diff options
author | David Walter Seikel | 2012-01-04 18:41:13 +1000 |
---|---|---|
committer | David Walter Seikel | 2012-01-04 18:41:13 +1000 |
commit | dd7595a3475407a7fa96a97393bae8c5220e8762 (patch) | |
tree | e341e911d7eb911a51684a7412ef7f7c7605d28e /libraries/embryo/src/bin/embryo_cc_sc2.c | |
parent | Add the skeleton. (diff) | |
download | SledjHamr-dd7595a3475407a7fa96a97393bae8c5220e8762.zip SledjHamr-dd7595a3475407a7fa96a97393bae8c5220e8762.tar.gz SledjHamr-dd7595a3475407a7fa96a97393bae8c5220e8762.tar.bz2 SledjHamr-dd7595a3475407a7fa96a97393bae8c5220e8762.tar.xz |
Add the base Enlightenment Foundation Libraries - eina, eet, evas, ecore, embryo, and edje.
Note that embryo wont be used, but I'm not sure yet if you can build edje without it.
Diffstat (limited to '')
-rw-r--r-- | libraries/embryo/src/bin/embryo_cc_sc2.c | 2779 |
1 files changed, 2779 insertions, 0 deletions
diff --git a/libraries/embryo/src/bin/embryo_cc_sc2.c b/libraries/embryo/src/bin/embryo_cc_sc2.c new file mode 100644 index 0000000..b3f4fae --- /dev/null +++ b/libraries/embryo/src/bin/embryo_cc_sc2.c | |||
@@ -0,0 +1,2779 @@ | |||
1 | /* Small compiler - File input, preprocessing and lexical analysis functions | ||
2 | * | ||
3 | * Copyright (c) ITB CompuPhase, 1997-2003 | ||
4 | * | ||
5 | * This software is provided "as-is", without any express or implied warranty. | ||
6 | * In no event will the authors be held liable for any damages arising from | ||
7 | * the use of this software. | ||
8 | * | ||
9 | * Permission is granted to anyone to use this software for any purpose, | ||
10 | * including commercial applications, and to alter it and redistribute it | ||
11 | * freely, subject to the following restrictions: | ||
12 | * | ||
13 | * 1. The origin of this software must not be misrepresented; you must not | ||
14 | * claim that you wrote the original software. If you use this software in | ||
15 | * a product, an acknowledgment in the product documentation would be | ||
16 | * appreciated but is not required. | ||
17 | * 2. Altered source versions must be plainly marked as such, and must not be | ||
18 | * misrepresented as being the original software. | ||
19 | * 3. This notice may not be removed or altered from any source distribution. | ||
20 | * | ||
21 | * Version: $Id: embryo_cc_sc2.c 62382 2011-08-12 12:39:29Z billiob $ | ||
22 | */ | ||
23 | |||
24 | |||
25 | #ifdef HAVE_CONFIG_H | ||
26 | # include <config.h> | ||
27 | #endif | ||
28 | |||
29 | #include <assert.h> | ||
30 | #include <stdio.h> | ||
31 | #include <stdlib.h> | ||
32 | #include <string.h> | ||
33 | #include <ctype.h> | ||
34 | #include <math.h> | ||
35 | #include "embryo_cc_sc.h" | ||
36 | #include "Embryo.h" | ||
37 | |||
38 | static int match(char *st, int end); | ||
39 | static cell litchar(char **lptr, int rawmode); | ||
40 | static int alpha(char c); | ||
41 | |||
42 | static int icomment; /* currently in multiline comment? */ | ||
43 | static int iflevel; /* nesting level if #if/#else/#endif */ | ||
44 | static int skiplevel; /* level at which we started skipping */ | ||
45 | static int elsedone; /* level at which we have seen an #else */ | ||
46 | static char term_expr[] = ""; | ||
47 | static int listline = -1; /* "current line" for the list file */ | ||
48 | |||
49 | /* pushstk & popstk | ||
50 | * | ||
51 | * Uses a LIFO stack to store information. The stack is used by doinclude(), | ||
52 | * doswitch() (to hold the state of "swactive") and some other routines. | ||
53 | * | ||
54 | * Porting note: I made the bold assumption that an integer will not be | ||
55 | * larger than a pointer (it may be smaller). That is, the stack element | ||
56 | * is typedef'ed as a pointer type, but I also store integers on it. See | ||
57 | * SC.H for "stkitem" | ||
58 | * | ||
59 | * Global references: stack,stkidx (private to pushstk() and popstk()) | ||
60 | */ | ||
61 | static stkitem stack[sSTKMAX]; | ||
62 | static int stkidx; | ||
63 | void | ||
64 | pushstk(stkitem val) | ||
65 | { | ||
66 | if (stkidx >= sSTKMAX) | ||
67 | error(102, "parser stack"); /* stack overflow (recursive include?) */ | ||
68 | stack[stkidx] = val; | ||
69 | stkidx += 1; | ||
70 | } | ||
71 | |||
72 | stkitem | ||
73 | popstk(void) | ||
74 | { | ||
75 | if (stkidx == 0) | ||
76 | return (stkitem) - 1; /* stack is empty */ | ||
77 | stkidx -= 1; | ||
78 | return stack[stkidx]; | ||
79 | } | ||
80 | |||
81 | int | ||
82 | plungequalifiedfile(char *name) | ||
83 | { | ||
84 | static char *extensions[] = { ".inc", ".sma", ".small" }; | ||
85 | FILE *fp; | ||
86 | char *ext; | ||
87 | int ext_idx; | ||
88 | |||
89 | ext_idx = 0; | ||
90 | do | ||
91 | { | ||
92 | fp = (FILE *) sc_opensrc(name); | ||
93 | ext = strchr(name, '\0'); /* save position */ | ||
94 | if (!fp) | ||
95 | { | ||
96 | /* try to append an extension */ | ||
97 | strcpy(ext, extensions[ext_idx]); | ||
98 | fp = (FILE *) sc_opensrc(name); | ||
99 | if (!fp) | ||
100 | *ext = '\0'; /* on failure, restore filename */ | ||
101 | } /* if */ | ||
102 | ext_idx++; | ||
103 | } | ||
104 | while ((!fp) && | ||
105 | (ext_idx < (int)(sizeof extensions / sizeof extensions[0]))); | ||
106 | if (!fp) | ||
107 | { | ||
108 | *ext = '\0'; /* restore filename */ | ||
109 | return FALSE; | ||
110 | } /* if */ | ||
111 | pushstk((stkitem) inpf); | ||
112 | pushstk((stkitem) inpfname); /* pointer to current file name */ | ||
113 | pushstk((stkitem) curlibrary); | ||
114 | pushstk((stkitem) iflevel); | ||
115 | assert(skiplevel == 0); | ||
116 | pushstk((stkitem) icomment); | ||
117 | pushstk((stkitem) fcurrent); | ||
118 | pushstk((stkitem) fline); | ||
119 | inpfname = strdup(name); /* set name of include file */ | ||
120 | if (!inpfname) | ||
121 | error(103); /* insufficient memory */ | ||
122 | inpf = fp; /* set input file pointer to include file */ | ||
123 | fnumber++; | ||
124 | fline = 0; /* set current line number to 0 */ | ||
125 | fcurrent = fnumber; | ||
126 | icomment = FALSE; | ||
127 | setfile(inpfname, fcurrent); | ||
128 | listline = -1; /* force a #line directive when changing the file */ | ||
129 | setactivefile(fcurrent); | ||
130 | return TRUE; | ||
131 | } | ||
132 | |||
133 | int | ||
134 | plungefile(char *name, int try_currentpath, int try_includepaths) | ||
135 | { | ||
136 | int result = FALSE; | ||
137 | int i; | ||
138 | char *ptr; | ||
139 | |||
140 | if (try_currentpath) | ||
141 | result = plungequalifiedfile(name); | ||
142 | |||
143 | if (try_includepaths && name[0] != DIRSEP_CHAR) | ||
144 | { | ||
145 | for (i = 0; !result && (ptr = get_path(i)); i++) | ||
146 | { | ||
147 | char path[PATH_MAX]; | ||
148 | |||
149 | strncpy(path, ptr, sizeof path); | ||
150 | path[sizeof path - 1] = '\0'; /* force '\0' termination */ | ||
151 | strncat(path, name, sizeof(path) - strlen(path)); | ||
152 | path[sizeof path - 1] = '\0'; | ||
153 | result = plungequalifiedfile(path); | ||
154 | } /* while */ | ||
155 | } /* if */ | ||
156 | return result; | ||
157 | } | ||
158 | |||
159 | static void | ||
160 | check_empty(char *lptr) | ||
161 | { | ||
162 | /* verifies that the string contains only whitespace */ | ||
163 | while (*lptr <= ' ' && *lptr != '\0') | ||
164 | lptr++; | ||
165 | if (*lptr != '\0') | ||
166 | error(38); /* extra characters on line */ | ||
167 | } | ||
168 | |||
169 | /* doinclude | ||
170 | * | ||
171 | * Gets the name of an include file, pushes the old file on the stack and | ||
172 | * sets some options. This routine doesn't use lex(), since lex() doesn't | ||
173 | * recognize file names (and directories). | ||
174 | * | ||
175 | * Global references: inpf (altered) | ||
176 | * inpfname (altered) | ||
177 | * fline (altered) | ||
178 | * lptr (altered) | ||
179 | */ | ||
180 | static void | ||
181 | doinclude(void) | ||
182 | { | ||
183 | char name[PATH_MAX], c; | ||
184 | int i, result; | ||
185 | |||
186 | while (*lptr <= ' ' && *lptr != 0) /* skip leading whitespace */ | ||
187 | lptr++; | ||
188 | if (*lptr == '<' || *lptr == '\"') | ||
189 | { | ||
190 | c = (char)((*lptr == '\"') ? '\"' : '>'); /* termination character */ | ||
191 | lptr++; | ||
192 | while (*lptr <= ' ' && *lptr != 0) /* skip whitespace after quote */ | ||
193 | lptr++; | ||
194 | } | ||
195 | else | ||
196 | { | ||
197 | c = '\0'; | ||
198 | } /* if */ | ||
199 | |||
200 | i = 0; | ||
201 | while ((*lptr != c) && (*lptr != '\0') && (i < (int)(sizeof(name) - 1))) /* find the end of the string */ | ||
202 | name[i++] = *lptr++; | ||
203 | while (i > 0 && name[i - 1] <= ' ') | ||
204 | i--; /* strip trailing whitespace */ | ||
205 | assert((i >= 0) && (i < (int)(sizeof(name)))); | ||
206 | name[i] = '\0'; /* zero-terminate the string */ | ||
207 | |||
208 | if (*lptr != c) | ||
209 | { /* verify correct string termination */ | ||
210 | error(37); /* invalid string */ | ||
211 | return; | ||
212 | } /* if */ | ||
213 | if (c != '\0') | ||
214 | check_empty(lptr + 1); /* verify that the rest of the line is whitespace */ | ||
215 | |||
216 | /* Include files between "..." or without quotes are read from the current | ||
217 | * directory, or from a list of "include directories". Include files | ||
218 | * between <...> are only read from the list of include directories. | ||
219 | */ | ||
220 | result = plungefile(name, (c != '>'), TRUE); | ||
221 | if (!result) | ||
222 | error(100, name); /* cannot read from ... (fatal error) */ | ||
223 | } | ||
224 | |||
225 | /* readline | ||
226 | * | ||
227 | * Reads in a new line from the input file pointed to by "inpf". readline() | ||
228 | * concatenates lines that end with a \ with the next line. If no more data | ||
229 | * can be read from the file, readline() attempts to pop off the previous file | ||
230 | * from the stack. If that fails too, it sets "freading" to 0. | ||
231 | * | ||
232 | * Global references: inpf,fline,inpfname,freading,icomment (altered) | ||
233 | */ | ||
234 | static void | ||
235 | readline(char *line) | ||
236 | { | ||
237 | int i, num, cont; | ||
238 | char *ptr; | ||
239 | |||
240 | if (lptr == term_expr) | ||
241 | return; | ||
242 | num = sLINEMAX; | ||
243 | cont = FALSE; | ||
244 | do | ||
245 | { | ||
246 | if (!inpf || sc_eofsrc(inpf)) | ||
247 | { | ||
248 | if (cont) | ||
249 | error(49); /* invalid line continuation */ | ||
250 | if (inpf && inpf != inpf_org) | ||
251 | sc_closesrc(inpf); | ||
252 | i = (int)(long)popstk(); | ||
253 | if (i == -1) | ||
254 | { /* All's done; popstk() returns "stack is empty" */ | ||
255 | freading = FALSE; | ||
256 | *line = '\0'; | ||
257 | /* when there is nothing more to read, the #if/#else stack should | ||
258 | * be empty and we should not be in a comment | ||
259 | */ | ||
260 | assert(iflevel >= 0); | ||
261 | if (iflevel > 0) | ||
262 | error(1, "#endif", "-end of file-"); | ||
263 | else if (icomment) | ||
264 | error(1, "*/", "-end of file-"); | ||
265 | return; | ||
266 | } /* if */ | ||
267 | fline = i; | ||
268 | fcurrent = (int)(long)popstk(); | ||
269 | icomment = (int)(long)popstk(); | ||
270 | assert(skiplevel == 0); /* skiplevel was not stored on stack, because it should always be zero at this point */ | ||
271 | iflevel = (int)(long)popstk(); | ||
272 | curlibrary = (constvalue *) popstk(); | ||
273 | free(inpfname); /* return memory allocated for the include file name */ | ||
274 | inpfname = (char *)popstk(); | ||
275 | inpf = (FILE *) popstk(); | ||
276 | setactivefile(fcurrent); | ||
277 | listline = -1; /* force a #line directive when changing the file */ | ||
278 | elsedone = 0; | ||
279 | } /* if */ | ||
280 | |||
281 | if (!sc_readsrc(inpf, line, num)) | ||
282 | { | ||
283 | *line = '\0'; /* delete line */ | ||
284 | cont = FALSE; | ||
285 | } | ||
286 | else | ||
287 | { | ||
288 | /* check whether to erase leading spaces */ | ||
289 | if (cont) | ||
290 | { | ||
291 | char *ptr = line; | ||
292 | |||
293 | while (*ptr == ' ' || *ptr == '\t') | ||
