Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 1 | /**************************************************************** |
| 2 | Copyright (C) Lucent Technologies 1997 |
| 3 | All Rights Reserved |
| 4 | |
| 5 | Permission to use, copy, modify, and distribute this software and |
| 6 | its documentation for any purpose and without fee is hereby |
| 7 | granted, provided that the above copyright notice appear in all |
| 8 | copies and that both that the copyright notice and this |
| 9 | permission notice and warranty disclaimer appear in supporting |
| 10 | documentation, and that the name Lucent Technologies or any of |
| 11 | its entities not be used in advertising or publicity pertaining |
| 12 | to distribution of the software without specific, written prior |
| 13 | permission. |
| 14 | |
| 15 | LUCENT DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, |
| 16 | INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. |
| 17 | IN NO EVENT SHALL LUCENT OR ANY OF ITS ENTITIES BE LIABLE FOR ANY |
| 18 | SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES |
| 19 | WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER |
| 20 | IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, |
| 21 | ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF |
| 22 | THIS SOFTWARE. |
| 23 | ****************************************************************/ |
| 24 | |
| 25 | #define DEBUG |
| 26 | #include <stdio.h> |
| 27 | #include <string.h> |
Michael Forney | 38e525f | 2020-12-15 04:46:30 -0800 | [diff] [blame] | 28 | #include <strings.h> |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 29 | #include <ctype.h> |
| 30 | #include <errno.h> |
| 31 | #include <stdlib.h> |
| 32 | #include <stdarg.h> |
Michael Forney | 6932571 | 2020-01-30 22:23:34 -0800 | [diff] [blame] | 33 | #include <limits.h> |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 34 | #include <math.h> |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 35 | #include "awk.h" |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 36 | |
zoulasc | 6a87709 | 2020-01-24 04:11:59 -0500 | [diff] [blame] | 37 | char EMPTY[] = { '\0' }; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 38 | FILE *infile = NULL; |
zoulasc | ffee778 | 2020-02-28 06:23:54 -0500 | [diff] [blame] | 39 | bool innew; /* true = infile has not been read by readrec */ |
zoulasc | 6a87709 | 2020-01-24 04:11:59 -0500 | [diff] [blame] | 40 | char *file = EMPTY; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 41 | char *record; |
| 42 | int recsize = RECSIZE; |
| 43 | char *fields; |
| 44 | int fieldssize = RECSIZE; |
| 45 | |
| 46 | Cell **fldtab; /* pointers to Cells */ |
zoulasc | 6a87709 | 2020-01-24 04:11:59 -0500 | [diff] [blame] | 47 | static size_t len_inputFS = 0; |
| 48 | static char *inputFS = NULL; /* FS at time of input, for field splitting */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 49 | |
| 50 | #define MAXFLD 2 |
| 51 | int nfields = MAXFLD; /* last allocated slot for $i */ |
| 52 | |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 53 | bool donefld; /* true = implies rec broken into fields */ |
| 54 | bool donerec; /* true = record is valid (no flds have changed) */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 55 | |
| 56 | int lastfld = 0; /* last used field */ |
| 57 | int argno = 1; /* current input argument number */ |
| 58 | extern Awkfloat *ARGC; |
| 59 | |
zoulasc | c2c8ecb | 2020-02-19 13:44:49 -0500 | [diff] [blame] | 60 | static Cell dollar0 = { OCELL, CFLD, NULL, EMPTY, 0.0, REC|STR|DONTFREE, NULL, NULL }; |
| 61 | static Cell dollar1 = { OCELL, CFLD, NULL, EMPTY, 0.0, FLD|STR|DONTFREE, NULL, NULL }; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 62 | |
| 63 | void recinit(unsigned int n) |
| 64 | { |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 65 | if ( (record = (char *) malloc(n)) == NULL |
| 66 | || (fields = (char *) malloc(n+1)) == NULL |
| 67 | || (fldtab = (Cell **) calloc(nfields+2, sizeof(*fldtab))) == NULL |
| 68 | || (fldtab[0] = (Cell *) malloc(sizeof(**fldtab))) == NULL) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 69 | FATAL("out of space for $0 and fields"); |
M. Warner Losh | 0939e33 | 2019-06-02 15:17:29 -0600 | [diff] [blame] | 70 | *record = '\0'; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 71 | *fldtab[0] = dollar0; |
| 72 | fldtab[0]->sval = record; |
| 73 | fldtab[0]->nval = tostring("0"); |
| 74 | makefields(1, nfields); |
| 75 | } |
| 76 | |
| 77 | void makefields(int n1, int n2) /* create $n1..$n2 inclusive */ |
| 78 | { |
| 79 | char temp[50]; |
| 80 | int i; |
| 81 | |
