token.h revision 1f5207b7604fb44407eb4342aff613f7c4508508
1#ifndef TOKEN_H
2#define TOKEN_H
3/*
4 * Basic tokenization structures. NOTE! Those tokens had better
5 * be pretty small, since we're going to keep them all in memory
6 * indefinitely.
7 *
8 * Copyright (C) 2003 Transmeta Corp.
9 *               2003 Linus Torvalds
10 *
11 * Permission is hereby granted, free of charge, to any person obtaining a copy
12 * of this software and associated documentation files (the "Software"), to deal
13 * in the Software without restriction, including without limitation the rights
14 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
15 * copies of the Software, and to permit persons to whom the Software is
16 * furnished to do so, subject to the following conditions:
17 *
18 * The above copyright notice and this permission notice shall be included in
19 * all copies or substantial portions of the Software.
20 *
21 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
22 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
23 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
24 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
25 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
26 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
27 * THE SOFTWARE.
28 */
29
30#include <sys/types.h>
31#include "lib.h"
32
33/*
34 * This describes the pure lexical elements (tokens), with
35 * no semantic meaning. In other words, an identifier doesn't
36 * have a type or meaning, it is only a specific string in
37 * the input stream.
38 *
39 * Semantic meaning is handled elsewhere.
40 */
41
42enum constantfile {
43  CONSTANT_FILE_MAYBE,    // To be determined, not inside any #ifs in this file
44  CONSTANT_FILE_IFNDEF,   // To be determined, currently inside #ifndef
45  CONSTANT_FILE_NOPE,     // No
46  CONSTANT_FILE_YES       // Yes
47};
48
49extern const char *includepath[];
50
51struct stream {
52	int fd;
53	const char *name;
54	const char *path;    // input-file path - see set_stream_include_path()
55	const char **next_path;
56
57	/* Use these to check for "already parsed" */
58	enum constantfile constant;
59	int dirty, next_stream, once;
60	struct ident *protect;
61	struct token *ifndef;
62	struct token *top_if;
63};
64
65extern int input_stream_nr;
66extern struct stream *input_streams;
67extern unsigned int tabstop;
68extern int no_lineno;
69extern int *hash_stream(const char *name);
70
71struct ident {
72	struct ident *next;	/* Hash chain of identifiers */
73	struct symbol *symbols;	/* Pointer to semantic meaning list */
74	unsigned char len;	/* Length of identifier name */
75	unsigned char tainted:1,
76	              reserved:1,
77		      keyword:1;
78	char name[];		/* Actual identifier */
79};
80
81enum token_type {
82	TOKEN_EOF,
83	TOKEN_ERROR,
84	TOKEN_IDENT,
85	TOKEN_ZERO_IDENT,
86	TOKEN_NUMBER,
87	TOKEN_CHAR,
88	TOKEN_CHAR_EMBEDDED_0,
89	TOKEN_CHAR_EMBEDDED_1,
90	TOKEN_CHAR_EMBEDDED_2,
91	TOKEN_CHAR_EMBEDDED_3,
92	TOKEN_WIDE_CHAR,
93	TOKEN_WIDE_CHAR_EMBEDDED_0,
94	TOKEN_WIDE_CHAR_EMBEDDED_1,
95	TOKEN_WIDE_CHAR_EMBEDDED_2,
96	TOKEN_WIDE_CHAR_EMBEDDED_3,
97	TOKEN_STRING,
98	TOKEN_WIDE_STRING,
99	TOKEN_SPECIAL,
100	TOKEN_STREAMBEGIN,
101	TOKEN_STREAMEND,
102	TOKEN_MACRO_ARGUMENT,
103	TOKEN_STR_ARGUMENT,
104	TOKEN_QUOTED_ARGUMENT,
105	TOKEN_CONCAT,
106	TOKEN_GNU_KLUDGE,
107	TOKEN_UNTAINT,
108	TOKEN_ARG_COUNT,
109	TOKEN_IF,
110	TOKEN_SKIP_GROUPS,
111	TOKEN_ELSE,
112};
113
114/* Combination tokens */
115#define COMBINATION_STRINGS {	\
116	"+=", "++",		\
117	"-=", "--", "->",	\
118	"*=",			\
119	"/=",			\
120	"%=",			\
121	"<=", ">=",		\
122	"==", "!=",		\
