c90标准(tc1).htm

ISO/IEC 9899:1990 C编程标准。 1992年的TECHNICAL CORRIGENDUM 1

<BLOCKQUOTE>The initial value of the object, including unnamed members, is that 
of the expression.
</BLOCKQUOTE>
<I>Page 80</I>
<BR>
<B>In subclause 6.6.6.4, lines 30-32, replace:</B>

<BLOCKQUOTE>If the expression has a type different from that of the function in 
which it appears, it is converted as if it were assigned to an object 
of that type.</BLOCKQUOTE>

<B>with:</B>

<BLOCKQUOTE>If the expression has a type different from the return type of the 
function in which it appears, the value is converted as if by assignment 
toan object having the return type of the function.<SUP>*</SUP>
<BR>
[Footnote *: The <TT><B>return</B></TT> statement is not an assignment. The 
overlap restriction in subclause 6.3.16.1 does not apply to the case 
of function return.]</BLOCKQUOTE>

<B>Add to subclause 6.6.6.4:</B>
<BR>
<B>Example</B>

<BLOCKQUOTE>In:</BLOCKQUOTE>

<PRE>   <TT><B>struct s {double i;} f(void);<BR>
     union {struct {int f1;
               struct s f2;} u1;
            struct {struct s f3;
               int f4;} u2;
     } g;
     struct s f(void)
          {
          return g.u1.f2;
          }
     /* </B></TT><I>...</I><TT> <B>*/
     g.u2.f3 = f();</B></TT></PRE>

<BLOCKQUOTE>the behavior is defined.</BLOCKQUOTE>

<I>Page 84</I>
<BR>
<B>Add to subclause 6.7.2, a second Example:</B>

<BLOCKQUOTE>If at the end of the translation unit containing
</BLOCKQUOTE>
<BLOCKQUOTE><TT><B> int i[]; </B></TT></BLOCKQUOTE>

<BLOCKQUOTE>the array <TT><B>i</B></TT> still has incomplete type, the array is assumed 
to have one element. This element is initialized to zero on program 
startup.</BLOCKQUOTE>

<I>Page 86</I>
<BR>
<B>Add to subclause 6.8, line 5, (<TT>Description</TT>): </B>

<BLOCKQUOTE>A new-line character ends the preprocessing directive even if it occurs 
within what would otherwise be an invocation of a function-like macro.</BLOCKQUOTE>

<B>Add to subclause 6.8, (<TT>Constraints</TT>):</B>

<BLOCKQUOTE>In the definition of an object-like macro, if the first character 
of a replacement list is not a character required by subclause 5.2.1, 
then there shall be white-space separation between the identifier 
and the replacement list.<SUP>*</SUP><BR>
[Footnote *: This allows an implementation to choose to interpret 
the directive:</BLOCKQUOTE>

<BLOCKQUOTE><TT><B> #define THIS$AND$THAT(a, b)	((a) + (b))</B></TT></BLOCKQUOTE>

<BLOCKQUOTE>as defining a function-like macro <TT><B>THIS$AND$THAT</B></TT>, rather than 
an object-like macro <TT><B>THIS</B></TT>. Whichever choice it makes, it must 
also issue a diagnostic.]
</BLOCKQUOTE>
<I>Page 90</I>
<BR>
<B>Add to subclause 6.8.3.3:</B>
<BR>
<B>Example</B>

<BLOCKQUOTE><TT><B>#define hash_hash # ## #<BR>
#define mkstr(a) # a<BR>
#define in_between(a) mkstr(a)<BR>
#define join(c, d) in_between(c hash_hash d)

char p[] = join(x, y); /*</B></TT> <I>equivalent to char p[] = "x ## y";</I> 
<TT><B>*/</B></TT></BLOCKQUOTE>

<BLOCKQUOTE>The expansion produces, at various stages:
</BLOCKQUOTE>
<BLOCKQUOTE><TT><B>join(x, y)<BR>
<BR>
in_between(x hash_hash y)<BR>
<BR>
in_between(x ## y)<BR>
<BR>
mkstr(x ## y)<BR>
<BR>
"x ## y"</B></TT></BLOCKQUOTE>

<BLOCKQUOTE>In other words, expanding <TT><B>hash_hash</B></TT> produces a new token, consisting 
of two adjacent sharp signs, but this new token is not the catenation 
operator.</BLOCKQUOTE>

<I>Page 96</I>
<BR>
<B>Add to subclause 7.1.2, (before <I>Forward references</I>):</B>

<BLOCKQUOTE>Any definition of an object-like macro described in this clause shall 
expand to code that is fully protected by parentheses where necessary, 
so that it g</BLOCKQUOTE>roups in an arbitrary expression as if it were a single 
identifier.

