File: | src/usr.bin/awk/lib.c |
Warning: | line 401, column 22 Although the value stored to 'sep' is used in the enclosing expression, the value is never actually read from 'sep' |
Press '?' to see keyboard shortcuts
Keyboard shortcuts:
1 | /* $OpenBSD: lib.c,v 1.55 2023/11/28 20:54:38 millert Exp $ */ |
2 | /**************************************************************** |
3 | Copyright (C) Lucent Technologies 1997 |
4 | All Rights Reserved |
5 | |
6 | Permission to use, copy, modify, and distribute this software and |
7 | its documentation for any purpose and without fee is hereby |
8 | granted, provided that the above copyright notice appear in all |
9 | copies and that both that the copyright notice and this |
10 | permission notice and warranty disclaimer appear in supporting |
11 | documentation, and that the name Lucent Technologies or any of |
12 | its entities not be used in advertising or publicity pertaining |
13 | to distribution of the software without specific, written prior |
14 | permission. |
15 | |
16 | LUCENT DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, |
17 | INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. |
18 | IN NO EVENT SHALL LUCENT OR ANY OF ITS ENTITIES BE LIABLE FOR ANY |
19 | SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES |
20 | WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER |
21 | IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, |
22 | ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF |
23 | THIS SOFTWARE. |
24 | ****************************************************************/ |
25 | |
26 | #define DEBUG |
27 | #include <stdio.h> |
28 | #include <string.h> |
29 | #include <strings.h> |
30 | #include <ctype.h> |
31 | #include <errno(*__errno()).h> |
32 | #include <stdlib.h> |
33 | #include <stdarg.h> |
34 | #include <limits.h> |
35 | #include <math.h> |
36 | #include "awk.h" |
37 | |
38 | extern int u8_nextlen(const char *s); |
39 | |
40 | char EMPTY[] = { '\0' }; |
41 | FILE *infile = NULL((void *)0); |
42 | bool_Bool innew; /* true = infile has not been read by readrec */ |
43 | char *file = EMPTY; |
44 | char *record; |
45 | int recsize = RECSIZE(8 * 1024); |
46 | char *fields; |
47 | int fieldssize = RECSIZE(8 * 1024); |
48 | |
49 | Cell **fldtab; /* pointers to Cells */ |
50 | static size_t len_inputFS = 0; |
51 | static char *inputFS = NULL((void *)0); /* FS at time of input, for field splitting */ |
52 | |
53 | #define MAXFLD2 2 |
54 | int nfields = MAXFLD2; /* last allocated slot for $i */ |
55 | |
56 | bool_Bool donefld; /* true = implies rec broken into fields */ |
57 | bool_Bool donerec; /* true = record is valid (no flds have changed) */ |
58 | |
59 | int lastfld = 0; /* last used field */ |
60 | int argno = 1; /* current input argument number */ |
61 | extern Awkfloat *ARGC; |
62 | |
63 | static Cell dollar0 = { OCELL1, CFLD1, NULL((void *)0), EMPTY, 0.0, REC0200|STR02|DONTFREE04, NULL((void *)0), NULL((void *)0) }; |
64 | static Cell dollar1 = { OCELL1, CFLD1, NULL((void *)0), EMPTY, 0.0, FLD0100|STR02|DONTFREE04, NULL((void *)0), NULL((void *)0) }; |
65 | |
66 | void recinit(unsigned int n) |
67 | { |
68 | if ( (record = (char *) malloc(n)) == NULL((void *)0) |
69 | || (fields = (char *) malloc(n+1)) == NULL((void *)0) |
70 | || (fldtab = (Cell **) calloc(nfields+2, sizeof(*fldtab))) == NULL((void *)0) |
71 | || (fldtab[0] = (Cell *) malloc(sizeof(**fldtab))) == NULL((void *)0)) |
72 | FATAL("out of space for $0 and fields"); |
73 | *record = '\0'; |
74 | *fldtab[0] = dollar0; |
75 | fldtab[0]->sval = record; |
76 | fldtab[0]->nval = tostring("0"); |
77 | makefields(1, nfields); |
78 | } |
79 | |
80 | void makefields(int n1, int n2) /* create $n1..$n2 inclusive */ |
81 | { |
82 | char temp[50]; |
83 | int i; |
84 | |
85 | for (i = n1; i <= n2; i++) { |
86 | fldtab[i] = (Cell *) malloc(sizeof(**fldtab)); |
87 | if (fldtab[i] == NULL((void *)0)) |
88 | FATAL("out of space in makefields %d", i); |
89 | *fldtab[i] = dollar1; |
90 | snprintf(temp, sizeof(temp), "%d", i); |
91 | fldtab[i]->nval = tostring(temp); |
92 | } |
93 | } |
94 | |
95 | void initgetrec(void) |
96 | { |
97 | int i; |
98 | char *p; |
99 | |
100 | for (i = 1; i < *ARGC; i++) { |
