/[dtapublic]/projs/trunk/shared_source/c_tcl_base_7_5_w_mods/tclscan.c
ViewVC logotype

Contents of /projs/trunk/shared_source/c_tcl_base_7_5_w_mods/tclscan.c

Parent Directory Parent Directory | Revision Log Revision Log


Revision 66 - (show annotations) (download)
Sun Oct 30 21:57:38 2016 UTC (8 years, 1 month ago) by dashley
File MIME type: text/plain
File size: 26137 byte(s)
Header and footer cleanup.
1 /* $Header$ */
2 /*
3 * tclScan.c --
4 *
5 * This file contains the implementation of the "scan" command.
6 *
7 * Copyright (c) 1998 by Scriptics Corporation.
8 *
9 * See the file "license.terms" for information on usage and redistribution
10 * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
11 *
12 * RCS: @(#) $Id: tclscan.c,v 1.1.1.1 2001/06/13 04:46:00 dtashley Exp $
13 */
14
15 #include "tclInt.h"
16
17 /*
18 * Flag values used by Tcl_ScanObjCmd.
19 */
20
21 #define SCAN_NOSKIP 0x1 /* Don't skip blanks. */
22 #define SCAN_SUPPRESS 0x2 /* Suppress assignment. */
23 #define SCAN_UNSIGNED 0x4 /* Read an unsigned value. */
24 #define SCAN_WIDTH 0x8 /* A width value was supplied. */
25
26 #define SCAN_SIGNOK 0x10 /* A +/- character is allowed. */
27 #define SCAN_NODIGITS 0x20 /* No digits have been scanned. */
28 #define SCAN_NOZERO 0x40 /* No zero digits have been scanned. */
29 #define SCAN_XOK 0x80 /* An 'x' is allowed. */
30 #define SCAN_PTOK 0x100 /* Decimal point is allowed. */
31 #define SCAN_EXPOK 0x200 /* An exponent is allowed. */
32
33
34 /*
35 * The following structure contains the information associated with
36 * a character set.
37 */
38
39 typedef struct CharSet {
40 int exclude; /* 1 if this is an exclusion set. */
41 int nchars;
42 Tcl_UniChar *chars;
43 int nranges;
44 struct Range {
45 Tcl_UniChar start;
46 Tcl_UniChar end;
47 } *ranges;
48 } CharSet;
49
50 /*
51 * Declarations for functions used only in this file.
52 */
53
54 static char * BuildCharSet _ANSI_ARGS_((CharSet *cset, char *format));
55 static int CharInSet _ANSI_ARGS_((CharSet *cset, int ch));
56 static void ReleaseCharSet _ANSI_ARGS_((CharSet *cset));
57 static int ValidateFormat _ANSI_ARGS_((Tcl_Interp *interp, char *format,
58 int numVars, int *totalVars));
59
60 /*
61 *----------------------------------------------------------------------
62 *
63 * BuildCharSet --
64 *
65 * This function examines a character set format specification
66 * and builds a CharSet containing the individual characters and
67 * character ranges specified.
68 *
69 * Results:
70 * Returns the next format position.
71 *
72 * Side effects:
73 * Initializes the charset.
74 *
75 *----------------------------------------------------------------------
76 */
77
78 static char *
79 BuildCharSet(cset, format)
80 CharSet *cset;
81 char *format; /* Points to first char of set. */
82 {
83 Tcl_UniChar ch, start;
84 int offset, nranges;
85 char *end;
86
87 memset(cset, 0, sizeof(CharSet));
88
89 offset = Tcl_UtfToUniChar(format, &ch);
90 if (ch == '^') {
91 cset->exclude = 1;
92 format += offset;
93 offset = Tcl_UtfToUniChar(format, &ch);
94 }
95 end = format + offset;
96
97 /*
98 * Find the close bracket so we can overallocate the set.
