Add support for PCRE2 pattern matching

* configure.ac: Check for libpcre2-8
* src/init.c (choices): Test for HAVE_LIBPCRE2
* src/main.c (main): Set regex compile and match functions
* src/options.h: Test for HAVE_LIBPCRE2
* src/utils.c: Include pcre2.h, add functions
  compile_pcre2_regex() and match_pcre2_regex()
* src/utils.h: Declare compile_pcre2_regex() and match_pcre2_regex()

Fixes #54677
Reported-by: Noël Köthe

configure.ac

@@ -715,18 +715,39 @@ AS_IF([test "x$uuid_mode" = x1], [
 ])
 
 dnl
-dnl Check for PCRE
+dnl Check for PCRE2 / PCRE
 dnl
 
+AC_ARG_ENABLE(pcre2, AC_HELP_STRING([--disable-pcre2],
+                                   [Disable PCRE2 style regular expressions]))
 AC_ARG_ENABLE(pcre, AC_HELP_STRING([--disable-pcre],
                                    [Disable PCRE style regular expressions]))
 
-AS_IF([test "X$enable_pcre" != "Xno"],[
+AS_IF([test "X$enable_pcre2" != "Xno"],[
+  enable_pcre2=no
+  PKG_CHECK_MODULES([PCRE2], libpcre2-8, [
+    CFLAGS="$PCRE2_CFLAGS $CFLAGS"
+    LIBS="$PCRE2_LIBS $LIBS"
+    AC_DEFINE([HAVE_LIBPCRE2], [1], [Define if libpcre2 is available.])
+    enable_pcre2=yes
+  ], [
+    AC_CHECK_HEADER(pcre2.h, [
+      AC_CHECK_LIB(pcre2-8, pcre2_compile_8, [
+        LIBS="${LIBS} -lpcre2-8"
+        AC_DEFINE([HAVE_LIBPCRE2], 1, [Define if libpcre2 is available.])
+        enable_pcre2=yes
+      ])
+    ])
+  ])
+])
+
+AS_IF([test "X$enable_pcre" != "Xno" && test "X$enable_pcre2" != "Xyes"],[
   PKG_CHECK_MODULES([PCRE], libpcre, [
     CFLAGS="$PCRE_CFLAGS $CFLAGS"
     AC_CHECK_HEADER(pcre.h, [
       LIBS="$PCRE_LIBS $LIBS"
       AC_DEFINE([HAVE_LIBPCRE], [1], [Define if libpcre is available.])
+      enable_pcre=yes
     ])
   ], [
     AC_CHECK_HEADER(pcre.h, [
@@ -738,6 +759,8 @@ AS_IF([test "X$enable_pcre" != "Xno"],[
   ])
 ])
 
+AS_IF([test "X$enable_pcre2" = Xyes], [PCRE_INFO="yes, via libpcre2"], [test "X$enable_pcre" = Xyes], [PCRE_INFO="yes, via libpcre"], [PCRE_INFO=no])
+
 dnl
 dnl Check for libcares (resolver library)
 dnl
@@ -817,6 +840,7 @@ AC_MSG_NOTICE([Summary of build options:
   SSL:               $with_ssl
   Zlib:              $with_zlib
   PSL:               $with_libpsl
+  PCRE:              $PCRE_INFO
   Digest:            $ENABLE_DIGEST
   NTLM:              $ENABLE_NTLM
   OPIE:              $ENABLE_OPIE

src/init.c

@@ -47,9 +47,6 @@ as that of the covered work.  */
 #endif
 
 #include <regex.h>
-#ifdef HAVE_LIBPCRE
-# include <pcre.h>
-#endif
 
 #ifdef HAVE_PWD_H
 # include <pwd.h>
@@ -1636,7 +1633,7 @@ cmd_spec_regex_type (const char *com, const char *val, void *place_ignored _GL_U
 {
   static const struct decode_item choices[] = {
     { "posix", regex_type_posix },
-#ifdef HAVE_LIBPCRE
+#if defined HAVE_LIBPCRE || defined HAVE_LIBPCRE2
     { "pcre",  regex_type_pcre },
 #endif
   };

