Replace my bounds-checked numeric conversion with Boost’s

Boost provides numeric_cast, which is much better than what I was using for
safe numeric type conversion.  This does introduce a Boost dependency, but that
tends to be true of most nontrivial C++ programs, so it’s pretty reasonable.

4 files changed, 291 insertions, 23 deletions

diff --git a/Makefile.am b/Makefile.am
index 3cc46e9..cba5a94 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -16,7 +16,8 @@ ACLOCAL_AMFLAGS = -I m4
 
 CPPFLAGS = \
 	-Wall \
-	-D_FORTIFY_SOURCE=2
+	-D_FORTIFY_SOURCE=2 \
+	$(BOOST_CPPFLAGS)
 
 CXXFLAGS = \
 	-std=c++11 \
@@ -53,7 +54,8 @@ liburweb_regex_la_DATA = \
 
 liburweb_regex_la_LIBADD = -lurweb
 liburweb_regex_la_LDFLAGS = \
-	-export-symbols-regex '^uw_Regex_'
+	-export-symbols-regex '^uw_Regex_' \
+	$(BOOST_LDFLAGS)
 
 liburweb_regex_ladir = $(datadir)/urweb/ur/regex
 
diff --git a/configure.ac b/configure.ac
index 9140509..763b8e4 100644
--- a/configure.ac
+++ b/configure.ac
@@ -24,6 +24,10 @@ AC_CONFIG_HEADERS([config.h])
 AC_PROG_CXX([clang++])
 AX_CXX_COMPILE_STDCXX_11([noext], [mandatory])
 