99 */
100
101 if (ch == ']') {
102 end += Tcl_UtfToUniChar(end, &ch);
103 }
104 nranges = 0;
105 while (ch != ']') {
106 if (ch == '-') {
107 nranges++;
108 }
109 end += Tcl_UtfToUniChar(end, &ch);
110 }
111
112 cset->chars = (Tcl_UniChar *) ckalloc(sizeof(Tcl_UniChar)
113 * (end - format - 1));
114 if (nranges > 0) {
115 cset->ranges = (struct Range *) ckalloc(sizeof(struct Range)*nranges);
116 } else {
117 cset->ranges = NULL;
118 }
119
120 /*
121 * Now build the character set.
122 */
123
124 cset->nchars = cset->nranges = 0;
125 format += Tcl_UtfToUniChar(format, &ch);
126 start = ch;
127 if (ch == ']' || ch == '-') {
128 cset->chars[cset->nchars++] = ch;
129 format += Tcl_UtfToUniChar(format, &ch);
130 }
131 while (ch != ']') {
132 if (*format == '-') {
133 /*
134 * This may be the first character of a range, so don't add
135 * it yet.
136 */
137
138 start = ch;
139 } else if (ch == '-') {
140 /*
141 * Check to see if this is the last character in the set, in which
142 * case it is not a range and we should add the previous character
143 * as well as the dash.
144 */
145
146 if (*format == ']') {
147 cset->chars[cset->nchars++] = start;
148 cset->chars[cset->nchars++] = ch;
149 } else {
150 format += Tcl_UtfToUniChar(format, &ch);
151
152 /*
153 * Check to see if the range is in reverse order.
154 */
155
156 if (start < ch) {
157 cset->ranges[cset->nranges].start = start;
158 cset->ranges[cset->nranges].end = ch;
159 } else {
160 cset->ranges[cset->nranges].start = ch;
161 cset->ranges[cset->nranges].end = start;
162 }
163 cset->nranges++;
164 }
165 } else {
166 cset->chars[cset->nchars++] = ch;
167 }
168 format += Tcl_UtfToUniChar(format, &ch);
169 }
170 return format;
171 }
172
173 /*
174 *----------------------------------------------------------------------
175 *
176 * CharInSet --
177 *
178 * Check to see if a character matches the given set.
179 *
180 * Results:
181 * Returns non-zero if the character matches the given set.
182 *
183 * Side effects:
184 * None.
185 *
186 *----------------------------------------------------------------------
187 */
188
189 static int
190 CharInSet(cset, c)
191 CharSet *cset;
192 int c; /* Character to test, passed as int because
193 * of non-ANSI prototypes. */
194 {
195 Tcl_UniChar ch = (Tcl_UniChar) c;
196 int i, match = 0;
197 for (i = 0; i < cset->nchars; i++) {
198 if (cset->chars[i] == ch) {
199 match = 1;
200 break;
201 }
202 }
203 if (!match) {
204 for (i = 0; i < cset->nranges; i++) {
205 if ((cset->ranges[i].start <= ch)
206 && (ch <= cset->ranges[i].end)) {
207 match = 1;
208 break;
209 }
210 }
211 }
212 return (cset->exclude ? !match : match);
213 }
214
215 /*
216 *----------------------------------------------------------------------
217 *
218 * ReleaseCharSet --
219 *
220 * Free the storage associated with a character set.
221 *
222 * Results:
223 * None.
224 *
225 * Side effects:
226 * None.
227 *
228 *----------------------------------------------------------------------
229 */
230
231 static void
232 ReleaseCharSet(cset)
233 CharSet *cset;
234 {
235 ckfree((char *)cset->chars);
236 if (cset->ranges) {
237 ckfree((char *)cset->ranges);
238 }
239 }
240
241 /*
242 *----------------------------------------------------------------------
243 *
244 * ValidateFormat --
245 *
246 * Parse the format string and verify that it is properly formed
247 * and that there are exactly enough variables on the command line.
248 *
249 * Results:
250 * A standard Tcl result.
251 *
252 * Side effects:
253 * May place an error in the interpreter result.