294 | ptr++; | ||
295 | if (ptr != line) | ||
296 | memmove(line, ptr, strlen(ptr) + 1); | ||
297 | } /* if */ | ||
298 | cont = FALSE; | ||
299 | /* check whether a full line was read */ | ||
300 | if (!strchr(line, '\n') && !sc_eofsrc(inpf)) | ||
301 | error(75); /* line too long */ | ||
302 | /* check if the next line must be concatenated to this line */ | ||
303 | if ((ptr = strchr(line, '\n')) && ptr > line) | ||
304 | { | ||
305 | assert(*(ptr + 1) == '\0'); /* '\n' should be last in the string */ | ||
306 | while (ptr > line | ||
307 | && (*ptr == '\n' || *ptr == ' ' || *ptr == '\t')) | ||
308 | ptr--; /* skip trailing whitespace */ | ||
309 | if (*ptr == '\\') | ||
310 | { | ||
311 | cont = TRUE; | ||
312 | /* set '\a' at the position of '\\' to make it possible to check | ||
313 | * for a line continuation in a single line comment (error 49) | ||
314 | */ | ||
315 | *ptr++ = '\a'; | ||
316 | *ptr = '\0'; /* erase '\n' (and any trailing whitespace) */ | ||
317 | } /* if */ | ||
318 | } /* if */ | ||
319 | num -= strlen(line); | ||
320 | line += strlen(line); | ||
321 | } /* if */ | ||
322 | fline += 1; | ||
323 | } | ||
324 | while (num >= 0 && cont); | ||
325 | } | ||
326 | |||
327 | /* stripcom | ||
328 | * | ||
329 | * Replaces all comments from the line by space characters. It updates | ||
330 | * a global variable ("icomment") for multiline comments. | ||
331 | * | ||
332 | * This routine also supports the C++ extension for single line comments. | ||
333 | * These comments are started with "//" and end at the end of the line. | ||
334 | * | ||
335 | * Global references: icomment (private to "stripcom") | ||
336 | */ | ||
337 | static void | ||
338 | stripcom(char *line) | ||
339 | { | ||
340 | char c; | ||
341 | |||
342 | while (*line) | ||
343 | { | ||
344 | if (icomment) | ||
345 | { | ||
346 | if (*line == '*' && *(line + 1) == '/') | ||
347 | { | ||
348 | icomment = FALSE; /* comment has ended */ | ||
349 | *line = ' '; /* replace '*' and '/' characters by spaces */ | ||
350 | *(line + 1) = ' '; | ||
351 | line += 2; | ||
352 | } | ||
353 | else | ||
354 | { | ||
355 | if (*line == '/' && *(line + 1) == '*') | ||
356 | error(216); /* nested comment */ | ||
357 | *line = ' '; /* replace comments by spaces */ | ||
358 | line += 1; | ||
359 | } /* if */ | ||
360 | } | ||
361 | else | ||
362 | { | ||
363 | if (*line == '/' && *(line + 1) == '*') | ||
364 | { | ||
365 | icomment = TRUE; /* start comment */ | ||
366 | *line = ' '; /* replace '/' and '*' characters by spaces */ | ||
367 | *(line + 1) = ' '; | ||
368 | line += 2; | ||
369 | } | ||
370 | else if (*line == '/' && *(line + 1) == '/') | ||
371 | { /* comment to end of line */ | ||
372 | if (strchr(line, '\a')) | ||
373 | error(49); /* invalid line continuation */ | ||
374 | *line++ = '\n'; /* put "newline" at first slash */ | ||
375 | *line = '\0'; /* put "zero-terminator" at second slash */ | ||
376 | } | ||
377 | else | ||
378 | { | ||
379 | if (*line == '\"' || *line == '\'') | ||
380 | { /* leave literals unaltered */ | ||
381 | c = *line; /* ending quote, single or double */ | ||
382 | line += 1; | ||
383 | while ((*line != c || *(line - 1) == '\\') | ||
384 | && *line != '\0') | ||
385 | line += 1; | ||
386 | line += 1; /* skip final quote */ | ||
387 | } | ||
388 | else | ||
389 | { | ||
390 | line += 1; | ||
391 | } /* if */ | ||
392 | } /* if */ | ||
393 | } /* if */ | ||
394 | } /* while */ | ||
395 | } | ||
396 | |||
397 | /* btoi | ||
398 | * | ||
399 | * Attempts to interpret a numeric symbol as a boolean value. On success | ||
400 | * it returns the number of characters processed (so the line pointer can be | ||
401 | * adjusted) and the value is stored in "val". Otherwise it returns 0 and | ||
402 | * "val" is garbage. | ||
403 | * | ||
404 | * A boolean value must start with "0b" | ||
405 | */ | ||
406 | static int | ||
407 | btoi(cell * val, char *curptr) | ||
408 | { | ||
409 | char *ptr; | ||
410 | |||
411 | *val = 0; | ||
412 | ptr = curptr; | ||
413 | if (*ptr == '0' && *(ptr + 1) == 'b') | ||
414 | { | ||
415 | ptr += 2; | ||
416 | while (*ptr == '0' || *ptr == '1' || *ptr == '_') | ||
417 | { | ||
418 | if (*ptr != '_') | ||
419 | *val = (*val << 1) | (*ptr - '0'); | ||
420 | ptr++; | ||
421 | } /* while */ | ||
422 | } | ||
423 | else | ||
424 | { | ||
425 | return 0; | ||
426 | } /* if */ | ||
427 | if (alphanum(*ptr)) /* number must be delimited by non-alphanumeric char */ | ||
428 | return 0; | ||
429 | else | ||
430 | return (int)(ptr - curptr); | ||
431 | } | ||
432 | |||
433 | /* dtoi | ||
434 | * | ||
435 | * Attempts to interpret a numeric symbol as a decimal value. On success | ||
436 | * it returns the number of characters processed and the value is stored in | ||
437 | * "val". Otherwise it returns 0 and "val" is garbage. | ||
438 | */ | ||
439 | static int | ||
440 | dtoi(cell * val, char *curptr) | ||
441 | { | ||
442 | char *ptr; | ||
443 | |||
444 | *val = 0; | ||
445 | ptr = curptr; | ||
446 | if (!isdigit(*ptr)) /* should start with digit */ | ||
447 | return 0; | ||
448 | while (isdigit(*ptr) || *ptr == '_') | ||
449 | { | ||
450 | if (*ptr != '_') | ||
451 | *val = (*val * 10) + (*ptr - '0'); | ||
452 | ptr++; | ||
453 | } /* while */ | ||
454 | if (alphanum(*ptr)) /* number must be delimited by non-alphanumerical */ | ||
455 | return 0; | ||
456 | if (*ptr == '.' && isdigit(*(ptr + 1))) | ||
457 | return 0; /* but a fractional part must not be present */ | ||
458 | return (int)(ptr - curptr); | ||
459 | } | ||
460 | |||
461 | /* htoi | ||
462 | * | ||
463 | * Attempts to interpret a numeric symbol as a hexadecimal value. On | ||
464 | * success it returns the number of characters processed and the value is | ||
465 | * stored in "val". Otherwise it return 0 and "val" is garbage. | ||
466 | */ | ||
467 | static int | ||
468 | htoi(cell * val, char *curptr) | ||
469 | { | ||
470 | char *ptr; | ||
471 | |||
472 | *val = 0; | ||
473 | ptr = curptr; | ||
474 | if (!isdigit(*ptr)) /* should start with digit */ | ||
475 | return 0; | ||
476 | if (*ptr == '0' && *(ptr + 1) == 'x') | ||
477 | { /* C style hexadecimal notation */ | ||
478 | ptr += 2; | ||
479 | while (isxdigit(*ptr) || *ptr == '_') | ||
480 | { | ||
481 | if (*ptr != '_') | ||
482 | { | ||
483 | assert(isxdigit(*ptr)); | ||
484 | *val = *val << 4; | ||
485 | if (isdigit(*ptr)) | ||
486 | *val += (*ptr - '0'); | ||
487 | else | ||
488 | *val += (tolower(*ptr) - 'a' + 10); | ||
489 | } /* if */ | ||
490 | ptr++; | ||
491 | } /* while */ | ||
492 | } | ||
493 | else | ||
494 | { | ||
495 | return 0; | ||
496 | } /* if */ | ||
497 | if (alphanum(*ptr)) | ||
498 | return 0; | ||
499 | else | ||
500 | return (int)(ptr - curptr); | ||
501 | } | ||
502 | |||
503 | #if defined LINUX | ||
504 | static double | ||
505 | pow10(int value) | ||
506 | { | ||
507 | double res = 1.0; | ||
508 | |||
509 | while (value >= 4) | ||
510 | { | ||
511 | res *= 10000.0; | ||
512 | value -= 5; | ||
513 | } /* while */ | ||
514 | while (value >= 2) | ||
515 | { | ||
516 | res *= 100.0; | ||
517 | value -= 2; | ||
518 | } /* while */ | ||
519 | while (value >= 1) | ||
520 | { | ||
521 | res *= 10.0; | ||
522 | value -= 1; | ||
523 | } /* while */ | ||
524 | return res; | ||
525 | } | ||
526 | #endif | ||
527 | |||
528 | /* ftoi | ||
529 | * | ||
530 | * Attempts to interpret a numeric symbol as a rational number, either as | ||
531 | * IEEE 754 single precision floating point or as a fixed point integer. | ||
532 | * On success it returns the number of characters processed and the value is | ||
533 | * stored in "val". Otherwise it returns 0 and "val" is unchanged. | ||
534 | * | ||
535 | * Small has stricter definition for floating point numbers than most: | ||
536 | * o the value must start with a digit; ".5" is not a valid number, you | ||
537 | * should write "0.5" | ||
538 | * o a period must appear in the value, even if an exponent is given; "2e3" | ||
539 | * is not a valid number, you should write "2.0e3" | ||
540 | * o at least one digit must follow the period; "6." is not a valid number, | ||
541 | * you should write "6.0" | ||
542 | */ | ||
543 | static int | ||
544 | ftoi(cell * val, char *curptr) | ||
545 | { | ||
546 | char *ptr; | ||
547 | double fnum, ffrac, fmult; | ||
548 | unsigned long dnum, dbase; | ||
549 | int i, ignore; | ||
550 | |||
551 | assert(rational_digits >= 0 && rational_digits < 9); | ||
552 | for (i = 0, dbase = 1; i < rational_digits; i++) | ||
553 | dbase *= 10; | ||
554 | fnum = 0.0; | ||
555 | dnum = 0L; | ||
556 | ptr = curptr; | ||
557 | if (!isdigit(*ptr)) /* should start with digit */ | ||
558 | return 0; | ||
559 | while (isdigit(*ptr) || *ptr == '_') | ||
560 | { | ||
561 | if (*ptr != '_') | ||
562 | { | ||
563 | fnum = (fnum * 10.0) + (*ptr - '0'); | ||
564 | dnum = (dnum * 10L) + (*ptr - '0') * dbase; | ||
565 | } /* if */ | ||
566 | ptr++; | ||
567 | } /* while */ | ||
568 | if (*ptr != '.') | ||
569 | return 0; /* there must be a period */ | ||
570 | ptr++; | ||
571 | if (!isdigit(*ptr)) /* there must be at least one digit after the dot */ | ||
572 | return 0; | ||
573 | ffrac = 0.0; | ||
574 | fmult = 1.0; | ||
575 | ignore = FALSE; | ||
576 | while (isdigit(*ptr) || *ptr == '_') | ||
577 | { | ||
578 | if (*ptr != '_') | ||
579 | { | ||
580 | ffrac = (ffrac * 10.0) + (*ptr - '0'); | ||
581 | fmult = fmult / 10.0; | ||
582 | dbase /= 10L; | ||
583 | dnum += (*ptr - '0') * dbase; | ||
584 | if (dbase == 0L && sc_rationaltag && rational_digits > 0 | ||
585 | && !ignore) | ||
586 | { | ||
587 | error(222); /* number of digits exceeds rational number precision */ | ||
588 | ignore = TRUE; | ||
589 | } /* if */ | ||
590 | } /* if */ | ||
591 | ptr++; | ||
592 | } /* while */ | ||
593 | fnum += ffrac * fmult; /* form the number so far */ | ||
594 | if (*ptr == 'e') | ||
595 | { /* optional fractional part */ | ||
596 | int exp, sign; | ||
597 | |||
598 | ptr++; | ||
599 | if (*ptr == '-') | ||
600 | { | ||
601 | sign = -1; | ||
602 | ptr++; | ||
603 | } | ||
604 | else | ||
605 | { | ||
606 | sign = 1; | ||
607 | } /* if */ | ||
608 | if (!isdigit(*ptr)) /* 'e' should be followed by a digit */ | ||
609 | return 0; | ||
610 | exp = 0; | ||
611 | while (isdigit(*ptr)) | ||
612 | { | ||
613 | exp = (exp * 10) + (*ptr - '0'); | ||
614 | ptr++; | ||
615 | } /* while */ | ||
616 | #if defined LINUX | ||
617 | fmult = pow10(exp * sign); | ||
618 | #else | ||
619 | fmult = pow(10, exp * sign); | ||
620 | #endif | ||
621 | fnum *= fmult; | ||
622 | dnum *= (unsigned long)(fmult + 0.5); | ||
623 | } /* if */ | ||
624 | |||
625 | /* decide how to store the number */ | ||
626 | if (sc_rationaltag == 0) | ||
627 | { | ||
628 | error(70); /* rational number support was not enabled */ | ||
629 | *val = 0; | ||
630 | } | ||
631 | else if (rational_digits == 0) | ||
632 | { | ||
633 | float f = (float) fnum; | ||
634 | /* floating point */ | ||
635 | *val = EMBRYO_FLOAT_TO_CELL(f); | ||
636 | #if !defined NDEBUG | ||
637 | /* I assume that the C/C++ compiler stores "float" values in IEEE 754 | ||
638 | * format (as mandated in the ANSI standard). Test this assumption anyway. | ||
639 | */ | ||
640 | { | ||
641 | float test1 = 0.0, test2 = 50.0; | ||
642 | Embryo_Cell c1 = EMBRYO_FLOAT_TO_CELL(test1); | ||
643 | Embryo_Cell c2 = EMBRYO_FLOAT_TO_CELL(test2); | ||
644 | |||
645 | if (c1 != 0x00000000L) | ||
646 | { | ||
647 | fprintf(stderr, | ||
648 | "embryo_cc: WARNING! you compiler does not SEEM to interpret IEEE floating\n" | ||
649 | "point math as embryo expects. this could be bad.\n" | ||
650 | "\n" | ||
651 | "(float 0.0 != 0x00000000 bitpattern, 0x%08x instead)\n" | ||
652 | "\n" | ||
653 | "this could be an issue with you compiling embryo with gcc 3.2.x that seems\n" | ||
654 | "to trigger this sanity check. we are not sure why yet, but gcc 3.3.x works\n" | ||
655 | , c1); | ||
656 | } | ||
