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 <math.h> |
| 28 | #include <ctype.h> |
| 29 | #include <string.h> |
| 30 | #include <stdlib.h> |
| 31 | #include "awk.h" |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 32 | |
| 33 | #define FULLTAB 2 /* rehash when table gets this x full */ |
| 34 | #define GROWTAB 4 /* grow table by this factor */ |
| 35 | |
| 36 | Array *symtab; /* main symbol table */ |
| 37 | |
| 38 | char **FS; /* initial field sep */ |
| 39 | char **RS; /* initial record sep */ |
| 40 | char **OFS; /* output field sep */ |
| 41 | char **ORS; /* output record sep */ |
| 42 | char **OFMT; /* output format for numbers */ |
| 43 | char **CONVFMT; /* format for conversions in getsval */ |
| 44 | Awkfloat *NF; /* number of fields in current record */ |
| 45 | Awkfloat *NR; /* number of current record */ |
| 46 | Awkfloat *FNR; /* number of current record in current file */ |
| 47 | char **FILENAME; /* current filename argument */ |
| 48 | Awkfloat *ARGC; /* number of arguments from command line */ |
| 49 | char **SUBSEP; /* subscript separator for a[i,j,k]; default \034 */ |
| 50 | Awkfloat *RSTART; /* start of re matched with ~; origin 1 (!) */ |
| 51 | Awkfloat *RLENGTH; /* length of same */ |
| 52 | |
| 53 | Cell *fsloc; /* FS */ |
| 54 | Cell *nrloc; /* NR */ |
| 55 | Cell *nfloc; /* NF */ |
| 56 | Cell *fnrloc; /* FNR */ |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 57 | Cell *ofsloc; /* OFS */ |
| 58 | Cell *orsloc; /* ORS */ |
| 59 | Cell *rsloc; /* RS */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 60 | Array *ARGVtab; /* symbol table containing ARGV[...] */ |
| 61 | Array *ENVtab; /* symbol table containing ENVIRON[...] */ |
| 62 | Cell *rstartloc; /* RSTART */ |
| 63 | Cell *rlengthloc; /* RLENGTH */ |
Cody Peter Mello | 97a4b7e | 2018-09-17 11:59:04 -0700 | [diff] [blame] | 64 | Cell *subseploc; /* SUBSEP */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 65 | Cell *symtabloc; /* SYMTAB */ |
| 66 | |
| 67 | Cell *nullloc; /* a guaranteed empty cell */ |
| 68 | Node *nullnode; /* zero&null, converted into a node for comparisons */ |
| 69 | Cell *literal0; |
| 70 | |
| 71 | extern Cell **fldtab; |
| 72 | |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 73 | static void |
| 74 | setfree(Cell *vp) |
| 75 | { |
| 76 | if (&vp->sval == FS || &vp->sval == RS || |
| 77 | &vp->sval == OFS || &vp->sval == ORS || |
| 78 | &vp->sval == OFMT || &vp->sval == CONVFMT || |
| 79 | &vp->sval == FILENAME || &vp->sval == SUBSEP) |
| 80 | vp->tval |= DONTFREE; |
| 81 | else |
| 82 | vp->tval &= ~DONTFREE; |
| 83 | } |
| 84 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 85 | void syminit(void) /* initialize symbol table with builtin vars */ |
| 86 | { |
| 87 | literal0 = setsymtab("0", "0", 0.0, NUM|STR|CON|DONTFREE, symtab); |
| 88 | /* this is used for if(x)... tests: */ |
| 89 | nullloc = setsymtab("$zero&null", "", 0.0, NUM|STR|CON|DONTFREE, symtab); |
| 90 | nullnode = celltonode(nullloc, CCON); |
| 91 | |
| 92 | fsloc = setsymtab("FS", " ", 0.0, STR|DONTFREE, symtab); |
| 93 | FS = &fsloc->sval; |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 94 | rsloc = setsymtab("RS", "\n", 0.0, STR|DONTFREE, symtab); |
| 95 | RS = &rsloc->sval; |
| 96 | ofsloc = setsymtab("OFS", " ", 0.0, STR|DONTFREE, symtab); |
| 97 | OFS = &ofsloc->sval; |
| 98 | orsloc = setsymtab("ORS", "\n", 0.0, STR|DONTFREE, symtab); |
| 99 | ORS = &orsloc->sval; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 100 | OFMT = &setsymtab("OFMT", "%.6g", 0.0, STR|DONTFREE, symtab)->sval; |
| 101 | CONVFMT = &setsymtab("CONVFMT", "%.6g", 0.0, STR|DONTFREE, symtab)->sval; |
| 102 | FILENAME = &setsymtab("FILENAME", "", 0.0, STR|DONTFREE, symtab)->sval; |
