mksyntax.c revision 221513 
1/*-
2 * Copyright (c) 1991, 1993
3 *	The Regents of the University of California.  All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Kenneth Almquist.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 *    notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 *    notice, this list of conditions and the following disclaimer in the
15 *    documentation and/or other materials provided with the distribution.
16 * 4. Neither the name of the University nor the names of its contributors
17 *    may be used to endorse or promote products derived from this software
18 *    without specific prior written permission.
19 *
20 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
30 * SUCH DAMAGE.
31 */
32
33#if 0
34#ifndef lint
35static char const copyright[] =
36"@(#) Copyright (c) 1991, 1993\n\
37	The Regents of the University of California.  All rights reserved.\n";
38#endif /* not lint */
39
40#ifndef lint
41static char sccsid[] = "@(#)mksyntax.c	8.2 (Berkeley) 5/4/95";
42#endif /* not lint */
43#endif
44#include <sys/cdefs.h>
45__FBSDID("$FreeBSD: head/bin/sh/mksyntax.c 221513 2011-05-05 20:55:55Z jilles $");
46
47/*
48 * This program creates syntax.h and syntax.c.
49 */
50
51#include <stdio.h>
52#include <stdlib.h>
53#include <string.h>
54#include "parser.h"
55
56
57struct synclass {
58	const char *name;
59	const char *comment;
60};
61
62/* Syntax classes */
63struct synclass synclass[] = {
64	{ "CWORD",	"character is nothing special" },
65	{ "CNL",	"newline character" },
66	{ "CBACK",	"a backslash character" },
67	{ "CSBACK",	"a backslash character in single quotes" },
68	{ "CSQUOTE",	"single quote" },
69	{ "CDQUOTE",	"double quote" },
70	{ "CENDQUOTE",	"a terminating quote" },
71	{ "CBQUOTE",	"backwards single quote" },
72	{ "CVAR",	"a dollar sign" },
73	{ "CENDVAR",	"a '}' character" },
74	{ "CLP",	"a left paren in arithmetic" },
75	{ "CRP",	"a right paren in arithmetic" },
76	{ "CEOF",	"end of file" },
77	{ "CCTL",	"like CWORD, except it must be escaped" },
78	{ "CSPCL",	"these terminate a word" },
79	{ "CIGN",       "character should be ignored" },
80	{ NULL,		NULL }
81};
82
83
84/*
85 * Syntax classes for is_ functions.  Warning:  if you add new classes
86 * you may have to change the definition of the is_in_name macro.
87 */
88struct synclass is_entry[] = {
89	{ "ISDIGIT",	"a digit" },
90	{ "ISUPPER",	"an upper case letter" },
91	{ "ISLOWER",	"a lower case letter" },
92	{ "ISUNDER",	"an underscore" },
93	{ "ISSPECL",	"the name of a special parameter" },
94	{ NULL, 	NULL }
95};
96
97static char writer[] = "\
98/*\n\
99 * This file was generated by the mksyntax program.\n\
100 */\n\
101\n";
102
103
104static FILE *cfile;
105static FILE *hfile;
106static const char *syntax[513];
107static int base;
108static int size;	/* number of values which a char variable can have */
109static int nbits;	/* number of bits in a character */
110static int digit_contig;/* true if digits are contiguous */
111
112static void filltable(const char *);
113static void init(void);
114static void add(const char *, const char *);
115static void print(const char *);
116static void output_type_macros(void);
117static void digit_convert(void);
118
119int
120main(int argc __unused, char **argv __unused)
121{
122	char c;
123	char d;
124	int sign;
125	int i;
126	char buf[80];
127	int pos;
128	static char digit[] = "0123456789";
129
130	/* Create output files */
131	if ((cfile = fopen("syntax.c", "w")) == NULL) {
132		perror("syntax.c");
133		exit(2);
134	}
135	if ((hfile = fopen("syntax.h", "w")) == NULL) {
136		perror("syntax.h");
137		exit(2);
138	}
139	fputs(writer, hfile);
140	fputs(writer, cfile);
