.c: pcre_compile -> pcre2_compile, pcre_exec -> pcre2_match,
  pcre_free -> pcre2_code_free, with proper error message retrieval
- cmake/FindPCRE.cmake -> cmake/FindPCRE2.cmake: find libpcre2-8 via
  pkg-config, pcre2-config, or manual search
- Update all CMakeLists.txt, config.h.cmake, configure.cmake, README.md

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

CMakeLists.txt

@@ -47,7 +47,7 @@ if (NOT WITH_EMBEDDED_SRC)
 endif ()
 
 # Find dependencies via Find modules
-find_package(PCRE REQUIRED)
+find_package(PCRE2 REQUIRED)
 find_package(MbedTLS REQUIRED)
 find_package(Sodium REQUIRED)
 find_package(Cares REQUIRED)

README.md

@@ -208,7 +208,7 @@ In general, you need the following build dependencies:
 * pkg-config
 * libmbedtls
 * libsodium (>= 1.0.4)
-* libpcre3 (old pcre library)
+* libpcre2
 * libev
 * libc-ares
 * asciidoc (for documentation only)
@@ -225,7 +225,7 @@ For some of the distributions, you might install build dependencies like this:
 ```bash
 # Installation of basic build dependencies
 ## Debian / Ubuntu
-sudo apt-get install --no-install-recommends build-essential cmake libpcre3-dev asciidoc xmlto libev-dev libc-ares-dev libmbedtls-dev libsodium-dev pkg-config
+sudo apt-get install --no-install-recommends build-essential cmake libpcre2-dev asciidoc xmlto libev-dev libc-ares-dev libmbedtls-dev libsodium-dev pkg-config
 ## CentOS / Fedora / RHEL
 sudo yum install gcc cmake make asciidoc xmlto c-ares-devel libev-devel
 ## Arch

cmake/FindPCRE.cmake

@@ -1,60 +0,0 @@
-# FindPCRE.cmake - Find PCRE library
-#
-# Sets:
-#   PCRE_FOUND
-#   PCRE_INCLUDE_DIRS
-#   PCRE_LIBRARIES
-
-include(FindPkgConfig)
-
-if(PKG_CONFIG_FOUND)
-    pkg_check_modules(_PCRE QUIET libpcre)
-endif()
-
-if(_PCRE_FOUND)
-    set(PCRE_INCLUDE_DIRS ${_PCRE_INCLUDE_DIRS})
-    set(PCRE_LIBRARIES ${_PCRE_LIBRARIES})
-    set(PCRE_FOUND TRUE)
-else()
-    # Try pcre-config
-    find_program(PCRE_CONFIG pcre-config)
-    if(PCRE_CONFIG)
-        execute_process(COMMAND ${PCRE_CONFIG} --cflags
-            OUTPUT_VARIABLE PCRE_CFLAGS
-            OUTPUT_STRIP_TRAILING_WHITESPACE)
-        execute_process(COMMAND ${PCRE_CONFIG} --libs
-            OUTPUT_VARIABLE PCRE_LDFLAGS
-            OUTPUT_STRIP_TRAILING_WHITESPACE)
-        string(REGEX REPLACE "-I" "" PCRE_INCLUDE_DIRS "${PCRE_CFLAGS}")
-        set(PCRE_LIBRARIES ${PCRE_LDFLAGS})
-        set(PCRE_FOUND TRUE)
-    else()
-        # Manual search
-        find_path(PCRE_INCLUDE_DIR
-            NAMES pcre.h
-            PATHS /usr/include /usr/local/include)
-        if(NOT PCRE_INCLUDE_DIR)
-            find_path(PCRE_INCLUDE_DIR
-                NAMES pcre/pcre.h
-                PATHS /usr/include /usr/local/include)
-        endif()
-
-        find_library(PCRE_LIBRARY
-            NAMES pcre
-            PATHS /usr/lib /usr/local/lib)
-
-        if(PCRE_INCLUDE_DIR AND PCRE_LIBRARY)
-            set(PCRE_INCLUDE_DIRS ${PCRE_INCLUDE_DIR})
-            set(PCRE_LIBRARIES ${PCRE_LIBRARY})
-            set(PCRE_FOUND TRUE)
-        else()
-            set(PCRE_FOUND FALSE)
-        endif()
-    endif()
-endif()
-
-if(PCRE_FOUND)
-    message(STATUS "Found PCRE: ${PCRE_LIBRARIES}")
-else()
-    message(FATAL_ERROR "Could not find PCRE library. Install libpcre3-dev or equivalent.")