254 *
255 *----------------------------------------------------------------------
256 */
257
258 static int
259 ValidateFormat(interp, format, numVars, totalSubs)
260 Tcl_Interp *interp; /* Current interpreter. */
261 char *format; /* The format string. */
262 int numVars; /* The number of variables passed to the
263 * scan command. */
264 int *totalSubs; /* The number of variables that will be
265 * required. */
266 {
267 #define STATIC_LIST_SIZE 16
268 int gotXpg, gotSequential, value, i, flags;
269 char *end;
270 Tcl_UniChar ch;
271 int staticAssign[STATIC_LIST_SIZE];
272 int *nassign = staticAssign;
273 int objIndex, xpgSize, nspace = STATIC_LIST_SIZE;
274
275 /*
276 * Initialize an array that records the number of times a variable
277 * is assigned to by the format string. We use this to detect if
278 * a variable is multiply assigned or left unassigned.
279 */
280
281 if (numVars > nspace) {
282 nassign = (int*)ckalloc(sizeof(int) * numVars);
283 nspace = numVars;
284 }
285 for (i = 0; i < nspace; i++) {
286 nassign[i] = 0;
287 }
288
289 xpgSize = objIndex = gotXpg = gotSequential = 0;
290
291 while (*format != '\0') {
292 format += Tcl_UtfToUniChar(format, &ch);
293
294 flags = 0;
295
296 if (ch != '%') {
297 continue;
298 }
299 format += Tcl_UtfToUniChar(format, &ch);
300 if (ch == '%') {
301 continue;
302 }
303 if (ch == '*') {
304 flags |= SCAN_SUPPRESS;
305 format += Tcl_UtfToUniChar(format, &ch);
306 goto xpgCheckDone;
307 }
308
309 if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
310 /*
311 * Check for an XPG3-style %n$ specification. Note: there
312 * must not be a mixture of XPG3 specs and non-XPG3 specs
313 * in the same format string.
314 */
315
316 value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */
317 if (*end != '$') {
318 goto notXpg;
319 }
320 format = end+1;
321 format += Tcl_UtfToUniChar(format, &ch);
322 gotXpg = 1;
323 if (gotSequential) {
324 goto mixedXPG;
325 }
326 objIndex = value - 1;
327 if ((objIndex < 0) || (numVars && (objIndex >= numVars))) {
328 goto badIndex;
329 } else if (numVars == 0) {
330 /*
331 * In the case where no vars are specified, the user can
332 * specify %9999$ legally, so we have to consider special
333 * rules for growing the assign array. 'value' is
334 * guaranteed to be > 0.
335 */
336 xpgSize = (xpgSize > value) ? xpgSize : value;
337 }
338 goto xpgCheckDone;
339 }
340
341 notXpg:
342 gotSequential = 1;
343 if (gotXpg) {
344 mixedXPG:
345 Tcl_SetResult(interp,
346 "cannot mix \"%\" and \"%n$\" conversion specifiers",
347 TCL_STATIC);
348 goto error;
349 }
350
351 xpgCheckDone:
352 /*
353 * Parse any width specifier.
354 */
355
356 if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
357 value = strtoul(format-1, &format, 10); /* INTL: "C" locale. */
358 flags |= SCAN_WIDTH;
359 format += Tcl_UtfToUniChar(format, &ch);
360 }
361
362 /*
363 * Ignore size specifier.
364 */
365
366 if ((ch == 'l') || (ch == 'L') || (ch == 'h')) {
367 format += Tcl_UtfToUniChar(format, &ch);
368 }
369
370 if (!(flags & SCAN_SUPPRESS) && numVars && (objIndex >= numVars)) {
371 goto badIndex;
372 }
373
374 /*
375 * Handle the various field types.