| 103 | nfloc = setsymtab("NF", "", 0.0, NUM, symtab); |
| 104 | NF = &nfloc->fval; |
| 105 | nrloc = setsymtab("NR", "", 0.0, NUM, symtab); |
| 106 | NR = &nrloc->fval; |
| 107 | fnrloc = setsymtab("FNR", "", 0.0, NUM, symtab); |
| 108 | FNR = &fnrloc->fval; |
Cody Peter Mello | 97a4b7e | 2018-09-17 11:59:04 -0700 | [diff] [blame] | 109 | subseploc = setsymtab("SUBSEP", "\034", 0.0, STR|DONTFREE, symtab); |
| 110 | SUBSEP = &subseploc->sval; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 111 | rstartloc = setsymtab("RSTART", "", 0.0, NUM, symtab); |
| 112 | RSTART = &rstartloc->fval; |
| 113 | rlengthloc = setsymtab("RLENGTH", "", 0.0, NUM, symtab); |
| 114 | RLENGTH = &rlengthloc->fval; |
| 115 | symtabloc = setsymtab("SYMTAB", "", 0.0, ARR, symtab); |
Arnold D. Robbins | 3358f3f | 2020-01-01 22:42:20 +0200 | [diff] [blame] | 116 | free(symtabloc->sval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 117 | symtabloc->sval = (char *) symtab; |
| 118 | } |
| 119 | |
| 120 | void arginit(int ac, char **av) /* set up ARGV and ARGC */ |
| 121 | { |
| 122 | Cell *cp; |
| 123 | int i; |
| 124 | char temp[50]; |
| 125 | |
| 126 | ARGC = &setsymtab("ARGC", "", (Awkfloat) ac, NUM, symtab)->fval; |
| 127 | cp = setsymtab("ARGV", "", 0.0, ARR, symtab); |
| 128 | ARGVtab = makesymtab(NSYMTAB); /* could be (int) ARGC as well */ |
Arnold D. Robbins | 3358f3f | 2020-01-01 22:42:20 +0200 | [diff] [blame] | 129 | free(cp->sval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 130 | cp->sval = (char *) ARGVtab; |
| 131 | for (i = 0; i < ac; i++) { |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 132 | double result; |
| 133 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 134 | sprintf(temp, "%d", i); |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 135 | if (is_number(*av, & result)) |
| 136 | setsymtab(temp, *av, result, STR|NUM, ARGVtab); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 137 | else |
| 138 | setsymtab(temp, *av, 0.0, STR, ARGVtab); |
| 139 | av++; |
| 140 | } |
| 141 | } |
| 142 | |
| 143 | void envinit(char **envp) /* set up ENVIRON variable */ |
| 144 | { |
| 145 | Cell *cp; |
| 146 | char *p; |
| 147 | |
| 148 | cp = setsymtab("ENVIRON", "", 0.0, ARR, symtab); |
| 149 | ENVtab = makesymtab(NSYMTAB); |
Arnold D. Robbins | 3358f3f | 2020-01-01 22:42:20 +0200 | [diff] [blame] | 150 | free(cp->sval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 151 | cp->sval = (char *) ENVtab; |
| 152 | for ( ; *envp; envp++) { |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 153 | double result; |
| 154 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 155 | if ((p = strchr(*envp, '=')) == NULL) |
| 156 | continue; |
| 157 | if( p == *envp ) /* no left hand side name in env string */ |
| 158 | continue; |
| 159 | *p++ = 0; /* split into two strings at = */ |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 160 | if (is_number(p, & result)) |
| 161 | setsymtab(*envp, p, result, STR|NUM, ENVtab); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 162 | else |
| 163 | setsymtab(*envp, p, 0.0, STR, ENVtab); |
| 164 | p[-1] = '='; /* restore in case env is passed down to a shell */ |
| 165 | } |
| 166 | } |
| 167 | |
| 168 | Array *makesymtab(int n) /* make a new symbol table */ |
| 169 | { |
| 170 | Array *ap; |
| 171 | Cell **tp; |
| 172 | |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 173 | ap = (Array *) malloc(sizeof(*ap)); |
| 174 | tp = (Cell **) calloc(n, sizeof(*tp)); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 175 | if (ap == NULL || tp == NULL) |
| 176 | FATAL("out of space in makesymtab"); |
| 177 | ap->nelem = 0; |
| 178 | ap->size = n; |
| 179 | ap->tab = tp; |
| 180 | return(ap); |
| 181 | } |