-endif()

cmake/FindPCRE2.cmake

@@ -0,0 +1,63 @@
+# FindPCRE2.cmake - Find PCRE2 library (8-bit)
+#
+# Sets:
+#   PCRE2_FOUND
+#   PCRE2_INCLUDE_DIRS
+#   PCRE2_LIBRARIES
+
+include(FindPkgConfig)
+
+if(PKG_CONFIG_FOUND)
+    pkg_check_modules(_PCRE2 QUIET libpcre2-8)
+endif()
+
+if(_PCRE2_FOUND)
+    set(PCRE2_INCLUDE_DIRS ${_PCRE2_INCLUDE_DIRS})
+    set(PCRE2_LIBRARIES ${_PCRE2_LIBRARIES})
+    set(PCRE2_FOUND TRUE)
+else()
+    # Try pcre2-config
+    find_program(PCRE2_CONFIG pcre2-config)
+    if(PCRE2_CONFIG)
+        execute_process(COMMAND ${PCRE2_CONFIG} --cflags
+            OUTPUT_VARIABLE PCRE2_CFLAGS
+            OUTPUT_STRIP_TRAILING_WHITESPACE)
+        execute_process(COMMAND ${PCRE2_CONFIG} --libs8
+            OUTPUT_VARIABLE PCRE2_LDFLAGS
+            OUTPUT_STRIP_TRAILING_WHITESPACE)
+        string(REGEX REPLACE "-I" "" PCRE2_INCLUDE_DIRS "${PCRE2_CFLAGS}")
+        set(PCRE2_LIBRARIES ${PCRE2_LDFLAGS})
+        set(PCRE2_FOUND TRUE)
+    else()
+        # Manual search
+        find_path(PCRE2_INCLUDE_DIR
+            NAMES pcre2.h
+            HINTS
+                /opt/homebrew/include
+                /usr/local/include
+                /usr/include
+        )
+
+        find_library(PCRE2_LIBRARY
+            NAMES pcre2-8
+            HINTS
+                /opt/homebrew/lib
+                /usr/local/lib
+                /usr/lib
+        )
+
+        if(PCRE2_INCLUDE_DIR AND PCRE2_LIBRARY)
+            set(PCRE2_INCLUDE_DIRS ${PCRE2_INCLUDE_DIR})
+            set(PCRE2_LIBRARIES ${PCRE2_LIBRARY})
+            set(PCRE2_FOUND TRUE)
+        else()
+            set(PCRE2_FOUND FALSE)
+        endif()
+    endif()
+endif()
+
+if(PCRE2_FOUND)
+    message(STATUS "Found PCRE2: ${PCRE2_LIBRARIES}")
+else()
+    message(FATAL_ERROR "Could not find PCRE2 library. Install libpcre2-dev or equivalent.")
+endif()

cmake/config.h.cmake

@@ -87,11 +87,8 @@
 /* Define to 1 if you have the <net/if.h> header file. */
 #cmakedefine HAVE_NET_IF_H 1
 
-/* Define to 1 if you have the <pcre.h> header file. */
-#cmakedefine HAVE_PCRE_H 1
-
-/* Define to 1 if you have the <pcre/pcre.h> header file. */
-#cmakedefine HAVE_PCRE_PCRE_H 1
+/* Define to 1 if you have the <pcre2.h> header file. */
+#cmakedefine HAVE_PCRE2_H 1
 
 /* Have PTHREAD_PRIO_INHERIT. */
 #cmakedefine HAVE_PTHREAD_PRIO_INHERIT 1

cmake/configure.cmake

@@ -56,8 +56,10 @@ else ()
 endif ()
 check_include_files(linux/tcp.h HAVE_LINUX_TCP_H)
 check_include_files(net/if.h HAVE_NET_IF_H)
-check_include_files(pcre.h HAVE_PCRE_H)
-check_include_files(pcre/pcre.h HAVE_PCRE_PCRE_H)
+set(CMAKE_REQUIRED_DEFINITIONS_SAVE ${CMAKE_REQUIRED_DEFINITIONS})
+set(CMAKE_REQUIRED_DEFINITIONS "-DPCRE2_CODE_UNIT_WIDTH=8")
+check_include_files(pcre2.h HAVE_PCRE2_H)
+set(CMAKE_REQUIRED_DEFINITIONS ${CMAKE_REQUIRED_DEFINITIONS_SAVE})
 check_symbol_exists(PTHREAD_PRIO_INHERIT pthread.h HAVE_PTHREAD_PRIO_INHERIT)
 