376 */
377
378 switch (ch) {
379 case 'n':
380 case 'd':
381 case 'i':
382 case 'o':
383 case 'x':
384 case 'u':
385 case 'f':
386 case 'e':
387 case 'g':
388 case 's':
389 break;
390 case 'c':
391 if (flags & SCAN_WIDTH) {
392 Tcl_SetResult(interp, "field width may not be specified in %c conversion", TCL_STATIC);
393 goto error;
394 }
395 break;
396 case '[':
397 if (*format == '\0') {
398 goto badSet;
399 }
400 format += Tcl_UtfToUniChar(format, &ch);
401 if (ch == '^') {
402 if (*format == '\0') {
403 goto badSet;
404 }
405 format += Tcl_UtfToUniChar(format, &ch);
406 }
407 if (ch == ']') {
408 if (*format == '\0') {
409 goto badSet;
410 }
411 format += Tcl_UtfToUniChar(format, &ch);
412 }
413 while (ch != ']') {
414 if (*format == '\0') {
415 goto badSet;
416 }
417 format += Tcl_UtfToUniChar(format, &ch);
418 }
419 break;
420 badSet:
421 Tcl_SetResult(interp, "unmatched [ in format string",
422 TCL_STATIC);
423 goto error;
424 default:
425 {
426 char buf[TCL_UTF_MAX+1];
427
428 buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
429 Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
430 "bad scan conversion character \"", buf, "\"", NULL);
431 goto error;
432 }
433 }
434 if (!(flags & SCAN_SUPPRESS)) {
435 if (objIndex >= nspace) {
436 /*
437 * Expand the nassign buffer. If we are using XPG specifiers,
438 * make sure that we grow to a large enough size. xpgSize is
439 * guaranteed to be at least one larger than objIndex.
440 */
441 value = nspace;
442 if (xpgSize) {
443 nspace = xpgSize;
444 } else {
445 nspace += STATIC_LIST_SIZE;
446 }
447 if (nassign == staticAssign) {
448 nassign = (void *)ckalloc(nspace * sizeof(int));
449 for (i = 0; i < STATIC_LIST_SIZE; ++i) {
450 nassign[i] = staticAssign[i];
451 }
452 } else {
453 nassign = (void *)ckrealloc((void *)nassign,
454 nspace * sizeof(int));
455 }
456 for (i = value; i < nspace; i++) {
457 nassign[i] = 0;
458 }
459 }
460 nassign[objIndex]++;
461 objIndex++;
462 }
463 }
464
465 /*
466 * Verify that all of the variable were assigned exactly once.
467 */
468
469 if (numVars == 0) {
470 if (xpgSize) {
471 numVars = xpgSize;
472 } else {
473 numVars = objIndex;
474 }
475 }
476 if (totalSubs) {
477 *totalSubs = numVars;
478 }
479 for (i = 0; i < numVars; i++) {
480 if (nassign[i] > 1) {
481 Tcl_SetResult(interp, "variable is assigned by multiple \"%n$\" conversion specifiers", TCL_STATIC);
482 goto error;
483 } else if (!xpgSize && (nassign[i] == 0)) {
484 /*
485 * If the space is empty, and xpgSize is 0 (means XPG wasn't
486 * used, and/or numVars != 0), then too many vars were given
487 */
488 Tcl_SetResult(interp, "variable is not assigned by any conversion specifiers", TCL_STATIC);
489 goto error;
490 }
491 }
492
493 if (nassign != staticAssign) {
494 ckfree((char *)nassign);
495 }
496 return TCL_OK;
497
498 badIndex:
499 if (gotXpg) {
500 Tcl_SetResult(interp, "\"%n$\" argument index out of range",
501 TCL_STATIC);
502 } else {
503 Tcl_SetResult(interp,
504 "different numbers of variable names and field specifiers",
505 TCL_STATIC);
506 }
507
508 error:
509 if (nassign != staticAssign) {
510 ckfree((char *)nassign);
511 }
512 return TCL_ERROR;
513 #undef STATIC_LIST_SIZE
514 }
515
516 /*
517 *----------------------------------------------------------------------
518 *
519 * Tcl_ScanObjCmd --
520 *
521 * This procedure is invoked to process the "scan" Tcl command.
522 * See the user documentation for details on what it does.
523 *
524 * Results:
525 * A standard Tcl result.
526 *
527 * Side effects:
528 * See the user documentation.