+AX_BOOST_BASE([1.48.0],
+  [],
+  [AC_MSG_FAILURE([Boost is required])])
+
 AC_CHECK_HEADERS([urweb/urweb_cpp.h],
   [],
   [AC_MSG_FAILURE([Ur/Web headers are required])])
diff --git a/m4/ax_boost_base.m4 b/m4/ax_boost_base.m4
new file mode 100644
index 0000000..8e6ee9a
--- /dev/null
+++ b/m4/ax_boost_base.m4
@@ -0,0 +1,272 @@
+# ===========================================================================
+#       http://www.gnu.org/software/autoconf-archive/ax_boost_base.html
+# ===========================================================================
+#
+# SYNOPSIS
+#
+#   AX_BOOST_BASE([MINIMUM-VERSION], [ACTION-IF-FOUND], [ACTION-IF-NOT-FOUND])
+#
+# DESCRIPTION
+#
+#   Test for the Boost C++ libraries of a particular version (or newer)
+#
+#   If no path to the installed boost library is given the macro searchs
+#   under /usr, /usr/local, /opt and /opt/local and evaluates the
+#   $BOOST_ROOT environment variable. Further documentation is available at
+#   <http://randspringer.de/boost/index.html>.
+#
+#   This macro calls:
+#
+#     AC_SUBST(BOOST_CPPFLAGS) / AC_SUBST(BOOST_LDFLAGS)
+#
+#   And sets:
+#
+#     HAVE_BOOST
+#
+# LICENSE
+#
+#   Copyright (c) 2008 Thomas Porschberg <thomas@randspringer.de>
+#   Copyright (c) 2009 Peter Adolphs
+#
+#   Copying and distribution of this file, with or without modification, are
+#   permitted in any medium without royalty provided the copyright notice
+#   and this notice are preserved. This file is offered as-is, without any
+#   warranty.
+
+#serial 23
+
+AC_DEFUN([AX_BOOST_BASE],
+[
+AC_ARG_WITH([boost],
+  [AS_HELP_STRING([--with-boost@<:@=ARG@:>@],
+    [use Boost library from a standard location (ARG=yes),
+     from the specified location (ARG=<path>),
+     or disable it (ARG=no)
+     @<:@ARG=yes@:>@ ])],
+    [
+    if test "$withval" = "no"; then
+        want_boost="no"
+    elif test "$withval" = "yes"; then
+        want_boost="yes"
+        ac_boost_path=""
+    else
+        want_boost="yes"
+        ac_boost_path="$withval"
+    fi
+    ],
+    [want_boost="yes"])
+
+
+AC_ARG_WITH([boost-libdir],
+        AS_HELP_STRING([--with-boost-libdir=LIB_DIR],
+        [Force given directory for boost libraries. Note that this will override library path detection, so use this parameter only if default library detection fails and you know exactly where your boost libraries are located.]),
+        [
+        if test -d "$withval"
+        then
+                ac_boost_lib_path="$withval"
+        else
+                AC_MSG_ERROR(--with-boost-libdir expected directory name)
+        fi
+        ],
+        [ac_boost_lib_path=""]
+)
+
+if test "x$want_boost" = "xyes"; then
+    boost_lib_version_req=ifelse([$1], ,1.20.0,$1)
+    boost_lib_version_req_shorten=`expr $boost_lib_version_req : '\([[0-9]]*\.[[0-9]]*\)'`
+    boost_lib_version_req_major=`expr $boost_lib_version_req : '\([[0-9]]*\)'`
+    boost_lib_version_req_minor=`expr $boost_lib_version_req : '[[0-9]]*\.\([[0-9]]*\)'`
+    boost_lib_version_req_sub_minor=`expr $boost_lib_version_req : '[[0-9]]*\.[[0-9]]*\.\([[0-9]]*\)'`
+    if test "x$boost_lib_version_req_sub_minor" = "x" ; then
+        boost_lib_version_req_sub_minor="0"
+        fi
+    WANT_BOOST_VERSION=`expr $boost_lib_version_req_major \* 100000 \+  $boost_lib_version_req_minor \* 100 \+ $boost_lib_version_req_sub_minor`
+    AC_MSG_CHECKING(for boostlib >= $boost_lib_version_req)
+    succeeded=no
+
+    dnl On 64-bit systems check for system libraries in both lib64 and lib.
+    dnl The former is specified by FHS, but e.g. Debian does not adhere to
+    dnl this (as it rises problems for generic multi-arch support).
+    dnl The last entry in the list is chosen by default when no libraries
+    dnl are found, e.g. when only header-only libraries are installed!
+    libsubdirs="lib"
+    ax_arch=`uname -m`
+    case $ax_arch in
+      x86_64|ppc64|s390x|sparc64|aarch64)
+        libsubdirs="lib64 lib lib64"
+        ;;
+    esac
+
+    dnl allow for real multi-arch paths e.g. /usr/lib/x86_64-linux-gnu. Give
+    dnl them priority over the other paths since, if libs are found there, they
+    dnl are almost assuredly the ones desired.
+    AC_REQUIRE([AC_CANONICAL_HOST])
+    libsubdirs="lib/${host_cpu}-${host_os} $libsubdirs"
+
+    case ${host_cpu} in
+      i?86)
+        libsubdirs="lib/i386-${host_os} $libsubdirs"
+        ;;
+    esac
+
+    dnl first we check the system location for boost libraries
+    dnl this location ist chosen if boost libraries are installed with the --layout=system option
+    dnl or if you install boost with RPM
+    if test "$ac_boost_path" != ""; then
+        BOOST_CPPFLAGS="-I$ac_boost_path/include"
+        for ac_boost_path_tmp in $libsubdirs; do
+                if test -d "$ac_boost_path"/"$ac_boost_path_tmp" ; then
+                        BOOST_LDFLAGS="-L$ac_boost_path/$ac_boost_path_tmp"
+                        break
+                fi
+        done
+    elif test "$cross_compiling" != yes; then
+        for ac_boost_path_tmp in /usr /usr/local /opt /opt/local ; do
+            if test -d "$ac_boost_path_tmp/include/boost" && test -r "$ac_boost_path_tmp/include/boost"; then
+                for libsubdir in $libsubdirs ; do
+                    if ls "$ac_boost_path_tmp/$libsubdir/libboost_"* >/dev/null 2>&1 ; then break; fi
+                done
+                BOOST_LDFLAGS="-L$ac_boost_path_tmp/$libsubdir"
+                BOOST_CPPFLAGS="-I$ac_boost_path_tmp/include"
+                break;
+            fi
+        done
+    fi
+
+    dnl overwrite ld flags if we have required special directory with
+    dnl --with-boost-libdir parameter
+    if test "$ac_boost_lib_path" != ""; then
+       BOOST_LDFLAGS="-L$ac_boost_lib_path"