| 82 | for (i = n1; i <= n2; i++) { |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 83 | fldtab[i] = (Cell *) malloc(sizeof(**fldtab)); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 84 | if (fldtab[i] == NULL) |
| 85 | FATAL("out of space in makefields %d", i); |
| 86 | *fldtab[i] = dollar1; |
zoulasc | 0d8778b | 2019-10-25 10:59:10 -0400 | [diff] [blame] | 87 | snprintf(temp, sizeof(temp), "%d", i); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 88 | fldtab[i]->nval = tostring(temp); |
| 89 | } |
| 90 | } |
| 91 | |
| 92 | void initgetrec(void) |
| 93 | { |
| 94 | int i; |
| 95 | char *p; |
| 96 | |
| 97 | for (i = 1; i < *ARGC; i++) { |
| 98 | p = getargv(i); /* find 1st real filename */ |
| 99 | if (p == NULL || *p == '\0') { /* deleted or zapped */ |
| 100 | argno++; |
| 101 | continue; |
| 102 | } |
| 103 | if (!isclvar(p)) { |
| 104 | setsval(lookup("FILENAME", symtab), p); |
| 105 | return; |
| 106 | } |
| 107 | setclvar(p); /* a commandline assignment before filename */ |
| 108 | argno++; |
| 109 | } |
| 110 | infile = stdin; /* no filenames, so use stdin */ |
zoulasc | ffee778 | 2020-02-28 06:23:54 -0500 | [diff] [blame] | 111 | innew = true; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 112 | } |
| 113 | |
Cody Peter Mello | b463680 | 2018-10-19 15:07:53 -0700 | [diff] [blame] | 114 | /* |
| 115 | * POSIX specifies that fields are supposed to be evaluated as if they were |
| 116 | * split using the value of FS at the time that the record's value ($0) was |
| 117 | * read. |
| 118 | * |
| 119 | * Since field-splitting is done lazily, we save the current value of FS |
| 120 | * whenever a new record is read in (implicitly or via getline), or when |
| 121 | * a new value is assigned to $0. |
| 122 | */ |
| 123 | void savefs(void) |
| 124 | { |
zoulasc | 6a87709 | 2020-01-24 04:11:59 -0500 | [diff] [blame] | 125 | size_t len; |
| 126 | if ((len = strlen(getsval(fsloc))) < len_inputFS) { |
| 127 | strcpy(inputFS, *FS); /* for subsequent field splitting */ |
| 128 | return; |
| 129 | } |
| 130 | |
| 131 | len_inputFS = len + 1; |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 132 | inputFS = (char *) realloc(inputFS, len_inputFS); |
zoulasc | 6a87709 | 2020-01-24 04:11:59 -0500 | [diff] [blame] | 133 | if (inputFS == NULL) |
Cody Peter Mello | b463680 | 2018-10-19 15:07:53 -0700 | [diff] [blame] | 134 | FATAL("field separator %.10s... is too long", *FS); |
zoulasc | 6a87709 | 2020-01-24 04:11:59 -0500 | [diff] [blame] | 135 | memcpy(inputFS, *FS, len_inputFS); |
Cody Peter Mello | b463680 | 2018-10-19 15:07:53 -0700 | [diff] [blame] | 136 | } |
| 137 | |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 138 | static bool firsttime = true; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 139 | |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 140 | int getrec(char **pbuf, int *pbufsize, bool isrecord) /* get next input record */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 141 | { /* note: cares whether buf == record */ |
| 142 | int c; |
| 143 | char *buf = *pbuf; |
| 144 | uschar saveb0; |
| 145 | int bufsize = *pbufsize, savebufsize = bufsize; |
| 146 | |
| 147 | if (firsttime) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 148 | firsttime = false; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 149 | initgetrec(); |
| 150 | } |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 151 | DPRINTF("RS=<%s>, FS=<%s>, ARGC=%g, FILENAME=%s\n", |
| 152 | *RS, *FS, *ARGC, *FILENAME); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 153 | if (isrecord) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 154 | donefld = false; |
| 155 | donerec = true; |
Cody Peter Mello | b463680 | 2018-10-19 15:07:53 -0700 | [diff] [blame] | 156 | savefs(); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 157 | } |
| 158 | saveb0 = buf[0]; |
| 159 | buf[0] = 0; |
| 160 | while (argno < *ARGC || infile == stdin) { |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 161 | DPRINTF("argno=%d, file=|%s|\n", argno, file); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 162 | if (infile == NULL) { /* have to open a new file */ |
| 163 | file = getargv(argno); |
| 164 | if (file == NULL || *file == '\0') { /* deleted or zapped */ |
| 165 | argno++; |
| 166 | continue; |
| 167 | } |
| 168 | if (isclvar(file)) { /* a var=value arg */ |
| 169 | setclvar(file); |
| 170 | argno++; |
| 171 | continue; |
| 172 | } |
| 173 | *FILENAME = file; |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 174 | DPRINTF("opening file %s\n", file); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 175 | if (*file == '-' && *(file+1) == '\0') |
