5 * Permission is hereby granted, free of charge, to any person obtaining a copy of
6 * this software and associated documentation files (the "Software"), to deal in
7 * the Software without restriction, including without limitation the rights to
8 * use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
9 * of the Software, and to permit persons to whom the Software is furnished to do
10 * so, subject to the following conditions:
12 * The above copyright notice and this permission notice shall be included in all
13 * copies or substantial portions of the Software.
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
29 * These are not lexical tokens: These are parse tree types. Most people
30 * perform tokenizing on language punctuation which is wrong. That stuff
31 * is technically already tokenized, it just needs to be parsed into a tree
33 #define PARSE_TYPE_DO 0
34 #define PARSE_TYPE_ELSE 1
35 #define PARSE_TYPE_IF 2
36 #define PARSE_TYPE_WHILE 3
37 #define PARSE_TYPE_BREAK 4
38 #define PARSE_TYPE_CONTINUE 5
39 #define PARSE_TYPE_RETURN 6
40 #define PARSE_TYPE_GOTO 7
41 #define PARSE_TYPE_FOR 8
42 #define PARSE_TYPE_VOID 9
43 #define PARSE_TYPE_STRING 10
44 #define PARSE_TYPE_FLOAT 11
45 #define PARSE_TYPE_VECTOR 12
46 #define PARSE_TYPE_ENTITY 13
47 #define PARSE_TYPE_LAND 14
48 #define PARSE_TYPE_LOR 15
49 #define PARSE_TYPE_LTEQ 16
50 #define PARSE_TYPE_GTEQ 17
51 #define PARSE_TYPE_EQEQ 18
52 #define PARSE_TYPE_LNEQ 19
53 #define PARSE_TYPE_COMMA 20
54 #define PARSE_TYPE_LNOT 21
55 #define PARSE_TYPE_STAR 22
56 #define PARSE_TYPE_DIVIDE 23
57 #define PARSE_TYPE_LPARTH 24
58 #define PARSE_TYPE_RPARTH 25
59 #define PARSE_TYPE_MINUS 26
60 #define PARSE_TYPE_ADD 27
61 #define PARSE_TYPE_EQUAL 28
62 #define PARSE_TYPE_LBS 29
63 #define PARSE_TYPE_RBS 30
64 #define PARSE_TYPE_ELIP 31
65 #define PARSE_TYPE_DOT 32
66 #define PARSE_TYPE_LT 33
67 #define PARSE_TYPE_GT 34
68 #define PARSE_TYPE_BAND 35
69 #define PARSE_TYPE_BOR 36
70 #define PARSE_TYPE_DONE 37
71 #define PARSE_TYPE_IDENT 38
74 * Adds a parse type to the parse tree, this is where all the hard
75 * work actually begins.
77 #define PARSE_TREE_ADD(X) \
79 parsetree->next = mem_a(sizeof(struct parsenode)); \
80 parsetree->next->next = NULL; \
81 parsetree->next->type = (X); \
82 parsetree = parsetree->next; \
86 * These are all the punctuation handled in the parser, these don't
87 * need tokens, they're already tokens.
90 "&&", "||", "<=", ">=", "==", "!=", ";", ",", "!", "*",
91 "/" , "(" , ")" , "-" , "+" , "=" , "[" , "]", "{", "}", "...",
92 "." , "<" , ">" , "&" , "|" ,
100 void parse_debug(struct parsenode *tree) {
101 while (tree && tree->next != NULL) {
103 if (tree->type == 0) {
108 switch (tree->type) {
109 case PARSE_TYPE_ADD: STORE("OPERATOR: ADD \n");
110 case PARSE_TYPE_BAND: STORE("OPERATOR: BITAND \n");
111 case PARSE_TYPE_BOR: STORE("OPERATOR: BITOR \n");
112 case PARSE_TYPE_COMMA: STORE("OPERATOR: SEPERATOR\n");
113 case PARSE_TYPE_DOT: STORE("OPERATOR: DOT\n");
114 case PARSE_TYPE_DIVIDE: STORE("OPERATOR: DIVIDE\n");
115 case PARSE_TYPE_EQUAL: STORE("OPERATOR: ASSIGNMENT\n");
117 case PARSE_TYPE_BREAK: STORE("STATEMENT: BREAK \n");