657 | else if (c2 != 0x42480000L) | ||
658 | { | ||
659 | fprintf(stderr, | ||
660 | "embryo_cc: WARNING! you compiler does not SEEM to interpret IEEE floating\n" | ||
661 | "point math as embryo expects. This could be bad.\n" | ||
662 | "\n" | ||
663 | "(float 50.0 != 0x42480000 bitpattern, 0x%08x instead)\n" | ||
664 | "\n" | ||
665 | "this could be an issue with you compiling embryo with gcc 3.2.x that seems\n" | ||
666 | "to trigger this sanity check. we are not sure why yet, but gcc 3.3.x works\n" | ||
667 | , c2); | ||
668 | } | ||
669 | } | ||
670 | #endif | ||
671 | } | ||
672 | else | ||
673 | { | ||
674 | /* fixed point */ | ||
675 | *val = (cell) dnum; | ||
676 | } /* if */ | ||
677 | |||
678 | return (int)(ptr - curptr); | ||
679 | } | ||
680 | |||
681 | /* number | ||
682 | * | ||
683 | * Reads in a number (binary, decimal or hexadecimal). It returns the number | ||
684 | * of characters processed or 0 if the symbol couldn't be interpreted as a | ||
685 | * number (in this case the argument "val" remains unchanged). This routine | ||
686 | * relies on the 'early dropout' implementation of the logical or (||) | ||
687 | * operator. | ||
688 | * | ||
689 | * Note: the routine doesn't check for a sign (+ or -). The - is checked | ||
690 | * for at "hier2()" (in fact, it is viewed as an operator, not as a | ||
691 | * sign) and the + is invalid (as in K&R C, and unlike ANSI C). | ||
692 | */ | ||
693 | static int | ||
694 | number(cell * val, char *curptr) | ||
695 | { | ||
696 | int i; | ||
697 | cell value; | ||
698 | |||
699 | if ((i = btoi(&value, curptr)) != 0 /* binary? */ | ||
700 | || (i = htoi(&value, curptr)) != 0 /* hexadecimal? */ | ||
701 | || (i = dtoi(&value, curptr)) != 0) /* decimal? */ | ||
702 | { | ||
703 | *val = value; | ||
704 | return i; | ||
705 | } | ||
706 | else | ||
707 | { | ||
708 | return 0; /* else not a number */ | ||
709 | } /* if */ | ||
710 | } | ||
711 | |||
712 | static void | ||
713 | chrcat(char *str, char chr) | ||
714 | { | ||
715 | str = strchr(str, '\0'); | ||
716 | *str++ = chr; | ||
717 | *str = '\0'; | ||
718 | } | ||
719 | |||
720 | static int | ||
721 | preproc_expr(cell * val, int *tag) | ||
722 | { | ||
723 | int result; | ||
724 | int index; | ||
725 | cell code_index; | ||
726 | char *term; | ||
727 | |||
728 | /* Disable staging; it should be disabled already because | ||
729 | * expressions may not be cut off half-way between conditional | ||
730 | * compilations. Reset the staging index, but keep the code | ||
731 | * index. | ||
732 | */ | ||
733 | if (stgget(&index, &code_index)) | ||
734 | { | ||
735 | error(57); /* unfinished expression */ | ||
736 | stgdel(0, code_index); | ||
737 | stgset(FALSE); | ||
738 | } /* if */ | ||
739 | /* append a special symbol to the string, so the expression | ||
740 | * analyzer won't try to read a next line when it encounters | ||
741 | * an end-of-line | ||
742 | */ | ||
743 | assert(strlen(pline) < sLINEMAX); | ||
744 | term = strchr(pline, '\0'); | ||
745 | assert(term != NULL); | ||
746 | chrcat(pline, PREPROC_TERM); /* the "DEL" code (see SC.H) */ | ||
747 | result = constexpr(val, tag); /* get value (or 0 on error) */ | ||
748 | *term = '\0'; /* erase the token (if still present) */ | ||
749 | lexclr(FALSE); /* clear any "pushed" tokens */ | ||
750 | return result; | ||
751 | } | ||
752 | |||
753 | /* getstring | ||
754 | * Returns returns a pointer behind the closing quote or to the other | ||
755 | * character that caused the input to be ended. | ||
756 | */ | ||
757 | static char * | ||
758 | getstring(char *dest, int max) | ||
759 | { | ||
760 | assert(dest != NULL); | ||
761 | *dest = '\0'; | ||
762 | while (*lptr <= ' ' && *lptr != '\0') | ||
763 | lptr++; /* skip whitespace */ | ||
764 | if (*lptr != '"') | ||
765 | { | ||
766 | error(37); /* invalid string */ | ||
767 | } | ||
768 | else | ||
769 | { | ||
770 | int len = 0; | ||
771 | |||
772 | lptr++; /* skip " */ | ||
773 | while (*lptr != '"' && *lptr != '\0') | ||
774 | { | ||
775 | if (len < max - 1) | ||
776 | dest[len++] = *lptr; | ||
777 | lptr++; | ||
778 | } /* if */ | ||
779 | dest[len] = '\0'; | ||
780 | if (*lptr == '"') | ||
781 | lptr++; /* skip closing " */ | ||
782 | else | ||
783 | error(37); /* invalid string */ | ||
784 | } /* if */ | ||
785 | return lptr; | ||
786 | } | ||
787 | |||
788 | enum | ||
789 | { | ||
790 | CMD_NONE, | ||
791 | CMD_TERM, | ||
792 | CMD_EMPTYLINE, | ||
793 | CMD_CONDFALSE, | ||
794 | CMD_INCLUDE, | ||
795 | CMD_DEFINE, | ||
796 | CMD_IF, | ||
797 | CMD_DIRECTIVE, | ||
798 | }; | ||
799 | |||
800 | /* command | ||
801 | * | ||
802 | * Recognizes the compiler directives. The function returns: | ||
803 | * CMD_NONE the line must be processed | ||
804 | * CMD_TERM a pending expression must be completed before processing further lines | ||
805 | * Other value: the line must be skipped, because: | ||
806 | * CMD_CONDFALSE false "#if.." code | ||
807 | * CMD_EMPTYLINE line is empty | ||
808 | * CMD_INCLUDE the line contains a #include directive | ||
809 | * CMD_DEFINE the line contains a #subst directive | ||
810 | * CMD_IF the line contains a #if/#else/#endif directive | ||
811 | * CMD_DIRECTIVE the line contains some other compiler directive | ||
812 | * | ||
813 | * Global variables: iflevel, skiplevel, elsedone (altered) | ||
814 | * lptr (altered) | ||
815 | */ | ||
816 | static int | ||
817 | command(void) | ||
818 | { | ||
819 | int tok, ret; | ||
820 | cell val; | ||
821 | char *str; | ||
822 | int index; | ||
823 | cell code_index; | ||
824 | |||
825 | while (*lptr <= ' ' && *lptr != '\0') | ||
826 | lptr += 1; | ||
827 | if (*lptr == '\0') | ||
828 | return CMD_EMPTYLINE; /* empty line */ | ||
829 | if (*lptr != '#') | ||
830 | return skiplevel > 0 ? CMD_CONDFALSE : CMD_NONE; /* it is not a compiler directive */ | ||
831 | /* compiler directive found */ | ||
832 | indent_nowarn = TRUE; /* allow loose indentation" */ | ||
833 | lexclr(FALSE); /* clear any "pushed" tokens */ | ||
834 | /* on a pending expression, force to return a silent ';' token and force to | ||
835 | * re-read the line | ||
836 | */ | ||
837 | if (!sc_needsemicolon && stgget(&index, &code_index)) | ||
838 | { | ||
839 | lptr = term_expr; | ||
840 | return CMD_TERM; | ||
841 | } /* if */ | ||
842 | tok = lex(&val, &str); | ||
843 | ret = skiplevel > 0 ? CMD_CONDFALSE : CMD_DIRECTIVE; /* preset 'ret' to CMD_DIRECTIVE (most common case) */ | ||
844 | switch (tok) | ||
845 | { | ||
846 | case tpIF: /* conditional compilation */ | ||
847 | ret = CMD_IF; | ||
848 | iflevel += 1; | ||
849 | if (skiplevel) | ||
850 | break; /* break out of switch */ | ||
851 | preproc_expr(&val, NULL); /* get value (or 0 on error) */ | ||
852 | if (!val) | ||
853 | skiplevel = iflevel; | ||
854 | check_empty(lptr); | ||
855 | break; | ||
856 | case tpELSE: | ||
857 | ret = CMD_IF; | ||
858 | if (iflevel == 0 && skiplevel == 0) | ||
859 | { | ||
860 | error(26); /* no matching #if */ | ||
861 | errorset(sRESET); | ||
862 | } | ||
863 | else | ||
864 | { | ||
865 | if (elsedone == iflevel) | ||
866 | error(60); /* multiple #else directives between #if ... #endif */ | ||
867 | elsedone = iflevel; | ||
868 | if (skiplevel == iflevel) | ||
869 | skiplevel = 0; | ||
870 | else if (skiplevel == 0) | ||
871 | skiplevel = iflevel; | ||
872 | } /* if */ | ||
873 | check_empty(lptr); | ||
874 | break; | ||
875 | #if 0 /* ??? *really* need to use a stack here */ | ||
876 | case tpELSEIF: | ||
877 | ret = CMD_IF; | ||
878 | if (iflevel == 0 && skiplevel == 0) | ||
879 | { | ||
880 | error(26); /* no matching #if */ | ||
881 | errorset(sRESET); | ||
882 | } | ||
883 | else if (elsedone == iflevel) | ||
884 | { | ||
885 | error(61); /* #elseif directive may not follow an #else */ | ||
886 | errorset(sRESET); | ||
887 | } | ||
888 | else | ||
889 | { | ||
890 | preproc_expr(&val, NULL); /* get value (or 0 on error) */ | ||
891 | if (skiplevel == 0) | ||
892 | skiplevel = iflevel; /* we weren't skipping, start skipping now */ | ||
893 | else if (val) | ||
894 | skiplevel = 0; /* we were skipping, condition is valid -> stop skipping */ | ||
895 | /* else: we were skipping and condition is invalid -> keep skipping */ | ||
896 | check_empty(lptr); | ||
897 | } /* if */ | ||
898 | break; | ||
899 | #endif | ||
900 | case tpENDIF: | ||
901 | ret = CMD_IF; | ||
902 | if (iflevel == 0 && skiplevel == 0) | ||
903 | { | ||
904 | error(26); | ||
905 | errorset(sRESET); | ||
906 | } | ||
907 | else | ||
908 | { | ||
909 | if (skiplevel == iflevel) | ||
910 | skiplevel = 0; | ||
911 | if (elsedone == iflevel) | ||
912 | elsedone = 0; /* ??? actually, should build a stack of #if/#endif and keep | ||
913 | * the state whether an #else was seen per nesting level */ | ||
914 | iflevel -= 1; | ||
915 | } /* if */ | ||
916 | check_empty(lptr); | ||
917 | break; | ||
918 | case tINCLUDE: /* #include directive */ | ||
919 | ret = CMD_INCLUDE; | ||
920 | if (skiplevel == 0) | ||
921 | doinclude(); | ||
922 | break; | ||
923 | case tpFILE: | ||
924 | if (skiplevel == 0) | ||
925 | { | ||
926 | char pathname[PATH_MAX]; | ||
927 | |||
928 | lptr = getstring(pathname, sizeof pathname); | ||
929 | if (pathname[0] != '\0') | ||
930 | { | ||
931 | free(inpfname); | ||
932 | inpfname = strdup(pathname); | ||
933 | if (!inpfname) | ||
934 | error(103); /* insufficient memory */ | ||
935 | } /* if */ | ||
936 | } /* if */ | ||
937 | check_empty(lptr); | ||
938 | break; | ||
939 | case tpLINE: | ||
940 | if (skiplevel == 0) | ||
941 | { | ||
942 | if (lex(&val, &str) != tNUMBER) | ||
943 | error(8); /* invalid/non-constant expression */ | ||
944 | fline = (int)val; | ||
945 | |||
946 | while (*lptr == ' ' && *lptr != '\0') | ||
947 | lptr++; /* skip whitespace */ | ||
948 | if (*lptr == '"') | ||
949 | { | ||
950 | char pathname[PATH_MAX]; | ||
951 | |||
952 | lptr = getstring(pathname, sizeof pathname); | ||
953 | if (pathname[0] != '\0') | ||
954 | { | ||
955 | free(inpfname); | ||
956 | inpfname = strdup(pathname); | ||
957 | if (!inpfname) | ||
958 | error(103); /* insufficient memory */ | ||
959 | } /* if */ | ||
960 | } | ||
961 | } /* if */ | ||
962 | check_empty(lptr); | ||
963 | break; | ||
964 | case tpASSERT: | ||
965 | if (skiplevel == 0 && (sc_debug & sCHKBOUNDS) != 0) | ||
966 | { | ||
967 | preproc_expr(&val, NULL); /* get constant expression (or 0 on error) */ | ||
968 | if (!val) | ||
969 | error(7); /* assertion failed */ | ||
970 | check_empty(lptr); | ||
971 | } /* if */ | ||
972 | break; | ||
973 | case tpPRAGMA: | ||
974 | if (skiplevel == 0) | ||
975 | { | ||
976 | if (lex(&val, &str) == tSYMBOL) | ||
977 | { | ||
978 | if (strcmp(str, "ctrlchar") == 0) | ||
979 | { | ||
980 | if (lex(&val, &str) != tNUMBER) | ||
981 | error(27); /* invalid character constant */ | ||
982 | sc_ctrlchar = (char)val; | ||
983 | } | ||
984 | else if (strcmp(str, "compress") == 0) | ||
985 | { | ||
986 | cell val; | ||
987 | |||
988 | preproc_expr(&val, NULL); | ||
989 | sc_compress = (int)val; /* switch code packing on/off */ | ||
990 | } | ||
991 | else if (strcmp(str, "dynamic") == 0) | ||
992 | { | ||
993 | preproc_expr(&sc_stksize, NULL); | ||
994 | } | ||
995 | else if (strcmp(str, "library") == 0) | ||
996 | { | ||
997 | char name[sNAMEMAX + 1]; | ||
998 | |||
999 | while (*lptr <= ' ' && *lptr != '\0') | ||
1000 | lptr++; | ||
1001 | if (*lptr == '"') | ||
1002 | { | ||
1003 | lptr = getstring(name, sizeof name); | ||
1004 | } | ||
1005 | else | ||
1006 | { | ||
1007 | int i; | ||
1008 | |||
1009 | for (i = 0; | ||
1010 | (i < (int)(sizeof(name))) && | ||
1011 | (alphanum(*lptr)); | ||
1012 | i++, lptr++) | ||
1013 | name[i] = *lptr; | ||
1014 | name[i] = '\0'; | ||
1015 | } /* if */ | ||
1016 | if (name[0] == '\0') | ||
1017 | { | ||
1018 | curlibrary = NULL; | ||
1019 | } | ||
1020 | else | ||
1021 | { | ||
1022 | if (strlen(name) > sEXPMAX) | ||
