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("ADD \n");
107 case PARSE_TYPE_BAND: STORE("BITAND \n");
108 case PARSE_TYPE_BOR: STORE("BITOR \n");
109 case PARSE_TYPE_BREAK: STORE("BREAK \n");
110 case PARSE_TYPE_COMMA: STORE("SEPERATOR\n");
111 case PARSE_TYPE_CONTINUE: STORE("CONTINUE\n");
112 case PARSE_TYPE_DIVIDE: STORE("DIVIDE\n");
113 case PARSE_TYPE_EQUAL: STORE("ASSIGNMENT\n");
114 case PARSE_TYPE_GOTO: STORE("GOTO\n");
115 case PARSE_TYPE_DOT: STORE("DOT\n");
118 case PARSE_TYPE_ELIP: STORE("DECLTYPE: VALIST\n");
119 case PARSE_TYPE_ENTITY: STORE("DECLTYPE: ENTITY\n");
120 case PARSE_TYPE_FLOAT: STORE("DECLTYPE: FLOAT\n");
122 case PARSE_TYPE_GT: STORE("TEST: GREATER THAN\n");
123 case PARSE_TYPE_LT: STORE("TEST: LESS THAN\n");
124 case PARSE_TYPE_GTEQ: STORE("TEST: GREATER THAN OR EQUAL\n");
125 case PARSE_TYPE_LTEQ: STORE("TEST: LESS THAN OR EQUAL\n");
126 case PARSE_TYPE_LNEQ: STORE("TEST: NOT EQUAL\n");
127 case PARSE_TYPE_EQEQ: STORE("TEST: EQUAL-EQUAL\n");
129 case PARSE_TYPE_LBS: break;
130 case PARSE_TYPE_RBS: break;
132 case PARSE_TYPE_LAND: STORE("LOGICAL: AND\n");
133 case PARSE_TYPE_LNOT: STORE("LOGICAL: NOT\n");
134 case PARSE_TYPE_LOR: STORE("LOGICAL: OR\n");
135 case PARSE_TYPE_LPARTH: STORE("PARTH: END\n");
136 case PARSE_TYPE_RPARTH: STORE("PARTH: BEG\n");
138 case PARSE_TYPE_FOR: STORE("LOOP: FOR\n");
139 case PARSE_TYPE_DO: STORE("LOOP: DO\n");
140 case PARSE_TYPE_ELSE: STORE("BLOCK: ELSE\n");
141 case PARSE_TYPE_IF: STORE("BLOCK: IF\n");
148 * This just skips the token and throws it in the parse tree for later
149 * checking / optimization / codegen, it doesn't do anything with it
150 * like syntax check for legal use -- like it should as it's a TODO item
151 * which is not implemented
153 #define PARSE_TODO(X) { \
154 token = lex_token(file); \
159 int parse(struct lex_file *file) {
160 struct parsenode *parsetree = NULL;
161 struct parsenode *parseroot = NULL;
164 * Allocate memory for our parse tree:
165 * the parse tree is just a singly linked list which will contain
166 * all the data for code generation.