118 case PARSE_TYPE_CONTINUE: STORE("STATEMENT: CONTINUE\n");
119 case PARSE_TYPE_GOTO: STORE("STATEMENT: GOTO\n");
120 case PARSE_TYPE_RETURN: STORE("STATEMENT: RETURN\n");
121 case PARSE_TYPE_DONE: STORE("STATEMENT: DONE\n");
123 case PARSE_TYPE_VOID: STORE("DECLTYPE: VOID\n");
124 case PARSE_TYPE_STRING: STORE("DECLTYPE: STRING\n");
125 case PARSE_TYPE_ELIP: STORE("DECLTYPE: VALIST\n");
126 case PARSE_TYPE_ENTITY: STORE("DECLTYPE: ENTITY\n");
127 case PARSE_TYPE_FLOAT: STORE("DECLTYPE: FLOAT\n");
128 case PARSE_TYPE_VECTOR: STORE("DECLTYPE: VECTOR\n");
130 case PARSE_TYPE_GT: STORE("TEST: GREATER THAN\n");
131 case PARSE_TYPE_LT: STORE("TEST: LESS THAN\n");
132 case PARSE_TYPE_GTEQ: STORE("TEST: GREATER THAN OR EQUAL\n");
133 case PARSE_TYPE_LTEQ: STORE("TEST: LESS THAN OR EQUAL\n");
134 case PARSE_TYPE_LNEQ: STORE("TEST: NOT EQUAL\n");
135 case PARSE_TYPE_EQEQ: STORE("TEST: EQUAL-EQUAL\n");
137 case PARSE_TYPE_LBS: STORE("BLOCK: BEG\n");
138 case PARSE_TYPE_RBS: STORE("BLOCK: END\n");
139 case PARSE_TYPE_ELSE: STORE("BLOCK: ELSE\n");
140 case PARSE_TYPE_IF: STORE("BLOCK: IF\n");
142 case PARSE_TYPE_LAND: STORE("LOGICAL: AND\n");
143 case PARSE_TYPE_LNOT: STORE("LOGICAL: NOT\n");
144 case PARSE_TYPE_LOR: STORE("LOGICAL: OR\n");
146 case PARSE_TYPE_LPARTH: STORE("PARTH: BEG\n");
147 case PARSE_TYPE_RPARTH: STORE("PARTH: END\n");
149 case PARSE_TYPE_WHILE: STORE("LOOP: WHILE\n");
150 case PARSE_TYPE_FOR: STORE("LOOP: FOR\n");
151 case PARSE_TYPE_DO: STORE("LOOP: DO\n");
153 case PARSE_TYPE_IDENT: STORE("IDENT: ???\n");
160 * This just skips the token and throws it in the parse tree for later
161 * checking / optimization / codegen, it doesn't do anything with it
162 * like syntax check for legal use -- like it should as it's a TODO item
163 * which is not implemented
165 #define PARSE_TODO(X) { \
166 token = lex_token(file); \
171 int parse(struct lex_file *file) {
172 struct parsenode *parsetree = NULL;
173 struct parsenode *parseroot = NULL;
176 * Allocate memory for our parse tree:
177 * the parse tree is just a singly linked list which will contain
178 * all the data for code generation.
181 parseroot = mem_a(sizeof(struct parsenode));
183 return error(ERROR_INTERNAL, "Ran out of memory", " ");
184 parsetree = parseroot;
188 while ((token = lex_token(file)) != ERROR_LEX && \
189 token != ERROR_COMPILER && \
190 token != ERROR_INTERNAL && \
191 token != ERROR_PARSE && \
192 token != ERROR_PREPRO && file->length >= 0) {
195 token = lex_token(file);
196 //while ((token == ' ' || token == '\n') && file->length >= 0)
197 // token = lex_token(file);
200 // error(ERROR_PARSE, "Expected `(` after if\n", "");
202 PARSE_TREE_ADD(PARSE_TYPE_IF);
205 token = lex_token(file);
206 //while ((token == ' ' || token == '\n') && file->length >= 0)
207 // token = lex_token(file);
209 PARSE_TREE_ADD(PARSE_TYPE_ELSE);
212 token = lex_token(file);
213 //while ((token == ' ' || token == '\n') && file->length >= 0)
214 // token = lex_token(file);
216 PARSE_TREE_ADD(PARSE_TYPE_FOR);
220 * This is a quick and easy way to do typedefs at parse time
221 * all power is in typedef_add(), in typedef.c. We handle
222 * the tokens accordingly here.