src/main.c

@@ -995,7 +995,7 @@ Recursive accept/reject:\n"),
        --accept-regex=REGEX        regex matching accepted URLs\n"),
     N_("\
        --reject-regex=REGEX        regex matching rejected URLs\n"),
-#ifdef HAVE_LIBPCRE
+#if defined HAVE_LIBPCRE || defined HAVE_LIBPCRE2
     N_("\
        --regex-type=TYPE           regex type (posix|pcre)\n"),
 #else
@@ -1784,6 +1784,12 @@ for details.\n\n"));
   /* Compile the regular expressions.  */
   switch (opt.regex_type)
     {
+#ifdef HAVE_LIBPCRE2
+      case regex_type_pcre:
+        opt.regex_compile_fun = compile_pcre2_regex;
+        opt.regex_match_fun = match_pcre2_regex;
+        break;
+#endif
 #ifdef HAVE_LIBPCRE
       case regex_type_pcre:
         opt.regex_compile_fun = compile_pcre_regex;

src/options.h

@@ -92,7 +92,7 @@ struct options
   void *acceptregex;            /* Patterns to accept (a regex struct). */
   void *rejectregex;            /* Patterns to reject (a regex struct). */
   enum {
-#ifdef HAVE_LIBPCRE
+#if defined HAVE_LIBPCRE || HAVE_LIBPCRE2
     regex_type_pcre,
 #endif
     regex_type_posix

src/utils.c

@@ -68,7 +68,10 @@ as that of the covered work.  */
 #include <setjmp.h>
 
 #include <regex.h>
-#ifdef HAVE_LIBPCRE
+#ifdef HAVE_LIBPCRE2
+# define PCRE2_CODE_UNIT_WIDTH 8
+# include <pcre2.h>
+#elif defined HAVE_LIBPCRE
 # include <pcre.h>
 #endif
 
@@ -2429,6 +2432,23 @@ wget_base64_decode (const char *base64, void *dest, size_t size)
   return n;
 }
 
+#ifdef HAVE_LIBPCRE2
+/* Compiles the PCRE2 regex. */
+void *
+compile_pcre2_regex (const char *str)
+{
+  int errornumber;
+  PCRE2_SIZE erroroffset;
+  pcre2_code *regex = pcre2_compile((PCRE2_SPTR) str, PCRE2_ZERO_TERMINATED, 0, &errornumber, &erroroffset, NULL);
+  if (! regex)
+    {
+      fprintf (stderr, _("Invalid regular expression %s, PCRE2 error %d\n"),
+               quote (str), errornumber);
+    }
+  return regex;
+}
+#endif
+
 #ifdef HAVE_LIBPCRE
 /* Compiles the PCRE regex. */
 void *
@@ -2441,7 +2461,6 @@ compile_pcre_regex (const char *str)
     {
       fprintf (stderr, _("Invalid regular expression %s, %s\n"),
                quote (str), errbuf);
-      return false;
     }
   return regex;
 }
@@ -2473,6 +2492,34 @@ compile_posix_regex (const char *str)
   return regex;
 }
 
+#ifdef HAVE_LIBPCRE2
+/* Matches a PCRE2 regex.  */
+bool
+match_pcre2_regex (const void *regex, const char *str)
+{
+  int rc;
+  pcre2_match_data *match_data;
+
+  match_data = pcre2_match_data_create_from_pattern(regex, NULL);
+
+  if (match_data)
+    {
+      rc = pcre2_match(regex, (PCRE2_SPTR) str, strlen(str), 0, 0, match_data, NULL);
+      pcre2_match_data_free(match_data);
+    }
+  else
+	  rc = PCRE2_ERROR_NOMEMORY;
+
+  if (rc < 0 && rc != PCRE2_ERROR_NOMATCH)
+    {
+      logprintf (LOG_VERBOSE, _("Error while matching %s: %d\n"),
+                 quote (str), rc);
+    }
+
+  return rc >= 0;
+}
+#endif
+
 #ifdef HAVE_LIBPCRE
 #define OVECCOUNT 30
 /* Matches a PCRE regex.  */

src/utils.h

@@ -150,6 +150,11 @@ void xsleep (double);
 size_t wget_base64_encode (const void *, size_t, char *);
 ssize_t wget_base64_decode (const char *, void *, size_t);
 
+#ifdef HAVE_LIBPCRE2
+void *compile_pcre2_regex (const char *);
+bool match_pcre2_regex (const void *, const char *);
+#endif
+
 #ifdef HAVE_LIBPCRE
 void *compile_pcre_regex (const char *);
 bool match_pcre_regex (const void *, const char *);