141
142	/* Determine the characteristics of chars. */
143	c = -1;
144	sign = (c > 0) ? 0 : 1;
145	for (nbits = 1 ; ; nbits++) {
146		d = (1 << nbits) - 1;
147		if (d == c)
148			break;
149	}
150#if 0
151	printf("%s %d bit chars\n", sign? "signed" : "unsigned", nbits);
152#endif
153	if (nbits > 9) {
154		fputs("Characters can't have more than 9 bits\n", stderr);
155		exit(2);
156	}
157	size = (1 << nbits) + 1;
158	base = 1;
159	if (sign)
160		base += 1 << (nbits - 1);
161	digit_contig = 1;
162	for (i = 0 ; i < 10 ; i++) {
163		if (digit[i] != '0' + i)
164			digit_contig = 0;
165	}
166
167	fputs("#include <sys/cdefs.h>\n", hfile);
168	fputs("#include <ctype.h>\n", hfile);
169
170	/* Generate the #define statements in the header file */
171	fputs("/* Syntax classes */\n", hfile);
172	for (i = 0 ; synclass[i].name ; i++) {
173		sprintf(buf, "#define %s %d", synclass[i].name, i);
174		fputs(buf, hfile);
175		for (pos = strlen(buf) ; pos < 32 ; pos = (pos + 8) & ~07)
176			putc('\t', hfile);
177		fprintf(hfile, "/* %s */\n", synclass[i].comment);
178	}
179	putc('\n', hfile);
180	fputs("/* Syntax classes for is_ functions */\n", hfile);
181	for (i = 0 ; is_entry[i].name ; i++) {
182		sprintf(buf, "#define %s %#o", is_entry[i].name, 1 << i);
183		fputs(buf, hfile);
184		for (pos = strlen(buf) ; pos < 32 ; pos = (pos + 8) & ~07)
185			putc('\t', hfile);
186		fprintf(hfile, "/* %s */\n", is_entry[i].comment);
187	}
188	putc('\n', hfile);
189	fprintf(hfile, "#define SYNBASE %d\n", base);
190	fprintf(hfile, "#define PEOF %d\n\n", -base);
191	putc('\n', hfile);
192	fputs("#define BASESYNTAX (basesyntax + SYNBASE)\n", hfile);
193	fputs("#define DQSYNTAX (dqsyntax + SYNBASE)\n", hfile);
194	fputs("#define SQSYNTAX (sqsyntax + SYNBASE)\n", hfile);
195	fputs("#define ARISYNTAX (arisyntax + SYNBASE)\n", hfile);
196	putc('\n', hfile);
197	output_type_macros();		/* is_digit, etc. */
198	putc('\n', hfile);
199
200	/* Generate the syntax tables. */
201	fputs("#include \"shell.h\"\n", cfile);
202	fputs("#include \"syntax.h\"\n\n", cfile);
203	init();
204	fputs("/* syntax table used when not in quotes */\n", cfile);
205	add("\n", "CNL");
206	add("\\", "CBACK");
207	add("'", "CSQUOTE");
208	add("\"", "CDQUOTE");
209	add("`", "CBQUOTE");
210	add("$", "CVAR");
211	add("}", "CENDVAR");
212	add("<>();&| \t", "CSPCL");
213	print("basesyntax");
214	init();
215	fputs("\n/* syntax table used when in double quotes */\n", cfile);
216	add("\n", "CNL");
217	add("\\", "CBACK");
218	add("\"", "CENDQUOTE");
219	add("`", "CBQUOTE");
220	add("$", "CVAR");
221	add("}", "CENDVAR");
222	/* ':/' for tilde expansion, '-' for [a\-x] pattern ranges */
223	add("!*?[=~:/-", "CCTL");
224	print("dqsyntax");
225	init();
226	fputs("\n/* syntax table used when in single quotes */\n", cfile);
227	add("\n", "CNL");
228	add("\\", "CSBACK");
229	add("'", "CENDQUOTE");
230	/* ':/' for tilde expansion, '-' for [a\-x] pattern ranges */
231	add("!*?[=~:/-", "CCTL");
232	print("sqsyntax");
233	init();
234	fputs("\n/* syntax table used when in arithmetic */\n", cfile);
235	add("\n", "CNL");
236	add("\\", "CBACK");
237	add("`", "CBQUOTE");
238	add("\"", "CIGN");
239	add("$", "CVAR");
240	add("}", "CENDVAR");
241	add("(", "CLP");
242	add(")", "CRP");
243	print("arisyntax");
244	filltable("0");
245	fputs("\n/* character classification table */\n", cfile);
246	add("0123456789", "ISDIGIT");
247	add("abcdefghijklmnopqrstuvwxyz", "ISLOWER");
248	add("ABCDEFGHIJKLMNOPQRSTUVWXYZ", "ISUPPER");
249	add("_", "ISUNDER");
250	add("#?$!-*@", "ISSPECL");
251	print("is_type");
252	if (! digit_contig)
253		digit_convert();
254	exit(0);
255}
256
257
258
259/*
260 * Clear the syntax table.