| 182 | |
| 183 | void freesymtab(Cell *ap) /* free a symbol table */ |
| 184 | { |
| 185 | Cell *cp, *temp; |
| 186 | Array *tp; |
| 187 | int i; |
| 188 | |
| 189 | if (!isarr(ap)) |
| 190 | return; |
| 191 | tp = (Array *) ap->sval; |
| 192 | if (tp == NULL) |
| 193 | return; |
| 194 | for (i = 0; i < tp->size; i++) { |
| 195 | for (cp = tp->tab[i]; cp != NULL; cp = temp) { |
| 196 | xfree(cp->nval); |
| 197 | if (freeable(cp)) |
| 198 | xfree(cp->sval); |
| 199 | temp = cp->cnext; /* avoids freeing then using */ |
Arnold D. Robbins | 795a06b | 2019-07-28 05:51:52 -0600 | [diff] [blame] | 200 | free(cp); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 201 | tp->nelem--; |
| 202 | } |
pfg | 5242194 | 2016-06-03 21:23:11 +0000 | [diff] [blame] | 203 | tp->tab[i] = NULL; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 204 | } |
| 205 | if (tp->nelem != 0) |
| 206 | WARNING("can't happen: inconsistent element count freeing %s", ap->nval); |
| 207 | free(tp->tab); |
| 208 | free(tp); |
| 209 | } |
| 210 | |
| 211 | void freeelem(Cell *ap, const char *s) /* free elem s from ap (i.e., ap["s"] */ |
| 212 | { |
| 213 | Array *tp; |
| 214 | Cell *p, *prev = NULL; |
| 215 | int h; |
Arnold D. Robbins | 795a06b | 2019-07-28 05:51:52 -0600 | [diff] [blame] | 216 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 217 | tp = (Array *) ap->sval; |
| 218 | h = hash(s, tp->size); |
| 219 | for (p = tp->tab[h]; p != NULL; prev = p, p = p->cnext) |
| 220 | if (strcmp(s, p->nval) == 0) { |
| 221 | if (prev == NULL) /* 1st one */ |
| 222 | tp->tab[h] = p->cnext; |
| 223 | else /* middle somewhere */ |
| 224 | prev->cnext = p->cnext; |
| 225 | if (freeable(p)) |
| 226 | xfree(p->sval); |
| 227 | free(p->nval); |
| 228 | free(p); |
| 229 | tp->nelem--; |
| 230 | return; |
| 231 | } |
| 232 | } |
| 233 | |
| 234 | Cell *setsymtab(const char *n, const char *s, Awkfloat f, unsigned t, Array *tp) |
| 235 | { |
| 236 | int h; |
| 237 | Cell *p; |
| 238 | |
| 239 | if (n != NULL && (p = lookup(n, tp)) != NULL) { |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 240 | DPRINTF("setsymtab found %p: n=%s s=\"%s\" f=%g t=%o\n", |
| 241 | (void*)p, NN(p->nval), NN(p->sval), p->fval, p->tval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 242 | return(p); |
| 243 | } |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 244 | p = (Cell *) malloc(sizeof(*p)); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 245 | if (p == NULL) |
| 246 | FATAL("out of space for symbol table at %s", n); |
| 247 | p->nval = tostring(n); |
| 248 | p->sval = s ? tostring(s) : tostring(""); |
| 249 | p->fval = f; |
| 250 | p->tval = t; |
| 251 | p->csub = CUNK; |
| 252 | p->ctype = OCELL; |
| 253 | tp->nelem++; |
| 254 | if (tp->nelem > FULLTAB * tp->size) |
| 255 | rehash(tp); |
| 256 | h = hash(n, tp->size); |
| 257 | p->cnext = tp->tab[h]; |
| 258 | tp->tab[h] = p; |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 259 | DPRINTF("setsymtab set %p: n=%s s=\"%s\" f=%g t=%o\n", |
| 260 | (void*)p, p->nval, p->sval, p->fval, p->tval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 261 | return(p); |
| 262 | } |
| 263 | |
| 264 | int hash(const char *s, int n) /* form hash value for string s */ |
| 265 | { |
| 266 | unsigned hashval; |
| 267 | |
| 268 | for (hashval = 0; *s != '\0'; s++) |
| 269 | hashval = (*s + 31 * hashval); |
| 270 | return hashval % n; |
| 271 | } |
| 272 | |
| 273 | void rehash(Array *tp) /* rehash items in small table into big one */ |
| 274 | { |
| 275 | int i, nh, nsz; |
| 276 | Cell *cp, *op, **np; |
| 277 | |
| 278 | nsz = GROWTAB * tp->size; |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 279 | np = (Cell **) calloc(nsz, sizeof(*np)); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 280 | if (np == NULL) /* can't do it, but can keep running. */ |
