diff options
| -rw-r--r-- | doc/info/chap/integer-types.texinfo | 49 |
1 files changed, 28 insertions, 21 deletions
diff --git a/doc/info/chap/integer-types.texinfo b/doc/info/chap/integer-types.texinfo index 12fc25a..787b9eb 100644 --- a/doc/info/chap/integer-types.texinfo +++ b/doc/info/chap/integer-types.texinfo @@ -1,6 +1,11 @@ @node Integer types @chapter Integer types +Many programmers have widely false assumption of +the integer types in C. You should read this chapter. +Integer types in C are not as straight-forward as +they are in other programming languages. + @menu * Builtin integer types:: Integer types provided by the compiler. @@ -31,7 +36,9 @@ made available by including the header file @lvindex CHAR_MIN @lvindex CHAR_MAX The smallest addressable unit of the machine -that can contain a basic character set. +that can contain a basic character set.@footnote{Because +of this, @code{char*} has been given a special property +in the C programming language: it may alias anything.} It is either signed or unsigned, depending on the implementation. It contains @code{CHAR_BIT} bits. @sc{POSIX} requires that @code{CHAR_BIT == 8}, @@ -48,16 +55,15 @@ Signed version of @code{char}. It is guaranteed to have a range of either [@math{-2}@sup{@code{CHAR_BIT}@math{~-~1}}@math{~+~1}, @math{2}@sup{@code{CHAR_BIT}@math{~-~1}}@math{~-~1}] -or -[@math{-2}@sup{@code{CHAR_BIT}@math{~-~1}}, +or [@math{-2}@sup{@code{CHAR_BIT}@math{~-~1}}, @math{2}@sup{@code{CHAR_BIT}@math{~-~1}}@math{~-~1}]. The exact range is [@code{SCHAR_MIN}, @code{SCHAR_MAX}]. @end iftex @ifnottex Signed version of @code{char}. It is guaranteed to have a range of at least either -[-2^(@code{CHAR_BIT} - 1) + 1, 2^(@code{CHAR_BIT} - 1) - 1] or -[-2^(@code{CHAR_BIT} - 1), 2^(@code{CHAR_BIT} - 1) - 1]. +[-2^(@code{CHAR_BIT} - 1) + 1, 2^(@code{CHAR_BIT} - 1) - 1] +or [-2^(@code{CHAR_BIT} - 1), 2^(@code{CHAR_BIT} - 1) - 1]. The exact range is [@code{SCHAR_MIN}, @code{SCHAR_MAX}]. @end ifnottex @@ -73,8 +79,7 @@ or equivalently [0, @code{UCHAR_MAX}]. @end iftex @ifnottex Unsigned version of @code{char}. It is guaranteed -to have a range of exactly -[0, 2^@code{CHAR_BIT} -1], +to have a range of exactly [0, 2^@code{CHAR_BIT} -1], or equivalently [0, @code{UCHAR_MAX}]. @end ifnottex @@ -88,15 +93,15 @@ or equivalently [0, @code{UCHAR_MAX}]. @lvindex SHORT_MAX Signed integer type of at least 16 bits. Its range is guaranteed contain at least [@math{-32767}, @math{32767}]. -The exact range [@code{SHORT_MIN}, @code{SHORT_MAX}]. +The exact range is [@code{SHORT_MIN}, @code{SHORT_MAX}]. @item unsigned short int @itemx unsigned short @tpindex unsigned short int @lvindex USHORT_MAX -Unsigned integer type of at least 16 bits. Its range is -guaranteed contain at least [0, @math{65535}]. -The exact range [0, @code{USHORT_MAX}]. +Unsigned integer type of at least 16 bits. Its range +is guaranteed contain at least [0, @math{65535}]. +The exact range is [0, @code{USHORT_MAX}]. @item signed int @itemx int @@ -106,21 +111,22 @@ The exact range [0, @code{USHORT_MAX}]. @lvindex INT_MAX Signed integer type of at least 16 bits. Its range is guaranteed contain at least [@math{-32767}, @math{32767}]. -The exact range [@code{INT_MIN}, @code{INT_MAX}]. +The exact range is [@code{INT_MIN}, @code{INT_MAX}]. This +integer type should not be used for indices! @item unsigned int @tpindex unsigned int @lvindex UINT_MAX -Unsigned integer type of at least 16 bits. Its range is -guaranteed contain at least [0, @math{65535}]. -The exact range [0, @code{UINT_MAX}]. +Unsigned integer type of at least 16 bits. Its range +is guaranteed contain at least [0, @math{65535}]. +The exact range is [0, @code{UINT_MAX}]. @item signed @tpindex signed Alias for @code{signed int} which is used when the valid value range is the least guaranteed range. Thus The range may exceed [@math{-32767}, @math{32767}], -but values outside that range [[@math{-32767}, +but values outside that range is [[@math{-32767}, @math{32767}]] should not be used. @item unsigned @@ -128,7 +134,8 @@ but values outside that range [[@math{-32767}, Alias for @code{unsigned int} which is used when the valid value range is the least guaranteed range. Thus The range may exceed [0, @math{65535}], but values -outside that range [[0, @math{65535}]] should not be used. +outside that range is [[0, @math{65535}]] should not +be used. @item long int @itemx long @@ -140,7 +147,7 @@ outside that range [[0, @math{65535}]] should not be used. @lvindex LONG_MAX Signed integer type of at least 32 bits. Its range is guaranteed contain at least [@math{-2147483647}, -@math{2147483647}]. The exact range [@code{LONG_MIN}, +@math{2147483647}]. The exact range is [@code{LONG_MIN}, @code{LONG_MAX}]. @item unsigned long int @@ -149,7 +156,7 @@ guaranteed contain at least [@math{-2147483647}, @lvindex ULONG_MAX Unsigned integer type of at least 32 bits. Its range is guaranteed contain at least [0, @math{4294967295}]. -The exact range [0, @code{ULONG_MAX}]. +The exact range is [0, @code{ULONG_MAX}]. @item long long int @itemx long long @@ -162,7 +169,7 @@ The exact range [0, @code{ULONG_MAX}]. Signed integer type of at least 64 bits. Its range is guaranteed contain at least [@math{-9223372036854775807}, @math{9223372036854775807}]. This type was added in C99. -The exact range [@code{LLONG_MIN}, @code{LLONG_MAX}]. +The exact range is [@code{LLONG_MIN}, @code{LLONG_MAX}]. @item unsigned long long int @itemx unsigned long long @@ -170,7 +177,7 @@ The exact range [@code{LLONG_MIN}, @code{LLONG_MAX}]. @lvindex ULLONG_MAX Unsigned integer type of at least 64 bits. Its range is guaranteed contain at least [0, @math{18446744073709551615}]. -This type was added in C99. The exact range [0, +This type was added in C99. The exact range is [0, @code{ULLONG_MAX}]. @end table |