101 | p = getargv(i); /* find 1st real filename */ |
102 | if (p == NULL((void *)0) || *p == '\0') { /* deleted or zapped */ |
103 | argno++; |
104 | continue; |
105 | } |
106 | if (!isclvar(p)) { |
107 | setsval(lookup("FILENAME", symtab), p); |
108 | return; |
109 | } |
110 | setclvar(p); /* a commandline assignment before filename */ |
111 | argno++; |
112 | } |
113 | infile = stdin(&__sF[0]); /* no filenames, so use stdin */ |
114 | innew = true1; |
115 | } |
116 | |
117 | /* |
118 | * POSIX specifies that fields are supposed to be evaluated as if they were |
119 | * split using the value of FS at the time that the record's value ($0) was |
120 | * read. |
121 | * |
122 | * Since field-splitting is done lazily, we save the current value of FS |
123 | * whenever a new record is read in (implicitly or via getline), or when |
124 | * a new value is assigned to $0. |
125 | */ |
126 | void savefs(void) |
127 | { |
128 | size_t len = strlen(getsval(fsloc)); |
129 | if (len >= len_inputFS) { |
130 | len_inputFS = len + 1; |
131 | inputFS = (char *) realloc(inputFS, len_inputFS); |
132 | if (inputFS == NULL((void *)0)) |
133 | FATAL("field separator %.10s... is too long", *FS); |
134 | } |
135 | if (strlcpy(inputFS, *FS, len_inputFS) >= len_inputFS) |
136 | FATAL("field separator %.10s... is too long", *FS); |
137 | } |
138 | |
139 | static bool_Bool firsttime = true1; |
140 | |
141 | int getrec(char **pbuf, int *pbufsize, bool_Bool isrecord) /* get next input record */ |
142 | { /* note: cares whether buf == record */ |
143 | int c; |
144 | char *buf = *pbuf; |
145 | uschar saveb0; |
146 | int bufsize = *pbufsize, savebufsize = bufsize; |
147 | |
148 | if (firsttime) { |
149 | firsttime = false0; |
150 | initgetrec(); |
151 | } |
152 | DPRINTF("RS=<%s>, FS=<%s>, ARGC=%g, FILENAME=%s\n",if (dbg) printf("RS=<%s>, FS=<%s>, ARGC=%g, FILENAME=%s\n" , *RS, *FS, *ARGC, *FILENAME) |
153 | *RS, *FS, *ARGC, *FILENAME)if (dbg) printf("RS=<%s>, FS=<%s>, ARGC=%g, FILENAME=%s\n" , *RS, *FS, *ARGC, *FILENAME); |
154 | saveb0 = buf[0]; |
155 | buf[0] = 0; |
156 | while (argno < *ARGC || infile == stdin(&__sF[0])) { |
157 | DPRINTF("argno=%d, file=|%s|\n", argno, file)if (dbg) printf("argno=%d, file=|%s|\n", argno, file); |
158 | if (infile == NULL((void *)0)) { /* have to open a new file */ |
159 | file = getargv(argno); |
160 | if (file == NULL((void *)0) || *file == '\0') { /* deleted or zapped */ |
161 | argno++; |
162 | continue; |
163 | } |
164 | if (isclvar(file)) { /* a var=value arg */ |
165 | setclvar(file); |
166 | argno++; |
167 | continue; |
168 | } |
169 | *FILENAME = file; |
170 | DPRINTF("opening file %s\n", file)if (dbg) printf("opening file %s\n", file); |
171 | if (*file == '-' && *(file+1) == '\0') |
172 | infile = stdin(&__sF[0]); |
173 | else if ((infile = fopen(file, "r")) == NULL((void *)0)) |
174 | FATAL("can't open file %s", file); |
175 | innew = true1; |
176 | setfval(fnrloc, 0.0); |
177 | } |
178 | c = readrec(&buf, &bufsize, infile, innew); |
179 | if (innew) |
180 | innew = false0; |
181 | if (c != 0 || buf[0] != '\0') { /* normal record */ |
182 | if (isrecord) { |
183 | double result; |
184 | |
185 | if (freeable(fldtab[0])( ((fldtab[0])->tval & (02|04)) == 02 )) |
186 | xfree(fldtab[0]->sval){ free((void *)(intptr_t)(fldtab[0]->sval)); (fldtab[0]-> sval) = ((void *)0); }; |
187 | fldtab[0]->sval = buf; /* buf == record */ |
188 | fldtab[0]->tval = REC0200 | STR02 | DONTFREE04; |
189 | if (is_number(fldtab[0]->sval, & result)is_valid_number(fldtab[0]->sval, 0, ((void *)0), & result )) { |
190 | fldtab[0]->fval = result; |
191 | fldtab[0]->tval |= NUM01; |
192 | } |
193 | donefld = false0; |
194 | donerec = true1; |
195 | savefs(); |
196 | } |
197 | setfval(nrloc, nrloc->fval+1); |
198 | setfval(fnrloc, fnrloc->fval+1); |
199 | *pbuf = buf; |
200 | *pbufsize = bufsize; |
201 | return 1; |
202 | } |
203 | /* EOF arrived on this file; set up next */ |
204 | if (infile != stdin(&__sF[0])) |
205 | fclose(infile); |
206 | infile = NULL((void *)0); |
207 | argno++; |
208 | } |
209 | buf[0] = saveb0; |
210 | *pbuf = buf; |
211 | *pbufsize = savebufsize; |
212 | return 0; /* true end of file */ |
213 | } |
214 | |
215 | void nextfile(void) |
216 | { |