| 281 | return; /* someone else will run out later. */ |
| 282 | for (i = 0; i < tp->size; i++) { |
| 283 | for (cp = tp->tab[i]; cp; cp = op) { |
| 284 | op = cp->cnext; |
| 285 | nh = hash(cp->nval, nsz); |
| 286 | cp->cnext = np[nh]; |
| 287 | np[nh] = cp; |
| 288 | } |
| 289 | } |
| 290 | free(tp->tab); |
| 291 | tp->tab = np; |
| 292 | tp->size = nsz; |
| 293 | } |
| 294 | |
| 295 | Cell *lookup(const char *s, Array *tp) /* look for s in tp */ |
| 296 | { |
| 297 | Cell *p; |
| 298 | int h; |
| 299 | |
| 300 | h = hash(s, tp->size); |
| 301 | for (p = tp->tab[h]; p != NULL; p = p->cnext) |
| 302 | if (strcmp(s, p->nval) == 0) |
| 303 | return(p); /* found it */ |
| 304 | return(NULL); /* not found */ |
| 305 | } |
| 306 | |
| 307 | Awkfloat setfval(Cell *vp, Awkfloat f) /* set float val of a Cell */ |
| 308 | { |
| 309 | int fldno; |
| 310 | |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 311 | f += 0.0; /* normalise negative zero to positive zero */ |
Arnold D. Robbins | 795a06b | 2019-07-28 05:51:52 -0600 | [diff] [blame] | 312 | if ((vp->tval & (NUM | STR)) == 0) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 313 | funnyvar(vp, "assign to"); |
| 314 | if (isfld(vp)) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 315 | donerec = false; /* mark $0 invalid */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 316 | fldno = atoi(vp->nval); |
| 317 | if (fldno > *NF) |
| 318 | newfld(fldno); |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 319 | DPRINTF("setting field %d to %g\n", fldno, f); |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 320 | } else if (&vp->fval == NF) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 321 | donerec = false; /* mark $0 invalid */ |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 322 | setlastfld(f); |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 323 | DPRINTF("setting NF to %g\n", f); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 324 | } else if (isrec(vp)) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 325 | donefld = false; /* mark $1... invalid */ |
| 326 | donerec = true; |
Cody Peter Mello | b463680 | 2018-10-19 15:07:53 -0700 | [diff] [blame] | 327 | savefs(); |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 328 | } else if (vp == ofsloc) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 329 | if (!donerec) |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 330 | recbld(); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 331 | } |
| 332 | if (freeable(vp)) |
| 333 | xfree(vp->sval); /* free any previous string */ |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 334 | vp->tval &= ~(STR|CONVC|CONVO); /* mark string invalid */ |
| 335 | vp->fmt = NULL; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 336 | vp->tval |= NUM; /* mark number ok */ |
| 337 | if (f == -0) /* who would have thought this possible? */ |
| 338 | f = 0; |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 339 | DPRINTF("setfval %p: %s = %g, t=%o\n", (void*)vp, NN(vp->nval), f, vp->tval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 340 | return vp->fval = f; |
| 341 | } |
| 342 | |
| 343 | void funnyvar(Cell *vp, const char *rw) |
| 344 | { |
| 345 | if (isarr(vp)) |
| 346 | FATAL("can't %s %s; it's an array name.", rw, vp->nval); |
| 347 | if (vp->tval & FCN) |
| 348 | FATAL("can't %s %s; it's a function.", rw, vp->nval); |
| 349 | WARNING("funny variable %p: n=%s s=\"%s\" f=%g t=%o", |
Arnold D. Robbins | 5068d20 | 2020-02-06 22:27:31 +0200 | [diff] [blame] | 350 | (void *)vp, vp->nval, vp->sval, vp->fval, vp->tval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 351 | } |
