xref: /openbsd/usr.bin/lex/sym.c (revision 20c29e2b)
1 /*	$OpenBSD: sym.c,v 1.10 2024/11/09 18:03:44 op Exp $	*/
2 
3 /* sym - symbol table routines */
4 
5 /*  Copyright (c) 1990 The Regents of the University of California. */
6 /*  All rights reserved. */
7 
8 /*  This code is derived from software contributed to Berkeley by */
9 /*  Vern Paxson. */
10 
11 /*  The United States Government has rights in this work pursuant */
12 /*  to contract no. DE-AC03-76SF00098 between the United States */
13 /*  Department of Energy and the University of California. */
14 
15 /*  This file is part of flex. */
16 
17 /*  Redistribution and use in source and binary forms, with or without */
18 /*  modification, are permitted provided that the following conditions */
19 /*  are met: */
20 
21 /*  1. Redistributions of source code must retain the above copyright */
22 /*     notice, this list of conditions and the following disclaimer. */
23 /*  2. Redistributions in binary form must reproduce the above copyright */
24 /*     notice, this list of conditions and the following disclaimer in the */
25 /*     documentation and/or other materials provided with the distribution. */
26 
27 /*  Neither the name of the University nor the names of its contributors */
28 /*  may be used to endorse or promote products derived from this software */
29 /*  without specific prior written permission. */
30 
31 /*  THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR */
32 /*  IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED */
33 /*  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR */
34 /*  PURPOSE. */
35 
36 #include "flexdef.h"
37 
38 /* Variables for symbol tables:
39  * sctbl - start-condition symbol table
40  * ndtbl - name-definition symbol table
41  * ccltab - character class text symbol table
42  */
43 
44 struct hash_entry {
45 	struct hash_entry *prev, *next;
46 	char   *name;
47 	char   *str_val;
48 	int     int_val;
49 };
50 
51 typedef struct hash_entry **hash_table;
52 
53 #define NAME_TABLE_HASH_SIZE 101
54 #define START_COND_HASH_SIZE 101
55 #define CCL_HASH_SIZE 101
56 
57 static struct hash_entry *ndtbl[NAME_TABLE_HASH_SIZE];
58 static struct hash_entry *sctbl[START_COND_HASH_SIZE];
59 static struct hash_entry *ccltab[CCL_HASH_SIZE];
60 
61 
62 /* declare functions that have forward references */
63 
64 static int addsym PROTO ((char[], char *, int, hash_table, int));
65 static struct hash_entry *findsym PROTO ((const char *sym,
66 					  hash_table table,
67 
68 					  int table_size));
69 static int hashfunct PROTO ((const char *, int));
70 
71 
72 /* addsym - add symbol and definitions to symbol table
73  *
74  * -1 is returned if the symbol already exists, and the change not made.
75  */
76 
addsym(char sym[],char * str_def,int int_def,hash_table table,int table_size)77 static int addsym (char sym[], char *str_def, int int_def, hash_table table,
78     int table_size)
79 {
80 	int     hash_val = hashfunct (sym, table_size);
81 	struct hash_entry *sym_entry = table[hash_val];
82 	struct hash_entry *new_entry;
83 	struct hash_entry *successor;
84 
85 	while (sym_entry) {
86 		if (!strcmp (sym, sym_entry->name)) {	/* entry already exists */
87 			return -1;
88 		}
89 
90 		sym_entry = sym_entry->next;
91 	}
92 
93 	/* create new entry */
94 	new_entry = (struct hash_entry *)
95 		malloc (sizeof (struct hash_entry));
96 
97 	if (new_entry == NULL)
98 		flexfatal (_("symbol table memory allocation failed"));
99 
100 	if ((successor = table[hash_val]) != 0) {
101 		new_entry->next = successor;
102 		successor->prev = new_entry;
103 	}
104 	else
105 		new_entry->next = NULL;
106 
107 	new_entry->prev = NULL;
108 	new_entry->name = sym;
109 	new_entry->str_val = str_def;
110 	new_entry->int_val = int_def;
111 
112 	table[hash_val] = new_entry;
113 
114 	return 0;
115 }
116 
117 
118 /* cclinstal - save the text of a character class */
119 
cclinstal(unsigned char ccltxt[],int cclnum)120 void    cclinstal (unsigned char ccltxt[], int cclnum)
121 {
122 	/* We don't bother checking the return status because we are not
123 	 * called unless the symbol is new.