217 | if (infile != NULL((void *)0) && infile != stdin(&__sF[0])) |
218 | fclose(infile); |
219 | infile = NULL((void *)0); |
220 | argno++; |
221 | } |
222 | |
223 | extern int readcsvrec(char **pbuf, int *pbufsize, FILE *inf, bool_Bool newflag); |
224 | |
225 | int readrec(char **pbuf, int *pbufsize, FILE *inf, bool_Bool newflag) /* read one record into buf */ |
226 | { |
227 | int sep, c, isrec; // POTENTIAL BUG? isrec is a macro in awk.h |
228 | char *rr = *pbuf, *buf = *pbuf; |
229 | int bufsize = *pbufsize; |
230 | char *rs = getsval(rsloc); |
231 | |
232 | if (CSV) { |
233 | c = readcsvrec(pbuf, pbufsize, inf, newflag); |
234 | isrec = (c == EOF(-1) && rr == buf) ? false0 : true1; |
235 | } else if (*rs && rs[1]) { |
236 | bool_Bool found; |
237 | |
238 | memset(buf, 0, bufsize); |
239 | fa *pfa = makedfa(rs, 1); |
240 | if (newflag) |
241 | found = fnematch(pfa, inf, &buf, &bufsize, recsize); |
242 | else { |
243 | int tempstat = pfa->initstat; |
244 | pfa->initstat = 2; |
245 | found = fnematch(pfa, inf, &buf, &bufsize, recsize); |
246 | pfa->initstat = tempstat; |
247 | } |
248 | if (found) |
249 | setptr(patbeg, '\0')(*(char *)(intptr_t)(patbeg)) = ('\0'); |
250 | isrec = (found == 0 && *buf == '\0') ? false0 : true1; |
251 | } else { |
252 | if ((sep = *rs) == 0) { |
253 | sep = '\n'; |
254 | while ((c=getc(inf)(!__isthreaded ? (--(inf)->_r < 0 ? __srget(inf) : (int )(*(inf)->_p++)) : (getc)(inf))) == '\n' && c != EOF(-1)) /* skip leading \n's */ |
255 | ; |
256 | if (c != EOF(-1)) |
257 | ungetc(c, inf); |
258 | } |
259 | for (rr = buf; ; ) { |
260 | for (; (c=getc(inf)(!__isthreaded ? (--(inf)->_r < 0 ? __srget(inf) : (int )(*(inf)->_p++)) : (getc)(inf))) != sep && c != EOF(-1); ) { |
261 | if (rr-buf+1 > bufsize) |
262 | if (!adjbuf(&buf, &bufsize, 1+rr-buf, |
263 | recsize, &rr, "readrec 1")) |
264 | FATAL("input record `%.30s...' too long", buf); |
265 | *rr++ = c; |
266 | } |
267 | if (*rs == sep || c == EOF(-1)) |
268 | break; |
269 | if ((c = getc(inf)(!__isthreaded ? (--(inf)->_r < 0 ? __srget(inf) : (int )(*(inf)->_p++)) : (getc)(inf))) == '\n' || c == EOF(-1)) /* 2 in a row */ |
270 | break; |
271 | if (!adjbuf(&buf, &bufsize, 2+rr-buf, recsize, &rr, |
272 | "readrec 2")) |
273 | FATAL("input record `%.30s...' too long", buf); |
274 | *rr++ = '\n'; |
275 | *rr++ = c; |
276 | } |
277 | if (!adjbuf(&buf, &bufsize, 1+rr-buf, recsize, &rr, "readrec 3")) |
278 | FATAL("input record `%.30s...' too long", buf); |
279 | *rr = 0; |
280 | isrec = (c == EOF(-1) && rr == buf) ? false0 : true1; |
281 | } |
282 | *pbuf = buf; |
283 | *pbufsize = bufsize; |
284 | DPRINTF("readrec saw <%s>, returns %d\n", buf, isrec)if (dbg) printf("readrec saw <%s>, returns %d\n", buf, isrec ); |
285 | return isrec; |
286 | } |
287 | |
288 | |
289 | /******************* |
290 | * loose ends here: |
291 | * \r\n should become \n |
292 | * what about bare \r? Excel uses that for embedded newlines |
293 | * can't have "" in unquoted fields, according to RFC 4180 |
294 | */ |
295 | |
296 | int readcsvrec(char **pbuf, int *pbufsize, FILE *inf, bool_Bool newflag) /* csv can have \n's */ |
297 | { /* so read a complete record that might be multiple lines */ |
298 | int sep, c; |
299 | char *rr = *pbuf, *buf = *pbuf; |
300 | int bufsize = *pbufsize; |
301 | bool_Bool in_quote = false0; |
302 | |
303 | sep = '\n'; /* the only separator; have to skip over \n embedded in "..." */ |
304 | rr = buf; |
305 | while ((c = getc(inf)(!__isthreaded ? (--(inf)->_r < 0 ? __srget(inf) : (int )(*(inf)->_p++)) : (getc)(inf))) != EOF(-1)) { |
306 | if (c == sep) { |
307 | if (! in_quote) |
308 | break; |
309 | if (rr > buf && rr[-1] == '\r') // remove \r if was \r\n |
310 | rr--; |
311 | } |
312 | |
313 | if (rr-buf+1 > bufsize) |
314 | if (!adjbuf(&buf, &bufsize, 1+rr-buf, |
315 | recsize, &rr, "readcsvrec 1")) |
316 | FATAL("input record `%.30s...' too long", buf); |
317 | *rr++ = c; |
318 | if (c == '"') |
319 | in_quote = ! in_quote; |
320 | } |
321 | if (c == '\n' && rr > buf && rr[-1] == '\r') // remove \r if was \r\n |
322 | rr--; |
323 | |
324 | if (!adjbuf(&buf, &bufsize, 1+rr-buf, recsize, &rr, "readcsvrec 4")) |
325 | FATAL("input record `%.30s...' too long", buf); |