| 352 | |
| 353 | char *setsval(Cell *vp, const char *s) /* set string val of a Cell */ |
| 354 | { |
| 355 | char *t; |
| 356 | int fldno; |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 357 | Awkfloat f; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 358 | |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 359 | DPRINTF("starting setsval %p: %s = \"%s\", t=%o, r,f=%d,%d\n", |
| 360 | (void*)vp, NN(vp->nval), s, vp->tval, donerec, donefld); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 361 | if ((vp->tval & (NUM | STR)) == 0) |
| 362 | funnyvar(vp, "assign to"); |
| 363 | if (isfld(vp)) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 364 | donerec = false; /* mark $0 invalid */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 365 | fldno = atoi(vp->nval); |
| 366 | if (fldno > *NF) |
| 367 | newfld(fldno); |
Chris | b785141 | 2020-08-07 18:10:20 +0800 | [diff] [blame] | 368 | DPRINTF("setting field %d to %s (%p)\n", fldno, s, (const void*)s); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 369 | } else if (isrec(vp)) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 370 | donefld = false; /* mark $1... invalid */ |
| 371 | donerec = true; |
Cody Peter Mello | b463680 | 2018-10-19 15:07:53 -0700 | [diff] [blame] | 372 | savefs(); |
Cody Peter Mello | 52566c0 | 2018-09-18 15:45:55 -0700 | [diff] [blame] | 373 | } else if (vp == ofsloc) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 374 | if (!donerec) |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 375 | recbld(); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 376 | } |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 377 | t = s ? tostring(s) : tostring(""); /* in case it's self-assign */ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 378 | if (freeable(vp)) |
| 379 | xfree(vp->sval); |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 380 | vp->tval &= ~(NUM|CONVC|CONVO); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 381 | vp->tval |= STR; |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 382 | vp->fmt = NULL; |
| 383 | setfree(vp); |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 384 | DPRINTF("setsval %p: %s = \"%s (%p) \", t=%o r,f=%d,%d\n", |
Chris | b785141 | 2020-08-07 18:10:20 +0800 | [diff] [blame] | 385 | (void*)vp, NN(vp->nval), t, (void*)t, vp->tval, donerec, donefld); |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 386 | vp->sval = t; |
| 387 | if (&vp->fval == NF) { |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 388 | donerec = false; /* mark $0 invalid */ |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 389 | f = getfval(vp); |
| 390 | setlastfld(f); |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 391 | DPRINTF("setting NF to %g\n", f); |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 392 | } |
| 393 | |
| 394 | return(vp->sval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 395 | } |
| 396 | |
| 397 | Awkfloat getfval(Cell *vp) /* get float val of a Cell */ |
| 398 | { |
| 399 | if ((vp->tval & (NUM | STR)) == 0) |
| 400 | funnyvar(vp, "read value of"); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 401 | if (isfld(vp) && !donefld) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 402 | fldbld(); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 403 | else if (isrec(vp) && !donerec) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 404 | recbld(); |
| 405 | if (!isnum(vp)) { /* not a number */ |
Arnold Robbins | cc9e9b6 | 2020-12-08 08:05:22 +0200 | [diff] [blame] | 406 | double fval; |
| 407 | bool no_trailing; |
| 408 | |
| 409 | if (is_valid_number(vp->sval, true, & no_trailing, & fval)) { |
| 410 | vp->fval = fval; |
| 411 | if (no_trailing && !(vp->tval&CON)) |
