Subversion Repositories tendra.SVN

Rev

Rev 2 | Details | Compare with Previous | Last modification | View Log | RSS feed

Rev Author Line No. Line
2 7u83 1 
<!-- Crown Copyright (c) 1998 -->
2 
<HTML>
3 
<HEAD>
4 
<TITLE>Constants</TITLE>
5 
</HEAD>
6 
<BODY TEXT="#000000" BGCOLOR="#FFFFFF" LINK="#0000FF" VLINK="#400080" ALINK="#FF0000">
7 
<A NAME=S75>
8 
<H1>TDF Guide, Issue 4.0 </H1>
9 
<A HREF="guide12.html">
10 
<H3>January 1998</H3>
11 
<IMG SRC="../images/next.gif" ALT="next section"></A>
12 
<A HREF="guide10.html">
13 
<IMG SRC="../images/prev.gif" ALT="previous section"></A>
14 
<A HREF="guide1.html"><IMG SRC="../images/top.gif" ALT="current document"></A>
15 
<A HREF="../index.html"><IMG SRC="../images/home.gif" ALT="TenDRA home page">
16 
</A>
17 
<IMG SRC="../images/no_index.gif" ALT="document index"><P>
18 
<HR>
19 
<DL>
20 
<DT><A HREF="#S76"><B>9.1 </B> - _cond constructors</A><DD>
21 
<DT><A HREF="#S77"><B>9.2 </B> - Primitive constant constructors</A><DD>
22 
</DL>
23 
<HR>
24 
<H1>9  <A NAME=0>Constants</H1>
25 
The representation of constants clearly has peculiar difficulties
26 
in any architecture neutral format. Leaving aside any problems of
27 
how numbers are to be represented, we also have the situation where
28 
a &quot;constant&quot; can have different values on different platforms.
29 
An obvious example would be the size of a structure which, although
30 
it is a constant of any particular run of a program, may have different
31 
values on different machines. Further, this constant is in general
32 
the result of some computation involving the sizes of its components
33 
which are not known until the platform is chosen. In TDF, sizes are
34 
always derived from some EXP OFFSET constructed using the various
35 
OFFSET arithmetic operations on primitives like shape_offset and offset_zero.
36 
Most such EXP OFFSETs produced are in fact constants of the platform;
37 
they include field displacements of structure as well as their sizes.
38 
TDF assumes that, if these EXPs can be evaluated at translate-time
39 
(i.e. when the sizes and alignments of primitive objects are known),
40 
then they must be evaluated there. An example of why this is so arises
41 
in make_compound; the SHAPE of its result EXP depends on its <I>arg1</I>
42 
EXP OFFSET parameter and all SHAPEs must be translate-time values.
43 
<P>
44 
An initialisation of a TAGDEF is a constant in this sense
45 
<A NAME=footnote76 HREF="footnote.html#76">*</A>; this allows one
46 
to ignore any difficulties about their order of evaluation in the
47 
UNIT and consequently the order of evaluation of UNITs. Once again
48 
all the EXPs which are initialisations must be evaluated before the
49 
program is run; this obviously includes any make_proc or make_general_proc.
50 
. The limitation on an initialisation EXP to ensure this is basically
51 
that one cannot take the contents of a variable declared outside the
52 
EXP after all tokens and conditional evaluation is taken into account.
53 
In other words, each TDF translator effectively has an TDF interpreter
54 
which can do evaluation of expressions (including conditionals etc.)
55 
involving only constants such as numbers, sizes and addresses of globals.
56 
This corresponds very roughly to the kind of initialisations of globals
57 
that are permissible in C; for a more precise definition, see (S7.3)</A>.<P>
58 
<A NAME=S76>
59 
<HR><H2>9.1. <A NAME=1>_cond constructors</H2>
60 
Another place where translate-time evaluation of constants is mandated
61 
is in the various _cond constructors which give a kind of &quot;conditional
62 
compilation&quot; facility; every SORT which has a SORTNAME, other
63 
that TAG, TOKEN and LABEL, has one of these constructors e.g. exp_cond:<P>
64 
<PRE>
65 
<I>	control</I>: 	EXP INTEGER(<I>v</I>)
66 
<I>	e1</I>: 	BITSTREAM EXP <I>x</I>
67 
<I>	e2</I>: 	BITSTREAM EXP <I>y</I>
68 
		   -&gt; 	EXP <I>x</I> or EXP <I>y</I>
69 
</PRE>
70 
The constant, <I>control</I>, is evaluated at translate time. If it
71 
is not zero the entire construction is replaced by the EXP in <I>e1</I>;
72 
otherwise it is replaced by the one in <I>e2</I>. In either case,
73 
the other BITSTREAM is totally ignored; it even does not need to be
74 
sensible TDF. This kind of construction is use extensively in C pre-processing
75 
directives e.g.:<P>
76 
<PRE>
77 
#if (sizeof(int) == sizeof(long)) ...
78 
 