529 *
530 *----------------------------------------------------------------------
531 */
532
533 /* ARGSUSED */
534 int
535 Tcl_ScanObjCmd(dummy, interp, objc, objv)
536 ClientData dummy; /* Not used. */
537 Tcl_Interp *interp; /* Current interpreter. */
538 int objc; /* Number of arguments. */
539 Tcl_Obj *CONST objv[]; /* Argument objects. */
540 {
541 char *format;
542 int numVars, nconversions, totalVars = -1;
543 int objIndex, offset, i, value, result, code;
544 char *string, *end, *baseString;
545 char op = 0;
546 int base = 0;
547 int underflow = 0;
548 size_t width;
549 long (*fn)() = NULL;
550 Tcl_UniChar ch, sch;
551 Tcl_Obj **objs = NULL, *objPtr = NULL;
552 int flags;
553 char buf[513]; /* Temporary buffer to hold scanned
554 * number strings before they are
555 * passed to strtoul. */
556
557 if (objc < 3) {
558 Tcl_WrongNumArgs(interp, 1, objv,
559 "string format ?varName varName ...?");
560 return TCL_ERROR;
561 }
562
563 format = Tcl_GetStringFromObj(objv[2], NULL);
564 numVars = objc-3;
565
566 /*
567 * Check for errors in the format string.
568 */
569
570 if (ValidateFormat(interp, format, numVars, &totalVars) == TCL_ERROR) {
571 return TCL_ERROR;
572 }
573
574 /*
575 * Allocate space for the result objects.
576 */
577
578 if (totalVars > 0) {
579 objs = (Tcl_Obj **) ckalloc(sizeof(Tcl_Obj*) * totalVars);
580 for (i = 0; i < totalVars; i++) {
581 objs[i] = NULL;
582 }
583 }
584
585 string = Tcl_GetStringFromObj(objv[1], NULL);
586 baseString = string;
587
588 /*
589 * Iterate over the format string filling in the result objects until
590 * we reach the end of input, the end of the format string, or there
591 * is a mismatch.
592 */
593
594 objIndex = 0;
595 nconversions = 0;
596 while (*format != '\0') {
597 format += Tcl_UtfToUniChar(format, &ch);
598
599 flags = 0;
600
601 /*
602 * If we see whitespace in the format, skip whitespace in the string.
603 */
604
605 if (Tcl_UniCharIsSpace(ch)) {
606 offset = Tcl_UtfToUniChar(string, &sch);
607 while (Tcl_UniCharIsSpace(sch)) {
608 if (*string == '\0') {
609 goto done;
610 }
611 string += offset;
612 offset = Tcl_UtfToUniChar(string, &sch);
613 }
614 continue;
615 }
616
617 if (ch != '%') {
618 literal:
619 if (*string == '\0') {
620 underflow = 1;
621 goto done;
622 }
623 string += Tcl_UtfToUniChar(string, &sch);
624 if (ch != sch) {
625 goto done;
626 }
627 continue;
628 }
629
630 format += Tcl_UtfToUniChar(format, &ch);
631 if (ch == '%') {
632 goto literal;
633 }
634
635 /*
636 * Check for assignment suppression ('*') or an XPG3-style
637 * assignment ('%n$').
638 */
639
640 if (ch == '*') {
641 flags |= SCAN_SUPPRESS;
642 format += Tcl_UtfToUniChar(format, &ch);
643 } else if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
644 value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */
645 if (*end == '$') {
646 format = end+1;
647 format += Tcl_UtfToUniChar(format, &ch);
648 objIndex = value - 1;
649 }
650 }
651
652 /*
653 * Parse any width specifier.
654 */
655
656 if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
657 width = strtoul(format-1, &format, 10); /* INTL: "C" locale. */
658 format += Tcl_UtfToUniChar(format, &ch);
659 } else {
660 width = 0;
661 }
662
663 /*
664 * Ignore size specifier.
665 */
666
667 if ((ch == 'l') || (ch == 'L') || (ch == 'h')) {
668 format += Tcl_UtfToUniChar(format, &ch);
669 }
670
671 /*
672 * Handle the various field types.