326 | *rr = 0; |
327 | *pbuf = buf; |
328 | *pbufsize = bufsize; |
329 | DPRINTF("readcsvrec saw <%s>, returns %d\n", buf, c)if (dbg) printf("readcsvrec saw <%s>, returns %d\n", buf , c); |
330 | return c; |
331 | } |
332 | |
333 | char *getargv(int n) /* get ARGV[n] */ |
334 | { |
335 | Cell *x; |
336 | char *s, temp[50]; |
337 | extern Array *ARGVtab; |
338 | |
339 | snprintf(temp, sizeof(temp), "%d", n); |
340 | if (lookup(temp, ARGVtab) == NULL((void *)0)) |
341 | return NULL((void *)0); |
342 | x = setsymtab(temp, "", 0.0, STR02, ARGVtab); |
343 | s = getsval(x); |
344 | DPRINTF("getargv(%d) returns |%s|\n", n, s)if (dbg) printf("getargv(%d) returns |%s|\n", n, s); |
345 | return s; |
346 | } |
347 | |
348 | void setclvar(char *s) /* set var=value from s */ |
349 | { |
350 | char *e, *p; |
351 | Cell *q; |
352 | double result; |
353 | |
354 | /* commit f3d9187d4e0f02294fb1b0e31152070506314e67 broke T.argv test */ |
355 | /* I don't understand why it was changed. */ |
356 | |
357 | for (p=s; *p != '='; p++) |
358 | ; |
359 | e = p; |
360 | *p++ = 0; |
361 | p = qstring(p, '\0'); |
362 | q = setsymtab(s, p, 0.0, STR02, symtab); |
363 | setsval(q, p); |
364 | if (is_number(q->sval, & result)is_valid_number(q->sval, 0, ((void *)0), & result)) { |
365 | q->fval = result; |
366 | q->tval |= NUM01; |
367 | } |
368 | DPRINTF("command line set %s to |%s|\n", s, p)if (dbg) printf("command line set %s to |%s|\n", s, p); |
369 | free(p); |
370 | *e = '='; |
371 | } |
372 | |
373 | |
374 | void fldbld(void) /* create fields from current record */ |
375 | { |
376 | /* this relies on having fields[] the same length as $0 */ |
377 | /* the fields are all stored in this one array with \0's */ |
378 | /* possibly with a final trailing \0 not associated with any field */ |
379 | char *r, *fr, sep; |
380 | Cell *p; |
381 | int i, j, n; |
382 | |
383 | if (donefld) |
384 | return; |
385 | if (!isstr(fldtab[0])((fldtab[0])->tval & 02)) |
386 | getsval(fldtab[0]); |
387 | r = fldtab[0]->sval; |
388 | n = strlen(r); |
389 | if (n > fieldssize) { |
390 | xfree(fields){ free((void *)(intptr_t)(fields)); (fields) = ((void *)0); }; |
391 | if ((fields = (char *) malloc(n+2)) == NULL((void *)0)) /* possibly 2 final \0s */ |
392 | FATAL("out of space for fields in fldbld %d", n); |
393 | fieldssize = n; |
394 | } |
395 | fr = fields; |
396 | i = 0; /* number of fields accumulated here */ |
397 | if (inputFS == NULL((void *)0)) /* make sure we have a copy of FS */ |
398 | savefs(); |
399 | if (!CSV && strlen(inputFS) > 1) { /* it's a regular expression */ |
400 | i = refldbld(r, inputFS); |
401 | } else if (!CSV && (sep = *inputFS) == ' ') { /* default whitespace */ |
Although the value stored to 'sep' is used in the enclosing expression, the value is never actually read from 'sep' | |
402 | for (i = 0; ; ) { |
403 | while (*r == ' ' || *r == '\t' || *r == '\n') |
404 | r++; |
405 | if (*r == 0) |
406 | break; |
407 | i++; |
408 | if (i > nfields) |
409 | growfldtab(i); |
410 | if (freeable(fldtab[i])( ((fldtab[i])->tval & (02|04)) == 02 )) |
411 | xfree(fldtab[i]->sval){ free((void *)(intptr_t)(fldtab[i]->sval)); (fldtab[i]-> sval) = ((void *)0); }; |
412 | fldtab[i]->sval = fr; |
413 | fldtab[i]->tval = FLD0100 | STR02 | DONTFREE04; |
414 | do |
415 | *fr++ = *r++; |
416 | while (*r != ' ' && *r != '\t' && *r != '\n' && *r != '\0'); |
417 | *fr++ = 0; |
418 | } |
419 | *fr = 0; |
420 | } else if (CSV) { /* CSV processing. no error handling */ |
421 | if (*r != 0) { |
422 | for (;;) { |
423 | i++; |
424 | if (i > nfields) |
425 | growfldtab(i); |
426 | if (freeable(fldtab[i])( ((fldtab[i])->tval & (02|04)) == 02 )) |
427 | xfree(fldtab[i]->sval){ free((void *)(intptr_t)(fldtab[i]->sval)); (fldtab[i]-> sval) = ((void *)0); }; |
428 | fldtab[i]->sval = fr; |
429 | fldtab[i]->tval = FLD0100 | STR02 | DONTFREE04; |
430 | if (*r == '"' ) { /* start of "..." */ |
431 | for (r++ ; *r != '\0'; ) { |
432 | if (*r == '"' && r[1] != '\0' && r[1] == '"') { |
433 | r += 2; /* doubled quote */ |
434 | *fr++ = '"'; |
435 | } else if (*r == '"' && (r[1] == '\0' || r[1] == ',')) { |
436 | r++; /* skip over closing quote */ |
437 | break; |
438 | } else { |
439 | *fr++ = *r++; |
440 | } |