224 case TOKEN_TYPEDEF: {
227 token = lex_token(file);
228 token = lex_token(file); f = strdup(file->lastok);
229 token = lex_token(file);
230 token = lex_token(file); t = strdup(file->lastok);
234 /* free stdup strings */
241 case TOKEN_DO: PARSE_TODO(PARSE_TYPE_DO);
242 case TOKEN_WHILE: PARSE_TODO(PARSE_TYPE_WHILE);
243 case TOKEN_BREAK: PARSE_TODO(PARSE_TYPE_BREAK);
244 case TOKEN_CONTINUE: PARSE_TODO(PARSE_TYPE_CONTINUE);
245 case TOKEN_RETURN: PARSE_TODO(PARSE_TYPE_RETURN);
246 case TOKEN_GOTO: PARSE_TODO(PARSE_TYPE_GOTO);
247 case TOKEN_VOID: PARSE_TODO(PARSE_TYPE_VOID);
248 case TOKEN_STRING: PARSE_TODO(PARSE_TYPE_STRING);
249 case TOKEN_FLOAT: PARSE_TODO(PARSE_TYPE_FLOAT);
250 case TOKEN_VECTOR: PARSE_TODO(PARSE_TYPE_VECTOR);
251 case TOKEN_ENTITY: PARSE_TODO(PARSE_TYPE_ENTITY);
254 * From here down is all language punctuation: There is no
255 * need to actual create tokens from these because they're already
256 * tokenized as these individual tokens (which are in a special area
257 * of the ascii table which doesn't conflict with our other tokens
258 * which are higer than the ascii table.)
262 * Skip the preprocessor for now: We'll implement our own
263 * eventually. For now we need to make sure directives are
264 * not accidently tokenized.
266 token = lex_token(file);
267 token = lex_token(file);
269 /* skip all tokens to end of directive */
270 while (token != '\n')
271 token = lex_token(file);
275 token = lex_token(file);
276 PARSE_TREE_ADD(PARSE_TYPE_LPARTH);
279 token = lex_token(file);
280 PARSE_TREE_ADD(PARSE_TYPE_RPARTH);
284 token = lex_token(file);
285 if (token == '&') { /* && */
286 token = lex_token(file);
287 PARSE_TREE_ADD(PARSE_TYPE_LAND);
290 PARSE_TREE_ADD(PARSE_TYPE_BAND);
293 token = lex_token(file);
294 if (token == '|') { /* || */
295 token = lex_token(file);
296 PARSE_TREE_ADD(PARSE_TYPE_LOR);
299 PARSE_TREE_ADD(PARSE_TYPE_BOR);
302 token = lex_token(file);
303 if (token == '=') { /* != */
304 token = lex_token(file);
305 PARSE_TREE_ADD(PARSE_TYPE_LNEQ);
308 PARSE_TREE_ADD(PARSE_TYPE_LNOT);
311 token = lex_token(file);
312 if (token == '=') { /* <= */
313 token = lex_token(file);
314 PARSE_TREE_ADD(PARSE_TYPE_LTEQ);
317 PARSE_TREE_ADD(PARSE_TYPE_LT);
320 token = lex_token(file);
321 if (token == '=') { /* >= */
322 token = lex_token(file);
323 PARSE_TREE_ADD(PARSE_TYPE_GTEQ);
326 PARSE_TREE_ADD(PARSE_TYPE_GT);
329 token = lex_token(file);
330 if (token == '=') { /* == */
331 token = lex_token(file);
332 PARSE_TREE_ADD(PARSE_TYPE_EQEQ);
335 PARSE_TREE_ADD(PARSE_TYPE_EQUAL);
338 token = lex_token(file);
339 PARSE_TREE_ADD(PARSE_TYPE_DONE);
342 token = lex_token(file);
343 PARSE_TREE_ADD(PARSE_TYPE_MINUS);
346 token = lex_token(file);
347 PARSE_TREE_ADD(PARSE_TYPE_ADD);
350 token = lex_token(file);
351 PARSE_TREE_ADD(PARSE_TYPE_LBS);
354 token = lex_token(file);
355 PARSE_TREE_ADD(PARSE_TYPE_RBS);
359 * TODO: Fix lexer to spit out ( ) as tokens, it seems the
360 * using '(' or ')' in parser doesn't work properly unless
361 * there are spaces before them to allow the lexer to properly
362 * seperate identifiers. -- otherwise it eats all of it.
365 token = lex_token(file);
366 PARSE_TREE_ADD(PARSE_TYPE_IDENT);
370 parse_debug(parseroot);