1023 | error(220, name, sEXPMAX); /* exported symbol is truncated */ | ||
1024 | /* add the name if it does not yet exist in the table */ | ||
1025 | if (!find_constval(&libname_tab, name, 0)) | ||
1026 | curlibrary = | ||
1027 | append_constval(&libname_tab, name, 0, 0); | ||
1028 | } /* if */ | ||
1029 | } | ||
1030 | else if (strcmp(str, "pack") == 0) | ||
1031 | { | ||
1032 | cell val; | ||
1033 | |||
1034 | preproc_expr(&val, NULL); /* default = packed/unpacked */ | ||
1035 | sc_packstr = (int)val; | ||
1036 | } | ||
1037 | else if (strcmp(str, "rational") == 0) | ||
1038 | { | ||
1039 | char name[sNAMEMAX + 1]; | ||
1040 | cell digits = 0; | ||
1041 | int i; | ||
1042 | |||
1043 | /* first gather all information, start with the tag name */ | ||
1044 | while ((*lptr <= ' ') && (*lptr != '\0')) | ||
1045 | lptr++; | ||
1046 | for (i = 0; | ||
1047 | (i < (int)(sizeof(name))) && | ||
1048 | (alphanum(*lptr)); | ||
1049 | i++, lptr++) | ||
1050 | name[i] = *lptr; | ||
1051 | name[i] = '\0'; | ||
1052 | /* then the precision (for fixed point arithmetic) */ | ||
1053 | while (*lptr <= ' ' && *lptr != '\0') | ||
1054 | lptr++; | ||
1055 | if (*lptr == '(') | ||
1056 | { | ||
1057 | preproc_expr(&digits, NULL); | ||
1058 | if (digits <= 0 || digits > 9) | ||
1059 | { | ||
1060 | error(68); /* invalid rational number precision */ | ||
1061 | digits = 0; | ||
1062 | } /* if */ | ||
1063 | if (*lptr == ')') | ||
1064 | lptr++; | ||
1065 | } /* if */ | ||
1066 | /* add the tag (make it public) and check the values */ | ||
1067 | i = sc_addtag(name); | ||
1068 | exporttag(i); | ||
1069 | if (sc_rationaltag == 0 | ||
1070 | || (sc_rationaltag == i | ||
1071 | && rational_digits == (int)digits)) | ||
1072 | { | ||
1073 | sc_rationaltag = i; | ||
1074 | rational_digits = (int)digits; | ||
1075 | } | ||
1076 | else | ||
1077 | { | ||
1078 | error(69); /* rational number format already set, can only be set once */ | ||
1079 | } /* if */ | ||
1080 | } | ||
1081 | else if (strcmp(str, "semicolon") == 0) | ||
1082 | { | ||
1083 | cell val; | ||
1084 | |||
1085 | preproc_expr(&val, NULL); | ||
1086 | sc_needsemicolon = (int)val; | ||
1087 | } | ||
1088 | else if (strcmp(str, "tabsize") == 0) | ||
1089 | { | ||
1090 | cell val; | ||
1091 | |||
1092 | preproc_expr(&val, NULL); | ||
1093 | sc_tabsize = (int)val; | ||
1094 | } | ||
1095 | else if (strcmp(str, "align") == 0) | ||
1096 | { | ||
1097 | sc_alignnext = TRUE; | ||
1098 | } | ||
1099 | else if (strcmp(str, "unused") == 0) | ||
1100 | { | ||
1101 | char name[sNAMEMAX + 1]; | ||
1102 | int i, comma; | ||
1103 | symbol *sym; | ||
1104 | |||
1105 | do | ||
1106 | { | ||
1107 | /* get the name */ | ||
1108 | while ((*lptr <= ' ') && (*lptr != '\0')) | ||
1109 | lptr++; | ||
1110 | for (i = 0; | ||
1111 | (i < (int)(sizeof(name))) && | ||
1112 | (isalpha(*lptr)); | ||
1113 | i++, lptr++) | ||
1114 | name[i] = *lptr; | ||
1115 | name[i] = '\0'; | ||
1116 | /* get the symbol */ | ||
1117 | sym = findloc(name); | ||
1118 | if (!sym) | ||
1119 | sym = findglb(name); | ||
1120 | if (sym) | ||
1121 | { | ||
1122 | sym->usage |= uREAD; | ||
1123 | if (sym->ident == iVARIABLE | ||
1124 | || sym->ident == iREFERENCE | ||
1125 | || sym->ident == iARRAY | ||
1126 | || sym->ident == iREFARRAY) | ||
1127 | sym->usage |= uWRITTEN; | ||
1128 | } | ||
1129 | else | ||
1130 | { | ||
1131 | error(17, name); /* undefined symbol */ | ||
1132 | } /* if */ | ||
1133 | /* see if a comma follows the name */ | ||
1134 | while (*lptr <= ' ' && *lptr != '\0') | ||
1135 | lptr++; | ||
1136 | comma = (*lptr == ','); | ||
1137 | if (comma) | ||
1138 | lptr++; | ||
1139 | } | ||
1140 | while (comma); | ||
1141 | } | ||
1142 | else | ||
1143 | { | ||
1144 | error(207); /* unknown #pragma */ | ||
1145 | } /* if */ | ||
1146 | } | ||
1147 | else | ||
1148 | { | ||
1149 | error(207); /* unknown #pragma */ | ||
1150 | } /* if */ | ||
1151 | check_empty(lptr); | ||
1152 | } /* if */ | ||
1153 | break; | ||
1154 | case tpENDINPUT: | ||
1155 | case tpENDSCRPT: | ||
1156 | if (skiplevel == 0) | ||
1157 | { | ||
1158 | check_empty(lptr); | ||
1159 | assert(inpf != NULL); | ||
1160 | if (inpf != inpf_org) | ||
1161 | sc_closesrc(inpf); | ||
1162 | inpf = NULL; | ||
1163 | } /* if */ | ||
1164 | break; | ||
1165 | #if !defined NOEMIT | ||
1166 | case tpEMIT: | ||
1167 | { | ||
1168 | /* write opcode to output file */ | ||
1169 | char name[40]; | ||
1170 | int i; | ||
1171 | |||
1172 | while (*lptr <= ' ' && *lptr != '\0') | ||
1173 | lptr++; | ||
1174 | for (i = 0; i < 40 && (isalpha(*lptr) || *lptr == '.'); i++, lptr++) | ||
1175 | name[i] = (char)tolower(*lptr); | ||
1176 | name[i] = '\0'; | ||
1177 | stgwrite("\t"); | ||
1178 | stgwrite(name); | ||
1179 | stgwrite(" "); | ||
1180 | code_idx += opcodes(1); | ||
1181 | /* write parameter (if any) */ | ||
1182 | while (*lptr <= ' ' && *lptr != '\0') | ||
1183 | lptr++; | ||
1184 | if (*lptr != '\0') | ||
1185 | { | ||
1186 | symbol *sym; | ||
1187 | |||
1188 | tok = lex(&val, &str); | ||
1189 | switch (tok) | ||
1190 | { | ||
1191 | case tNUMBER: | ||
1192 | case tRATIONAL: | ||
1193 | outval(val, FALSE); | ||
1194 | code_idx += opargs(1); | ||
1195 | break; | ||
1196 | case tSYMBOL: | ||
1197 | sym = findloc(str); | ||
1198 | if (!sym) | ||
1199 | sym = findglb(str); | ||
1200 | if (!sym || (sym->ident != iFUNCTN | ||
1201 | && sym->ident != iREFFUNC | ||
1202 | && (sym->usage & uDEFINE) == 0)) | ||
1203 | { | ||
1204 | error(17, str); /* undefined symbol */ | ||
1205 | } | ||
1206 | else | ||
1207 | { | ||
1208 | outval(sym->addr, FALSE); | ||
1209 | /* mark symbol as "used", unknown whether for read or write */ | ||
1210 | markusage(sym, uREAD | uWRITTEN); | ||
1211 | code_idx += opargs(1); | ||
1212 | } /* if */ | ||
1213 | break; | ||
1214 | default: | ||
1215 | { | ||
1216 | char s2[20]; | ||
1217 | extern char *sc_tokens[]; /* forward declaration */ | ||
1218 | |||
1219 | if (tok < 256) | ||
1220 | sprintf(s2, "%c", (char)tok); | ||
1221 | else | ||
1222 | strcpy(s2, sc_tokens[tok - tFIRST]); | ||
1223 | error(1, sc_tokens[tSYMBOL - tFIRST], s2); | ||
1224 | break; | ||
1225 | } /* case */ | ||
1226 | } /* switch */ | ||
1227 | } /* if */ | ||
1228 | stgwrite("\n"); | ||
1229 | check_empty(lptr); | ||
1230 | break; | ||
1231 | } /* case */ | ||
1232 | #endif | ||
1233 | #if !defined NO_DEFINE | ||
1234 | case tpDEFINE: | ||
1235 | { | ||
1236 | ret = CMD_DEFINE; | ||
1237 | if (skiplevel == 0) | ||
1238 | { | ||
1239 | char *pattern, *substitution; | ||
1240 | char *start, *end; | ||
1241 | int count, prefixlen; | ||
1242 | stringpair *def; | ||
1243 | |||
1244 | /* find the pattern to match */ | ||
1245 | while (*lptr <= ' ' && *lptr != '\0') | ||
1246 | lptr++; | ||
1247 | start = lptr; /* save starting point of the match pattern */ | ||
1248 | count = 0; | ||
1249 | while (*lptr > ' ' && *lptr != '\0') | ||
1250 | { | ||
1251 | litchar(&lptr, FALSE); /* litchar() advances "lptr" and handles escape characters */ | ||
1252 | count++; | ||
1253 | } /* while */ | ||
1254 | end = lptr; | ||
1255 | /* check pattern to match */ | ||
1256 | if (!isalpha(*start) && *start != '_') | ||
1257 | { | ||
1258 | error(74); /* pattern must start with an alphabetic character */ | ||
1259 | break; | ||
1260 | } /* if */ | ||
1261 | /* store matched pattern */ | ||
1262 | pattern = malloc(count + 1); | ||
1263 | if (!pattern) | ||
1264 | error(103); /* insufficient memory */ | ||
1265 | lptr = start; | ||
1266 | count = 0; | ||
1267 | while (lptr != end) | ||
1268 | { | ||
1269 | assert(lptr < end); | ||
1270 | assert(*lptr != '\0'); | ||
1271 | pattern[count++] = (char)litchar(&lptr, FALSE); | ||
1272 | } /* while */ | ||
1273 | pattern[count] = '\0'; | ||
1274 | /* special case, erase trailing variable, because it could match anything */ | ||
1275 | if (count >= 2 && isdigit(pattern[count - 1]) | ||
1276 | && pattern[count - 2] == '%') | ||
1277 | pattern[count - 2] = '\0'; | ||
1278 | /* find substitution string */ | ||
1279 | while (*lptr <= ' ' && *lptr != '\0') | ||
1280 | lptr++; | ||
1281 | start = lptr; /* save starting point of the match pattern */ | ||
1282 | count = 0; | ||
1283 | end = NULL; | ||
1284 | while (*lptr != '\0') | ||
1285 | { | ||
1286 | /* keep position of the start of trailing whitespace */ | ||
1287 | if (*lptr <= ' ') | ||
1288 | { | ||
1289 | if (!end) | ||
1290 | end = lptr; | ||
1291 | } | ||
1292 | else | ||
1293 | { | ||
1294 | end = NULL; | ||
1295 | } /* if */ | ||
1296 | count++; | ||
1297 | lptr++; | ||
1298 | } /* while */ | ||
1299 | if (!end) | ||
1300 | end = lptr; | ||
1301 | /* store matched substitution */ | ||
1302 | substitution = malloc(count + 1); /* +1 for '\0' */ | ||
1303 | if (!substitution) | ||
1304 | error(103); /* insufficient memory */ | ||
1305 | lptr = start; | ||
1306 | count = 0; | ||
1307 | while (lptr != end) | ||
1308 | { | ||
1309 | assert(lptr < end); | ||
1310 | assert(*lptr != '\0'); | ||
1311 | substitution[count++] = *lptr++; | ||
1312 | } /* while */ | ||
1313 | substitution[count] = '\0'; | ||
1314 | /* check whether the definition already exists */ | ||
1315 | for (prefixlen = 0, start = pattern; | ||
1316 | isalpha(*start) || isdigit(*start) || *start == '_'; | ||
1317 | prefixlen++, start++) | ||
1318 | /* nothing */ ; | ||
1319 | assert(prefixlen > 0); | ||
1320 | if ((def = find_subst(pattern, prefixlen))) | ||
1321 | { | ||
1322 | if (strcmp(def->first, pattern) != 0 | ||
1323 | || strcmp(def->second, substitution) != 0) | ||
1324 | error(201, pattern); /* redefinition of macro (non-identical) */ | ||
1325 | delete_subst(pattern, prefixlen); | ||
1326 | } /* if */ | ||
1327 | /* add the pattern/substitution pair to the list */ | ||
1328 | assert(pattern[0] != '\0'); | ||
1329 | insert_subst(pattern, substitution, prefixlen); | ||
1330 | free(pattern); | ||
1331 | free(substitution); | ||
1332 | } /* if */ | ||
1333 | break; | ||
1334 | } /* case */ | ||
1335 | case tpUNDEF: | ||
1336 | if (skiplevel == 0) | ||
1337 | { | ||
1338 | if (lex(&val, &str) == tSYMBOL) | ||
1339 | { | ||
1340 | if (!delete_subst(str, strlen(str))) | ||
1341 | error(17, str); /* undefined symbol */ | ||
1342 | } | ||
1343 | else | ||
1344 | { | ||
1345 | error(20, str); /* invalid symbol name */ | ||
1346 | } /* if */ | ||
1347 | check_empty(lptr); | ||
1348 | } /* if */ | ||
1349 | break; | ||
1350 | #endif | ||
1351 | default: | ||
1352 | error(31); /* unknown compiler directive */ | ||
1353 | ret = skiplevel > 0 ? CMD_DIRECTIVE : CMD_NONE; /* line must be processed (if skiplevel==0) */ | ||
1354 | } /* switch */ | ||
1355 | return ret; | ||
1356 | } | ||
1357 | |||
1358 | #if !defined NO_DEFINE | ||
1359 | static int | ||
1360 | is_startstring(char *string) | ||
1361 | { | ||
1362 | if (*string == '\"' || *string == '\'') | ||
1363 | return TRUE; /* "..." */ | ||
1364 | |||
1365 | if (*string == '!') | ||
1366 | { | ||
1367 | string++; | ||
1368 | if (*string == '\"' || *string == '\'') | ||
1369 | return TRUE; /* !"..." */ | ||
1370 | if (*string == sc_ctrlchar) | ||
1371 | { | ||
1372 | string++; | ||
1373 | if (*string == '\"' || *string == '\'') | ||
1374 | return TRUE; /* !\"..." */ | ||
1375 | } /* if */ | ||
1376 | } | ||
1377 | else if (*string == sc_ctrlchar) | ||
1378 | { | ||
1379 | string++; | ||
1380 | if (*string == '\"' || *string == '\'') | ||
1381 | return TRUE; /* \"..." */ | ||
1382 | if (*string == '!') | ||
1383 | { | ||
1384 | string++; | ||
1385 | if (*string == '\"' || *string == '\'') | ||
1386 | return TRUE; /* \!"..." */ | ||
1387 | } /* if */ | ||
1388 | } /* if */ | ||
1389 | |||
1390 | return FALSE; | ||
1391 | } | ||
1392 | |||
1393 | static char * | ||
1394 | skipstring(char *string) | ||
1395 | { | ||
1396 | char endquote; | ||
1397 | int rawstring = FALSE; | ||
1398 | |||
1399 | while (*string == '!' || *string == sc_ctrlchar) | ||
1400 | { | ||
1401 | rawstring = (*string == sc_ctrlchar); | ||