441 | } |
442 | *fr++ = 0; |
443 | } else { /* unquoted field */ |
444 | while (*r != ',' && *r != '\0') |
445 | *fr++ = *r++; |
446 | *fr++ = 0; |
447 | } |
448 | if (*r++ == 0) |
449 | break; |
450 | |
451 | } |
452 | } |
453 | *fr = 0; |
454 | } else if ((sep = *inputFS) == 0) { /* new: FS="" => 1 char/field */ |
455 | for (i = 0; *r != '\0'; ) { |
456 | char buf[10]; |
457 | i++; |
458 | if (i > nfields) |
459 | growfldtab(i); |
460 | if (freeable(fldtab[i])( ((fldtab[i])->tval & (02|04)) == 02 )) |
461 | xfree(fldtab[i]->sval){ free((void *)(intptr_t)(fldtab[i]->sval)); (fldtab[i]-> sval) = ((void *)0); }; |
462 | n = u8_nextlen(r); |
463 | for (j = 0; j < n; j++) |
464 | buf[j] = *r++; |
465 | buf[j] = '\0'; |
466 | fldtab[i]->sval = tostring(buf); |
467 | fldtab[i]->tval = FLD0100 | STR02; |
468 | } |
469 | *fr = 0; |
470 | } else if (*r != 0) { /* if 0, it's a null field */ |
471 | /* subtle case: if length(FS) == 1 && length(RS > 0) |
472 | * \n is NOT a field separator (cf awk book 61,84). |
473 | * this variable is tested in the inner while loop. |
474 | */ |
475 | int rtest = '\n'; /* normal case */ |
476 | if (strlen(*RS) > 0) |
477 | rtest = '\0'; |
478 | for (;;) { |
479 | i++; |
480 | if (i > nfields) |
481 | growfldtab(i); |
482 | if (freeable(fldtab[i])( ((fldtab[i])->tval & (02|04)) == 02 )) |
483 | xfree(fldtab[i]->sval){ free((void *)(intptr_t)(fldtab[i]->sval)); (fldtab[i]-> sval) = ((void *)0); }; |
484 | fldtab[i]->sval = fr; |
485 | fldtab[i]->tval = FLD0100 | STR02 | DONTFREE04; |
486 | while (*r != sep && *r != rtest && *r != '\0') /* \n is always a separator */ |
487 | *fr++ = *r++; |
488 | *fr++ = 0; |
489 | if (*r++ == 0) |
490 | break; |
491 | } |
492 | *fr = 0; |
493 | } |
494 | if (i > nfields) |
495 | FATAL("record `%.30s...' has too many fields; can't happen", r); |
496 | cleanfld(i+1, lastfld); /* clean out junk from previous record */ |
497 | lastfld = i; |
498 | donefld = true1; |
499 | for (j = 1; j <= lastfld; j++) { |
500 | double result; |
501 | |
502 | p = fldtab[j]; |
503 | if(is_number(p->sval, & result)is_valid_number(p->sval, 0, ((void *)0), & result)) { |
504 | p->fval = result; |
505 | p->tval |= NUM01; |
506 | } |
507 | } |
508 | setfval(nfloc, (Awkfloat) lastfld); |
509 | donerec = true1; /* restore */ |
510 | if (dbg) { |
511 | for (j = 0; j <= lastfld; j++) { |
512 | p = fldtab[j]; |
513 | printf("field %d (%s): |%s|\n", j, p->nval, p->sval); |
514 | } |
515 | } |
516 | } |
517 | |
518 | void cleanfld(int n1, int n2) /* clean out fields n1 .. n2 inclusive */ |
519 | { /* nvals remain intact */ |
520 | Cell *p; |
521 | int i; |
522 | |
523 | for (i = n1; i <= n2; i++) { |
524 | p = fldtab[i]; |
525 | if (freeable(p)( ((p)->tval & (02|04)) == 02 )) |
526 | xfree(p->sval){ free((void *)(intptr_t)(p->sval)); (p->sval) = ((void *)0); }; |
527 | p->sval = EMPTY, |
528 | p->tval = FLD0100 | STR02 | DONTFREE04; |
529 | } |
530 | } |
531 | |
532 | void newfld(int n) /* add field n after end of existing lastfld */ |
533 | { |
534 | if (n > nfields) |
535 | growfldtab(n); |
536 | cleanfld(lastfld+1, n); |
537 | lastfld = n; |
538 | setfval(nfloc, (Awkfloat) n); |
539 | } |
540 | |
541 | void setlastfld(int n) /* set lastfld cleaning fldtab cells if necessary */ |
542 | { |
543 | if (n < 0) |
544 | FATAL("cannot set NF to a negative value"); |
545 | if (n > nfields) |
546 | growfldtab(n); |
547 | |
548 | if (lastfld < n) |
549 | cleanfld(lastfld+1, n); |
550 | else |
551 | cleanfld(n+1, lastfld); |
552 | |
553 | lastfld = n; |
554 | } |
555 | |
556 | Cell *fieldadr(int n) /* get nth field */ |
557 | { |
558 | if (n < 0) |
559 | FATAL("trying to access out of range field %d", n); |
560 | if (n > nfields) /* fields after NF are empty */ |
561 | growfldtab(n); /* but does not increase NF */ |
562 | return(fldtab[n]); |
563 | } |
564 | |
565 | void growfldtab(int n) /* make new fields up to at least $n */ |
566 | { |
567 | int nf = 2 * nfields; |
568 | size_t s; |
569 | |
570 | if (n > nf) |
571 | nf = n; |
572 | s = (nf+1) * (sizeof (struct Cell *)); /* freebsd: how much do we need? */ |
573 | if (s / sizeof(struct Cell *) - 1 == (size_t)nf) /* didn't overflow */ |
574 | fldtab = (Cell **) realloc(fldtab, s); |
