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 // extension
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_LSS 29 // left subscript
63 #define PARSE_TYPE_RSS 30
64 #define PARSE_TYPE_LBS 31 // left bracket scope
65 #define PARSE_TYPE_RBS 32 // right bracket scope
66 #define PARSE_TYPE_ELIP 33 // ...
67 #define PARSE_TYPE_DOT 34
68 #define PARSE_TYPE_LT 35
69 #define PARSE_TYPE_GT 36
70 #define PARSE_TYPE_BAND 37
71 #define PARSE_TYPE_BOR 38
72 #define PARSE_TYPE_DONE 39 // finished statement
75 * Adds a parse type to the parse tree, this is where all the hard
76 * work actually begins.
78 #define PARSE_TREE_ADD(X) \
80 parsetree->next = mem_a(sizeof(struct parsenode)); \
81 parsetree->next->next = NULL; \
82 parsetree->next->type = (X); \
83 parsetree = parsetree->next; \
86 static const char *const parse_punct[] = {
87 "&&", "||", "<=", ">=", "==", "!=", ";", ",", "!", "*",
88 "/" , "(" , ")" , "-" , "+" , "=" , "[" , "]", "{", "}", "...",
89 "." , "<" , ">" , "&" , "|" , NULL
97 void parse_debug(struct parsenode *tree) {
98 while (tree && tree->next != NULL) {
100 if (tree->type == 0) {
105 switch (tree->type) {
106 case PARSE_TYPE_ADD: STORE("OPERATOR: ADD \n");
107 case PARSE_TYPE_BAND: STORE("OPERATOR: BITAND \n");
108 case PARSE_TYPE_BOR: STORE("OPERATOR: BITOR \n");
109 case PARSE_TYPE_COMMA: STORE("OPERATOR: SEPERATOR\n");
110 case PARSE_TYPE_DOT: STORE("OPERATOR: DOT\n");
111 case PARSE_TYPE_DIVIDE: STORE("OPERATOR: DIVIDE\n");
112 case PARSE_TYPE_EQUAL: STORE("OPERATOR: ASSIGNMENT\n");
114 case PARSE_TYPE_BREAK: STORE("STATEMENT: BREAK \n");
115 case PARSE_TYPE_CONTINUE: STORE("STATEMENT: CONTINUE\n");
116 case PARSE_TYPE_GOTO: STORE("STATEMENT: GOTO\n");
119 case PARSE_TYPE_ELIP: STORE("DECLTYPE: VALIST\n");
120 case PARSE_TYPE_ENTITY: STORE("DECLTYPE: ENTITY\n");
121 case PARSE_TYPE_FLOAT: STORE("DECLTYPE: FLOAT\n");
123 case PARSE_TYPE_GT: STORE("TEST: GREATER THAN\n");
124 case PARSE_TYPE_LT: STORE("TEST: LESS THAN\n");
125 case PARSE_TYPE_GTEQ: STORE("TEST: GREATER THAN OR EQUAL\n");
126 case PARSE_TYPE_LTEQ: STORE("TEST: LESS THAN OR EQUAL\n");
127 case PARSE_TYPE_LNEQ: STORE("TEST: NOT EQUAL\n");
128 case PARSE_TYPE_EQEQ: STORE("TEST: EQUAL-EQUAL\n");
130 case PARSE_TYPE_LBS: STORE("BLOCK: BEG\n");
131 case PARSE_TYPE_RBS: STORE("BLOCK: END\n");
133 case PARSE_TYPE_LAND: STORE("LOGICAL: AND\n");
134 case PARSE_TYPE_LNOT: STORE("LOGICAL: NOT\n");
135 case PARSE_TYPE_LOR: STORE("LOGICAL: OR\n");
137 case PARSE_TYPE_LPARTH: STORE("PARTH: BEG\n");
138 case PARSE_TYPE_RPARTH: STORE("PARTH: END\n");
140 case PARSE_TYPE_FOR: STORE("LOOP: FOR\n");
141 case PARSE_TYPE_DO: STORE("LOOP: DO\n");
143 case PARSE_TYPE_ELSE: STORE("BLOCK: ELSE\n");
144 case PARSE_TYPE_IF: STORE("BLOCK: IF\n");
151 * This just skips the token and throws it in the parse tree for later
152 * checking / optimization / codegen, it doesn't do anything with it
153 * like syntax check for legal use -- like it should as it's a TODO item
154 * which is not implemented
156 #define PARSE_TODO(X) { \
157 token = lex_token(file); \
162 int parse(struct lex_file *file) {
163 struct parsenode *parsetree = NULL;
164 struct parsenode *parseroot = NULL;
167 * Allocate memory for our parse tree:
168 * the parse tree is just a singly linked list which will contain
169 * all the data for code generation.