| 176 | infile = stdin; |
| 177 | else if ((infile = fopen(file, "r")) == NULL) |
| 178 | FATAL("can't open file %s", file); |
| 179 | setfval(fnrloc, 0.0); |
| 180 | } |
zoulasc | ffee778 | 2020-02-28 06:23:54 -0500 | [diff] [blame] | 181 | c = readrec(&buf, &bufsize, infile, innew); |
| 182 | if (innew) |
| 183 | innew = false; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 184 | if (c != 0 || buf[0] != '\0') { /* normal record */ |
| 185 | if (isrecord) { |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 186 | double result; |
| 187 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 188 | if (freeable(fldtab[0])) |
| 189 | xfree(fldtab[0]->sval); |
| 190 | fldtab[0]->sval = buf; /* buf == record */ |
| 191 | fldtab[0]->tval = REC | STR | DONTFREE; |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 192 | if (is_number(fldtab[0]->sval, & result)) { |
| 193 | fldtab[0]->fval = result; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 194 | fldtab[0]->tval |= NUM; |
| 195 | } |
| 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) |
| 205 | fclose(infile); |
| 206 | infile = NULL; |
| 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 && infile != stdin) |
| 218 | fclose(infile); |
| 219 | infile = NULL; |
| 220 | argno++; |
| 221 | } |
| 222 | |
zoulasc | ffee778 | 2020-02-28 06:23:54 -0500 | [diff] [blame] | 223 | int readrec(char **pbuf, int *pbufsize, FILE *inf, bool newflag) /* read one record into buf */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 224 | { |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 225 | int sep, c, isrec; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 226 | char *rr, *buf = *pbuf; |
| 227 | int bufsize = *pbufsize; |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 228 | char *rs = getsval(rsloc); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 229 | |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 230 | if (*rs && rs[1]) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 231 | bool found; |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 232 | |
| 233 | fa *pfa = makedfa(rs, 1); |
zoulasc | ffee778 | 2020-02-28 06:23:54 -0500 | [diff] [blame] | 234 | if (newflag) |
| 235 | found = fnematch(pfa, inf, &buf, &bufsize, recsize); |
| 236 | else { |
| 237 | int tempstat = pfa->initstat; |
| 238 | pfa->initstat = 2; |
| 239 | found = fnematch(pfa, inf, &buf, &bufsize, recsize); |
| 240 | pfa->initstat = tempstat; |
| 241 | } |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 242 | if (found) |
zoulasc | 6589208 | 2019-10-24 09:40:15 -0400 | [diff] [blame] | 243 | setptr(patbeg, '\0'); |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 244 | } else { |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 245 | if ((sep = *rs) == 0) { |
| 246 | sep = '\n'; |
| 247 | while ((c=getc(inf)) == '\n' && c != EOF) /* skip leading \n's */ |
| 248 | ; |
| 249 | if (c != EOF) |
| 250 | ungetc(c, inf); |
| 251 | } |
| 252 | for (rr = buf; ; ) { |
| 253 | for (; (c=getc(inf)) != sep && c != EOF; ) { |
| 254 | if (rr-buf+1 > bufsize) |
| 255 | if (!adjbuf(&buf, &bufsize, 1+rr-buf, |
| 256 | recsize, &rr, "readrec 1")) |
| 257 | FATAL("input record `%.30s...' too long", buf); |
| 258 | *rr++ = c; |
| 259 | } |
| 260 | if (*rs == sep || c == EOF) |
| 261 | break; |
| 262 | if ((c = getc(inf)) == '\n' || c == EOF) /* 2 in a row */ |
| 263 | break; |
| 264 | if (!adjbuf(&buf, &bufsize, 2+rr-buf, recsize, &rr, |
| 265 | "readrec 2")) |
| 266 | FATAL("input record `%.30s...' too long", buf); |
| 267 | *rr++ = '\n'; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 268 | *rr++ = c; |
| 269 | } |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 270 | if (!adjbuf(&buf, &bufsize, 1+rr-buf, recsize, &rr, "readrec 3")) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 271 | FATAL("input record `%.30s...' too long", buf); |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 272 | *rr = 0; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 273 | } |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 274 | *pbuf = buf; |
| 275 | *pbufsize = bufsize; |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 276 | isrec = *buf || !feof(inf); |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 277 | DPRINTF("readrec saw <%s>, returns %d\n", buf, isrec); |
Arnold D. Robbins | 643a5a3 | 2019-07-28 20:12:05 +0300 | [diff] [blame] | 278 | return isrec; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 279 | } |
| 280 | |
| 281 | char *getargv(int n) /* get ARGV[n] */ |
| 282 | { |
| 283 | Cell *x; |
| 284 | char *s, temp[50]; |
| 285 | extern Array *ARGVtab; |
