diff --git a/include/boost/detail/lcast_precision.hpp b/include/boost/detail/lcast_precision.hpp
new file mode 100644
index 0000000..49345d5
--- /dev/null
+++ b/include/boost/detail/lcast_precision.hpp
@@ -0,0 +1,113 @@
+// Copyright Alexander Nasonov & Paul A. Bristow 2006.
+
+// Use, modification and distribution are subject to the
+// Boost Software License, Version 1.0.
+// (See accompanying file LICENSE_1_0.txt
+// or copy at http://www.boost.org/LICENSE_1_0.txt)
+
+#ifndef BOOST_DETAIL_LCAST_SET_PRECISION_HPP_INCLUDED
+#define BOOST_DETAIL_LCAST_SET_PRECISION_HPP_INCLUDED
+
+#include <climits>
+#include <ios>
+#include <limits>
+
+#include <boost/config.hpp>
+#include <boost/integer_traits.hpp>
+
+#ifdef BOOST_NO_LIMITS_COMPILE_TIME_CONSTANTS
+#include <boost/assert.hpp>
+#define BOOST_DETAIL_LCAST_ASSERT(cond) BOOST_ASSERT(cond)
+#else
+#include <boost/static_assert.hpp>
+#define BOOST_DETAIL_LCAST_ASSERT(cond) BOOST_STATIC_ASSERT(cond)
+// Remember, static_cast is evaluated even inside dead branches.
+#endif
+
+namespace boost { namespace detail {
+
+#ifdef _MSC_VER
+#pragma warning (push)
+// conditional expression is constant (stream precision)
+#pragma warning (disable : 4127)
+#endif
+
+template<class T>
+inline std::streamsize lcast_get_precision()
+{
+    typedef std::numeric_limits<T> limits;
+
+    bool const is_floating = limits::is_specialized && !limits::is_exact;
+    if(is_floating)
+    { // Includes all built-in floating-point types, float, double ...
+      // and UDT types for which digits (significand bits) is defined (not zero)
+
+        bool const is_specialized_binary = is_floating &&
+            limits::radix == 2 && limits::digits > 0;
+        bool const is_specialized_decimal = is_floating &&
+            limits::radix == 10 && limits::digits10 > 0;
+        std::streamsize const precision_maxarg =
+            integer_traits<std::streamsize>::const_max;
+        if(is_specialized_binary)
+        { // Floating-point types with
+          // limist::digits defined by the specialization.
+
+            unsigned long const digits = limits::digits;
+            unsigned long const precision = 2UL + digits * 30103UL / 100000UL;
+            // unsigned long is selected because it is at least 32-bits
+            // and thus ULONG_MAX / 30103UL is big enough for all types.
+            BOOST_DETAIL_LCAST_ASSERT(!is_specialized_binary ||
+                    digits < ULONG_MAX / 30103UL &&
+                    precision > limits::digits10 + 0UL &&
+                    precision <= precision_maxarg + 0UL
+                );
+            return precision;
+        }
+        else if(is_specialized_decimal)
+        {   // Decimal Floating-point type, most likely a User Defined Type
+            // rather than a real floating-point hardware type.
+            unsigned int const precision = limits::digits10 + 1U;
+            BOOST_DETAIL_LCAST_ASSERT(!is_specialized_decimal ||
+                    precision <= precision_maxarg + 0UL
+                );
+            return precision;
+        }
+    }
+
+    // Integral type (for which precision has no effect)
+    // or type T for which limits is NOT specialized,
+    // so assume stream precision remains the default 6 decimal digits.
+    // Warning: if your User-defined Floating-point type T is NOT specialized,
+    // then you may lose accuracy by only using 6 decimal digits.
+    // To avoid this, you need to specialize T with either
+    // radix == 2 and digits == the number of significand bits,
+    // OR
+    // radix = 10 and digits10 == the number of decimal digits.
+
+    return 6;
+} // template<class T> bool lcast_set_precision(std::ios_base& stream)
+
+template<class T>
+inline void lcast_set_precision(std::ios_base& stream)
+{
+    stream.precision(lcast_get_precision<T>());
+}
+
+template<class Source, class Target>
+inline void lcast_set_precision(std::ios_base& stream)
+{
+    std::streamsize const s = lcast_get_precision<Source>();
+    std::streamsize const t = lcast_get_precision<Target>();
+    stream.precision(s > t ? s : t);
+}
+
+#ifdef _MSC_VER
+#pragma warning (pop)
+#endif
+
+}}
+
+#undef BOOST_DETAIL_LCAST_ASSERT
+
+#endif //  BOOST_DETAIL_LCAST_SET_PRECISION_HPP_INCLUDED
+