1402 | string++; | ||
1403 | } /* while */ | ||
1404 | |||
1405 | endquote = *string; | ||
1406 | assert(endquote == '\"' || endquote == '\''); | ||
1407 | string++; /* skip open quote */ | ||
1408 | while (*string != endquote && *string != '\0') | ||
1409 | litchar(&string, rawstring); | ||
1410 | return string; | ||
1411 | } | ||
1412 | |||
1413 | static char * | ||
1414 | skippgroup(char *string) | ||
1415 | { | ||
1416 | int nest = 0; | ||
1417 | char open = *string; | ||
1418 | char close; | ||
1419 | |||
1420 | switch (open) | ||
1421 | { | ||
1422 | case '(': | ||
1423 | close = ')'; | ||
1424 | break; | ||
1425 | case '{': | ||
1426 | close = '}'; | ||
1427 | break; | ||
1428 | case '[': | ||
1429 | close = ']'; | ||
1430 | break; | ||
1431 | case '<': | ||
1432 | close = '>'; | ||
1433 | break; | ||
1434 | default: | ||
1435 | assert(0); | ||
1436 | close = '\0'; /* only to avoid a compiler warning */ | ||
1437 | } /* switch */ | ||
1438 | |||
1439 | string++; | ||
1440 | while (*string != close || nest > 0) | ||
1441 | { | ||
1442 | if (*string == open) | ||
1443 | nest++; | ||
1444 | else if (*string == close) | ||
1445 | nest--; | ||
1446 | else if (is_startstring(string)) | ||
1447 | string = skipstring(string); | ||
1448 | if (*string == '\0') | ||
1449 | break; | ||
1450 | string++; | ||
1451 | } /* while */ | ||
1452 | return string; | ||
1453 | } | ||
1454 | |||
1455 | static char * | ||
1456 | strdel(char *str, size_t len) | ||
1457 | { | ||
1458 | size_t length = strlen(str); | ||
1459 | |||
1460 | if (len > length) | ||
1461 | len = length; | ||
1462 | memmove(str, str + len, length - len + 1); /* include EOS byte */ | ||
1463 | return str; | ||
1464 | } | ||
1465 | |||
1466 | static char * | ||
1467 | strins(char *dest, char *src, size_t srclen) | ||
1468 | { | ||
1469 | size_t destlen = strlen(dest); | ||
1470 | |||
1471 | assert(srclen <= strlen(src)); | ||
1472 | memmove(dest + srclen, dest, destlen + 1); /* include EOS byte */ | ||
1473 | memcpy(dest, src, srclen); | ||
1474 | return dest; | ||
1475 | } | ||
1476 | |||
1477 | static int | ||
1478 | substpattern(char *line, size_t buffersize, char *pattern, char *substitution) | ||
1479 | { | ||
1480 | int prefixlen; | ||
1481 | char *p, *s, *e, *args[10]; | ||
1482 | int match, arg, len; | ||
1483 | |||
1484 | memset(args, 0, sizeof args); | ||
1485 | |||
1486 | /* check the length of the prefix */ | ||
1487 | for (prefixlen = 0, s = pattern; isalpha(*s) || isdigit(*s) || *s == '_'; | ||
1488 | prefixlen++, s++) | ||
1489 | /* nothing */ ; | ||
1490 | assert(prefixlen > 0); | ||
1491 | assert(strncmp(line, pattern, prefixlen) == 0); | ||
1492 | |||
1493 | /* pattern prefix matches; match the rest of the pattern, gather | ||
1494 | * the parameters | ||
1495 | */ | ||
1496 | s = line + prefixlen; | ||
1497 | p = pattern + prefixlen; | ||
1498 | match = TRUE; /* so far, pattern matches */ | ||
1499 | while (match && *s != '\0' && *p != '\0') | ||
1500 | { | ||
1501 | if (*p == '%') | ||
1502 | { | ||
1503 | p++; /* skip '%' */ | ||
1504 | if (isdigit(*p)) | ||
1505 | { | ||
1506 | arg = *p - '0'; | ||
1507 | assert(arg >= 0 && arg <= 9); | ||
1508 | p++; /* skip parameter id */ | ||
1509 | assert(*p != '\0'); | ||
1510 | /* match the source string up to the character after the digit | ||
1511 | * (skipping strings in the process | ||
1512 | */ | ||
1513 | e = s; | ||
1514 | while (*e != *p && *e != '\0' && *e != '\n') | ||
1515 | { | ||
1516 | if (is_startstring(e)) /* skip strings */ | ||
1517 | e = skipstring(e); | ||
1518 | else if (strchr("({[", *e)) /* skip parenthized groups */ | ||
1519 | e = skippgroup(e); | ||
1520 | if (*e != '\0') | ||
1521 | e++; /* skip non-alphapetic character (or closing quote of | ||
1522 | * a string, or the closing paranthese of a group) */ | ||
1523 | } /* while */ | ||
1524 | /* store the parameter (overrule any earlier) */ | ||
1525 | if (args[arg]) | ||
1526 | free(args[arg]); | ||
1527 | len = (int)(e - s); | ||
1528 | args[arg] = malloc(len + 1); | ||
1529 | if (!args[arg]) | ||
1530 | error(103); /* insufficient memory */ | ||
1531 | strncpy(args[arg], s, len); | ||
1532 | args[arg][len] = '\0'; | ||
1533 | /* character behind the pattern was matched too */ | ||
1534 | if (*e == *p) | ||
1535 | { | ||
1536 | s = e + 1; | ||
1537 | } | ||
1538 | else if (*e == '\n' && *p == ';' && *(p + 1) == '\0' | ||
1539 | && !sc_needsemicolon) | ||
1540 | { | ||
1541 | s = e; /* allow a trailing ; in the pattern match to end of line */ | ||
1542 | } | ||
1543 | else | ||
1544 | { | ||
1545 | assert(*e == '\0' || *e == '\n'); | ||
1546 | match = FALSE; | ||
1547 | s = e; | ||
1548 | } /* if */ | ||
1549 | p++; | ||
1550 | } | ||
1551 | else | ||
1552 | { | ||
1553 | match = FALSE; | ||
1554 | } /* if */ | ||
1555 | } | ||
1556 | else if (*p == ';' && *(p + 1) == '\0' && !sc_needsemicolon) | ||
1557 | { | ||
1558 | /* source may be ';' or end of the line */ | ||
1559 | while (*s <= ' ' && *s != '\0') | ||
1560 | s++; /* skip white space */ | ||
1561 | if (*s != ';' && *s != '\0') | ||
1562 | match = FALSE; | ||
1563 | p++; /* skip the semicolon in the pattern */ | ||
1564 | } | ||
1565 | else | ||
1566 | { | ||
1567 | cell ch; | ||
1568 | |||
1569 | /* skip whitespace between two non-alphanumeric characters, except | ||
1570 | * for two identical symbols | ||
1571 | */ | ||
1572 | assert(p > pattern); | ||
1573 | if (!alphanum(*p) && *(p - 1) != *p) | ||
1574 | while (*s <= ' ' && *s != '\0') | ||
1575 | s++; /* skip white space */ | ||
1576 | ch = litchar(&p, FALSE); /* this increments "p" */ | ||
1577 | if (*s != ch) | ||
1578 | match = FALSE; | ||
1579 | else | ||
1580 | s++; /* this character matches */ | ||
1581 | } /* if */ | ||
1582 | } /* while */ | ||
1583 | |||
1584 | if (match && *p == '\0') | ||
1585 | { | ||
1586 | /* if the last character to match is an alphanumeric character, the | ||
1587 | * current character in the source may not be alphanumeric | ||
1588 | */ | ||
1589 | assert(p > pattern); | ||
1590 | if (alphanum(*(p - 1)) && alphanum(*s)) | ||
1591 | match = FALSE; | ||
1592 | } /* if */ | ||
1593 | |||
1594 | if (match) | ||
1595 | { | ||
1596 | /* calculate the length of the substituted string */ | ||
1597 | for (e = substitution, len = 0; *e != '\0'; e++) | ||
1598 | { | ||
1599 | if (*e == '%' && isdigit(*(e + 1))) | ||
1600 | { | ||
1601 | arg = *(e + 1) - '0'; | ||
1602 | assert(arg >= 0 && arg <= 9); | ||
1603 | if (args[arg]) | ||
1604 | len += strlen(args[arg]); | ||
1605 | e++; /* skip %, digit is skipped later */ | ||
1606 | } | ||
1607 | else | ||
1608 | { | ||
1609 | len++; | ||
1610 | } /* if */ | ||
1611 | } /* for */ | ||
1612 | /* check length of the string after substitution */ | ||
1613 | if (strlen(line) + len - (int)(s - line) > buffersize) | ||
1614 | { | ||
1615 | error(75); /* line too long */ | ||
1616 | } | ||
1617 | else | ||
1618 | { | ||
1619 | /* substitute pattern */ | ||
1620 | strdel(line, (int)(s - line)); | ||
1621 | for (e = substitution, s = line; *e != '\0'; e++) | ||
1622 | { | ||
1623 | if (*e == '%' && isdigit(*(e + 1))) | ||
1624 | { | ||
1625 | arg = *(e + 1) - '0'; | ||
1626 | assert(arg >= 0 && arg <= 9); | ||
1627 | if (args[arg]) | ||
1628 | { | ||
1629 | strins(s, args[arg], strlen(args[arg])); | ||
1630 | s += strlen(args[arg]); | ||
1631 | } /* if */ | ||
1632 | e++; /* skip %, digit is skipped later */ | ||
1633 | } | ||
1634 | else | ||
1635 | { | ||
1636 | strins(s, e, 1); | ||
1637 | s++; | ||
1638 | } /* if */ | ||
1639 | } /* for */ | ||
1640 | } /* if */ | ||
1641 | } /* if */ | ||
1642 | |||
1643 | for (arg = 0; arg < 10; arg++) | ||
1644 | if (args[arg]) | ||
1645 | free(args[arg]); | ||
1646 | |||
1647 | return match; | ||
1648 | } | ||
1649 | |||
1650 | static void | ||
1651 | substallpatterns(char *line, int buffersize) | ||
1652 | { | ||
1653 | char *start, *end; | ||
1654 | int prefixlen; | ||
1655 | stringpair *subst; | ||
1656 | |||
1657 | start = line; | ||
1658 | while (*start != '\0') | ||
1659 | { | ||
1660 | /* find the start of a prefix (skip all non-alphabetic characters), | ||
1661 | * also skip strings | ||
1662 | */ | ||
1663 | while (!isalpha(*start) && *start != '_' && *start != '\0') | ||
1664 | { | ||
1665 | /* skip strings */ | ||
1666 | if (is_startstring(start)) | ||
1667 | { | ||
1668 | start = skipstring(start); | ||
1669 | if (*start == '\0') | ||
1670 | break; /* abort loop on error */ | ||
1671 | } /* if */ | ||
1672 | start++; /* skip non-alphapetic character (or closing quote of a string) */ | ||
1673 | } /* while */ | ||
1674 | if (*start == '\0') | ||
1675 | break; /* abort loop on error */ | ||
1676 | /* get the prefix (length), look for a matching definition */ | ||
1677 | prefixlen = 0; | ||
1678 | end = start; | ||
1679 | while (isalpha(*end) || isdigit(*end) || *end == '_') | ||
1680 | { | ||
1681 | prefixlen++; | ||
1682 | end++; | ||
1683 | } /* while */ | ||
1684 | assert(prefixlen > 0); | ||
1685 | subst = find_subst(start, prefixlen); | ||
1686 | if (subst) | ||
1687 | { | ||
1688 | /* properly match the pattern and substitute */ | ||
1689 | if (!substpattern | ||
1690 | (start, buffersize - (start - line), subst->first, | ||
1691 | subst->second)) | ||
1692 | start = end; /* match failed, skip this prefix */ | ||
1693 | /* match succeeded: do not update "start", because the substitution text | ||
1694 | * may be matched by other macros | ||
1695 | */ | ||
1696 | } | ||
1697 | else | ||
1698 | { | ||
1699 | start = end; /* no macro with this prefix, skip this prefix */ | ||
1700 | } /* if */ | ||
1701 | } /* while */ | ||
1702 | } | ||
1703 | #endif | ||
1704 | |||
1705 | /* preprocess | ||
1706 | * | ||
1707 | * Reads a line by readline() into "pline" and performs basic preprocessing: | ||
1708 | * deleting comments, skipping lines with false "#if.." code and recognizing | ||
1709 | * other compiler directives. There is an indirect recursion: lex() calls | ||
1710 | * preprocess() if a new line must be read, preprocess() calls command(), | ||
1711 | * which at his turn calls lex() to identify the token. | ||
1712 | * | ||
1713 | * Global references: lptr (altered) | ||
1714 | * pline (altered) | ||
1715 | * freading (referred to only) | ||
1716 | */ | ||
1717 | void | ||
1718 | preprocess(void) | ||
1719 | { | ||
1720 | int iscommand; | ||
1721 | |||
1722 | if (!freading) | ||
1723 | return; | ||
1724 | do | ||
1725 | { | ||
1726 | readline(pline); | ||
1727 | stripcom(pline); /* ??? no need for this when reading back from list file (in the second pass) */ | ||
1728 | lptr = pline; /* set "line pointer" to start of the parsing buffer */ | ||
1729 | iscommand = command(); | ||
1730 | if (iscommand != CMD_NONE) | ||
1731 | errorset(sRESET); /* reset error flag ("panic mode") on empty line or directive */ | ||
1732 | #if !defined NO_DEFINE | ||
1733 | if (iscommand == CMD_NONE) | ||
1734 | { | ||
1735 | assert(lptr != term_expr); | ||
1736 | substallpatterns(pline, sLINEMAX); | ||
1737 | lptr = pline; /* reset "line pointer" to start of the parsing buffer */ | ||
1738 | } /* if */ | ||
1739 | #endif | ||
1740 | } | ||
1741 | while (iscommand != CMD_NONE && iscommand != CMD_TERM && freading); /* enddo */ | ||
1742 | } | ||
1743 | |||
1744 | static char * | ||
1745 | unpackedstring(char *lptr, int rawstring) | ||
1746 | { | ||
1747 | while (*lptr != '\0') | ||
1748 | { | ||
1749 | /* check for doublequotes indicating the end of the string */ | ||
1750 | if (*lptr == '\"') | ||
1751 | { | ||
1752 | /* check whether there's another pair of quotes following. | ||
1753 | * If so, paste the two strings together, thus | ||
1754 | * "pants""off" becomes "pantsoff" | ||
1755 | */ | ||
1756 | if (*(lptr + 1) == '\"') | ||
1757 | lptr += 2; | ||
1758 | else | ||
1759 | break; | ||
1760 | } | ||
1761 | |||
1762 | if (*lptr == '\a') | ||
1763 | { /* ignore '\a' (which was inserted at a line concatenation) */ | ||