673 */
674
675 switch (ch) {
676 case 'n':
677 if (!(flags & SCAN_SUPPRESS)) {
678 objPtr = Tcl_NewIntObj(string - baseString);
679 Tcl_IncrRefCount(objPtr);
680 objs[objIndex++] = objPtr;
681 }
682 nconversions++;
683 continue;
684
685 case 'd':
686 op = 'i';
687 base = 10;
688 fn = (long (*)())strtol;
689 break;
690 case 'i':
691 op = 'i';
692 base = 0;
693 fn = (long (*)())strtol;
694 break;
695 case 'o':
696 op = 'i';
697 base = 8;
698 fn = (long (*)())strtol;
699 break;
700 case 'x':
701 op = 'i';
702 base = 16;
703 fn = (long (*)())strtol;
704 break;
705 case 'u':
706 op = 'i';
707 base = 10;
708 flags |= SCAN_UNSIGNED;
709 fn = (long (*)())strtoul;
710 break;
711
712 case 'f':
713 case 'e':
714 case 'g':
715 op = 'f';
716 break;
717
718 case 's':
719 op = 's';
720 break;
721
722 case 'c':
723 op = 'c';
724 flags |= SCAN_NOSKIP;
725 break;
726 case '[':
727 op = '[';
728 flags |= SCAN_NOSKIP;
729 break;
730 }
731
732 /*
733 * At this point, we will need additional characters from the
734 * string to proceed.
735 */
736
737 if (*string == '\0') {
738 underflow = 1;
739 goto done;
740 }
741
742 /*
743 * Skip any leading whitespace at the beginning of a field unless
744 * the format suppresses this behavior.
745 */
746
747 if (!(flags & SCAN_NOSKIP)) {
748 while (*string != '\0') {
749 offset = Tcl_UtfToUniChar(string, &sch);
750 if (!Tcl_UniCharIsSpace(sch)) {
751 break;
752 }
753 string += offset;
754 }
755 if (*string == '\0') {
756 underflow = 1;
757 goto done;
758 }
759 }
760
761 /*
762 * Perform the requested scanning operation.
763 */
764
765 switch (op) {
766 case 's':
767 /*
768 * Scan a string up to width characters or whitespace.
769 */
770
771 if (width == 0) {
772 width = (size_t) ~0;
773 }
774 end = string;
775 while (*end != '\0') {
776 offset = Tcl_UtfToUniChar(end, &sch);
777 if (Tcl_UniCharIsSpace(sch)) {
778 break;
779 }
780 end += offset;
781 if (--width == 0) {
782 break;
783 }
784 }
785 if (!(flags & SCAN_SUPPRESS)) {
786 objPtr = Tcl_NewStringObj(string, end-string);
787 Tcl_IncrRefCount(objPtr);
788 objs[objIndex++] = objPtr;
789 }
790 string = end;
791 break;
792
793 case '[': {
794 CharSet cset;
795
796 if (width == 0) {
797 width = (size_t) ~0;
798 }
799 end = string;
800
801 format = BuildCharSet(&cset, format);
802 while (*end != '\0') {
803 offset = Tcl_UtfToUniChar(end, &sch);
804 if (!CharInSet(&cset, (int)sch)) {
805 break;
806 }
807 end += offset;
808 if (--width == 0) {
809 break;
810 }
811 }
812 ReleaseCharSet(&cset);
813
814 if (string == end) {
815 /*
816 * Nothing matched the range, stop processing
817 */
818 goto done;
819 }
820 if (!(flags & SCAN_SUPPRESS)) {
821 objPtr = Tcl_NewStringObj(string, end-string);
822 Tcl_IncrRefCount(objPtr);
823 objs[objIndex++] = objPtr;
824 }
825 string = end;
826
827 break;
828 }
829 case 'c':
830 /*
831 * Scan a single Unicode character.
832 */
833
834 string += Tcl_UtfToUniChar(string, &sch);
835 if (!(flags & SCAN_SUPPRESS)) {
836 objPtr = Tcl_NewIntObj((int)sch);
837 Tcl_IncrRefCount(objPtr);
838 objs[objIndex++] = objPtr;
839 }
840 break;
841
842 case 'i':
843 /*
844 * Scan an unsigned or signed integer.
845 */
846
847 if ((width == 0) || (width > sizeof(buf) - 1)) {
848 width = sizeof(buf) - 1;
849 }
850 flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_NOZERO;
851 for (end = buf; width > 0; width--) {
852 switch (*string) {
853 /*
854 * The 0 digit has special meaning at the beginning of
855 * a number. If we are unsure of the base, it
856 * indicates that we are in base 8 or base 16 (if it is
857 * followed by an 'x').