 check_function_exists(select HAVE_SELECT)

src/CMakeLists.txt

@@ -92,7 +92,7 @@ set(SS_WERROR_FLAGS -Werror)
 # that may be in /opt/homebrew/include (e.g. mbedtls 4.x vs 3.x)
 include_directories(BEFORE ${MBEDTLS_INCLUDE_DIRS})
 include_directories(${SODIUM_INCLUDE_DIRS})
-include_directories(${PCRE_INCLUDE_DIRS})
+include_directories(${PCRE2_INCLUDE_DIRS})
 include_directories(${CARES_INCLUDE_DIRS})
 
 # Extract mbedTLS lib dir to ensure static/shared libs come from same version
@@ -106,7 +106,7 @@ find_library(LIBMBEDCRYPTO libmbedcrypto.a HINTS ${_MBEDTLS_LIB_DIR} NO_DEFAULT_
 find_library(LIBMBEDCRYPTO libmbedcrypto.a)
 find_library(LIBEV libev.a)
 find_library(LIBUDNS libcares.a)
-find_library(LIBPCRE libpcre.a)
+find_library(LIBPCRE2 libpcre2-8.a)
 
 # Dependencies we need for static and shared
 list(APPEND DEPS
@@ -114,7 +114,7 @@ list(APPEND DEPS
         bloom
         ${LIBEV}
         ${LIBUDNS}
-        ${LIBPCRE}
+        ${LIBPCRE2}
         ${LIBSODIUM}
         ${LIBMBEDTLS}
         ${LIBMBEDCRYPTO}
@@ -122,7 +122,7 @@ list(APPEND DEPS
 
 if (MINGW)
 list(APPEND DEPS ws2_32 iphlpapi)
-add_compile_definitions(CARES_STATICLIB PCRE_STATIC)
+add_compile_definitions(CARES_STATICLIB PCRE2_STATIC)
 endif ()
 endif ()
 
@@ -133,7 +133,7 @@ find_library(LIBMBEDCRYPTO_SHARED mbedcrypto HINTS ${_MBEDTLS_LIB_DIR} NO_DEFAUL
 find_library(LIBMBEDCRYPTO_SHARED mbedcrypto)
 find_library(LIBEV_SHARED ev)
 find_library(LIBUDNS_SHARED cares)
-find_library(LIBPCRE_SHARED pcre)
+find_library(LIBPCRE2_SHARED pcre2-8)
 
 if (WITH_EMBEDDED_SRC)
 list(APPEND DEPS_SHARED
@@ -143,7 +143,7 @@ list(APPEND DEPS_SHARED
         ipset-shared
         ${LIBEV_SHARED}
         ${LIBUDNS_SHARED}
-        ${LIBPCRE_SHARED}
+        ${LIBPCRE2_SHARED}
         ${LIBSODIUM_SHARED}
         ${LIBMBEDTLS_SHARED}
         ${LIBMBEDCRYPTO_SHARED}
@@ -159,7 +159,7 @@ list(APPEND DEPS_SHARED
         ${LIBCORKIPSET_SHARED}
         ${LIBEV_SHARED}
         ${LIBUDNS_SHARED}
-        ${LIBPCRE_SHARED}
+        ${LIBPCRE2_SHARED}
         ${LIBSODIUM_SHARED}
         ${LIBMBEDTLS_SHARED}
         ${LIBMBEDCRYPTO_SHARED}

src/rule.c

@@ -78,16 +78,21 @@ int
 init_rule(rule_t *rule)
 {
     if (rule->pattern_re == NULL) {
-        const char *reerr;
-        int reerroffset;
+        int errcode;
+        PCRE2_SIZE erroffset;
 
         rule->pattern_re =
-            pcre_compile(rule->pattern, 0, &reerr, &reerroffset, NULL);
+            pcre2_compile((PCRE2_SPTR)rule->pattern, PCRE2_ZERO_TERMINATED,
+                          0, &errcode, &erroffset, NULL);
         if (rule->pattern_re == NULL) {
+            PCRE2_UCHAR errbuf[256];
+            pcre2_get_error_message(errcode, errbuf, sizeof(errbuf));
             LOGE("Regex compilation of \"%s\" failed: %s, offset %d",
-                 rule->pattern, reerr, reerroffset);
+                 rule->pattern, errbuf, (int)erroffset);