| 286 | |
zoulasc | 0d8778b | 2019-10-25 10:59:10 -0400 | [diff] [blame] | 287 | snprintf(temp, sizeof(temp), "%d", n); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 288 | if (lookup(temp, ARGVtab) == NULL) |
| 289 | return NULL; |
| 290 | x = setsymtab(temp, "", 0.0, STR, ARGVtab); |
| 291 | s = getsval(x); |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 292 | DPRINTF("getargv(%d) returns |%s|\n", n, s); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 293 | return s; |
| 294 | } |
| 295 | |
| 296 | void setclvar(char *s) /* set var=value from s */ |
| 297 | { |
| 298 | char *p; |
| 299 | Cell *q; |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 300 | double result; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 301 | |
| 302 | for (p=s; *p != '='; p++) |
| 303 | ; |
| 304 | *p++ = 0; |
| 305 | p = qstring(p, '\0'); |
| 306 | q = setsymtab(s, p, 0.0, STR, symtab); |
| 307 | setsval(q, p); |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 308 | if (is_number(q->sval, & result)) { |
| 309 | q->fval = result; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 310 | q->tval |= NUM; |
| 311 | } |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 312 | DPRINTF("command line set %s to |%s|\n", s, p); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 313 | } |
| 314 | |
| 315 | |
| 316 | void fldbld(void) /* create fields from current record */ |
| 317 | { |
| 318 | /* this relies on having fields[] the same length as $0 */ |
| 319 | /* the fields are all stored in this one array with \0's */ |
| 320 | /* possibly with a final trailing \0 not associated with any field */ |
| 321 | char *r, *fr, sep; |
| 322 | Cell *p; |
| 323 | int i, j, n; |
| 324 | |
| 325 | if (donefld) |
| 326 | return; |
| 327 | if (!isstr(fldtab[0])) |
| 328 | getsval(fldtab[0]); |
| 329 | r = fldtab[0]->sval; |
| 330 | n = strlen(r); |
| 331 | if (n > fieldssize) { |
| 332 | xfree(fields); |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 333 | if ((fields = (char *) malloc(n+2)) == NULL) /* possibly 2 final \0s */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 334 | FATAL("out of space for fields in fldbld %d", n); |
| 335 | fieldssize = n; |
| 336 | } |
| 337 | fr = fields; |
| 338 | i = 0; /* number of fields accumulated here */ |
Arnold D. Robbins | 754cf93 | 2020-06-05 12:25:15 +0300 | [diff] [blame] | 339 | if (inputFS == NULL) /* make sure we have a copy of FS */ |
| 340 | savefs(); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 341 | if (strlen(inputFS) > 1) { /* it's a regular expression */ |
| 342 | i = refldbld(r, inputFS); |
| 343 | } else if ((sep = *inputFS) == ' ') { /* default whitespace */ |
| 344 | for (i = 0; ; ) { |
| 345 | while (*r == ' ' || *r == '\t' || *r == '\n') |
| 346 | r++; |
| 347 | if (*r == 0) |
| 348 | break; |
| 349 | i++; |
| 350 | if (i > nfields) |
| 351 | growfldtab(i); |
| 352 | if (freeable(fldtab[i])) |
| 353 | xfree(fldtab[i]->sval); |
| 354 | fldtab[i]->sval = fr; |
| 355 | fldtab[i]->tval = FLD | STR | DONTFREE; |
| 356 | do |
| 357 | *fr++ = *r++; |
| 358 | while (*r != ' ' && *r != '\t' && *r != '\n' && *r != '\0'); |
| 359 | *fr++ = 0; |
| 360 | } |
| 361 | *fr = 0; |
| 362 | } else if ((sep = *inputFS) == 0) { /* new: FS="" => 1 char/field */ |
Arnold D. Robbins | c879fbf | 2019-11-08 14:40:18 +0200 | [diff] [blame] | 363 | for (i = 0; *r != '\0'; r += n) { |
Michael Forney | 6932571 | 2020-01-30 22:23:34 -0800 | [diff] [blame] | 364 | char buf[MB_LEN_MAX + 1]; |
Arnold D. Robbins | c879fbf | 2019-11-08 14:40:18 +0200 | [diff] [blame] | 365 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 366 | i++; |
| 367 | if (i > nfields) |
| 368 | growfldtab(i); |
| 369 | if (freeable(fldtab[i])) |
| 370 | xfree(fldtab[i]->sval); |
Michael Forney | 6932571 | 2020-01-30 22:23:34 -0800 | [diff] [blame] | 371 | n = mblen(r, MB_LEN_MAX); |
Arnold D. Robbins | c879fbf | 2019-11-08 14:40:18 +0200 | [diff] [blame] | 372 | if (n < 0) |
| 373 | n = 1; |
| 374 | memcpy(buf, r, n); |
| 375 | buf[n] = '\0'; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 376 | fldtab[i]->sval = tostring(buf); |
| 377 | fldtab[i]->tval = FLD | STR; |
| 378 | } |
| 379 | *fr = 0; |
| 380 | } else if (*r != 0) { /* if 0, it's a null field */ |
| 381 | /* subtlecase : if length(FS) == 1 && length(RS > 0) |
| 382 | * \n is NOT a field separator (cf awk book 61,84). |
| 383 | * this variable is tested in the inner while loop. |
| 384 | */ |
| 385 | int rtest = '\n'; /* normal case */ |
| 386 | if (strlen(*RS) > 0) |
| 387 | rtest = '\0'; |
| 388 | for (;;) { |
| 389 | i++; |
| 390 | if (i > nfields) |