858 */
859 case '0':
860 if (base == 0) {
861 base = 8;
862 flags |= SCAN_XOK;
863 }
864 if (flags & SCAN_NOZERO) {
865 flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS
866 | SCAN_NOZERO);
867 } else {
868 flags &= ~(SCAN_SIGNOK | SCAN_XOK
869 | SCAN_NODIGITS);
870 }
871 goto addToInt;
872
873 case '1': case '2': case '3': case '4':
874 case '5': case '6': case '7':
875 if (base == 0) {
876 base = 10;
877 }
878 flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
879 goto addToInt;
880
881 case '8': case '9':
882 if (base == 0) {
883 base = 10;
884 }
885 if (base <= 8) {
886 break;
887 }
888 flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
889 goto addToInt;
890
891 case 'A': case 'B': case 'C':
892 case 'D': case 'E': case 'F':
893 case 'a': case 'b': case 'c':
894 case 'd': case 'e': case 'f':
895 if (base <= 10) {
896 break;
897 }
898 flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
899 goto addToInt;
900
901 case '+': case '-':
902 if (flags & SCAN_SIGNOK) {
903 flags &= ~SCAN_SIGNOK;
904 goto addToInt;
905 }
906 break;
907
908 case 'x': case 'X':
909 if ((flags & SCAN_XOK) && (end == buf+1)) {
910 base = 16;
911 flags &= ~SCAN_XOK;
912 goto addToInt;
913 }
914 break;
915 }
916
917 /*
918 * We got an illegal character so we are done accumulating.
919 */
920
921 break;
922
923 addToInt:
924 /*
925 * Add the character to the temporary buffer.
926 */
927
928 *end++ = *string++;
929 if (*string == '\0') {
930 break;
931 }
932 }
933
934 /*
935 * Check to see if we need to back up because we only got a
936 * sign or a trailing x after a 0.
937 */
938
939 if (flags & SCAN_NODIGITS) {
940 if (*string == '\0') {
941 underflow = 1;
942 }
943 goto done;
944 } else if (end[-1] == 'x' || end[-1] == 'X') {
945 end--;
946 string--;
947 }
948
949
950 /*
951 * Scan the value from the temporary buffer. If we are
952 * returning a large unsigned value, we have to convert it back
953 * to a string since Tcl only supports signed values.
954 */
955
956 if (!(flags & SCAN_SUPPRESS)) {
957 *end = '\0';
958 value = (int) (*fn)(buf, NULL, base);
959 if ((flags & SCAN_UNSIGNED) && (value < 0)) {
960 sprintf(buf, "%u", value); /* INTL: ISO digit */
961 objPtr = Tcl_NewStringObj(buf, -1);
962 } else {
963 objPtr = Tcl_NewIntObj(value);
964 }
965 Tcl_IncrRefCount(objPtr);
966 objs[objIndex++] = objPtr;
967 }
968
969 break;
970
971 case 'f':
972 /*
973 * Scan a floating point number
974 */
975
976 if ((width == 0) || (width > sizeof(buf) - 1)) {
977 width = sizeof(buf) - 1;
978 }
979 flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_PTOK | SCAN_EXPOK;
980 for (end = buf; width > 0; width--) {
981 switch (*string) {
982 case '0': case '1': case '2': case '3':
983 case '4': case '5': case '6': case '7':
984 case '8': case '9':
985 flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS);
986 goto addToFloat;
987 case '+': case '-':
988 if (flags & SCAN_SIGNOK) {
989 flags &= ~SCAN_SIGNOK;
990 goto addToFloat;
991 }
992 break;
993 case '.':
994 if (flags & SCAN_PTOK) {
995 flags &= ~(SCAN_SIGNOK | SCAN_PTOK);
996 goto addToFloat;
997 }
998 break;
999 case 'e': case 'E':
1000 /*
1001 * An exponent is not allowed until there has
1002 * been at least one digit.