172 parseroot = mem_a(sizeof(struct parsenode));
174 return error(ERROR_INTERNAL, "Ran out of memory", " ");
175 parsetree = parseroot;
176 parsetree = parseroot;
180 while ((token = lex_token(file)) != ERROR_LEX && \
181 token != ERROR_COMPILER && \
182 token != ERROR_INTERNAL && \
183 token != ERROR_PARSE && \
184 token != ERROR_PREPRO && file->length >= 0) {
187 token = lex_token(file);
188 while ((token == ' ' || token == '\n') && file->length >= 0)
189 token = lex_token(file);
192 error(ERROR_PARSE, "Expected `(` after if\n", "");
194 PARSE_TREE_ADD(PARSE_TYPE_IF);
197 token = lex_token(file);
198 while ((token == ' ' || token == '\n') && file->length >= 0)
199 token = lex_token(file);
201 PARSE_TREE_ADD(PARSE_TYPE_ELSE);
204 token = lex_token(file);
205 while ((token == ' ' || token == '\n') && file->length >= 0)
206 token = lex_token(file);
208 PARSE_TREE_ADD(PARSE_TYPE_FOR);
212 token = lex_token(file);
213 printf("FOO: %s\n", file->lastok);
216 case TOKEN_TYPEDEF: {
219 token = lex_token(file);
220 token = lex_token(file); f = strdup(file->lastok);
221 token = lex_token(file);
222 token = lex_token(file); t = strdup(file->lastok);
226 /* free new strings */
233 case TOKEN_DO: PARSE_TODO(PARSE_TYPE_DO);
234 case TOKEN_WHILE: PARSE_TODO(PARSE_TYPE_WHILE);
235 case TOKEN_BREAK: PARSE_TODO(PARSE_TYPE_BREAK);
236 case TOKEN_CONTINUE: PARSE_TODO(PARSE_TYPE_CONTINUE);
237 case TOKEN_RETURN: PARSE_TODO(PARSE_TYPE_RETURN);
238 case TOKEN_GOTO: PARSE_TODO(PARSE_TYPE_GOTO);
239 case TOKEN_VOID: PARSE_TODO(PARSE_TYPE_VOID);
240 case TOKEN_STRING: PARSE_TODO(PARSE_TYPE_STRING);
241 case TOKEN_FLOAT: PARSE_TODO(PARSE_TYPE_FLOAT);
242 case TOKEN_VECTOR: PARSE_TODO(PARSE_TYPE_VECTOR);
243 case TOKEN_ENTITY: PARSE_TODO(PARSE_TYPE_ENTITY);
246 * From here down is all language punctuation: There is no
247 * need to actual create tokens from these because they're already
248 * tokenized as these individual tokens (which are in a special area
249 * of the ascii table which doesn't conflict with our other tokens
250 * which are higer than the ascii table.
253 token = lex_token(file);
254 if (token == '&') { /* && */
255 token = lex_token(file);
256 PARSE_TREE_ADD(PARSE_TYPE_LAND);
259 PARSE_TREE_ADD(PARSE_TYPE_BAND);
262 token = lex_token(file);
263 if (token == '|') { /* || */
264 token = lex_token(file);
265 PARSE_TREE_ADD(PARSE_TYPE_LOR);
268 PARSE_TREE_ADD(PARSE_TYPE_BOR);
271 token = lex_token(file);
272 if (token == '=') { /* != */
273 token = lex_token(file);
274 PARSE_TREE_ADD(PARSE_TYPE_LNEQ);
277 PARSE_TREE_ADD(PARSE_TYPE_LNOT);
280 token = lex_token(file);
281 if (token == '=') { /* <= */
282 token = lex_token(file);
283 PARSE_TREE_ADD(PARSE_TYPE_LTEQ);
286 PARSE_TREE_ADD(PARSE_TYPE_LT);
289 token = lex_token(file);
290 if (token == '=') { /* >= */
291 token = lex_token(file);
292 PARSE_TREE_ADD(PARSE_TYPE_GTEQ);
295 PARSE_TREE_ADD(PARSE_TYPE_GT);
298 token = lex_token(file);
299 if (token == '=') { /* == */
300 token = lex_token(file);
301 PARSE_TREE_ADD(PARSE_TYPE_EQEQ);
304 PARSE_TREE_ADD(PARSE_TYPE_EQUAL);
307 token = lex_token(file);
308 PARSE_TREE_ADD(PARSE_TYPE_DONE);
311 token = lex_token(file);
312 PARSE_TREE_ADD(PARSE_TYPE_MINUS);
315 token = lex_token(file);
316 PARSE_TREE_ADD(PARSE_TYPE_ADD);
319 token = lex_token(file);
320 PARSE_TREE_ADD(PARSE_TYPE_LPARTH);
323 token = lex_token(file);
324 PARSE_TREE_ADD(PARSE_TYPE_RPARTH);
327 token = lex_token(file);
328 PARSE_TREE_ADD(PARSE_TYPE_LBS);
331 token = lex_token(file);
332 PARSE_TREE_ADD(PARSE_TYPE_RBS);
336 parse_debug(parseroot);