| 391 | growfldtab(i); |
| 392 | if (freeable(fldtab[i])) |
| 393 | xfree(fldtab[i]->sval); |
| 394 | fldtab[i]->sval = fr; |
| 395 | fldtab[i]->tval = FLD | STR | DONTFREE; |
| 396 | while (*r != sep && *r != rtest && *r != '\0') /* \n is always a separator */ |
| 397 | *fr++ = *r++; |
| 398 | *fr++ = 0; |
| 399 | if (*r++ == 0) |
| 400 | break; |
| 401 | } |
| 402 | *fr = 0; |
| 403 | } |
| 404 | if (i > nfields) |
| 405 | FATAL("record `%.30s...' has too many fields; can't happen", r); |
| 406 | cleanfld(i+1, lastfld); /* clean out junk from previous record */ |
| 407 | lastfld = i; |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 408 | donefld = true; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 409 | for (j = 1; j <= lastfld; j++) { |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 410 | double result; |
| 411 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 412 | p = fldtab[j]; |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 413 | if(is_number(p->sval, & result)) { |
| 414 | p->fval = result; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 415 | p->tval |= NUM; |
| 416 | } |
| 417 | } |
| 418 | setfval(nfloc, (Awkfloat) lastfld); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 419 | donerec = true; /* restore */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 420 | if (dbg) { |
| 421 | for (j = 0; j <= lastfld; j++) { |
| 422 | p = fldtab[j]; |
| 423 | printf("field %d (%s): |%s|\n", j, p->nval, p->sval); |
| 424 | } |
| 425 | } |
| 426 | } |
| 427 | |
| 428 | void cleanfld(int n1, int n2) /* clean out fields n1 .. n2 inclusive */ |
| 429 | { /* nvals remain intact */ |
| 430 | Cell *p; |
| 431 | int i; |
| 432 | |
| 433 | for (i = n1; i <= n2; i++) { |
| 434 | p = fldtab[i]; |
| 435 | if (freeable(p)) |
| 436 | xfree(p->sval); |
zoulasc | 6a87709 | 2020-01-24 04:11:59 -0500 | [diff] [blame] | 437 | p->sval = EMPTY, |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 438 | p->tval = FLD | STR | DONTFREE; |
| 439 | } |
| 440 | } |
| 441 | |
| 442 | void newfld(int n) /* add field n after end of existing lastfld */ |
| 443 | { |
| 444 | if (n > nfields) |
| 445 | growfldtab(n); |
| 446 | cleanfld(lastfld+1, n); |
| 447 | lastfld = n; |
| 448 | setfval(nfloc, (Awkfloat) n); |
| 449 | } |
| 450 | |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 451 | void setlastfld(int n) /* set lastfld cleaning fldtab cells if necessary */ |
| 452 | { |
Cody Peter Mello | 179536a | 2018-09-25 21:19:49 -0700 | [diff] [blame] | 453 | if (n < 0) |
| 454 | FATAL("cannot set NF to a negative value"); |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 455 | if (n > nfields) |
| 456 | growfldtab(n); |
| 457 | |
| 458 | if (lastfld < n) |
| 459 | cleanfld(lastfld+1, n); |
| 460 | else |
| 461 | cleanfld(n+1, lastfld); |
| 462 | |
| 463 | lastfld = n; |
| 464 | } |
| 465 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 466 | Cell *fieldadr(int n) /* get nth field */ |
| 467 | { |
| 468 | if (n < 0) |
| 469 | FATAL("trying to access out of range field %d", n); |
| 470 | if (n > nfields) /* fields after NF are empty */ |
| 471 | growfldtab(n); /* but does not increase NF */ |
| 472 | return(fldtab[n]); |
| 473 | } |
| 474 | |
| 475 | void growfldtab(int n) /* make new fields up to at least $n */ |
| 476 | { |
| 477 | int nf = 2 * nfields; |
| 478 | size_t s; |
| 479 | |
| 480 | if (n > nf) |
| 481 | nf = n; |
| 482 | s = (nf+1) * (sizeof (struct Cell *)); /* freebsd: how much do we need? */ |
zoulasc | c2c8ecb | 2020-02-19 13:44:49 -0500 | [diff] [blame] | 483 | if (s / sizeof(struct Cell *) - 1 == (size_t)nf) /* didn't overflow */ |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 484 | fldtab = (Cell **) realloc(fldtab, s); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 485 | else /* overflow sizeof int */ |
| 486 | xfree(fldtab); /* make it null */ |
| 487 | if (fldtab == NULL) |
| 488 | FATAL("out of space creating %d fields", nf); |
| 489 | makefields(nfields+1, nf); |
| 490 | nfields = nf; |
| 491 | } |
| 492 | |
| 493 | int refldbld(const char *rec, const char *fs) /* build fields from reg expr in FS */ |
| 494 | { |
| 495 | /* this relies on having fields[] the same length as $0 */ |
| 496 | /* the fields are all stored in this one array with \0's */ |
| 497 | char *fr; |
| 498 | int i, tempstat, n; |
| 499 | fa *pfa; |
| 500 | |
| 501 | n = strlen(rec); |
| 502 | if (n > fieldssize) { |
| 503 | xfree(fields); |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 504 | if ((fields = (char *) malloc(n+1)) == NULL) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 505 | FATAL("out of space for fields in refldbld %d", n); |