1003 */
1004
1005 if ((flags & (SCAN_NODIGITS | SCAN_EXPOK))
1006 == SCAN_EXPOK) {
1007 flags = (flags & ~(SCAN_EXPOK|SCAN_PTOK))
1008 | SCAN_SIGNOK | SCAN_NODIGITS;
1009 goto addToFloat;
1010 }
1011 break;
1012 }
1013
1014 /*
1015 * We got an illegal character so we are done accumulating.
1016 */
1017
1018 break;
1019
1020 addToFloat:
1021 /*
1022 * Add the character to the temporary buffer.
1023 */
1024
1025 *end++ = *string++;
1026 if (*string == '\0') {
1027 break;
1028 }
1029 }
1030
1031 /*
1032 * Check to see if we need to back up because we saw a
1033 * trailing 'e' or sign.
1034 */
1035
1036 if (flags & SCAN_NODIGITS) {
1037 if (flags & SCAN_EXPOK) {
1038 /*
1039 * There were no digits at all so scanning has
1040 * failed and we are done.
1041 */
1042 if (*string == '\0') {
1043 underflow = 1;
1044 }
1045 goto done;
1046 }
1047
1048 /*
1049 * We got a bad exponent ('e' and maybe a sign).
1050 */
1051
1052 end--;
1053 string--;
1054 if (*end != 'e' && *end != 'E') {
1055 end--;
1056 string--;
1057 }
1058 }
1059
1060 /*
1061 * Scan the value from the temporary buffer.
1062 */
1063
1064 if (!(flags & SCAN_SUPPRESS)) {
1065 double dvalue;
1066 *end = '\0';
1067 dvalue = strtod(buf, NULL);
1068 objPtr = Tcl_NewDoubleObj(dvalue);
1069 Tcl_IncrRefCount(objPtr);
1070 objs[objIndex++] = objPtr;
1071 }
1072 break;
1073 }
1074 nconversions++;
1075 }
1076
1077 done:
1078 result = 0;
1079 code = TCL_OK;
1080
1081 if (numVars) {
1082 /*
1083 * In this case, variables were specified (classic scan)
1084 */
1085 for (i = 0; i < totalVars; i++) {
1086 if (objs[i] != NULL) {
1087 result++;
1088 if (Tcl_ObjSetVar2(interp, objv[i+3], NULL,
1089 objs[i], 0) == NULL) {
1090 Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
1091 "couldn't set variable \"",
1092 Tcl_GetString(objv[i+3]), "\"", (char *) NULL);
1093 code = TCL_ERROR;
1094 }
1095 Tcl_DecrRefCount(objs[i]);
1096 }
1097 }
1098 } else {
1099 /*
1100 * Here no vars were specified, we want a list returned (inline scan)
1101 */
1102 objPtr = Tcl_NewObj();
1103 for (i = 0; i < totalVars; i++) {
1104 if (objs[i] != NULL) {
1105 Tcl_ListObjAppendElement(NULL, objPtr, objs[i]);
1106 Tcl_DecrRefCount(objs[i]);
1107 } else {
1108 /*
1109 * More %-specifiers than matching chars, so we
1110 * just spit out empty strings for these
1111 */
1112 Tcl_ListObjAppendElement(NULL, objPtr, Tcl_NewObj());
1113 }
1114 }
1115 }
1116 ckfree((char*) objs);
1117 if (code == TCL_OK) {
1118 if (underflow && (nconversions == 0)) {
1119 if (numVars) {
1120 objPtr = Tcl_NewIntObj(-1);
1121 } else {
1122 if (objPtr) {
1123 Tcl_SetListObj(objPtr, 0, NULL);
1124 } else {
1125 objPtr = Tcl_NewObj();
1126 }
1127 }
1128 } else if (numVars) {
1129 objPtr = Tcl_NewIntObj(result);
1130 }
1131 Tcl_SetObjResult(interp, objPtr);
1132 }
1133 return code;
1134 }
1135
1136 /* End of tclscan.c */

Properties

Name Value
svn:keywords Header

dashley@gmail.com
ViewVC Help
Powered by ViewVC 1.1.25