+    fi
+
+    CPPFLAGS_SAVED="$CPPFLAGS"
+    CPPFLAGS="$CPPFLAGS $BOOST_CPPFLAGS"
+    export CPPFLAGS
+
+    LDFLAGS_SAVED="$LDFLAGS"
+    LDFLAGS="$LDFLAGS $BOOST_LDFLAGS"
+    export LDFLAGS
+
+    AC_REQUIRE([AC_PROG_CXX])
+    AC_LANG_PUSH(C++)
+        AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
+    @%:@include <boost/version.hpp>
+    ]], [[
+    #if BOOST_VERSION >= $WANT_BOOST_VERSION
+    // Everything is okay
+    #else
+    #  error Boost version is too old
+    #endif
+    ]])],[
+        AC_MSG_RESULT(yes)
+    succeeded=yes
+    found_system=yes
+        ],[
+        ])
+    AC_LANG_POP([C++])
+
+
+
+    dnl if we found no boost with system layout we search for boost libraries
+    dnl built and installed without the --layout=system option or for a staged(not installed) version
+    if test "x$succeeded" != "xyes"; then
+        _version=0
+        if test "$ac_boost_path" != ""; then
+            if test -d "$ac_boost_path" && test -r "$ac_boost_path"; then
+                for i in `ls -d $ac_boost_path/include/boost-* 2>/dev/null`; do
+                    _version_tmp=`echo $i | sed "s#$ac_boost_path##" | sed 's/\/include\/boost-//' | sed 's/_/./'`
+                    V_CHECK=`expr $_version_tmp \> $_version`
+                    if test "$V_CHECK" = "1" ; then
+                        _version=$_version_tmp
+                    fi
+                    VERSION_UNDERSCORE=`echo $_version | sed 's/\./_/'`
+                    BOOST_CPPFLAGS="-I$ac_boost_path/include/boost-$VERSION_UNDERSCORE"
+                done
+            fi
+        else
+            if test "$cross_compiling" != yes; then
+                for ac_boost_path in /usr /usr/local /opt /opt/local ; do
+                    if test -d "$ac_boost_path" && test -r "$ac_boost_path"; then
+                        for i in `ls -d $ac_boost_path/include/boost-* 2>/dev/null`; do
+                            _version_tmp=`echo $i | sed "s#$ac_boost_path##" | sed 's/\/include\/boost-//' | sed 's/_/./'`
+                            V_CHECK=`expr $_version_tmp \> $_version`
+                            if test "$V_CHECK" = "1" ; then
+                                _version=$_version_tmp
+                                best_path=$ac_boost_path
+                            fi
+                        done
+                    fi
+                done
+
+                VERSION_UNDERSCORE=`echo $_version | sed 's/\./_/'`
+                BOOST_CPPFLAGS="-I$best_path/include/boost-$VERSION_UNDERSCORE"
+                if test "$ac_boost_lib_path" = ""; then
+                    for libsubdir in $libsubdirs ; do
+                        if ls "$best_path/$libsubdir/libboost_"* >/dev/null 2>&1 ; then break; fi
+                    done
+                    BOOST_LDFLAGS="-L$best_path/$libsubdir"
+                fi
+            fi
+
+            if test "x$BOOST_ROOT" != "x"; then
+                for libsubdir in $libsubdirs ; do
+                    if ls "$BOOST_ROOT/stage/$libsubdir/libboost_"* >/dev/null 2>&1 ; then break; fi
+                done
+                if test -d "$BOOST_ROOT" && test -r "$BOOST_ROOT" && test -d "$BOOST_ROOT/stage/$libsubdir" && test -r "$BOOST_ROOT/stage/$libsubdir"; then
+                    version_dir=`expr //$BOOST_ROOT : '.*/\(.*\)'`
+                    stage_version=`echo $version_dir | sed 's/boost_//' | sed 's/_/./g'`
+                        stage_version_shorten=`expr $stage_version : '\([[0-9]]*\.[[0-9]]*\)'`
+                    V_CHECK=`expr $stage_version_shorten \>\= $_version`
+                    if test "$V_CHECK" = "1" -a "$ac_boost_lib_path" = "" ; then
+                        AC_MSG_NOTICE(We will use a staged boost library from $BOOST_ROOT)
+                        BOOST_CPPFLAGS="-I$BOOST_ROOT"
+                        BOOST_LDFLAGS="-L$BOOST_ROOT/stage/$libsubdir"
+                    fi
+                fi
+            fi
+        fi
+
+        CPPFLAGS="$CPPFLAGS $BOOST_CPPFLAGS"
+        export CPPFLAGS
+        LDFLAGS="$LDFLAGS $BOOST_LDFLAGS"
+        export LDFLAGS
+
+        AC_LANG_PUSH(C++)
+            AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
+        @%:@include <boost/version.hpp>
+        ]], [[
+        #if BOOST_VERSION >= $WANT_BOOST_VERSION
+        // Everything is okay
+        #else
+        #  error Boost version is too old
+        #endif
+        ]])],[
+            AC_MSG_RESULT(yes)
+        succeeded=yes
+        found_system=yes
+            ],[
+            ])
+        AC_LANG_POP([C++])
+    fi
+
+    if test "$succeeded" != "yes" ; then
+        if test "$_version" = "0" ; then
+            AC_MSG_NOTICE([[We could not detect the boost libraries (version $boost_lib_version_req_shorten or higher). If you have a staged boost library (still not installed) please specify \$BOOST_ROOT in your environment and do not give a PATH to --with-boost option.  If you are sure you have boost installed, then check your version number looking in <boost/version.hpp>. See http://randspringer.de/boost for more documentation.]])
+        else
+            AC_MSG_NOTICE([Your boost libraries seems to old (version $_version).])
+        fi
+        # execute ACTION-IF-NOT-FOUND (if present):
+        ifelse([$3], , :, [$3])
+    else
+        AC_SUBST(BOOST_CPPFLAGS)
+        AC_SUBST(BOOST_LDFLAGS)
+        AC_DEFINE(HAVE_BOOST,,[define if the Boost library is available])
+        # execute ACTION-IF-FOUND (if present):
+        ifelse([$2], , :, [$2])
+    fi
+
+    CPPFLAGS="$CPPFLAGS_SAVED"
+    LDFLAGS="$LDFLAGS_SAVED"
+fi
+
+])
diff --git a/src/regex__FFI.cc b/src/regex__FFI.cc
index 0ea3455..c2f8a4b 100644
--- a/src/regex__FFI.cc
+++ b/src/regex__FFI.cc
@@ -14,12 +14,10 @@
 