| 506 | fieldssize = n; |
| 507 | } |
| 508 | fr = fields; |
| 509 | *fr = '\0'; |
| 510 | if (*rec == '\0') |
| 511 | return 0; |
| 512 | pfa = makedfa(fs, 1); |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 513 | DPRINTF("into refldbld, rec = <%s>, pat = <%s>\n", rec, fs); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 514 | tempstat = pfa->initstat; |
| 515 | for (i = 1; ; i++) { |
| 516 | if (i > nfields) |
| 517 | growfldtab(i); |
| 518 | if (freeable(fldtab[i])) |
| 519 | xfree(fldtab[i]->sval); |
| 520 | fldtab[i]->tval = FLD | STR | DONTFREE; |
| 521 | fldtab[i]->sval = fr; |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 522 | DPRINTF("refldbld: i=%d\n", i); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 523 | if (nematch(pfa, rec)) { |
| 524 | pfa->initstat = 2; /* horrible coupling to b.c */ |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 525 | DPRINTF("match %s (%d chars)\n", patbeg, patlen); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 526 | strncpy(fr, rec, patbeg-rec); |
| 527 | fr += patbeg - rec + 1; |
| 528 | *(fr-1) = '\0'; |
| 529 | rec = patbeg + patlen; |
| 530 | } else { |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 531 | DPRINTF("no match %s\n", rec); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 532 | strcpy(fr, rec); |
| 533 | pfa->initstat = tempstat; |
| 534 | break; |
| 535 | } |
| 536 | } |
Arnold D. Robbins | 795a06b | 2019-07-28 05:51:52 -0600 | [diff] [blame] | 537 | return i; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 538 | } |
| 539 | |
| 540 | void recbld(void) /* create $0 from $1..$NF if necessary */ |
| 541 | { |
| 542 | int i; |
| 543 | char *r, *p; |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 544 | char *sep = getsval(ofsloc); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 545 | |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 546 | if (donerec) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 547 | return; |
| 548 | r = record; |
| 549 | for (i = 1; i <= *NF; i++) { |
| 550 | p = getsval(fldtab[i]); |
| 551 | if (!adjbuf(&record, &recsize, 1+strlen(p)+r-record, recsize, &r, "recbld 1")) |
| 552 | FATAL("created $0 `%.30s...' too long", record); |
| 553 | while ((*r = *p++) != 0) |
| 554 | r++; |
| 555 | if (i < *NF) { |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 556 | if (!adjbuf(&record, &recsize, 2+strlen(sep)+r-record, recsize, &r, "recbld 2")) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 557 | FATAL("created $0 `%.30s...' too long", record); |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 558 | for (p = sep; (*r = *p++) != 0; ) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 559 | r++; |
| 560 | } |
| 561 | } |
| 562 | if (!adjbuf(&record, &recsize, 2+r-record, recsize, &r, "recbld 3")) |
| 563 | FATAL("built giant record `%.30s...'", record); |
| 564 | *r = '\0'; |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 565 | DPRINTF("in recbld inputFS=%s, fldtab[0]=%p\n", inputFS, (void*)fldtab[0]); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 566 | |
| 567 | if (freeable(fldtab[0])) |
| 568 | xfree(fldtab[0]->sval); |
| 569 | fldtab[0]->tval = REC | STR | DONTFREE; |
| 570 | fldtab[0]->sval = record; |
| 571 | |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 572 | DPRINTF("in recbld inputFS=%s, fldtab[0]=%p\n", inputFS, (void*)fldtab[0]); |
| 573 | DPRINTF("recbld = |%s|\n", record); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 574 | donerec = true; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 575 | } |
| 576 | |
| 577 | int errorflag = 0; |
| 578 | |
| 579 | void yyerror(const char *s) |
| 580 | { |
| 581 | SYNTAX("%s", s); |
| 582 | } |
| 583 | |
| 584 | void SYNTAX(const char *fmt, ...) |
| 585 | { |
| 586 | extern char *cmdname, *curfname; |
| 587 | static int been_here = 0; |
| 588 | va_list varg; |
| 589 | |
| 590 | if (been_here++ > 2) |
| 591 | return; |
| 592 | fprintf(stderr, "%s: ", cmdname); |
| 593 | va_start(varg, fmt); |
| 594 | vfprintf(stderr, fmt, varg); |
| 595 | va_end(varg); |
| 596 | fprintf(stderr, " at source line %d", lineno); |
| 597 | if (curfname != NULL) |
| 598 | fprintf(stderr, " in function %s", curfname); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 599 | if (compile_time == COMPILING && cursource() != NULL) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 600 | fprintf(stderr, " source file %s", cursource()); |
| 601 | fprintf(stderr, "\n"); |
| 602 | errorflag = 2; |
| 603 | eprint(); |
| 604 | } |
| 605 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 606 | extern int bracecnt, brackcnt, parencnt; |