1764 | lptr++; | ||
1765 | continue; | ||
1766 | } /* if */ | ||
1767 | stowlit(litchar(&lptr, rawstring)); /* litchar() alters "lptr" */ | ||
1768 | } /* while */ | ||
1769 | stowlit(0); /* terminate string */ | ||
1770 | return lptr; | ||
1771 | } | ||
1772 | |||
1773 | static char * | ||
1774 | packedstring(char *lptr, int rawstring) | ||
1775 | { | ||
1776 | int i; | ||
1777 | ucell val, c; | ||
1778 | |||
1779 | i = sizeof(ucell) - (charbits / 8); /* start at most significant byte */ | ||
1780 | val = 0; | ||
1781 | while (*lptr != '\0') | ||
1782 | { | ||
1783 | /* check for doublequotes indicating the end of the string */ | ||
1784 | if (*lptr == '\"') | ||
1785 | { | ||
1786 | /* check whether there's another pair of quotes following. | ||
1787 | * If so, paste the two strings together, thus | ||
1788 | * "pants""off" becomes "pantsoff" | ||
1789 | */ | ||
1790 | if (*(lptr + 1) == '\"') | ||
1791 | lptr += 2; | ||
1792 | else | ||
1793 | break; | ||
1794 | } | ||
1795 | |||
1796 | if (*lptr == '\a') | ||
1797 | { /* ignore '\a' (which was inserted at a line concatenation) */ | ||
1798 | lptr++; | ||
1799 | continue; | ||
1800 | } /* if */ | ||
1801 | c = litchar(&lptr, rawstring); /* litchar() alters "lptr" */ | ||
1802 | if (c >= (ucell) (1 << charbits)) | ||
1803 | error(43); /* character constant exceeds range */ | ||
1804 | val |= (c << 8 * i); | ||
1805 | if (i == 0) | ||
1806 | { | ||
1807 | stowlit(val); | ||
1808 | val = 0; | ||
1809 | } /* if */ | ||
1810 | i = (i + sizeof(ucell) - (charbits / 8)) % sizeof(ucell); | ||
1811 | } /* if */ | ||
1812 | /* save last code; make sure there is at least one terminating zero character */ | ||
1813 | if (i != (int)(sizeof(ucell) - (charbits / 8))) | ||
1814 | stowlit(val); /* at least one zero character in "val" */ | ||
1815 | else | ||
1816 | stowlit(0); /* add full cell of zeros */ | ||
1817 | return lptr; | ||
1818 | } | ||
1819 | |||
1820 | /* lex(lexvalue,lexsym) Lexical Analysis | ||
1821 | * | ||
1822 | * lex() first deletes leading white space, then checks for multi-character | ||
1823 | * operators, keywords (including most compiler directives), numbers, | ||
1824 | * labels, symbols and literals (literal characters are converted to a number | ||
1825 | * and are returned as such). If every check fails, the line must contain | ||
1826 | * a single-character operator. So, lex() returns this character. In the other | ||
1827 | * case (something did match), lex() returns the number of the token. All | ||
1828 | * these tokens have been assigned numbers above 255. | ||
1829 | * | ||
1830 | * Some tokens have "attributes": | ||
1831 | * tNUMBER the value of the number is return in "lexvalue". | ||
1832 | * tRATIONAL the value is in IEEE 754 encoding or in fixed point | ||
1833 | * encoding in "lexvalue". | ||
1834 | * tSYMBOL the first sNAMEMAX characters of the symbol are | ||
1835 | * stored in a buffer, a pointer to this buffer is | ||
1836 | * returned in "lexsym". | ||
1837 | * tLABEL the first sNAMEMAX characters of the label are | ||
1838 | * stored in a buffer, a pointer to this buffer is | ||
1839 | * returned in "lexsym". | ||
1840 | * tSTRING the string is stored in the literal pool, the index | ||
1841 | * in the literal pool to this string is stored in | ||
1842 | * "lexvalue". | ||
1843 | * | ||
1844 | * lex() stores all information (the token found and possibly its attribute) | ||
1845 | * in global variables. This allows a token to be examined twice. If "_pushed" | ||
1846 | * is true, this information is returned. | ||
1847 | * | ||
1848 | * Global references: lptr (altered) | ||
1849 | * fline (referred to only) | ||
1850 | * litidx (referred to only) | ||
1851 | * _lextok, _lexval, _lexstr | ||
1852 | * _pushed | ||
1853 | */ | ||
1854 | |||
1855 | static int _pushed; | ||
1856 | static int _lextok; | ||
1857 | static cell _lexval; | ||
1858 | static char _lexstr[sLINEMAX + 1]; | ||
1859 | static int _lexnewline; | ||
1860 | |||
1861 | void | ||
1862 | lexinit(void) | ||
1863 | { | ||
1864 | stkidx = 0; /* index for pushstk() and popstk() */ | ||
1865 | iflevel = 0; /* preprocessor: nesting of "#if" */ | ||
1866 | skiplevel = 0; /* preprocessor: skipping lines or compiling lines */ | ||
1867 | icomment = FALSE; /* currently not in a multiline comment */ | ||
1868 | _pushed = FALSE; /* no token pushed back into lex */ | ||
1869 | _lexnewline = FALSE; | ||
1870 | } | ||
1871 | |||
1872 | char *sc_tokens[] = { | ||
1873 | "*=", "/=", "%=", "+=", "-=", "<<=", ">>>=", ">>=", "&=", "^=", "|=", | ||
1874 | "||", "&&", "==", "!=", "<=", ">=", "<<", ">>>", ">>", "++", "--", | ||
1875 | "...", "..", | ||
1876 | "assert", "break", "case", "char", "const", "continue", "default", | ||
1877 | "defined", "do", "else", "enum", "exit", "for", "forward", "goto", | ||
1878 | "if", "native", "new", "operator", "public", "return", "sizeof", | ||
1879 | "sleep", "static", "stock", "switch", "tagof", "while", | ||
1880 | "#assert", "#define", "#else", "#emit", "#endif", "#endinput", | ||
1881 | "#endscript", "#file", "#if", "#include", "#line", "#pragma", "#undef", | ||
1882 | ";", ";", "-integer value-", "-rational value-", "-identifier-", | ||
1883 | "-label-", "-string-" | ||
1884 | }; | ||
1885 | |||
1886 | int | ||
1887 | lex(cell * lexvalue, char **lexsym) | ||
1888 | { | ||
1889 | int i, toolong, newline, rawstring; | ||
1890 | char **tokptr; | ||
1891 | |||
1892 | if (_pushed) | ||
1893 | { | ||
1894 | _pushed = FALSE; /* reset "_pushed" flag */ | ||
1895 | *lexvalue = _lexval; | ||
1896 | *lexsym = _lexstr; | ||
1897 | return _lextok; | ||
1898 | } /* if */ | ||
1899 | |||
1900 | _lextok = 0; /* preset all values */ | ||
1901 | _lexval = 0; | ||
1902 | _lexstr[0] = '\0'; | ||
1903 | *lexvalue = _lexval; | ||
1904 | *lexsym = _lexstr; | ||
1905 | _lexnewline = FALSE; | ||
1906 | if (!freading) | ||
1907 | return 0; | ||
1908 | |||
1909 | newline = (lptr == pline); /* does lptr point to start of line buffer */ | ||
1910 | while (*lptr <= ' ') | ||
1911 | { /* delete leading white space */ | ||
1912 | if (*lptr == '\0') | ||
1913 | { | ||
1914 | preprocess(); /* preprocess resets "lptr" */ | ||
1915 | if (!freading) | ||
1916 | return 0; | ||
1917 | if (lptr == term_expr) /* special sequence to terminate a pending expression */ | ||
1918 | return (_lextok = tENDEXPR); | ||
1919 | _lexnewline = TRUE; /* set this after preprocess(), because | ||
1920 | * preprocess() calls lex() recursively */ | ||
1921 | newline = TRUE; | ||
1922 | } | ||
1923 | else | ||
1924 | { | ||
1925 | lptr += 1; | ||
1926 | } /* if */ | ||
1927 | } /* while */ | ||
1928 | if (newline) | ||
1929 | { | ||
1930 | stmtindent = 0; | ||
1931 | for (i = 0; i < (int)(lptr - pline); i++) | ||
1932 | if (pline[i] == '\t' && sc_tabsize > 0) | ||
1933 | stmtindent += | ||
1934 | (int)(sc_tabsize - (stmtindent + sc_tabsize) % sc_tabsize); | ||
1935 | else | ||
1936 | stmtindent++; | ||
1937 | } /* if */ | ||
1938 | |||
1939 | i = tFIRST; | ||
1940 | tokptr = sc_tokens; | ||
1941 | while (i <= tMIDDLE) | ||
1942 | { /* match multi-character operators */ | ||
1943 | if (match(*tokptr, FALSE)) | ||
1944 | { | ||
1945 | _lextok = i; | ||
1946 | return _lextok; | ||
1947 | } /* if */ | ||
1948 | i += 1; | ||
1949 | tokptr += 1; | ||
1950 | } /* while */ | ||
1951 | while (i <= tLAST) | ||
1952 | { /* match reserved words and compiler directives */ | ||
1953 | if (match(*tokptr, TRUE)) | ||
1954 | { | ||
1955 | _lextok = i; | ||
1956 | errorset(sRESET); /* reset error flag (clear the "panic mode") */ | ||
1957 | return _lextok; | ||
1958 | } /* if */ | ||
1959 | i += 1; | ||
1960 | tokptr += 1; | ||
1961 | } /* while */ | ||
1962 | |||
1963 | if ((i = number(&_lexval, lptr)) != 0) | ||
1964 | { /* number */ | ||
1965 | _lextok = tNUMBER; | ||
1966 | *lexvalue = _lexval; | ||
1967 | lptr += i; | ||
1968 | } | ||
1969 | else if ((i = ftoi(&_lexval, lptr)) != 0) | ||
1970 | { | ||
1971 | _lextok = tRATIONAL; | ||
1972 | *lexvalue = _lexval; | ||
1973 | lptr += i; | ||
1974 | } | ||
1975 | else if (alpha(*lptr)) | ||
1976 | { /* symbol or label */ | ||
1977 | /* Note: only sNAMEMAX characters are significant. The compiler | ||
1978 | * generates a warning if a symbol exceeds this length. | ||
1979 | */ | ||
1980 | _lextok = tSYMBOL; | ||
1981 | i = 0; | ||
1982 | toolong = 0; | ||
1983 | while (alphanum(*lptr)) | ||
1984 | { | ||
1985 | _lexstr[i] = *lptr; | ||
1986 | lptr += 1; | ||
1987 | if (i < sNAMEMAX) | ||
1988 | i += 1; | ||
1989 | else | ||
1990 | toolong = 1; | ||
1991 | } /* while */ | ||
1992 | _lexstr[i] = '\0'; | ||
1993 | if (toolong) | ||
1994 | error(200, _lexstr, sNAMEMAX); /* symbol too long, truncated to sNAMEMAX chars */ | ||
1995 | if (_lexstr[0] == PUBLIC_CHAR && _lexstr[1] == '\0') | ||
1996 | { | ||
1997 | _lextok = PUBLIC_CHAR; /* '@' all alone is not a symbol, it is an operator */ | ||
1998 | } | ||
1999 | else if (_lexstr[0] == '_' && _lexstr[1] == '\0') | ||
2000 | { | ||
2001 | _lextok = '_'; /* '_' by itself is not a symbol, it is a placeholder */ | ||
2002 | } /* if */ | ||
2003 | if (*lptr == ':' && sc_allowtags && _lextok != PUBLIC_CHAR) | ||
2004 | { | ||
2005 | _lextok = tLABEL; /* it wasn't a normal symbol, it was a label/tagname */ | ||
2006 | lptr += 1; /* skip colon */ | ||
2007 | } /* if */ | ||
2008 | } | ||
2009 | else if (*lptr == '\"' || (*lptr == sc_ctrlchar && *(lptr + 1) == '\"')) | ||
2010 | { /* unpacked string literal */ | ||
2011 | _lextok = tSTRING; | ||
2012 | rawstring = (*lptr == sc_ctrlchar); | ||
2013 | *lexvalue = _lexval = litidx; | ||
2014 | lptr += 1; /* skip double quote */ | ||
2015 | if (rawstring) | ||
2016 | lptr += 1; /* skip "escape" character too */ | ||
2017 | lptr = | ||
2018 | sc_packstr ? packedstring(lptr, rawstring) : unpackedstring(lptr, | ||
2019 | rawstring); | ||
2020 | if (*lptr == '\"') | ||
2021 | lptr += 1; /* skip final quote */ | ||
2022 | else | ||
2023 | error(37); /* invalid (non-terminated) string */ | ||
2024 | } | ||
2025 | else if ((*lptr == '!' && *(lptr + 1) == '\"') | ||
2026 | || (*lptr == '!' && *(lptr + 1) == sc_ctrlchar && *(lptr + 2) == '\"') | ||
2027 | || (*lptr == sc_ctrlchar && *(lptr + 1) == '!' | ||
2028 | && *(lptr + 2) == '\"')) | ||
2029 | { /* packed string literal */ | ||
2030 | _lextok = tSTRING; | ||
2031 | rawstring = (*lptr == sc_ctrlchar || *(lptr + 1) == sc_ctrlchar); | ||
2032 | *lexvalue = _lexval = litidx; | ||
2033 | lptr += 2; /* skip exclamation point and double quote */ | ||
2034 | if (rawstring) | ||
2035 | lptr += 1; /* skip "escape" character too */ | ||
2036 | lptr = | ||
2037 | sc_packstr ? unpackedstring(lptr, rawstring) : packedstring(lptr, | ||
2038 | rawstring); | ||
2039 | if (*lptr == '\"') | ||
2040 | lptr += 1; /* skip final quote */ | ||
2041 | else | ||
2042 | error(37); /* invalid (non-terminated) string */ | ||
2043 | } | ||
2044 | else if (*lptr == '\'') | ||
2045 | { /* character literal */ | ||
2046 | lptr += 1; /* skip quote */ | ||
2047 | _lextok = tNUMBER; | ||
2048 | *lexvalue = _lexval = litchar(&lptr, FALSE); | ||
2049 | if (*lptr == '\'') | ||
2050 | lptr += 1; /* skip final quote */ | ||
2051 | else | ||
2052 | error(27); /* invalid character constant (must be one character) */ | ||
2053 | } | ||
2054 | else if (*lptr == ';') | ||
2055 | { /* semicolumn resets "error" flag */ | ||
2056 | _lextok = ';'; | ||
2057 | lptr += 1; | ||
2058 | errorset(sRESET); /* reset error flag (clear the "panic mode") */ | ||
2059 | } | ||
2060 | else | ||
2061 | { | ||
2062 | _lextok = *lptr; /* if every match fails, return the character */ | ||
2063 | lptr += 1; /* increase the "lptr" pointer */ | ||
2064 | } /* if */ | ||
2065 | return _lextok; | ||
2066 | } | ||
2067 | |||
2068 | /* lexpush | ||
2069 | * | ||
2070 | * Pushes a token back, so the next call to lex() will return the token | ||