<B>In subclause 7.1.2, lines 32-33, change:</B>

<BLOCKQUOTE>However, if the identifier is declared or defined in more than one 
header,</BLOCKQUOTE>

<B>to:</B>

<BLOCKQUOTE>However, if an identifier is declared or defined in more than one 
header,</BLOCKQUOTE>

<I>Page 120</I>
<BR>
<B>In subclause 7.7, lines 14-16, change:</B>

<BLOCKQUOTE>and the following, each of which expands to a positive integral constant 
expression that is the signal number corresponding to the specified 
condition:</BLOCKQUOTE>

<B>to:</B>

<BLOCKQUOTE>and the following, which expand to positive integral constant expressions 
with distinct values that are the signal numbers, each corresponding 
to the specified condition:
</BLOCKQUOTE>
<I>Page 132</I>
<BR>
<B>In subclause 7.9.6.1, lines 37-38, change:</B>

<BLOCKQUOTE>For <TT><B>o</B></TT> conversion, it increases the precision to force the first 
digit of the result to be a zero.
</BLOCKQUOTE>
<B>to:</B>

<BLOCKQUOTE>For <TT><B>o</B></TT> conversion, it increases the precision, if and only if 
necessary, to force the first digit of the result to be a zero.</BLOCKQUOTE>

<I>Page 135</I>
<BR>
<B>In subclause 7.9.6.2, lines 31-33, change:</B>

<BLOCKQUOTE>An input item is defined as the longest matching sequence of input 
characters, unless that exceeds a specified field width, in which 
case it is the initial subsequence of that length in the sequence.
</BLOCKQUOTE>
<B>to:</B>

<BLOCKQUOTE>An input item is defined as the longest sequence of input characters 
which does not exceed any specified field width and which is, or is 
a prefix of, a matching input sequence.
</BLOCKQUOTE>
<I>Page 137</I>
<BR>
<B>In subclause 7.9.6.2, delete:</B>

<BLOCKQUOTE>If conversion terminates on a conflicting input character, the offending 
input character is left unread in the input stream.</BLOCKQUOTE>

<B>Add to subclause 7.9.6.2, line 4 (the</B> <TT><B>n</B></TT> <B>conversion specifier):</B>

<BLOCKQUOTE>No argument is converted, but one is consumed. If the conversion specification 
with this conversion specifier is not one of <TT><B>%n</B></TT>, <TT><B>%ln</B></TT>, 
or <TT><B>%hn</B></TT>, the behavior is undefined.</BLOCKQUOTE>

<B>Add to subclause 7.9.6.2:</B>

<BLOCKQUOTE>If conversion terminates on a conflicting input character, the offending 
input character is left unread in the input stream.<SUP>*</SUP><BR> [Footnote *: 
<TT><B>fscanf</B></TT> pushes back at most one input character onto the input 
stream. Therefore, some sequences that are acceptable to <TT><B>strtod</B></TT>, 
<TT><B>strtol</B></TT>, or <TT><B>strtoul</B></TT> are unacceptable to <TT><B>fscanf</B></TT>.]
</BLOCKQUOTE>
<I>Page 138</I>
<BR>
<B>Add to subclause 7.9.6.2, another Example:</B>
<BR>
<BLOCKQUOTE>In:</BLOCKQUOTE>
<BLOCKQUOTE><TT><B>#include &lt;stdio.h&gt;<BR>
/*</B></TT><I> ...</I> <TT><B>*/<BR>
int d1, d2, n1, n2, i;<BR>
i = sscanf("123", "%d%n%n%d", &amp;d1, &amp;n1,&amp;n2, &amp;d2);</B></TT>
</BLOCKQUOTE>
<BLOCKQUOTE>the value <B><TT>123</TT></B> is assigned to <TT><B>d1</B></TT> and the value 3 to <TT><B>n1</B></TT>. 
Because <TT><B>%n</B></TT> can never get an input failure the value of 3 is 
also assigned to <TT><B>n2</B></TT>. The value of <TT><B>d2</B></TT> is not affected. 
The value 3 is assigned to <TT><B>i</B></TT>.</BLOCKQUOTE>