79 
</PRE>
80 
<A NAME=S77>
81 
<HR><H2>9.2. Primitive constant constructors</H2>
82 
Integer constants are constructed using make_int:<P>
83 
<PRE>
84 
<I>	v</I>:	VARIETY
85 
<I>	value</I>: 	SIGNED_NAT
86 
		   -&gt; 	EXP INTEGER(<I>v</I>)
87 
</PRE>
88 
The SIGNED_NAT <I>value</I> is an encoding of the binary value required
89 
for the integer; this value must lie within the limits given by <I>v</I>.
90 
I have been rather slip-shod in writing down examples of integer constants
91 
earlier in this document; where I have written 1 as an integer EXP,
92 
for example, I should have written make_int(v, 1) where v is some
93 
appropriate VARIETY.<P>
94 
Constants for both floats and strings use STRINGs. A constant string
95 
is just an particular example of make_nof_int:<P>
96 
<PRE>
97 
<I>	v</I>:	VARIETY
98 
<I>	str</I>: 	STRING(<I>k, n</I>)
99 
		   -&gt; 	EXP NOF(<I>n</I>, INTEGER(<I>v</I>))
100 
</PRE>
101 
Each unsigned integer in <I>str</I> must lie in the variety <I>v</I>
102 
and the result is the constant array whose elements are the integers
103 
considered to be of VARIETY <I>v</I>. An ASCI-C constant string might
104 
have <I>v</I> = variety(-128,127) and <I>k</I> = 7; however, make_nof_int
105 
can be used to make strings of any INTEGER VARIETY; a the elements
106 
of a Unicode string would be integers of size 16 bits.<P>
107 
A floating constant uses a STRING which contains the ASCI characters
108 
of a expansion of the number to some base in make_floating:<P>
109 
<PRE>
110 
<I>	f</I>: 	FLOATING_VARIETY
111 
<I>	rm</I>: 	ROUNDING_MODE
112 
<I>	sign</I>:	BOOL
113 
<I>	mantissa</I>: 	STRING(<I>k, n</I>)
114 
<I>	base</I>:	NAT
115 
<I>	exponent</I>: 	SIGNED_NAT
116 
		   -&gt; 	EXP FLOATING(<I>f</I>)
117 
</PRE>
118 
For a normal floating point number, each integer in <I>mantissa</I>
119 
is either the ASCI `.'-symbol or the ASCI representation of a digit
120 
of the representation in the given <I>base</I>; i.e. if c is the ASCI
121 
symbol, the digit value is c-'0'. The resulting floating point number
122 
has SHAPE FLOATING(f) and value <I>mantissa</I> *<I> base</I><I> exponent</I>
123 
rounded according to <I>rm</I>. Usually the base will be 10 (sometimes
124 
2) and the rounding mode to_nearest. Any floating-point evaluation
125 
of expressions done at translate-time will be done to an accuracy
126 
greater that implied by the FLOATING_VARIETY involved, so that floating
127 
constants will be as accurate as the platform permits.<P>
128 
The make_floating construct does not apply apply to a complex FLOATING_VARIETY
129 
<I>f</I>; to construct a complex constant use make_complex with two
130 
make_floating arguments.<P>
131 
Constants are also provided to give unique null values for pointers,
132 
label values and procs i.e.: make_null_ptr, make_null_local_lv and
133 
make_null_proc. Any significant use of these values (e.g. taking the
134 
contents of a null pointer) is undefined, but they can be assigned
135 
and used in tests in the normal way.<P>
136 
<HR>
137 
<P><I>Part of the <A HREF="../index.html">TenDRA Web</A>.<BR>Crown
138 
Copyright &copy; 1998.</I></P>
139 
</BODY>
140 
</HTML>