575 | else /* overflow sizeof int */ |
576 | xfree(fldtab){ free((void *)(intptr_t)(fldtab)); (fldtab) = ((void *)0); }; /* make it null */ |
577 | if (fldtab == NULL((void *)0)) |
578 | FATAL("out of space creating %d fields", nf); |
579 | makefields(nfields+1, nf); |
580 | nfields = nf; |
581 | } |
582 | |
583 | int refldbld(const char *rec, const char *fs) /* build fields from reg expr in FS */ |
584 | { |
585 | /* this relies on having fields[] the same length as $0 */ |
586 | /* the fields are all stored in this one array with \0's */ |
587 | char *fr; |
588 | int i, tempstat, n; |
589 | fa *pfa; |
590 | |
591 | n = strlen(rec); |
592 | if (n > fieldssize) { |
593 | xfree(fields){ free((void *)(intptr_t)(fields)); (fields) = ((void *)0); }; |
594 | if ((fields = (char *) malloc(n+1)) == NULL((void *)0)) |
595 | FATAL("out of space for fields in refldbld %d", n); |
596 | fieldssize = n; |
597 | } |
598 | fr = fields; |
599 | *fr = '\0'; |
600 | if (*rec == '\0') |
601 | return 0; |
602 | pfa = makedfa(fs, 1); |
603 | DPRINTF("into refldbld, rec = <%s>, pat = <%s>\n", rec, fs)if (dbg) printf("into refldbld, rec = <%s>, pat = <%s>\n" , rec, fs); |
604 | tempstat = pfa->initstat; |
605 | for (i = 1; ; i++) { |
606 | const size_t fss_rem = fields + fieldssize + 1 - fr; |
607 | if (i > nfields) |
608 | growfldtab(i); |
609 | if (freeable(fldtab[i])( ((fldtab[i])->tval & (02|04)) == 02 )) |
610 | xfree(fldtab[i]->sval){ free((void *)(intptr_t)(fldtab[i]->sval)); (fldtab[i]-> sval) = ((void *)0); }; |
611 | fldtab[i]->tval = FLD0100 | STR02 | DONTFREE04; |
612 | fldtab[i]->sval = fr; |
613 | DPRINTF("refldbld: i=%d\n", i)if (dbg) printf("refldbld: i=%d\n", i); |
614 | if (nematch(pfa, rec)) { |
615 | const size_t reclen = patbeg - rec; |
616 | pfa->initstat = 2; /* horrible coupling to b.c */ |
617 | DPRINTF("match %s (%d chars)\n", patbeg, patlen)if (dbg) printf("match %s (%d chars)\n", patbeg, patlen); |
618 | if (reclen >= fss_rem) |
619 | FATAL("out of space for fields in refldbld"); |
620 | memcpy(fr, rec, reclen); |
621 | fr += reclen; |
622 | *fr++ = '\0'; |
623 | rec = patbeg + patlen; |
624 | } else { |
625 | DPRINTF("no match %s\n", rec)if (dbg) printf("no match %s\n", rec); |
626 | if (strlcpy(fr, rec, fss_rem) >= fss_rem) |
627 | FATAL("out of space for fields in refldbld"); |
628 | pfa->initstat = tempstat; |
629 | break; |
630 | } |
631 | } |
632 | return i; |
633 | } |
634 | |
635 | void recbld(void) /* create $0 from $1..$NF if necessary */ |
636 | { |
637 | int i; |
638 | char *r, *p; |
639 | char *sep = getsval(ofsloc); |
640 | |
641 | if (donerec) |
642 | return; |
643 | r = record; |
644 | for (i = 1; i <= *NF; i++) { |
645 | p = getsval(fldtab[i]); |
646 | if (!adjbuf(&record, &recsize, 1+strlen(p)+r-record, recsize, &r, "recbld 1")) |
647 | FATAL("created $0 `%.30s...' too long", record); |
648 | while ((*r = *p++) != 0) |
649 | r++; |
650 | if (i < *NF) { |
651 | if (!adjbuf(&record, &recsize, 2+strlen(sep)+r-record, recsize, &r, "recbld 2")) |
652 | FATAL("created $0 `%.30s...' too long", record); |
653 | for (p = sep; (*r = *p++) != 0; ) |
654 | r++; |
655 | } |
656 | } |
657 | if (!adjbuf(&record, &recsize, 2+r-record, recsize, &r, "recbld 3")) |
658 | FATAL("built giant record `%.30s...'", record); |
659 | *r = '\0'; |
660 | DPRINTF("in recbld inputFS=%s, fldtab[0]=%p\n", inputFS, (void*)fldtab[0])if (dbg) printf("in recbld inputFS=%s, fldtab[0]=%p\n", inputFS , (void*)fldtab[0]); |
661 | |
662 | if (freeable(fldtab[0])( ((fldtab[0])->tval & (02|04)) == 02 )) |
663 | xfree(fldtab[0]->sval){ free((void *)(intptr_t)(fldtab[0]->sval)); (fldtab[0]-> sval) = ((void *)0); }; |
664 | fldtab[0]->tval = REC0200 | STR02 | DONTFREE04; |
665 | fldtab[0]->sval = record; |
666 | |
667 | DPRINTF("in recbld inputFS=%s, fldtab[0]=%p\n", inputFS, (void*)fldtab[0])if (dbg) printf("in recbld inputFS=%s, fldtab[0]=%p\n", inputFS , (void*)fldtab[0]); |
668 | DPRINTF("recbld = |%s|\n", record)if (dbg) printf("recbld = |%s|\n", record); |
669 | donerec = true1; |
670 | } |
671 | |
672 | int errorflag = 0; |
673 | |
674 | void yyerror(const char *s) |
675 | { |
676 | SYNTAX("%s", s); |
677 | } |
678 | |
679 | void SYNTAX(const char *fmt, ...) |