| 412 | vp->tval |= NUM; /* make NUM only sparingly */ |
| 413 | } else |
| 414 | vp->fval = 0.0; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 415 | } |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 416 | DPRINTF("getfval %p: %s = %g, t=%o\n", |
| 417 | (void*)vp, NN(vp->nval), vp->fval, vp->tval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 418 | return(vp->fval); |
| 419 | } |
| 420 | |
Arnold D. Robbins | c0f4e97 | 2021-02-15 20:33:15 +0200 | [diff] [blame] | 421 | static const char *get_inf_nan(double d) |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 422 | { |
| 423 | if (isinf(d)) { |
| 424 | return (d < 0 ? "-inf" : "+inf"); |
| 425 | } else if (isnan(d)) { |
| 426 | return (signbit(d) != 0 ? "-nan" : "+nan"); |
| 427 | } else |
| 428 | return NULL; |
| 429 | } |
| 430 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 431 | static char *get_str_val(Cell *vp, char **fmt) /* get string val of a Cell */ |
| 432 | { |
Cody Peter Mello | e059b3b | 2018-09-14 19:56:34 -0700 | [diff] [blame] | 433 | char s[256]; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 434 | double dtemp; |
Arnold D. Robbins | c0f4e97 | 2021-02-15 20:33:15 +0200 | [diff] [blame] | 435 | const char *p; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 436 | |
| 437 | if ((vp->tval & (NUM | STR)) == 0) |
| 438 | funnyvar(vp, "read value of"); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 439 | if (isfld(vp) && ! donefld) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 440 | fldbld(); |
Arnold D. Robbins | 108224b | 2019-11-10 21:19:18 +0200 | [diff] [blame] | 441 | else if (isrec(vp) && ! donerec) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 442 | recbld(); |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 443 | |
| 444 | /* |
| 445 | * ADR: This is complicated and more fragile than is desirable. |
| 446 | * Retrieving a string value for a number associates the string |
| 447 | * value with the scalar. Previously, the string value was |
| 448 | * sticky, meaning if converted via OFMT that became the value |
| 449 | * (even though POSIX wants it to be via CONVFMT). Or if CONVFMT |
| 450 | * changed after a string value was retrieved, the original value |
| 451 | * was maintained and used. Also not per POSIX. |
| 452 | * |
| 453 | * We work around this design by adding two additional flags, |
| 454 | * CONVC and CONVO, indicating how the string value was |
| 455 | * obtained (via CONVFMT or OFMT) and _also_ maintaining a copy |
| 456 | * of the pointer to the xFMT format string used for the |
| 457 | * conversion. This pointer is only read, **never** dereferenced. |
| 458 | * The next time we do a conversion, if it's coming from the same |
| 459 | * xFMT as last time, and the pointer value is different, we |
| 460 | * know that the xFMT format string changed, and we need to |
| 461 | * redo the conversion. If it's the same, we don't have to. |
| 462 | * |
| 463 | * There are also several cases where we don't do a conversion, |
| 464 | * such as for a field (see the checks below). |
| 465 | */ |
| 466 | |
| 467 | /* Don't duplicate the code for actually updating the value */ |
| 468 | #define update_str_val(vp) \ |
| 469 | { \ |
| 470 | if (freeable(vp)) \ |
| 471 | xfree(vp->sval); \ |
Arnold D. Robbins | 8909e00 | 2020-12-18 11:57:48 +0200 | [diff] [blame] | 472 | if ((p = get_inf_nan(vp->fval)) != NULL) \ |
| 473 | strcpy(s, p); \ |
| 474 | else if (modf(vp->fval, &dtemp) == 0) /* it's integral */ \ |
Cody Peter Mello | e059b3b | 2018-09-14 19:56:34 -0700 | [diff] [blame] | 475 | snprintf(s, sizeof (s), "%.30g", vp->fval); \ |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 476 | else \ |