| 607 | |
| 608 | void bracecheck(void) |
| 609 | { |
| 610 | int c; |
| 611 | static int beenhere = 0; |
| 612 | |
| 613 | if (beenhere++) |
| 614 | return; |
| 615 | while ((c = input()) != EOF && c != '\0') |
| 616 | bclass(c); |
| 617 | bcheck2(bracecnt, '{', '}'); |
| 618 | bcheck2(brackcnt, '[', ']'); |
| 619 | bcheck2(parencnt, '(', ')'); |
| 620 | } |
| 621 | |
| 622 | void bcheck2(int n, int c1, int c2) |
| 623 | { |
| 624 | if (n == 1) |
| 625 | fprintf(stderr, "\tmissing %c\n", c2); |
| 626 | else if (n > 1) |
| 627 | fprintf(stderr, "\t%d missing %c's\n", n, c2); |
| 628 | else if (n == -1) |
| 629 | fprintf(stderr, "\textra %c\n", c2); |
| 630 | else if (n < -1) |
| 631 | fprintf(stderr, "\t%d extra %c's\n", -n, c2); |
| 632 | } |
| 633 | |
| 634 | void FATAL(const char *fmt, ...) |
| 635 | { |
| 636 | extern char *cmdname; |
| 637 | va_list varg; |
| 638 | |
| 639 | fflush(stdout); |
| 640 | fprintf(stderr, "%s: ", cmdname); |
| 641 | va_start(varg, fmt); |
| 642 | vfprintf(stderr, fmt, varg); |
| 643 | va_end(varg); |
| 644 | error(); |
| 645 | if (dbg > 1) /* core dump if serious debugging on */ |
| 646 | abort(); |
| 647 | exit(2); |
| 648 | } |
| 649 | |
| 650 | void WARNING(const char *fmt, ...) |
| 651 | { |
| 652 | extern char *cmdname; |
| 653 | va_list varg; |
| 654 | |
| 655 | fflush(stdout); |
| 656 | fprintf(stderr, "%s: ", cmdname); |
| 657 | va_start(varg, fmt); |
| 658 | vfprintf(stderr, fmt, varg); |
| 659 | va_end(varg); |
| 660 | error(); |
| 661 | } |
| 662 | |
| 663 | void error() |
| 664 | { |
| 665 | extern Node *curnode; |
| 666 | |
| 667 | fprintf(stderr, "\n"); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 668 | if (compile_time != ERROR_PRINTING) { |
| 669 | if (NR && *NR > 0) { |
| 670 | fprintf(stderr, " input record number %d", (int) (*FNR)); |
| 671 | if (strcmp(*FILENAME, "-") != 0) |
| 672 | fprintf(stderr, ", file %s", *FILENAME); |
| 673 | fprintf(stderr, "\n"); |
| 674 | } |
| 675 | if (curnode) |
| 676 | fprintf(stderr, " source line number %d", curnode->lineno); |
| 677 | else if (lineno) |
| 678 | fprintf(stderr, " source line number %d", lineno); |
Todd C. Miller | feb247a | 2020-12-03 10:30:36 -0700 | [diff] [blame] | 679 | if (compile_time == COMPILING && cursource() != NULL) |
| 680 | fprintf(stderr, " source file %s", cursource()); |
| 681 | fprintf(stderr, "\n"); |
| 682 | eprint(); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 683 | } |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 684 | } |
| 685 | |
| 686 | void eprint(void) /* try to print context around error */ |
| 687 | { |
| 688 | char *p, *q; |
| 689 | int c; |
| 690 | static int been_here = 0; |
| 691 | extern char ebuf[], *ep; |
| 692 | |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 693 | if (compile_time != COMPILING || been_here++ > 0 || ebuf == ep) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 694 | return; |
Brian Kernighan | 3ed9e24 | 2018-08-15 10:45:03 -0400 | [diff] [blame] | 695 | if (ebuf == ep) |
| 696 | return; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 697 | p = ep - 1; |
| 698 | if (p > ebuf && *p == '\n') |
| 699 | p--; |
| 700 | for ( ; p > ebuf && *p != '\n' && *p != '\0'; p--) |
| 701 | ; |
| 702 | while (*p == '\n') |
| 703 | p++; |
| 704 | fprintf(stderr, " context is\n\t"); |
| 705 | for (q=ep-1; q>=p && *q!=' ' && *q!='\t' && *q!='\n'; q--) |
| 706 | ; |
| 707 | for ( ; p < q; p++) |
| 708 | if (*p) |
| 709 | putc(*p, stderr); |
| 710 | fprintf(stderr, " >>> "); |
| 711 | for ( ; p < ep; p++) |
| 712 | if (*p) |
| 713 | putc(*p, stderr); |
| 714 | fprintf(stderr, " <<< "); |
| 715 | if (*ep) |
| 716 | while ((c = input()) != '\n' && c != '\0' && c != EOF) { |
| 717 | putc(c, stderr); |
| 718 | bclass(c); |
| 719 | } |
| 720 | putc('\n', stderr); |
| 721 | ep = ebuf; |
| 722 | } |
| 723 | |
| 724 | void bclass(int c) |
| 725 | { |
| 726 | switch (c) { |
| 727 | case '{': bracecnt++; break; |
| 728 | case '}': bracecnt--; break; |
| 729 | case '[': brackcnt++; break; |
| 730 | case ']': brackcnt--; break; |
| 731 | case '(': parencnt++; break; |
| 732 | case ')': parencnt--; break; |
| 733 | } |
| 734 | } |
| 735 | |
| 736 | double errcheck(double x, const char *s) |
| 737 | { |
| 738 | |
| 739 | if (errno == EDOM) { |
| 740 | errno = 0; |
| 741 | WARNING("%s argument out of domain", s); |
| 742 | x = 1; |
| 743 | } else if (errno == ERANGE) { |
| 744 | errno = 0; |
| 745 | WARNING("%s result out of range", s); |
| 746 | x = 1; |
| 747 | } |
| 748 | return x; |
| 749 | } |
| 750 | |
| 751 | int isclvar(const char *s) /* is s of form var=something ? */ |