             return 0;
         }
+        rule->match_data = pcre2_match_data_create_from_pattern(
+            rule->pattern_re, NULL);
     }
 
     return 1;
@@ -105,8 +110,8 @@ lookup_rule(const struct cork_dllist *rules, const char *name, size_t name_len)
 
     cork_dllist_foreach_void(rules, curr, next) {
         rule_t *rule = cork_container_of(curr, rule_t, entries);
-        if (pcre_exec(rule->pattern_re, NULL,
-                      name, name_len, 0, 0, NULL, 0) >= 0)
+        if (pcre2_match(rule->pattern_re, (PCRE2_SPTR)name,
+                        name_len, 0, 0, rule->match_data, NULL) >= 0)
             return rule;
     }
 
@@ -127,7 +132,9 @@ free_rule(rule_t *rule)
         return;
 
     ss_free(rule->pattern);
+    if (rule->match_data != NULL)
+        pcre2_match_data_free(rule->match_data);
     if (rule->pattern_re != NULL)
-        pcre_free(rule->pattern_re);
+        pcre2_code_free(rule->pattern_re);
     ss_free(rule);
 }

src/rule.h

@@ -33,17 +33,15 @@
 
 #include <libcork/ds.h>
 
-#ifdef HAVE_PCRE_H
-#include <pcre.h>
-#elif HAVE_PCRE_PCRE_H
-#include <pcre/pcre.h>
-#endif
+#define PCRE2_CODE_UNIT_WIDTH 8
+#include <pcre2.h>
 
 typedef struct rule {
     char *pattern;
 
     /* Runtime fields */
-    pcre *pattern_re;
+    pcre2_code *pattern_re;
+    pcre2_match_data *match_data;
 
     struct cork_dllist_item entries;
 } rule_t;

tests/CMakeLists.txt

@@ -3,7 +3,7 @@ include_directories(BEFORE ${MBEDTLS_INCLUDE_DIRS})
 include_directories(${PROJECT_SOURCE_DIR}/src)
 include_directories(${PROJECT_BINARY_DIR}/src)
 include_directories(${SODIUM_INCLUDE_DIRS})
-include_directories(${PCRE_INCLUDE_DIRS})
+include_directories(${PCRE2_INCLUDE_DIRS})
 include_directories(${CARES_INCLUDE_DIRS})
 
 if(WITH_EMBEDDED_SRC)
@@ -69,7 +69,7 @@ ss_add_test(test_ppbloom
 
 # test_rule - needs pcre and cork
 set(TEST_RULE_LIBS
-    ${LIBPCRE_SHARED}
+    ${LIBPCRE2_SHARED}
     ${LIBSODIUM_SHARED}
 )
 if(WITH_EMBEDDED_SRC)

tests/test_rule.c

@@ -52,8 +52,10 @@ test_init_rule(void)
     (void)ret;
     assert(rule->pattern_re != NULL);
 
+    if (rule->match_data)
+        pcre2_match_data_free(rule->match_data);
     if (rule->pattern_re)
-        pcre_free(rule->pattern_re);
+        pcre2_code_free(rule->pattern_re);
     free(rule->pattern);
     free(rule);
 }
@@ -102,10 +104,12 @@ test_lookup_rule(void)
     (void)found;
 
     /* Clean up */
-    if (rule1->pattern_re) pcre_free(rule1->pattern_re);
+    if (rule1->match_data) pcre2_match_data_free(rule1->match_data);
+    if (rule1->pattern_re) pcre2_code_free(rule1->pattern_re);
     free(rule1->pattern);
     free(rule1);
-    if (rule2->pattern_re) pcre_free(rule2->pattern_re);
+    if (rule2->match_data) pcre2_match_data_free(rule2->match_data);
+    if (rule2->pattern_re) pcre2_code_free(rule2->pattern_re);
     free(rule2->pattern);
     free(rule2);
 }