123	"&&", "&=",		\
124	"||", "|=",		\
125	"^=", "##",		\
126	"<<", ">>", "..",	\
127	"<<=", ">>=", "...",	\
128	"",			\
129	"<", ">", "<=", ">="	\
130}
131
132extern unsigned char combinations[][4];
133
134enum special_token {
135	SPECIAL_BASE = 256,
136	SPECIAL_ADD_ASSIGN = SPECIAL_BASE,
137	SPECIAL_INCREMENT,
138	SPECIAL_SUB_ASSIGN,
139	SPECIAL_DECREMENT,
140	SPECIAL_DEREFERENCE,
141	SPECIAL_MUL_ASSIGN,
142	SPECIAL_DIV_ASSIGN,
143	SPECIAL_MOD_ASSIGN,
144	SPECIAL_LTE,
145	SPECIAL_GTE,
146	SPECIAL_EQUAL,
147	SPECIAL_NOTEQUAL,
148	SPECIAL_LOGICAL_AND,
149	SPECIAL_AND_ASSIGN,
150	SPECIAL_LOGICAL_OR,
151	SPECIAL_OR_ASSIGN,
152	SPECIAL_XOR_ASSIGN,
153	SPECIAL_HASHHASH,
154	SPECIAL_LEFTSHIFT,
155	SPECIAL_RIGHTSHIFT,
156	SPECIAL_DOTDOT,
157	SPECIAL_SHL_ASSIGN,
158	SPECIAL_SHR_ASSIGN,
159	SPECIAL_ELLIPSIS,
160	SPECIAL_ARG_SEPARATOR,
161	SPECIAL_UNSIGNED_LT,
162	SPECIAL_UNSIGNED_GT,
163	SPECIAL_UNSIGNED_LTE,
164	SPECIAL_UNSIGNED_GTE,
165};
166
167struct string {
168	unsigned int length:31;
169	unsigned int immutable:1;
170	char data[];
171};
172
173/* will fit into 32 bits */
174struct argcount {
175	unsigned normal:10;
176	unsigned quoted:10;
177	unsigned str:10;
178	unsigned vararg:1;
179};
180
181/*
182 * This is a very common data structure, it should be kept
183 * as small as humanly possible. Big (rare) types go as
184 * pointers.
185 */
186struct token {
187	struct position pos;
188	struct token *next;
189	union {
190		const char *number;
191		struct ident *ident;
192		unsigned int special;
193		struct string *string;
194		int argnum;
195		struct argcount count;
196		char embedded[4];
197	};
198};
199
200#define MAX_STRING 8191
201
202static inline struct token *containing_token(struct token **p)
203{
204	void *addr = (char *)p - ((char *)&((struct token *)0)->next - (char *)0);
205	return addr;
206}
207
208#define token_type(x) ((x)->pos.type)
209
210/*
211 * Last token in the stream - points to itself.
212 * This allows us to not test for NULL pointers
213 * when following the token->next chain..
214 */
215extern struct token eof_token_entry;
216#define eof_token(x) ((x) == &eof_token_entry)
217
218extern int init_stream(const char *, int fd, const char **next_path);
219extern const char *stream_name(int stream);
220extern struct ident *hash_ident(struct ident *);
221extern struct ident *built_in_ident(const char *);
222extern struct token *built_in_token(int, struct ident *);
223extern const char *show_special(int);
224extern const char *show_ident(const struct ident *);
225extern const char *show_string(const struct string *string);
226extern const char *show_token(const struct token *);
227extern const char *quote_token(const struct token *);
228extern struct token * tokenize(const char *, int, struct token *, const char **next_path);
229extern struct token * tokenize_buffer(void *, unsigned long, struct token **);
230
231extern void show_identifier_stats(void);
232extern void init_include_path(void);
233extern struct token *preprocess(struct token *);
234
235extern void store_all_tokens(struct token *token);
236extern struct token *pos_get_token(struct position pos);
237extern char *pos_ident(struct position pos);
238
239extern void store_macro_pos(struct token *);
240extern char *get_macro_name(struct position pos);
241
242static inline int match_op(struct token *token, unsigned int op)
243{
244	return token->pos.type == TOKEN_SPECIAL && token->special == op;
245}
246
247static inline int match_ident(struct token *token, struct ident *id)
248{
249	return token->pos.type == TOKEN_IDENT && token->ident == id;
250}
251
252#endif
253