169 parseroot = mem_a(sizeof(struct parsenode));
171 return error(ERROR_INTERNAL, "Ran out of memory", " ");
172 parsetree = parseroot;
173 parsetree = parseroot;
177 while ((token = lex_token(file)) != ERROR_LEX && \
178 token != ERROR_COMPILER && \
179 token != ERROR_INTERNAL && \
180 token != ERROR_PARSE && \
181 token != ERROR_PREPRO && file->length >= 0) {
184 token = lex_token(file);
185 while ((token == ' ' || token == '\n') && file->length >= 0)
186 token = lex_token(file);
189 error(ERROR_PARSE, "Expected `(` after if\n", "");
191 PARSE_TREE_ADD(PARSE_TYPE_IF);
194 token = lex_token(file);
195 while ((token == ' ' || token == '\n') && file->length >= 0)
196 token = lex_token(file);
198 PARSE_TREE_ADD(PARSE_TYPE_ELSE);
201 token = lex_token(file);
202 while ((token == ' ' || token == '\n') && file->length >= 0)
203 token = lex_token(file);
205 PARSE_TREE_ADD(PARSE_TYPE_FOR);
209 token = lex_token(file);
210 printf("FOO: %s\n", file->lastok);
213 case TOKEN_TYPEDEF: {
216 token = lex_token(file);
217 token = lex_token(file); f = strdup(file->lastok);
218 token = lex_token(file);
219 token = lex_token(file); t = strdup(file->lastok);
223 /* free new strings */
230 case TOKEN_DO: PARSE_TODO(PARSE_TYPE_DO);
231 case TOKEN_WHILE: PARSE_TODO(PARSE_TYPE_WHILE);
232 case TOKEN_BREAK: PARSE_TODO(PARSE_TYPE_BREAK);
233 case TOKEN_CONTINUE: PARSE_TODO(PARSE_TYPE_CONTINUE);
234 case TOKEN_RETURN: PARSE_TODO(PARSE_TYPE_RETURN);
235 case TOKEN_GOTO: PARSE_TODO(PARSE_TYPE_GOTO);
236 case TOKEN_VOID: PARSE_TODO(PARSE_TYPE_VOID);
237 case TOKEN_STRING: PARSE_TODO(PARSE_TYPE_STRING);
238 case TOKEN_FLOAT: PARSE_TODO(PARSE_TYPE_FLOAT);
239 case TOKEN_VECTOR: PARSE_TODO(PARSE_TYPE_VECTOR);
240 case TOKEN_ENTITY: PARSE_TODO(PARSE_TYPE_ENTITY);
243 * From here down is all language punctuation: There is no
244 * need to actual create tokens from these because they're already
245 * tokenized as these individual tokens (which are in a special area
246 * of the ascii table which doesn't conflict with our other tokens
247 * which are higer than the ascii table.
250 token = lex_token(file);
251 if (token == '&') { /* && */
252 token = lex_token(file);
253 PARSE_TREE_ADD(PARSE_TYPE_LAND);
256 PARSE_TREE_ADD(PARSE_TYPE_BAND);
259 token = lex_token(file);
260 if (token == '|') { /* || */
261 token = lex_token(file);
262 PARSE_TREE_ADD(PARSE_TYPE_LOR);
265 PARSE_TREE_ADD(PARSE_TYPE_BOR);
268 token = lex_token(file);
269 if (token == '=') { /* != */
270 token = lex_token(file);
271 PARSE_TREE_ADD(PARSE_TYPE_LNEQ);
274 PARSE_TREE_ADD(PARSE_TYPE_LNOT);
277 token = lex_token(file);
278 if (token == '=') { /* <= */
279 token = lex_token(file);
280 PARSE_TREE_ADD(PARSE_TYPE_LTEQ);
283 PARSE_TREE_ADD(PARSE_TYPE_LT);
286 token = lex_token(file);
287 if (token == '=') { /* >= */
288 token = lex_token(file);
289 PARSE_TREE_ADD(PARSE_TYPE_GTEQ);
292 PARSE_TREE_ADD(PARSE_TYPE_GT);
295 token = lex_token(file);
296 if (token == '=') { /* == */
297 token = lex_token(file);
298 PARSE_TREE_ADD(PARSE_TYPE_EQEQ);
301 PARSE_TREE_ADD(PARSE_TYPE_EQUAL);
304 token = lex_token(file);
305 PARSE_TREE_ADD(PARSE_TYPE_DONE);
308 token = lex_token(file);
309 PARSE_TREE_ADD(PARSE_TYPE_MINUS);
312 token = lex_token(file);
313 PARSE_TREE_ADD(PARSE_TYPE_ADD);
316 token = lex_token(file);
317 PARSE_TREE_ADD(PARSE_TYPE_LPARTH);
320 token = lex_token(file);
321 PARSE_TREE_ADD(PARSE_TYPE_RPARTH);
324 token = lex_token(file);
325 PARSE_TREE_ADD(PARSE_TYPE_LBS);
328 token = lex_token(file);
329 PARSE_TREE_ADD(PARSE_TYPE_RBS);
333 parse_debug(parseroot);