124 	 */
125 
126 	(void) addsym ((char *) copy_unsigned_string (ccltxt),
127 		       (char *) 0, cclnum, ccltab, CCL_HASH_SIZE);
128 }
129 
130 
131 /* ccllookup - lookup the number associated with character class text
132  *
133  * Returns 0 if there's no CCL associated with the text.
134  */
135 
ccllookup(unsigned char ccltxt[])136 int     ccllookup (unsigned char ccltxt[])
137 {
138 	return findsym ((char *) ccltxt, ccltab, CCL_HASH_SIZE)->int_val;
139 }
140 
141 
142 /* findsym - find symbol in symbol table */
143 
findsym(const char * sym,hash_table table,int table_size)144 static struct hash_entry *findsym (const char *sym, hash_table table,
145     int table_size)
146 {
147 	static struct hash_entry empty_entry = {
148 		(struct hash_entry *) 0, (struct hash_entry *) 0,
149 		(char *) 0, (char *) 0, 0,
150 	};
151 	struct hash_entry *sym_entry =
152 
153 		table[hashfunct (sym, table_size)];
154 
155 	while (sym_entry) {
156 		if (!strcmp (sym, sym_entry->name))
157 			return sym_entry;
158 		sym_entry = sym_entry->next;
159 	}
160 
161 	return &empty_entry;
162 }
163 
164 /* hashfunct - compute the hash value for "str" and hash size "hash_size" */
165 
hashfunct(const char * str,int hash_size)166 static int hashfunct (const char *str, int hash_size)
167 {
168 	int hashval;
169 	int locstr;
170 
171 	hashval = 0;
172 	locstr = 0;
173 
174 	while (str[locstr]) {
175 		hashval = (hashval << 1) + (unsigned char) str[locstr++];
176 		hashval %= hash_size;
177 	}
178 
179 	return hashval;
180 }
181 
182 
183 /* ndinstal - install a name definition */
184 
ndinstal(const char * name,unsigned char definition[])185 void    ndinstal (const char *name, unsigned char definition[])
186 {
187 	if (addsym (copy_string (name),
188 		    (char *) copy_unsigned_string (definition), 0,
189 		    ndtbl, NAME_TABLE_HASH_SIZE))
190 			synerr (_("name defined twice"));
191 }
192 
193 
194 /* ndlookup - lookup a name definition
195  *
196  * Returns a nil pointer if the name definition does not exist.
197  */
198 
ndlookup(const char * nd)199 u_char   *ndlookup (const char *nd)
200 {
201 	return (u_char *) findsym (nd, ndtbl, NAME_TABLE_HASH_SIZE)->str_val;
202 }
203 
204 
205 /* scextend - increase the maximum number of start conditions */
206 
scextend(void)207 void    scextend (void)
208 {
209 	current_max_scs += MAX_SCS_INCREMENT;
210 
211 	++num_reallocs;
212 
213 	scset = reallocate_integer_array (scset, current_max_scs);
214 	scbol = reallocate_integer_array (scbol, current_max_scs);
215 	scxclu = reallocate_integer_array (scxclu, current_max_scs);
216 	sceof = reallocate_integer_array (sceof, current_max_scs);
217 	scname = reallocate_char_ptr_array (scname, current_max_scs);
218 }
219 
220 
221 /* scinstal - make a start condition
222  *
223  * NOTE
224  *    The start condition is "exclusive" if xcluflg is true.
225  */
226 
scinstal(const char * str,int xcluflg)227 void    scinstal (const char *str, int xcluflg)
228 {
229 	if (++lastsc >= current_max_scs)
230 		scextend ();
231 
232 	scname[lastsc] = copy_string (str);
233 
234 	if (addsym (scname[lastsc], (char *) 0, lastsc,
235 		    sctbl, START_COND_HASH_SIZE))
236 			format_pinpoint_message (_
237 						 ("start condition %s declared twice"),
238 str);
239 
240 	scset[lastsc] = mkstate (SYM_EPSILON);
241 	scbol[lastsc] = mkstate (SYM_EPSILON);
242 	scxclu[lastsc] = xcluflg;
243 	sceof[lastsc] = false;
244 }
245 
246 
247 /* sclookup - lookup the number associated with a start condition
248  *
249  * Returns 0 if no such start condition.
250  */
251 
sclookup(const char * str)252 int     sclookup (const char *str)
253 {
254 	return findsym (str, sctbl, START_COND_HASH_SIZE)->int_val;
255 }
256