680 | { |
681 | extern char *cmdname, *curfname; |
682 | static int been_here = 0; |
683 | va_list varg; |
684 | |
685 | if (been_here++ > 2) |
686 | return; |
687 | fprintf(stderr(&__sF[2]), "%s: ", cmdname); |
688 | va_start(varg, fmt)__builtin_va_start((varg), fmt); |
689 | vfprintf(stderr(&__sF[2]), fmt, varg); |
690 | va_end(varg)__builtin_va_end((varg)); |
691 | fprintf(stderr(&__sF[2]), " at source line %d", lineno); |
692 | if (curfname != NULL((void *)0)) |
693 | fprintf(stderr(&__sF[2]), " in function %s", curfname); |
694 | if (compile_time == COMPILING && cursource() != NULL((void *)0)) |
695 | fprintf(stderr(&__sF[2]), " source file %s", cursource()); |
696 | fprintf(stderr(&__sF[2]), "\n"); |
697 | errorflag = 2; |
698 | eprint(); |
699 | } |
700 | |
701 | extern int bracecnt, brackcnt, parencnt; |
702 | |
703 | void bracecheck(void) |
704 | { |
705 | int c; |
706 | static int beenhere = 0; |
707 | |
708 | if (beenhere++) |
709 | return; |
710 | while ((c = input()) != EOF(-1) && c != '\0') |
711 | bclass(c); |
712 | bcheck2(bracecnt, '{', '}'); |
713 | bcheck2(brackcnt, '[', ']'); |
714 | bcheck2(parencnt, '(', ')'); |
715 | } |
716 | |
717 | void bcheck2(int n, int c1, int c2) |
718 | { |
719 | if (n == 1) |
720 | fprintf(stderr(&__sF[2]), "\tmissing %c\n", c2); |
721 | else if (n > 1) |
722 | fprintf(stderr(&__sF[2]), "\t%d missing %c's\n", n, c2); |
723 | else if (n == -1) |
724 | fprintf(stderr(&__sF[2]), "\textra %c\n", c2); |
725 | else if (n < -1) |
726 | fprintf(stderr(&__sF[2]), "\t%d extra %c's\n", -n, c2); |
727 | } |
728 | |
729 | void FATAL(const char *fmt, ...) |
730 | { |
731 | extern char *cmdname; |
732 | va_list varg; |
733 | |
734 | fflush(stdout(&__sF[1])); |
735 | fprintf(stderr(&__sF[2]), "%s: ", cmdname); |
736 | va_start(varg, fmt)__builtin_va_start((varg), fmt); |
737 | vfprintf(stderr(&__sF[2]), fmt, varg); |
738 | va_end(varg)__builtin_va_end((varg)); |
739 | error(); |
740 | if (dbg > 1) /* core dump if serious debugging on */ |
741 | abort(); |
742 | exit(2); |
743 | } |
744 | |
745 | void WARNING(const char *fmt, ...) |
746 | { |
747 | extern char *cmdname; |
748 | va_list varg; |
749 | |
750 | fflush(stdout(&__sF[1])); |
751 | fprintf(stderr(&__sF[2]), "%s: ", cmdname); |
752 | va_start(varg, fmt)__builtin_va_start((varg), fmt); |
753 | vfprintf(stderr(&__sF[2]), fmt, varg); |
754 | va_end(varg)__builtin_va_end((varg)); |
755 | error(); |
756 | } |
757 | |
758 | void error() |
759 | { |
760 | extern Node *curnode; |
761 | |
762 | fprintf(stderr(&__sF[2]), "\n"); |
763 | if (compile_time != ERROR_PRINTING) { |
764 | if (NR && *NR > 0) { |
765 | fprintf(stderr(&__sF[2]), " input record number %d", (int) (*FNR)); |
766 | if (strcmp(*FILENAME, "-") != 0) |
767 | fprintf(stderr(&__sF[2]), ", file %s", *FILENAME); |
768 | fprintf(stderr(&__sF[2]), "\n"); |
769 | } |
770 | if (curnode) |
771 | fprintf(stderr(&__sF[2]), " source line number %d", curnode->lineno); |
772 | else if (lineno) |
773 | fprintf(stderr(&__sF[2]), " source line number %d", lineno); |
774 | if (compile_time == COMPILING && cursource() != NULL((void *)0)) |
775 | fprintf(stderr(&__sF[2]), " source file %s", cursource()); |
776 | fprintf(stderr(&__sF[2]), "\n"); |
777 | eprint(); |
778 | } |
779 | } |
780 | |
781 | void eprint(void) /* try to print context around error */ |
782 | { |
783 | char *p, *q; |
784 | int c; |
785 | static int been_here = 0; |
786 | extern char ebuf[], *ep; |
787 | |
788 | if (compile_time != COMPILING || been_here++ > 0 || ebuf == ep) |
789 | return; |
790 | p = ep - 1; |
791 | if (p > ebuf && *p == '\n') |
792 | p--; |
793 | for ( ; p > ebuf && *p != '\n' && *p != '\0'; p--) |
794 | ; |
795 | while (*p == '\n') |
796 | p++; |
797 | fprintf(stderr(&__sF[2]), " context is\n\t"); |
798 | for (q=ep-1; q>=p && *q!=' ' && *q!='\t' && *q!='\n'; q--) |
799 | ; |
800 | for ( ; p < q; p++) |
801 | if (*p) |
802 | putc(*p, stderr)(!__isthreaded ? __sputc(*p, (&__sF[2])) : (putc)(*p, (& __sF[2]))); |
803 | fprintf(stderr(&__sF[2]), " >>> "); |
804 | for ( ; p < ep; p++) |
805 | if (*p) |
806 | putc(*p, stderr)(!__isthreaded ? __sputc(*p, (&__sF[2])) : (putc)(*p, (& __sF[2]))); |
807 | fprintf(stderr(&__sF[2]), " <<< "); |