Cody Peter Mello | e059b3b | 2018-09-14 19:56:34 -0700 | [diff] [blame] | 477 | snprintf(s, sizeof (s), *fmt, vp->fval); \ |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 478 | vp->sval = tostring(s); \ |
| 479 | vp->tval &= ~DONTFREE; \ |
| 480 | vp->tval |= STR; \ |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 481 | } |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 482 | |
| 483 | if (isstr(vp) == 0) { |
| 484 | update_str_val(vp); |
| 485 | if (fmt == OFMT) { |
| 486 | vp->tval &= ~CONVC; |
| 487 | vp->tval |= CONVO; |
| 488 | } else { |
| 489 | /* CONVFMT */ |
| 490 | vp->tval &= ~CONVO; |
| 491 | vp->tval |= CONVC; |
| 492 | } |
| 493 | vp->fmt = *fmt; |
| 494 | } else if ((vp->tval & DONTFREE) != 0 || ! isnum(vp) || isfld(vp)) { |
| 495 | goto done; |
| 496 | } else if (isstr(vp)) { |
| 497 | if (fmt == OFMT) { |
| 498 | if ((vp->tval & CONVC) != 0 |
| 499 | || ((vp->tval & CONVO) != 0 && vp->fmt != *fmt)) { |
| 500 | update_str_val(vp); |
| 501 | vp->tval &= ~CONVC; |
| 502 | vp->tval |= CONVO; |
| 503 | vp->fmt = *fmt; |
| 504 | } |
| 505 | } else { |
| 506 | /* CONVFMT */ |
| 507 | if ((vp->tval & CONVO) != 0 |
| 508 | || ((vp->tval & CONVC) != 0 && vp->fmt != *fmt)) { |
| 509 | update_str_val(vp); |
| 510 | vp->tval &= ~CONVO; |
| 511 | vp->tval |= CONVC; |
| 512 | vp->fmt = *fmt; |
| 513 | } |
| 514 | } |
| 515 | } |
| 516 | done: |
Todd C. Miller | 292d39f | 2020-06-25 12:32:34 -0600 | [diff] [blame] | 517 | DPRINTF("getsval %p: %s = \"%s (%p)\", t=%o\n", |
Chris | b785141 | 2020-08-07 18:10:20 +0800 | [diff] [blame] | 518 | (void*)vp, NN(vp->nval), vp->sval, (void*)vp->sval, vp->tval); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 519 | return(vp->sval); |
| 520 | } |
| 521 | |
| 522 | char *getsval(Cell *vp) /* get string val of a Cell */ |
| 523 | { |
| 524 | return get_str_val(vp, CONVFMT); |
| 525 | } |
| 526 | |
| 527 | char *getpssval(Cell *vp) /* get string val of a Cell for print */ |
| 528 | { |
| 529 | return get_str_val(vp, OFMT); |
| 530 | } |
| 531 | |
| 532 | |
| 533 | char *tostring(const char *s) /* make a copy of string s */ |
| 534 | { |
zoulasc | 6589208 | 2019-10-24 09:40:15 -0400 | [diff] [blame] | 535 | char *p = strdup(s); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 536 | if (p == NULL) |
| 537 | FATAL("out of space in tostring on %s", s); |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 538 | return(p); |
| 539 | } |
| 540 | |
Arnold D. Robbins | 5068d20 | 2020-02-06 22:27:31 +0200 | [diff] [blame] | 541 | char *tostringN(const char *s, size_t n) /* make a copy of string s */ |
| 542 | { |
| 543 | char *p; |
| 544 | |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 545 | p = (char *) malloc(n); |
Arnold D. Robbins | 5068d20 | 2020-02-06 22:27:31 +0200 | [diff] [blame] | 546 | if (p == NULL) |
| 547 | FATAL("out of space in tostring on %s", s); |
| 548 | strcpy(p, s); |
| 549 | return(p); |
| 550 | } |
| 551 | |
Arnold D. Robbins | c95b960 | 2019-07-28 20:09:24 +0300 | [diff] [blame] | 552 | Cell *catstr(Cell *a, Cell *b) /* concatenate a and b */ |
| 553 | { |
| 554 | Cell *c; |
| 555 | char *p; |
| 556 | char *sa = getsval(a); |
| 557 | char *sb = getsval(b); |
| 558 | size_t l = strlen(sa) + strlen(sb) + 1; |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 559 | p = (char *) malloc(l); |
Arnold D. Robbins | c95b960 | 2019-07-28 20:09:24 +0300 | [diff] [blame] | 560 | if (p == NULL) |
| 561 | FATAL("out of space concatenating %s and %s", sa, sb); |
| 562 | snprintf(p, l, "%s%s", sa, sb); |
Arnold D. Robbins | 944989b | 2020-01-06 00:01:46 -0700 | [diff] [blame] | 563 | |
| 564 | l++; // add room for ' ' |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 565 | char *newbuf = (char *) malloc(l); |