261 */
262
263static void
264filltable(const char *dftval)
265{
266	int i;
267
268	for (i = 0 ; i < size ; i++)
269		syntax[i] = dftval;
270}
271
272
273/*
274 * Initialize the syntax table with default values.
275 */
276
277static void
278init(void)
279{
280	filltable("CWORD");
281	syntax[0] = "CEOF";
282	syntax[base + CTLESC] = "CCTL";
283	syntax[base + CTLVAR] = "CCTL";
284	syntax[base + CTLENDVAR] = "CCTL";
285	syntax[base + CTLBACKQ] = "CCTL";
286	syntax[base + CTLBACKQ + CTLQUOTE] = "CCTL";
287	syntax[base + CTLARI] = "CCTL";
288	syntax[base + CTLENDARI] = "CCTL";
289	syntax[base + CTLQUOTEMARK] = "CCTL";
290	syntax[base + CTLQUOTEEND] = "CCTL";
291}
292
293
294/*
295 * Add entries to the syntax table.
296 */
297
298static void
299add(const char *p, const char *type)
300{
301	while (*p)
302		syntax[*p++ + base] = type;
303}
304
305
306
307/*
308 * Output the syntax table.
309 */
310
311static void
312print(const char *name)
313{
314	int i;
315	int col;
316
317	fprintf(hfile, "extern const char %s[];\n", name);
318	fprintf(cfile, "const char %s[%d] = {\n", name, size);
319	col = 0;
320	for (i = 0 ; i < size ; i++) {
321		if (i == 0) {
322			fputs("      ", cfile);
323		} else if ((i & 03) == 0) {
324			fputs(",\n      ", cfile);
325			col = 0;
326		} else {
327			putc(',', cfile);
328			while (++col < 9 * (i & 03))
329				putc(' ', cfile);
330		}
331		fputs(syntax[i], cfile);
332		col += strlen(syntax[i]);
333	}
334	fputs("\n};\n", cfile);
335}
336
337
338
339/*
340 * Output character classification macros (e.g. is_digit).  If digits are
341 * contiguous, we can test for them quickly.
342 */
343
344static const char *macro[] = {
345	"#define is_digit(c)\t((is_type+SYNBASE)[(int)c] & ISDIGIT)",
346	"#define is_eof(c)\t((c) == PEOF)",
347	"#define is_alpha(c)\t((is_type+SYNBASE)[(int)c] & (ISUPPER|ISLOWER))",
348	"#define is_name(c)\t((is_type+SYNBASE)[(int)c] & (ISUPPER|ISLOWER|ISUNDER))",
349	"#define is_in_name(c)\t((is_type+SYNBASE)[(int)c] & (ISUPPER|ISLOWER|ISUNDER|ISDIGIT))",
350	"#define is_special(c)\t((is_type+SYNBASE)[(int)c] & (ISSPECL|ISDIGIT))",
351	NULL
352};
353
354static void
355output_type_macros(void)
356{
357	const char **pp;
358
359	if (digit_contig)
360		macro[0] = "#define is_digit(c)\t((unsigned int)((c) - '0') <= 9)";
361	for (pp = macro ; *pp ; pp++)
362		fprintf(hfile, "%s\n", *pp);
363	if (digit_contig)
364		fputs("#define digit_val(c)\t((c) - '0')\n", hfile);
365	else
366		fputs("#define digit_val(c)\t(digit_value[c])\n", hfile);
367}
368
369
370
371/*
372 * Output digit conversion table (if digits are not contiguous).
373 */
374
375static void
376digit_convert(void)
377{
378	int maxdigit;
379	static char digit[] = "0123456789";
380	char *p;
381	int i;
382
383	maxdigit = 0;
384	for (p = digit ; *p ; p++)
385		if (*p > maxdigit)
386			maxdigit = *p;
387	fputs("extern const char digit_value[];\n", hfile);
388	fputs("\n\nconst char digit_value[] = {\n", cfile);
389	for (i = 0 ; i <= maxdigit ; i++) {
390		for (p = digit ; *p && *p != i ; p++);
391		if (*p == '\0')
392			p = digit;
393		fprintf(cfile, "      %d,\n", (int)(p - digit));
394	}
395	fputs("};\n", cfile);
396}
397