Mercurial > repo

comparison interps/clc-intercal/CLC-INTERCAL-Docs-1.-94.-2/doc/html/input_output.html @ 996:859f9b4339e6

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
<Gregor> tar xf egobot.tar.xz
author HackBot
date Sun, 09 Dec 2012 19:30:08 +0000
parents
children
comparison
equal deleted inserted replaced
995:6883f5911eb7 996:859f9b4339e6
1 <HTML>
2 <HEAD>
3 <TITLE>CLC-INTERCAL Reference</TITLE>
4 </HEAD>
5 <BODY>
6 <H1>CLC-INTERCAL Reference</H1>
7 <H2>... Input/Output</H2>
8
9 <P>
10 We agree with the designers of C-INTERCAL that the I/O capabilities of
11 INTERCAL-72 are insufficient. However, we don't agree with their implementation
12 so we shamelessly go and do something entirely incompatible. However, if you
13 select the <I>ick</I> compiler for CLC-INTERCAL you get back their I/O behaviour.
14 </P>
15
16 <P>
17 Table of contents:
18 <UL>
19 <LI><A HREF="index.html">Parent directory</A>
20 <LI><A HREF="#numeric">Numeric I/O</A>
21 <LI><A HREF="#alphanumeric">Alphanumeric I/O</A>
22 <LI><A HREF="#binary">Binary I/O</A>
23 <LI><A HREF="#class">Class I/O</A>
24 </UL>
25 </P>
26
27 <H2><A NAME="numeric">Numeric I/O</A></H2>
28
29 <P>
30 This is the standard INTERCAL-72 I/O. It is used with numeric variables
31 (one spot or two spots).
32 </P>
33
34 <P>
35 On input, one line is read and its contents must be a sequence of digits
36 spelled out in full in a selection or popular languages, for example
37 "ONE THREE TWO ONE ZERO" (English for 13210) or "COIG CEITHIR TRI" (Scottish
38 Gaelic for 543). Spaces between digits are ignored.
39 </P>
40
41 <P>
42 On output, butchered Roman numerals are used. However, we differ from
43 INTERCAL-72 slightly in that we use backslashes before numerals to multiply
44 them by 1,000,000 - however the actual format of the Roman numerals is
45 controlled by the compiler you use: if it is <I>sick</I>, you get the
46 backslashes; if it is <I>ick</I> or <I>1972</I>, you get the overline
47 like C-INTERCAL does; you can also write your own compiler (or modify
48 an existing one) and have another one of the seven or eight formats available.
49 </P>
50
51 <P>
52 For example, 1,234,567,890 is written
53 <PRE>
54 \M\C\C\X\X\X\I\VdlxviiDCCCXC
55 </PRE>
56 by a program compiled with <I>sick</I> or
57 <PRE>
58 ________
59 MCCXXXIVdlxviiDCCCXC
60 </PRE>
61 by a program compiled with <I>ick</I>.
62 </P>
63
64 <H2><A NAME="alphanumeric">Alphanumeric I/O</A></H2>
65
66 <P>
67 This is the kind of input/output the compiler itself uses to read the program
68 source and produce program listings. It applies to 16 bit arrays (tails),
69 which must be dimensioned. Multidimensional arrays are "flattened".
70 </P>
71
72 <P>
73 On input, text is obtained from the next line, and it is converted to
74 an extended Baudot code. This is essentially the same as standard Baudot,
75 except that selecting "letters" when you are already in "letters" will cause
76 a shift to lowercase, and selecting "figures" when you are already in
77 "figures" causes a shift to other symbols. See <A HREF="charset.html">the
78 chapter on Character Sets</A> for a complete description of extended Baudot.
79 Please note that the insertion of shift codes means that the Baudot code
80 can be as much as three times as long as the original ASCII, so you must
81 dimension your array accordingly.
82 </P>
83
84 <P>
85 On output, extended Baudot is converted to ASCII and sent to the virtual
86 line printer.
87 </P>
88
89 <H2><A NAME="binary">Binary I/O</A></H2>
90
91 <P>
92 This is the simplest, yet most powerful, form of input/output. It applies
93 to 32 bit arrays, which must be dimensioned. Multidimensional arrays are
94 "flattened".