Arnold D. Robbins | c7eeb57 | 2020-01-05 21:18:36 +0200 | [diff] [blame] | 566 | if (newbuf == NULL) |
| 567 | FATAL("out of space concatenating %s and %s", sa, sb); |
| 568 | // See string() in lex.c; a string "xx" is stored in the symbol |
| 569 | // table as "xx ". |
Arnold D. Robbins | 944989b | 2020-01-06 00:01:46 -0700 | [diff] [blame] | 570 | snprintf(newbuf, l, "%s ", p); |
Arnold D. Robbins | c7eeb57 | 2020-01-05 21:18:36 +0200 | [diff] [blame] | 571 | c = setsymtab(newbuf, p, 0.0, CON|STR|DONTFREE, symtab); |
Arnold D. Robbins | c95b960 | 2019-07-28 20:09:24 +0300 | [diff] [blame] | 572 | free(p); |
Arnold D. Robbins | c7eeb57 | 2020-01-05 21:18:36 +0200 | [diff] [blame] | 573 | free(newbuf); |
Arnold D. Robbins | c95b960 | 2019-07-28 20:09:24 +0300 | [diff] [blame] | 574 | return c; |
| 575 | } |
| 576 | |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 577 | char *qstring(const char *is, int delim) /* collect string up to next delim */ |
| 578 | { |
| 579 | const char *os = is; |
| 580 | int c, n; |
zoulasc | 6589208 | 2019-10-24 09:40:15 -0400 | [diff] [blame] | 581 | const uschar *s = (const uschar *) is; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 582 | uschar *buf, *bp; |
| 583 | |
Arnold D. Robbins | 3b42cfa | 2020-10-13 20:52:43 +0300 | [diff] [blame] | 584 | if ((buf = (uschar *) malloc(strlen(is)+3)) == NULL) |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 585 | FATAL( "out of space in qstring(%s)", s); |
| 586 | for (bp = buf; (c = *s) != delim; s++) { |
| 587 | if (c == '\n') |
| 588 | SYNTAX( "newline in string %.20s...", os ); |
| 589 | else if (c != '\\') |
| 590 | *bp++ = c; |
| 591 | else { /* \something */ |
| 592 | c = *++s; |
| 593 | if (c == 0) { /* \ at end */ |
| 594 | *bp++ = '\\'; |
| 595 | break; /* for loop */ |
Arnold D. Robbins | 795a06b | 2019-07-28 05:51:52 -0600 | [diff] [blame] | 596 | } |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 597 | switch (c) { |
| 598 | case '\\': *bp++ = '\\'; break; |
| 599 | case 'n': *bp++ = '\n'; break; |
| 600 | case 't': *bp++ = '\t'; break; |
| 601 | case 'b': *bp++ = '\b'; break; |
| 602 | case 'f': *bp++ = '\f'; break; |
| 603 | case 'r': *bp++ = '\r'; break; |
Martijn Dekker | 5b602ca | 2019-07-26 10:46:58 +0200 | [diff] [blame] | 604 | case 'v': *bp++ = '\v'; break; |
| 605 | case 'a': *bp++ = '\a'; break; |
Brian Kernighan | 87b9493 | 2012-12-22 10:35:39 -0500 | [diff] [blame] | 606 | default: |
| 607 | if (!isdigit(c)) { |
| 608 | *bp++ = c; |
| 609 | break; |
| 610 | } |
| 611 | n = c - '0'; |
| 612 | if (isdigit(s[1])) { |
| 613 | n = 8 * n + *++s - '0'; |
| 614 | if (isdigit(s[1])) |
| 615 | n = 8 * n + *++s - '0'; |
| 616 | } |
| 617 | *bp++ = n; |
| 618 | break; |
| 619 | } |
| 620 | } |
| 621 | } |
| 622 | *bp++ = 0; |
| 623 | return (char *) buf; |
| 624 | } |
Arnold D. Robbins | 32093f5 | 2018-08-22 20:40:26 +0300 | [diff] [blame] | 625 | |
| 626 | const char *flags2str(int flags) |
| 627 | { |
| 628 | static const struct ftab { |
| 629 | const char *name; |
| 630 | int value; |
| 631 | } flagtab[] = { |
| 632 | { "NUM", NUM }, |
| 633 | { "STR", STR }, |
| 634 | { "DONTFREE", DONTFREE }, |
| 635 | { "CON", CON }, |
| 636 | { "ARR", ARR }, |
| 637 | { "FCN", FCN }, |
| 638 | { "FLD", FLD }, |
| 639 | { "REC", REC }, |
| 640 | { "CONVC", CONVC }, |
| 641 | { "CONVO", CONVO }, |
| 642 | { NULL, 0 } |
| 643 | }; |
| 644 | static char buf[100]; |
| 645 | int i; |
| 646 | char *cp = buf; |
| 647 | |
| 648 | for (i = 0; flagtab[i].name != NULL; i++) { |
| 649 | if ((flags & flagtab[i].value) != 0) { |
| 650 | if (cp > buf) |
| 651 | *cp++ = '|'; |
| 652 | strcpy(cp, flagtab[i].name); |
| 653 | cp += strlen(cp); |
| 654 | } |
| 655 | } |
| 656 | |
| 657 | return buf; |
| 658 | } |