2071 | * last examined, instead of a new token. | ||
2072 | * | ||
2073 | * Only one token can be pushed back. | ||
2074 | * | ||
2075 | * In fact, lex() already stores the information it finds into global | ||
2076 | * variables, so all that is to be done is set a flag that informs lex() | ||
2077 | * to read and return the information from these variables, rather than | ||
2078 | * to read in a new token from the input file. | ||
2079 | */ | ||
2080 | void | ||
2081 | lexpush(void) | ||
2082 | { | ||
2083 | assert(_pushed == FALSE); | ||
2084 | _pushed = TRUE; | ||
2085 | } | ||
2086 | |||
2087 | /* lexclr | ||
2088 | * | ||
2089 | * Sets the variable "_pushed" to 0 to make sure lex() will read in a new | ||
2090 | * symbol (a not continue with some old one). This is required upon return | ||
2091 | * from Assembler mode. | ||
2092 | */ | ||
2093 | void | ||
2094 | lexclr(int clreol) | ||
2095 | { | ||
2096 | _pushed = FALSE; | ||
2097 | if (clreol) | ||
2098 | { | ||
2099 | lptr = strchr(pline, '\0'); | ||
2100 | assert(lptr != NULL); | ||
2101 | } /* if */ | ||
2102 | } | ||
2103 | |||
2104 | /* matchtoken | ||
2105 | * | ||
2106 | * This routine is useful if only a simple check is needed. If the token | ||
2107 | * differs from the one expected, it is pushed back. | ||
2108 | */ | ||
2109 | int | ||
2110 | matchtoken(int token) | ||
2111 | { | ||
2112 | cell val; | ||
2113 | char *str; | ||
2114 | int tok; | ||
2115 | |||
2116 | tok = lex(&val, &str); | ||
2117 | if (tok == token || (token == tTERM && (tok == ';' || tok == tENDEXPR))) | ||
2118 | { | ||
2119 | return 1; | ||
2120 | } | ||
2121 | else if (!sc_needsemicolon && token == tTERM && (_lexnewline || !freading)) | ||
2122 | { | ||
2123 | lexpush(); /* push "tok" back, we use the "hidden" newline token */ | ||
2124 | return 1; | ||
2125 | } | ||
2126 | else | ||
2127 | { | ||
2128 | lexpush(); | ||
2129 | return 0; | ||
2130 | } /* if */ | ||
2131 | } | ||
2132 | |||
2133 | /* tokeninfo | ||
2134 | * | ||
2135 | * Returns additional information of a token after using "matchtoken()" | ||
2136 | * or needtoken(). It does no harm using this routine after a call to | ||
2137 | * "lex()", but lex() already returns the same information. | ||
2138 | * | ||
2139 | * The token itself is the return value. Normally, this one is already known. | ||
2140 | */ | ||
2141 | int | ||
2142 | tokeninfo(cell * val, char **str) | ||
2143 | { | ||
2144 | /* if the token was pushed back, tokeninfo() returns the token and | ||
2145 | * parameters of the *next* token, not of the *current* token. | ||
2146 | */ | ||
2147 | assert(!_pushed); | ||
2148 | *val = _lexval; | ||
2149 | *str = _lexstr; | ||
2150 | return _lextok; | ||
2151 | } | ||
2152 | |||
2153 | /* needtoken | ||
2154 | * | ||
2155 | * This routine checks for a required token and gives an error message if | ||
2156 | * it isn't there (and returns FALSE in that case). | ||
2157 | * | ||
2158 | * Global references: _lextok; | ||
2159 | */ | ||
2160 | int | ||
2161 | needtoken(int token) | ||
2162 | { | ||
2163 | char s1[20], s2[20]; | ||
2164 | |||
2165 | if (matchtoken(token)) | ||
2166 | { | ||
2167 | return TRUE; | ||
2168 | } | ||
2169 | else | ||
2170 | { | ||
2171 | /* token already pushed back */ | ||
2172 | assert(_pushed); | ||
2173 | if (token < 256) | ||
2174 | sprintf(s1, "%c", (char)token); /* single character token */ | ||
2175 | else | ||
2176 | strcpy(s1, sc_tokens[token - tFIRST]); /* multi-character symbol */ | ||
2177 | if (!freading) | ||
2178 | strcpy(s2, "-end of file-"); | ||
2179 | else if (_lextok < 256) | ||
2180 | sprintf(s2, "%c", (char)_lextok); | ||
2181 | else | ||
2182 | strcpy(s2, sc_tokens[_lextok - tFIRST]); | ||
2183 | error(1, s1, s2); /* expected ..., but found ... */ | ||
2184 | return FALSE; | ||
2185 | } /* if */ | ||
2186 | } | ||
2187 | |||
2188 | /* match | ||
2189 | * | ||
2190 | * Compares a series of characters from the input file with the characters | ||
2191 | * in "st" (that contains a token). If the token on the input file matches | ||
2192 | * "st", the input file pointer "lptr" is adjusted to point to the next | ||
2193 | * token, otherwise "lptr" remains unaltered. | ||
2194 | * | ||
2195 | * If the parameter "end: is true, match() requires that the first character | ||
2196 | * behind the recognized token is non-alphanumeric. | ||
2197 | * | ||
2198 | * Global references: lptr (altered) | ||
2199 | */ | ||
2200 | static int | ||
2201 | match(char *st, int end) | ||
2202 | { | ||
2203 | int k; | ||
2204 | char *ptr; | ||
2205 | |||
2206 | k = 0; | ||
2207 | ptr = lptr; | ||
2208 | while (st[k]) | ||
2209 | { | ||
2210 | if (st[k] != *ptr) | ||
2211 | return 0; | ||
2212 | k += 1; | ||
2213 | ptr += 1; | ||
2214 | } /* while */ | ||
2215 | if (end) | ||
2216 | { /* symbol must terminate with non-alphanumeric char */ | ||
2217 | if (alphanum(*ptr)) | ||
2218 | return 0; | ||
2219 | } /* if */ | ||
2220 | lptr = ptr; /* match found, skip symbol */ | ||
2221 | return 1; | ||
2222 | } | ||
2223 | |||
2224 | /* stowlit | ||
2225 | * | ||
2226 | * Stores a value into the literal queue. The literal queue is used for | ||
2227 | * literal strings used in functions and for initializing array variables. | ||
2228 | * | ||
2229 | * Global references: litidx (altered) | ||
2230 | * litq (altered) | ||
2231 | */ | ||
2232 | void | ||
2233 | stowlit(cell value) | ||
2234 | { | ||
2235 | if (litidx >= litmax) | ||
2236 | { | ||
2237 | cell *p; | ||
2238 | |||
2239 | litmax += sDEF_LITMAX; | ||
2240 | p = (cell *) realloc(litq, litmax * sizeof(cell)); | ||
2241 | if (!p) | ||
2242 | error(102, "literal table"); /* literal table overflow (fatal error) */ | ||
2243 | litq = p; | ||
2244 | } /* if */ | ||
2245 | assert(litidx < litmax); | ||
2246 | litq[litidx++] = value; | ||
2247 | } | ||
2248 | |||
2249 | /* litchar | ||
2250 | * | ||
2251 | * Return current literal character and increase the pointer to point | ||
2252 | * just behind this literal character. | ||
2253 | * | ||
2254 | * Note: standard "escape sequences" are suported, but the backslash may be | ||
2255 | * replaced by another character; the syntax '\ddd' is supported, | ||
2256 | * but ddd must be decimal! | ||
2257 | */ | ||
2258 | static cell | ||
2259 | litchar(char **lptr, int rawmode) | ||
2260 | { | ||
2261 | cell c = 0; | ||
2262 | unsigned char *cptr; | ||
2263 | |||
2264 | cptr = (unsigned char *)*lptr; | ||
2265 | if (rawmode || *cptr != sc_ctrlchar) | ||
2266 | { /* no escape character */ | ||
2267 | c = *cptr; | ||
2268 | cptr += 1; | ||
2269 | } | ||
2270 | else | ||
2271 | { | ||
2272 | cptr += 1; | ||
2273 | if (*cptr == sc_ctrlchar) | ||
2274 | { | ||
2275 | c = *cptr; /* \\ == \ (the escape character itself) */ | ||
2276 | cptr += 1; | ||
2277 | } | ||
2278 | else | ||
2279 | { | ||
2280 | switch (*cptr) | ||
2281 | { | ||
2282 | case 'a': /* \a == audible alarm */ | ||
2283 | c = 7; | ||
2284 | cptr += 1; | ||
2285 | break; | ||
2286 | case 'b': /* \b == backspace */ | ||
2287 | c = 8; | ||
2288 | cptr += 1; | ||
2289 | break; | ||
2290 | case 'e': /* \e == escape */ | ||
2291 | c = 27; | ||
2292 | cptr += 1; | ||
2293 | break; | ||
2294 | case 'f': /* \f == form feed */ | ||
2295 | c = 12; | ||
2296 | cptr += 1; | ||
2297 | break; | ||
2298 | case 'n': /* \n == NewLine character */ | ||
2299 | c = 10; | ||
2300 | cptr += 1; | ||
2301 | break; | ||
2302 | case 'r': /* \r == carriage return */ | ||
2303 | c = 13; | ||
2304 | cptr += 1; | ||
2305 | break; | ||
2306 | case 't': /* \t == horizontal TAB */ | ||
2307 | c = 9; | ||
2308 | cptr += 1; | ||
2309 | break; | ||
2310 | case 'v': /* \v == vertical TAB */ | ||
2311 | c = 11; | ||
2312 | cptr += 1; | ||
2313 | break; | ||
2314 | case '\'': /* \' == ' (single quote) */ | ||
2315 | case '"': /* \" == " (single quote) */ | ||
2316 | case '%': /* \% == % (percent) */ | ||
2317 | c = *cptr; | ||
2318 | cptr += 1; | ||
2319 | break; | ||
2320 | default: | ||
2321 | if (isdigit(*cptr)) | ||
2322 | { /* \ddd */ | ||
2323 | c = 0; | ||
2324 | while (*cptr >= '0' && *cptr <= '9') /* decimal! */ | ||
2325 | c = c * 10 + *cptr++ - '0'; | ||
2326 | if (*cptr == ';') | ||
2327 | cptr++; /* swallow a trailing ';' */ | ||
2328 | } | ||
2329 | else | ||
2330 | { | ||
2331 | error(27); /* invalid character constant */ | ||
2332 | } /* if */ | ||
2333 | } /* switch */ | ||
2334 | } /* if */ | ||
2335 | } /* if */ | ||
2336 | *lptr = (char *)cptr; | ||
2337 | assert(c >= 0 && c < 256); | ||
2338 | return c; | ||
2339 | } | ||
2340 | |||
2341 | /* alpha | ||
2342 | * | ||
2343 | * Test if character "c" is alphabetic ("a".."z"), an underscore ("_") | ||
2344 | * or an "at" sign ("@"). The "@" is an extension to standard C. | ||
2345 | */ | ||
2346 | static int | ||
2347 | alpha(char c) | ||
2348 | { | ||
2349 | return (isalpha(c) || c == '_' || c == PUBLIC_CHAR); | ||
2350 | } | ||
2351 | |||
2352 | /* alphanum | ||
2353 | * | ||
2354 | * Test if character "c" is alphanumeric ("a".."z", "0".."9", "_" or "@") | ||
2355 | */ | ||
2356 | int | ||
2357 | alphanum(char c) | ||
2358 | { | ||
2359 | return (alpha(c) || isdigit(c)); | ||
2360 | } | ||
2361 | |||
2362 | /* The local variable table must be searched backwards, so that the deepest | ||
2363 | * nesting of local variables is searched first. The simplest way to do | ||
2364 | * this is to insert all new items at the head of the list. | ||
2365 | * In the global list, the symbols are kept in sorted order, so that the | ||
2366 | * public functions are written in sorted order. | ||
2367 | */ | ||
2368 | static symbol * | ||
2369 | add_symbol(symbol * root, symbol * entry, int sort) | ||
2370 | { | ||
2371 | symbol *newsym; | ||
2372 | |||
2373 | if (sort) | ||
2374 | while (root->next && strcmp(entry->name, root->next->name) > 0) | ||
2375 | root = root->next; | ||
2376 | |||
2377 | if (!(newsym = (symbol *)malloc(sizeof(symbol)))) | ||
2378 | { | ||
2379 | error(103); | ||
2380 | return NULL; | ||
2381 | } /* if */ | ||
2382 | memcpy(newsym, entry, sizeof(symbol)); | ||
2383 | newsym->next = root->next; | ||
2384 | root->next = newsym; | ||
2385 | return newsym; | ||
2386 | } | ||
2387 | |||
2388 | static void | ||
2389 | free_symbol(symbol * sym) | ||
2390 | { | ||
2391 | arginfo *arg; | ||
2392 | |||
2393 | /* free all sub-symbol allocated memory blocks, depending on the | ||
2394 | * kind of the symbol | ||
2395 | */ | ||
2396 | assert(sym != NULL); | ||
2397 | if (sym->ident == iFUNCTN) | ||
2398 | { | ||
2399 | /* run through the argument list; "default array" arguments | ||
2400 | * must be freed explicitly; the tag list must also be freed */ | ||
2401 | assert(sym->dim.arglist != NULL); | ||
2402 | for (arg = sym->dim.arglist; arg->ident != 0; arg++) | ||
2403 | { | ||
2404 | if (arg->ident == iREFARRAY && arg->hasdefault) | ||
2405 | free(arg->defvalue.array.data); | ||
2406 | else if (arg->ident == iVARIABLE | ||
2407 | && ((arg->hasdefault & uSIZEOF) != 0 | ||
2408 | || (arg->hasdefault & uTAGOF) != 0)) | ||
2409 | free(arg->defvalue.size.symname); | ||
2410 | assert(arg->tags != NULL); | ||
2411 | free(arg->tags); | ||
2412 | } /* for */ | ||
2413 | free(sym->dim.arglist); | ||
2414 | } /* if */ | ||
2415 | assert(sym->refer != NULL); | ||
2416 | free(sym->refer); | ||
2417 | free(sym); | ||
2418 | } | ||
2419 | |||
2420 | void | ||
2421 | delete_symbol(symbol * root, symbol * sym) | ||
2422 | { | ||
2423 | /* find the symbol and its predecessor | ||
2424 | * (this function assumes that you will never delete a symbol that is not | ||
2425 | * in the table pointed at by "root") | ||
2426 | */ | ||
2427 | assert(root != sym); | ||
2428 | while (root->next != sym) | ||
2429 | { | ||
2430 | root = root->next; | ||
2431 | assert(root != NULL); | ||
2432 | } /* while */ | ||
2433 | |||
2434 | /* unlink it, then free it */ | ||
2435 | root->next = sym->next; | ||
2436 | free_symbol(sym); | ||