95 </P>
96
97 <P>
98 The input is assumed to be a stream of bytes, which is not interpreted in
99 any way. The number of bytes written in from the input is the same as the
100 size of the array specified. To compute the data to store in the array, a
101 simple algorithm is applied to every pair of consecutive elements. Since
102 this would give one less element, a #172 is inserted at the start. If the
103 two numbers in a pair are in <CODE>.1</CODE> (left) and <CODE>.2</CODE>
104 (right), the following fragment produces the value to be stored and leaves
105 it in <CODE>:1</CODE>
106 <PRE>
107 DO :1 &lt;- '.2~.1'&#162;"'"&#165;'#65535&#162;.2'"~"#0&#162;#65535"'~'"&#165;'#65535&#162;.1'"~"#0&#162;#65535"'"
108 </PRE>
109 </P>
110
111 <P>
112 We know it's an insult to the reader's intelligence to explain what it does,
113 but maybe somebody is in a hurry and doesn't have the two milliseconds to parse
114 that, so here it goes: the result is obtained by computing two numbers and
115 interleaving them. The first number computed is the second input element
116 selected by the first input element (<CODE>.2~.1</CODE>). The second number
117 computed is the same selection but applied to the bit-complement of the two
118 input numbers. Before applying the algorithm, the numbers are extended from
119 8 to 16 bits by padding them with a random value. It is clear that the result
120 has all the bits of the corresponding input value, permuted in a predictable
121 order, so all the information is there.
122 </P>
123
124 <P>
125 A value of zero is inserted to indicate end-of-file. The padding is guaranteed
126 to have at least one bit set, so the values stored for any input value is never
127 zero.
128 </P>
129
130 <P>
131 For output, the same algorithm is applied in reverse, except that zeros are
132 skipped. If you need to use a value zero, pad it to 32 bits with ones. The
133 current implementation does not check that the padding is sufficiently random,
134 although this might well change in future.
135 </P>
136
137 <H2><A NAME="class">Class I/O</A></H2>
138
139 <P>
140 If you use a class name (for example, <CODE>@1</CODE>) for I/O, nothing gets
141 actually input or output by the class. However, class registers have associated
142 filehandles and mentioning the class instructs the runtime to use the filehandle
143 on the current WRITE IN or READ OUT statement. For example:
144 <PRE>
145 DO READ OUT .1 + @3 + .2 + @2 + .3
146 </PRE>
147 produces the value of .1 on the standard read filehandle (the default
148 filehandle for READ OUT), then the value of .2 on the standard splat#
149 filehandle (@3), finally .3 back on the standard read (@2).
150 </P>
151
152 <P>
153 The runtime system predefines a small number of filehandles:
154 <UL>
155 <LI>@1 - standard write; the default for WRITE IN
156 <LI>@2 - standard read; the default for READ OUT
157 <LI>@3 - standard splat; error messages go here
158 <LI>@7 - null filehandle; if used in WRITE IN, always
159 returns an end-of-file condition; if used in READ OUT,
160 discards anything you send to it
161 <LI>@9 - trace filehandle; this is only available if
162 tracing is enabled
163 </UL>
164 </P>
165
166 <P>
167 When running as a standalone program, or from inside the command-line
168 compiler tool, @1 corresponds to the system standard input, @2
169 to the system standard output and @3 to the system standard error;
170 if tracing is enabled, @9 corresponds to the standard error, unless
171 the user has specified <I>--stdtrace=FILE</I> on the command line,
172 in which case @9 refers to that file.
173 </P>
174
175 <P>
176 When running inside the calculator, these filehandles are mapped to
177 something appropriate for the user interface in use. Usually, @2
178 and @3 send their output to the calculator's display, @9 stores the
179 output separately (use the "trace" entry from the "Windows" menu to
180 see it) and @1 causes a dialog to pop up and asking the user to type
181 something.
182 </P>
183
184 <P>
185 The <A HREF="syscall.html">system call interface</A> allows to
186 open files and TCP sockets and associate them with filehandles.
187 </P>
188
189 <P>
190 Note that if a program steals a class register from another program,
191 it will use the corresponding filehandle as well.
192 </P>
193
194 </BODY>
195 </HTML>