| 752 | { |
| 753 | const char *os = s; |
| 754 | |
| 755 | if (!isalpha((uschar) *s) && *s != '_') |
| 756 | return 0; |
| 757 | for ( ; *s; s++) |
| 758 | if (!(isalnum((uschar) *s) || *s == '_')) |
| 759 | break; |
Arnold D. Robbins | 4189ef5 | 2019-05-29 21:04:18 +0300 | [diff] [blame] | 760 | return *s == '=' && s > os; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 761 | } |
| 762 | |
| 763 | /* strtod is supposed to be a proper test of what's a valid number */ |
| 764 | /* appears to be broken in gcc on linux: thinks 0x123 is a valid FP number */ |
| 765 | /* wrong: violates 4.10.1.4 of ansi C standard */ |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 766 | |
Arnold D. Robbins | cef5180 | 2020-06-12 14:30:03 +0300 | [diff] [blame] | 767 | /* well, not quite. As of C99, hex floating point is allowed. so this is |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 768 | * a bit of a mess. We work around the mess by checking for a hexadecimal |
| 769 | * value and disallowing it. Similarly, we now follow gawk and allow only |
| 770 | * +nan, -nan, +inf, and -inf for NaN and infinity values. |
Arnold D. Robbins | cef5180 | 2020-06-12 14:30:03 +0300 | [diff] [blame] | 771 | */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 772 | |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 773 | /* |
| 774 | * This routine now has a more complicated interface, the main point |
| 775 | * being to avoid the double conversion of a string to double, and |
| 776 | * also to convey out, if requested, the information that the numeric |
| 777 | * value was a leading string or is all of the string. The latter bit |
| 778 | * is used in getfval(). |
| 779 | */ |
| 780 | |
| 781 | bool is_valid_number(const char *s, bool trailing_stuff_ok, |
| 782 | bool *no_trailing, double *result) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 783 | { |
| 784 | double r; |
| 785 | char *ep; |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 786 | bool retval = false; |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 787 | bool is_nan = false; |
| 788 | bool is_inf = false; |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 789 | |
| 790 | if (no_trailing) |
| 791 | *no_trailing = false; |
| 792 | |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 793 | while (isspace(*s)) |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 794 | s++; |
| 795 | |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 796 | // no hex floating point, sorry |
| 797 | if (s[0] == '0' && tolower(s[1]) == 'x') |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 798 | return false; |
| 799 | |
| 800 | // allow +nan, -nan, +inf, -inf, any other letter, no |
| 801 | if (s[0] == '+' || s[0] == '-') { |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 802 | is_nan = (strncasecmp(s+1, "nan", 3) == 0); |
| 803 | is_inf = (strncasecmp(s+1, "inf", 3) == 0); |
| 804 | if ((is_nan || is_inf) |
| 805 | && (isspace(s[4]) || s[4] == '\0')) |
| 806 | goto convert; |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 807 | else if (! isdigit(s[1]) && s[1] != '.') |
| 808 | return false; |
| 809 | } |
| 810 | else if (! isdigit(s[0]) && s[0] != '.') |
| 811 | return false; |
| 812 | |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 813 | convert: |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 814 | errno = 0; |
| 815 | r = strtod(s, &ep); |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 816 | if (ep == s || errno == ERANGE) |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 817 | return false; |
| 818 | |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 819 | if (isnan(r) && s[0] == '-' && signbit(r) == 0) |
| 820 | r = -r; |
| 821 | |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 822 | if (result != NULL) |
| 823 | *result = r; |
| 824 | |
ozan s. yigit | 1fd5fa3 | 2021-01-06 18:37:48 -0500 | [diff] [blame] | 825 | /* |
| 826 | * check for trailing stuff |
| 827 | */ |
| 828 | while (isspace(*ep)) |
| 829 | ep++; |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 830 | |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 831 | if (no_trailing != NULL) |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 832 | *no_trailing = (*ep == '\0'); |
| 833 | |
ozan s. yigit | 1fd5fa3 | 2021-01-06 18:37:48 -0500 | [diff] [blame] | 834 | // return true if found the end, or trailing stuff is allowed |
| 835 | retval = *ep == '\0' || trailing_stuff_ok; |
| 836 | |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 837 | return retval; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 838 | } |