2437 | } | ||
2438 | |||
2439 | void | ||
2440 | delete_symbols(symbol * root, int level, int delete_labels, | ||
2441 | int delete_functions) | ||
2442 | { | ||
2443 | symbol *sym; | ||
2444 | |||
2445 | /* erase only the symbols with a deeper nesting level than the | ||
2446 | * specified nesting level */ | ||
2447 | while (root->next) | ||
2448 | { | ||
2449 | sym = root->next; | ||
2450 | if (sym->compound < level) | ||
2451 | break; | ||
2452 | if ((delete_labels || sym->ident != iLABEL) | ||
2453 | && (delete_functions || sym->ident != iFUNCTN | ||
2454 | || (sym->usage & uNATIVE) != 0) && (delete_functions | ||
2455 | || sym->ident != iCONSTEXPR | ||
2456 | || (sym->usage & uPREDEF) == | ||
2457 | 0) && (delete_functions | ||
2458 | || (sym->ident != | ||
2459 | iVARIABLE | ||
2460 | && sym->ident != | ||
2461 | iARRAY))) | ||
2462 | { | ||
2463 | root->next = sym->next; | ||
2464 | free_symbol(sym); | ||
2465 | } | ||
2466 | else | ||
2467 | { | ||
2468 | /* if the function was prototyped, but not implemented in this source, | ||
2469 | * mark it as such, so that its use can be flagged | ||
2470 | */ | ||
2471 | if (sym->ident == iFUNCTN && (sym->usage & uDEFINE) == 0) | ||
2472 | sym->usage |= uMISSING; | ||
2473 | if (sym->ident == iFUNCTN || sym->ident == iVARIABLE | ||
2474 | || sym->ident == iARRAY) | ||
2475 | sym->usage &= ~uDEFINE; /* clear "defined" flag */ | ||
2476 | /* for user defined operators, also remove the "prototyped" flag, as | ||
2477 | * user-defined operators *must* be declared before use | ||
2478 | */ | ||
2479 | if (sym->ident == iFUNCTN && !isalpha(*sym->name) | ||
2480 | && *sym->name != '_' && *sym->name != PUBLIC_CHAR) | ||
2481 | sym->usage &= ~uPROTOTYPED; | ||
2482 | root = sym; /* skip the symbol */ | ||
2483 | } /* if */ | ||
2484 | } /* if */ | ||
2485 | } | ||
2486 | |||
2487 | /* The purpose of the hash is to reduce the frequency of a "name" | ||
2488 | * comparison (which is costly). There is little interest in avoiding | ||
2489 | * clusters in similar names, which is why this function is plain simple. | ||
2490 | */ | ||
2491 | unsigned int | ||
2492 | namehash(char *name) | ||
2493 | { | ||
2494 | unsigned char *ptr = (unsigned char *)name; | ||
2495 | int len = strlen(name); | ||
2496 | |||
2497 | if (len == 0) | ||
2498 | return 0L; | ||
2499 | assert(len < 256); | ||
2500 | return (len << 24Lu) + (ptr[0] << 16Lu) + (ptr[len - 1] << 8Lu) + | ||
2501 | (ptr[len >> 1Lu]); | ||
2502 | } | ||
2503 | |||
2504 | static symbol * | ||
2505 | find_symbol(symbol * root, char *name, int fnumber) | ||
2506 | { | ||
2507 | symbol *ptr = root->next; | ||
2508 | unsigned long hash = namehash(name); | ||
2509 | |||
2510 | while (ptr) | ||
2511 | { | ||
2512 | if (hash == ptr->hash && strcmp(name, ptr->name) == 0 | ||
2513 | && !ptr->parent && (ptr->fnumber < 0 | ||
2514 | || ptr->fnumber == fnumber)) | ||
2515 | return ptr; | ||
2516 | ptr = ptr->next; | ||
2517 | } /* while */ | ||
2518 | return NULL; | ||
2519 | } | ||
2520 | |||
2521 | static symbol * | ||
2522 | find_symbol_child(symbol * root, symbol * sym) | ||
2523 | { | ||
2524 | symbol *ptr = root->next; | ||
2525 | |||
2526 | while (ptr) | ||
2527 | { | ||
2528 | if (ptr->parent == sym) | ||
2529 | return ptr; | ||
2530 | ptr = ptr->next; | ||
2531 | } /* while */ | ||
2532 | return NULL; | ||
2533 | } | ||
2534 | |||
2535 | /* Adds "bywhom" to the list of referrers of "entry". Typically, | ||
2536 | * bywhom will be the function that uses a variable or that calls | ||
2537 | * the function. | ||
2538 | */ | ||
2539 | int | ||
2540 | refer_symbol(symbol * entry, symbol * bywhom) | ||
2541 | { | ||
2542 | int count; | ||
2543 | |||
2544 | assert(bywhom != NULL); /* it makes no sense to add a "void" referrer */ | ||
2545 | assert(entry != NULL); | ||
2546 | assert(entry->refer != NULL); | ||
2547 | |||
2548 | /* see if it is already there */ | ||
2549 | for (count = 0; count < entry->numrefers && entry->refer[count] != bywhom; | ||
2550 | count++) | ||
2551 | /* nothing */ ; | ||
2552 | if (count < entry->numrefers) | ||
2553 | { | ||
2554 | assert(entry->refer[count] == bywhom); | ||
2555 | return TRUE; | ||
2556 | } /* if */ | ||
2557 | |||
2558 | /* see if there is an empty spot in the referrer list */ | ||
2559 | for (count = 0; count < entry->numrefers && entry->refer[count]; | ||
2560 | count++) | ||
2561 | /* nothing */ ; | ||
2562 | assert(count <= entry->numrefers); | ||
2563 | if (count == entry->numrefers) | ||
2564 | { | ||
2565 | symbol **refer; | ||
2566 | int newsize = 2 * entry->numrefers; | ||
2567 | |||
2568 | assert(newsize > 0); | ||
2569 | /* grow the referrer list */ | ||
2570 | refer = (symbol **) realloc(entry->refer, newsize * sizeof(symbol *)); | ||
2571 | if (!refer) | ||
2572 | return FALSE; /* insufficient memory */ | ||
2573 | /* initialize the new entries */ | ||
2574 | entry->refer = refer; | ||
2575 | for (count = entry->numrefers; count < newsize; count++) | ||
2576 | entry->refer[count] = NULL; | ||
2577 | count = entry->numrefers; /* first empty spot */ | ||
2578 | entry->numrefers = newsize; | ||
2579 | } /* if */ | ||
2580 | |||
2581 | /* add the referrer */ | ||
2582 | assert(entry->refer[count] == NULL); | ||
2583 | entry->refer[count] = bywhom; | ||
2584 | return TRUE; | ||
2585 | } | ||
2586 | |||
2587 | void | ||
2588 | markusage(symbol * sym, int usage) | ||
2589 | { | ||
2590 | sym->usage |= (char)usage; | ||
2591 | /* check if (global) reference must be added to the symbol */ | ||
2592 | if ((usage & (uREAD | uWRITTEN)) != 0) | ||
2593 | { | ||
2594 | /* only do this for global symbols */ | ||
2595 | if (sym->vclass == sGLOBAL) | ||
2596 | { | ||
2597 | /* "curfunc" should always be valid, since statements may not occurs | ||
2598 | * outside functions; in the case of syntax errors, however, the | ||
2599 | * compiler may arrive through this function | ||
2600 | */ | ||
2601 | if (curfunc) | ||
2602 | refer_symbol(sym, curfunc); | ||
2603 | } /* if */ | ||
2604 | } /* if */ | ||
2605 | } | ||
2606 | |||
2607 | /* findglb | ||
2608 | * | ||
2609 | * Returns a pointer to the global symbol (if found) or NULL (if not found) | ||
2610 | */ | ||
2611 | symbol * | ||
2612 | findglb(char *name) | ||
2613 | { | ||
2614 | return find_symbol(&glbtab, name, fcurrent); | ||
2615 | } | ||
2616 | |||
2617 | /* findloc | ||
2618 | * | ||
2619 | * Returns a pointer to the local symbol (if found) or NULL (if not found). | ||
2620 | * See add_symbol() how the deepest nesting level is searched first. | ||
2621 | */ | ||
2622 | symbol * | ||
2623 | findloc(char *name) | ||
2624 | { | ||
2625 | return find_symbol(&loctab, name, -1); | ||
2626 | } | ||
2627 | |||
2628 | symbol * | ||
2629 | findconst(char *name) | ||
2630 | { | ||
2631 | symbol *sym; | ||
2632 | |||
2633 | sym = find_symbol(&loctab, name, -1); /* try local symbols first */ | ||
2634 | if (!sym || sym->ident != iCONSTEXPR) /* not found, or not a constant */ | ||
2635 | sym = find_symbol(&glbtab, name, fcurrent); | ||
2636 | if (!sym || sym->ident != iCONSTEXPR) | ||
2637 | return NULL; | ||
2638 | assert(sym->parent == NULL); /* constants have no hierarchy */ | ||
2639 | return sym; | ||
2640 | } | ||
2641 | |||
2642 | symbol * | ||
2643 | finddepend(symbol * parent) | ||
2644 | { | ||
2645 | symbol *sym; | ||
2646 | |||
2647 | sym = find_symbol_child(&loctab, parent); /* try local symbols first */ | ||
2648 | if (!sym) /* not found */ | ||
2649 | sym = find_symbol_child(&glbtab, parent); | ||
2650 | return sym; | ||
2651 | } | ||
2652 | |||
2653 | /* addsym | ||
2654 | * | ||
2655 | * Adds a symbol to the symbol table (either global or local variables, | ||
2656 | * or global and local constants). | ||
2657 | */ | ||
2658 | symbol * | ||
2659 | addsym(char *name, cell addr, int ident, int vclass, int tag, int usage) | ||
2660 | { | ||
2661 | symbol entry, **refer; | ||
2662 | |||
2663 | /* global variables/constants/functions may only be defined once */ | ||
2664 | assert(!(ident == iFUNCTN || ident == iCONSTEXPR) || vclass != sGLOBAL | ||
2665 | || findglb(name) == NULL); | ||
2666 | /* labels may only be defined once */ | ||
2667 | assert(ident != iLABEL || findloc(name) == NULL); | ||
2668 | |||
2669 | /* create an empty referrer list */ | ||
2670 | if (!(refer = (symbol **)malloc(sizeof(symbol *)))) | ||
2671 | { | ||
2672 | error(103); /* insufficient memory */ | ||
2673 | return NULL; | ||
2674 | } /* if */ | ||
2675 | *refer = NULL; | ||
2676 | |||
2677 | /* first fill in the entry */ | ||
2678 | strcpy(entry.name, name); | ||
2679 | entry.hash = namehash(name); | ||
2680 | entry.addr = addr; | ||
2681 | entry.vclass = (char)vclass; | ||
2682 | entry.ident = (char)ident; | ||
2683 | entry.tag = tag; | ||
2684 | entry.usage = (char)usage; | ||
2685 | entry.compound = 0; /* may be overridden later */ | ||
2686 | entry.fnumber = -1; /* assume global visibility (ignored for local symbols) */ | ||
2687 | entry.numrefers = 1; | ||
2688 | entry.refer = refer; | ||
2689 | entry.parent = NULL; | ||
2690 | |||
2691 | /* then insert it in the list */ | ||
2692 | if (vclass == sGLOBAL) | ||
2693 | return add_symbol(&glbtab, &entry, TRUE); | ||
2694 | else | ||
2695 | return add_symbol(&loctab, &entry, FALSE); | ||
2696 | } | ||
2697 | |||
2698 | symbol * | ||
2699 | addvariable(char *name, cell addr, int ident, int vclass, int tag, | ||
2700 | int dim[], int numdim, int idxtag[]) | ||
2701 | { | ||
2702 | symbol *sym, *parent, *top; | ||
2703 | int level; | ||
2704 | |||
2705 | /* global variables may only be defined once */ | ||
2706 | assert(vclass != sGLOBAL || (sym = findglb(name)) == NULL | ||
2707 | || (sym->usage & uDEFINE) == 0); | ||
2708 | |||
2709 | if (ident == iARRAY || ident == iREFARRAY) | ||
2710 | { | ||
2711 | parent = NULL; | ||
2712 | sym = NULL; /* to avoid a compiler warning */ | ||
2713 | for (level = 0; level < numdim; level++) | ||
2714 | { | ||
2715 | top = addsym(name, addr, ident, vclass, tag, uDEFINE); | ||
2716 | top->dim.array.length = dim[level]; | ||
2717 | top->dim.array.level = (short)(numdim - level - 1); | ||
2718 | top->x.idxtag = idxtag[level]; | ||
2719 | top->parent = parent; | ||
2720 | parent = top; | ||
2721 | if (level == 0) | ||
2722 | sym = top; | ||
2723 | } /* for */ | ||
2724 | } | ||
2725 | else | ||
2726 | { | ||
2727 | sym = addsym(name, addr, ident, vclass, tag, uDEFINE); | ||
2728 | } /* if */ | ||
2729 | return sym; | ||
2730 | } | ||
2731 | |||
2732 | /* getlabel | ||
2733 | * | ||
2734 | * Return next available internal label number. | ||
2735 | */ | ||
2736 | int | ||
2737 | getlabel(void) | ||
2738 | { | ||
2739 | return labnum++; | ||
2740 | } | ||
2741 | |||
2742 | /* itoh | ||
2743 | * | ||
2744 | * Converts a number to a hexadecimal string and returns a pointer to that | ||
2745 | * string. | ||
2746 | */ | ||
2747 | char * | ||
2748 | itoh(ucell val) | ||
2749 | { | ||
2750 | static char itohstr[15]; /* hex number is 10 characters long at most */ | ||
2751 | char *ptr; | ||
2752 | int i, nibble[8]; /* a 32-bit hexadecimal cell has 8 nibbles */ | ||
2753 | int max; | ||
2754 | |||
2755 | #if defined(BIT16) | ||
2756 | max = 4; | ||
2757 | #else | ||
2758 | max = 8; | ||
2759 | #endif | ||
2760 | ptr = itohstr; | ||
2761 | for (i = 0; i < max; i += 1) | ||
2762 | { | ||
2763 | nibble[i] = (int)(val & 0x0f); /* nibble 0 is lowest nibble */ | ||
2764 | val >>= 4; | ||
2765 | } /* endfor */ | ||
2766 | i = max - 1; | ||
2767 | while (nibble[i] == 0 && i > 0) /* search for highest non-zero nibble */ | ||
2768 | i -= 1; | ||
2769 | while (i >= 0) | ||
2770 | { | ||
2771 | if (nibble[i] >= 10) | ||
2772 | *ptr++ = (char)('a' + (nibble[i] - 10)); | ||
2773 | else | ||
2774 | *ptr++ = (char)('0' + nibble[i]); | ||
2775 | i -= 1; | ||
2776 | } /* while */ | ||
2777 | *ptr = '\0'; /* and a zero-terminator */ | ||
2778 | return itohstr; | ||
2779 | } | ||