808 | if (*ep) |
809 | while ((c = input()) != '\n' && c != '\0' && c != EOF(-1)) { |
810 | putc(c, stderr)(!__isthreaded ? __sputc(c, (&__sF[2])) : (putc)(c, (& __sF[2]))); |
811 | bclass(c); |
812 | } |
813 | putc('\n', stderr)(!__isthreaded ? __sputc('\n', (&__sF[2])) : (putc)('\n', (&__sF[2]))); |
814 | ep = ebuf; |
815 | } |
816 | |
817 | void bclass(int c) |
818 | { |
819 | switch (c) { |
820 | case '{': bracecnt++; break; |
821 | case '}': bracecnt--; break; |
822 | case '[': brackcnt++; break; |
823 | case ']': brackcnt--; break; |
824 | case '(': parencnt++; break; |
825 | case ')': parencnt--; break; |
826 | } |
827 | } |
828 | |
829 | double errcheck(double x, const char *s) |
830 | { |
831 | |
832 | if (errno(*__errno()) == EDOM33) { |
833 | errno(*__errno()) = 0; |
834 | WARNING("%s argument out of domain", s); |
835 | x = 1; |
836 | } else if (errno(*__errno()) == ERANGE34) { |
837 | errno(*__errno()) = 0; |
838 | WARNING("%s result out of range", s); |
839 | x = 1; |
840 | } |
841 | return x; |
842 | } |
843 | |
844 | int isclvar(const char *s) /* is s of form var=something ? */ |
845 | { |
846 | const char *os = s; |
847 | |
848 | if (!isalpha((uschar) *s) && *s != '_') |
849 | return 0; |
850 | for ( ; *s; s++) |
851 | if (!(isalnum((uschar) *s) || *s == '_')) |
852 | break; |
853 | return *s == '=' && s > os; |
854 | } |
855 | |
856 | /* strtod is supposed to be a proper test of what's a valid number */ |
857 | /* appears to be broken in gcc on linux: thinks 0x123 is a valid FP number */ |
858 | /* wrong: violates 4.10.1.4 of ansi C standard */ |
859 | |
860 | /* well, not quite. As of C99, hex floating point is allowed. so this is |
861 | * a bit of a mess. We work around the mess by checking for a hexadecimal |
862 | * value and disallowing it. Similarly, we now follow gawk and allow only |
863 | * +nan, -nan, +inf, and -inf for NaN and infinity values. |
864 | */ |
865 | |
866 | /* |
867 | * This routine now has a more complicated interface, the main point |
868 | * being to avoid the double conversion of a string to double, and |
869 | * also to convey out, if requested, the information that the numeric |
870 | * value was a leading string or is all of the string. The latter bit |
871 | * is used in getfval(). |
872 | */ |
873 | |
874 | bool_Bool is_valid_number(const char *s, bool_Bool trailing_stuff_ok, |
875 | bool_Bool *no_trailing, double *result) |
876 | { |
877 | double r; |
878 | char *ep; |
879 | bool_Bool retval = false0; |
880 | bool_Bool is_nan = false0; |
881 | bool_Bool is_inf = false0; |
882 | |
883 | if (no_trailing) |
884 | *no_trailing = false0; |
885 | |
886 | while (isspace((uschar)*s)) |
887 | s++; |
888 | |
889 | /* no hex floating point, sorry */ |
890 | if (s[0] == '0' && tolower((uschar)s[1]) == 'x') |
891 | return false0; |
892 | |
893 | /* allow +nan, -nan, +inf, -inf, any other letter, no */ |
894 | if (s[0] == '+' || s[0] == '-') { |
895 | is_nan = (strncasecmp(s+1, "nan", 3) == 0); |
896 | is_inf = (strncasecmp(s+1, "inf", 3) == 0); |
897 | if ((is_nan || is_inf) |
898 | && (isspace((uschar)s[4]) || s[4] == '\0')) |
899 | goto convert; |
900 | else if (! isdigit((uschar)s[1]) && s[1] != '.') |
901 | return false0; |
902 | } |
903 | else if (! isdigit((uschar)s[0]) && s[0] != '.') |
904 | return false0; |
905 | |
906 | convert: |
907 | errno(*__errno()) = 0; |
908 | r = strtod(s, &ep); |
909 | if (ep == s || errno(*__errno()) == ERANGE34) |
910 | return false0; |
911 | |
912 | if (isnan(r)((sizeof (r) == sizeof (float)) ? __isnanf(r) : (sizeof (r) == sizeof (double)) ? __isnan(r) : __isnanl(r)) && s[0] == '-' && signbit(r)((sizeof (r) == sizeof (float)) ? __signbitf(r) : (sizeof (r) == sizeof (double)) ? __signbit(r) : __signbitl(r)) == 0) |
913 | r = -r; |
914 | |
915 | if (result != NULL((void *)0)) |
916 | *result = r; |
917 | |
918 | /* |
919 | * check for trailing stuff |
920 | */ |
921 | while (isspace((uschar)*ep)) |
922 | ep++; |
923 | |
924 | if (no_trailing != NULL((void *)0)) |
925 | *no_trailing = (*ep == '\0'); |
926 | |
927 | /* return true if found the end, or trailing stuff is allowed */ |
928 | retval = *ep == '\0' || trailing_stuff_ok; |
929 | |
930 | return retval; |
931 | } |