 #include "regex__FFI.h"
 
-#include <regex.h>
-
 #include <cstring>
-#include <limits>
 #include <regex>
 
+#include <boost/numeric/conversion/cast.hpp>
 extern "C" {
 #include <urweb/urweb_cpp.h>
 }
@@ -54,18 +52,13 @@ void DeleteMatchResults(void* match_result,
 }
 
 // Bounds-checked numeric type conversion
-template<typename T, typename U>
-U Number(uw_context* const context, const T input) {
-  Assert(context, input <= std::numeric_limits<U>::max(),
-         "regex: detected overflow during numeric conversion");
-  if (std::numeric_limits<T>::is_signed == std::numeric_limits<U>::is_signed) {
-    Assert(context, std::numeric_limits<U>::lowest() <= input,
-           "regex: detected underflow during numeric conversion");
-  } else if (std::numeric_limits<T>::is_signed) {
-    Assert(context, 0 <= input,
-           "regex: detected underflow during numeric conversion");
+template<typename Target, typename Source>
+Target Number(uw_context* const context, Source arg) {
+  try {
+    return boost::numeric_cast<Target>(arg);
+  } catch (const boost::numeric::bad_numeric_cast& e) {
+    uw_error(context, FATAL, "regex: %s", e.what());
   }
-  return static_cast<U>(input);
 }
 
 }  // namespace
@@ -90,7 +83,7 @@ uw_Basis_int uw_Regex__FFI_n_subexpression_matches(
   } else {
     // At least one match occurred.  Compute the number of parenthesized
     // subexpressions that got matched, and return it.
-    return Number<std::cmatch::size_type, uw_Basis_int>(context, n_matches) - 1;
+    return Number<uw_Basis_int>(context, n_matches) - 1;
   }
 }
 
@@ -101,18 +94,15 @@ uw_Basis_string uw_Regex__FFI_subexpression_match(
   const std::cmatch* const match_result =
       reinterpret_cast<std::cmatch*>(match.result);
   const std::size_t match_index =
-      Number<uw_Basis_int, std::size_t>(context, match_index_signed);
+      Number<std::size_t>(context, match_index_signed);
   Assert(context, match_index < match_result->size(),
          "regex: match does not exist");
   const auto matched_substring = (*match_result)[match_index + 1];
   // Save the matched substring.
   const std::size_t result_length =
-      Number<std::csub_match::difference_type, std::size_t>(
-          context,
-          matched_substring.length());
+      Number<std::size_t>(context,matched_substring.length());
   uw_Basis_string result =
-      reinterpret_cast<uw_Basis_string>(
-          uw_malloc(context, result_length + 1));
+      reinterpret_cast<uw_Basis_string>(uw_malloc(context, result_length + 1));
   std::strcpy(result, matched_substring.str().c_str());
   return result;
 }