<I>Page 145</I>
<BR>
<B>In subclause 7.9.9.2, lines 39-40, change:</B>

<BLOCKQUOTE>a value returned by an earlier call to the <TT><B>ftell</B></TT> function
</BLOCKQUOTE>
<B>to:</B>
<BLOCKQUOTE>a value returned by an earlier successful call to the <TT><B>ftell</B></TT> 
function</BLOCKQUOTE>
<I>Page 146</I>
<BR>
<B>In subclause 7.9.9.3, lines 10-11, change:</B>

<BLOCKQUOTE>a value obtained from an earlier call to the <TT><B>fgetpos</B></TT> function</BLOCKQUOTE>

<B>to:</B>

<BLOCKQUOTE>a value obtained from an earlier successful call to the <TT><B>fgetpos</B></TT> 
function</BLOCKQUOTE>
<BR>
<I>Page 162</I>
<BR>
<B>Add to subclause 7.11.1:</B>

<BLOCKQUOTE>Where an argument declared as <TT><B>size_t n</B></TT> specifies the length 
of the array for a function, <TT><B>n</B></TT> can have the value zero on a 
call to that function. Unless explicitly stated otherwise in the description 
of a particular function in this subclause, pointer arguments on such 
a call must still have valid values, as described in subclause 7.1.7. 
On such a call, a function that locates a character finds no occurrence, 
a function that compares two character sequences returns zero, and 
a function that copies characters copies zero characters.</BLOCKQUOTE>

<I>Page 172</I>
<BR>
<B>In subclause 7.12.2.3, line 16, change:</B>
<BLOCKQUOTE><PRE>
<TT><B>if (mktime(&amp;time_str) == -1)</B></TT>
</PRE></BLOCKQUOTE>
<B>to:</B>
<PRE>
   <TT><B>if (mktime(&amp;ime_str) == (time_t)-1)</B></TT>
</PRE>
<I>Page 200</I>
<BR>
<B>Add to subclause G.2:</B>
<BR>
 <BLOCKQUOTE>A program contains no function called <TT><B>main</B></TT> (5.1.2.2.1).</BLOCKQUOTE>

<I>Page 201</I>
<BR>
<B>Add to subclause G.2:</B>
<BR>
 <BLOCKQUOTE>A storage-class specifier or type qualifier modifies the keyword 
<TT><B>void</B></TT> as a function parameter type list (6.5.4.3).</BLOCKQUOTE>
<BR>
<B>Add to subclause G.2:</B>

 <BLOCKQUOTE>An array subscript is out of range, even if an object is apparently 
accessible with the given subscript (as in the lvalue expression <TT><B>a[1][7]</B></TT> 
given the declaration <TT><B>int a[4][5]</B></TT>) (6.3.6).
</BLOCKQUOTE>
<I>Page 202</I>
<BR>
<B>Add to subclause G.2:</B>

<BLOCKQUOTE>A fully expanded macro replacement list contains a function-like 
macro name as its last preprocessing token (6.8.3).</BLOCKQUOTE>

<I>Page 203</I>
<BR>
<B>Add to subclause G.2:</B>

<BLOCKQUOTE> A call to a library function exceeds an <B>environmental limit</B> 
(7.9.2, 7.9.3, 7.9.4.4, 7.9.6.1, 7.10.2.1).
</BLOCKQUOTE>
<I>Page 217</I>
<BR>
<B>In the index, change:</B>
<BR>
<BLOCKQUOTE><TT><B>static</B></TT> storage-class specifier, 3.1.2.2, 6.1.2.4, <B>6.5.1</B>, 
6.7</BLOCKQUOTE>

<B>to:</B>

<BLOCKQUOTE><TT><B>static</B></TT> storage-class specifier, 6.1.2.2, 6.1.2.4, <B>6.5.1</B>, 6.7</BLOCKQUOTE>
</BODY>
</HTML>