|
sl@0
|
1 |
/*
|
|
sl@0
|
2 |
* tclScan.c --
|
|
sl@0
|
3 |
*
|
|
sl@0
|
4 |
* This file contains the implementation of the "scan" command.
|
|
sl@0
|
5 |
*
|
|
sl@0
|
6 |
* Copyright (c) 1998 by Scriptics Corporation.
|
|
sl@0
|
7 |
*
|
|
sl@0
|
8 |
* See the file "license.terms" for information on usage and redistribution
|
|
sl@0
|
9 |
* of this file, and for a DISCLAIMER OF ALL WARRANTIES.
|
|
sl@0
|
10 |
*
|
|
sl@0
|
11 |
* RCS: @(#) $Id: tclScan.c,v 1.12.2.2 2005/10/23 22:01:30 msofer Exp $
|
|
sl@0
|
12 |
*/
|
|
sl@0
|
13 |
|
|
sl@0
|
14 |
#include "tclInt.h"
|
|
sl@0
|
15 |
/*
|
|
sl@0
|
16 |
* For strtoll() and strtoull() declarations on some platforms...
|
|
sl@0
|
17 |
*/
|
|
sl@0
|
18 |
#include "tclPort.h"
|
|
sl@0
|
19 |
|
|
sl@0
|
20 |
/*
|
|
sl@0
|
21 |
* Flag values used by Tcl_ScanObjCmd.
|
|
sl@0
|
22 |
*/
|
|
sl@0
|
23 |
|
|
sl@0
|
24 |
#define SCAN_NOSKIP 0x1 /* Don't skip blanks. */
|
|
sl@0
|
25 |
#define SCAN_SUPPRESS 0x2 /* Suppress assignment. */
|
|
sl@0
|
26 |
#define SCAN_UNSIGNED 0x4 /* Read an unsigned value. */
|
|
sl@0
|
27 |
#define SCAN_WIDTH 0x8 /* A width value was supplied. */
|
|
sl@0
|
28 |
|
|
sl@0
|
29 |
#define SCAN_SIGNOK 0x10 /* A +/- character is allowed. */
|
|
sl@0
|
30 |
#define SCAN_NODIGITS 0x20 /* No digits have been scanned. */
|
|
sl@0
|
31 |
#define SCAN_NOZERO 0x40 /* No zero digits have been scanned. */
|
|
sl@0
|
32 |
#define SCAN_XOK 0x80 /* An 'x' is allowed. */
|
|
sl@0
|
33 |
#define SCAN_PTOK 0x100 /* Decimal point is allowed. */
|
|
sl@0
|
34 |
#define SCAN_EXPOK 0x200 /* An exponent is allowed. */
|
|
sl@0
|
35 |
|
|
sl@0
|
36 |
#define SCAN_LONGER 0x400 /* Asked for a wide value. */
|
|
sl@0
|
37 |
|
|
sl@0
|
38 |
/*
|
|
sl@0
|
39 |
* The following structure contains the information associated with
|
|
sl@0
|
40 |
* a character set.
|
|
sl@0
|
41 |
*/
|
|
sl@0
|
42 |
|
|
sl@0
|
43 |
typedef struct CharSet {
|
|
sl@0
|
44 |
int exclude; /* 1 if this is an exclusion set. */
|
|
sl@0
|
45 |
int nchars;
|
|
sl@0
|
46 |
Tcl_UniChar *chars;
|
|
sl@0
|
47 |
int nranges;
|
|
sl@0
|
48 |
struct Range {
|
|
sl@0
|
49 |
Tcl_UniChar start;
|
|
sl@0
|
50 |
Tcl_UniChar end;
|
|
sl@0
|
51 |
} *ranges;
|
|
sl@0
|
52 |
} CharSet;
|
|
sl@0
|
53 |
|
|
sl@0
|
54 |
/*
|
|
sl@0
|
55 |
* Declarations for functions used only in this file.
|
|
sl@0
|
56 |
*/
|
|
sl@0
|
57 |
|
|
sl@0
|
58 |
static char * BuildCharSet _ANSI_ARGS_((CharSet *cset, char *format));
|
|
sl@0
|
59 |
static int CharInSet _ANSI_ARGS_((CharSet *cset, int ch));
|
|
sl@0
|
60 |
static void ReleaseCharSet _ANSI_ARGS_((CharSet *cset));
|
|
sl@0
|
61 |
static int ValidateFormat _ANSI_ARGS_((Tcl_Interp *interp, char *format,
|
|
sl@0
|
62 |
int numVars, int *totalVars));
|
|
sl@0
|
63 |
|
|
sl@0
|
64 |
/*
|
|
sl@0
|
65 |
*----------------------------------------------------------------------
|
|
sl@0
|
66 |
*
|
|
sl@0
|
67 |
* BuildCharSet --
|
|
sl@0
|
68 |
*
|
|
sl@0
|
69 |
* This function examines a character set format specification
|
|
sl@0
|
70 |
* and builds a CharSet containing the individual characters and
|
|
sl@0
|
71 |
* character ranges specified.
|
|
sl@0
|
72 |
*
|
|
sl@0
|
73 |
* Results:
|
|
sl@0
|
74 |
* Returns the next format position.
|
|
sl@0
|
75 |
*
|
|
sl@0
|
76 |
* Side effects:
|
|
sl@0
|
77 |
* Initializes the charset.
|
|
sl@0
|
78 |
*
|
|
sl@0
|
79 |
*----------------------------------------------------------------------
|
|
sl@0
|
80 |
*/
|
|
sl@0
|
81 |
|
|
sl@0
|
82 |
static char *
|
|
sl@0
|
83 |
BuildCharSet(cset, format)
|
|
sl@0
|
84 |
CharSet *cset;
|
|
sl@0
|
85 |
char *format; /* Points to first char of set. */
|
|
sl@0
|
86 |
{
|
|
sl@0
|
87 |
Tcl_UniChar ch, start;
|
|
sl@0
|
88 |
int offset, nranges;
|
|
sl@0
|
89 |
char *end;
|
|
sl@0
|
90 |
|
|
sl@0
|
91 |
memset(cset, 0, sizeof(CharSet));
|
|
sl@0
|
92 |
|
|
sl@0
|
93 |
offset = Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
94 |
if (ch == '^') {
|
|
sl@0
|
95 |
cset->exclude = 1;
|
|
sl@0
|
96 |
format += offset;
|
|
sl@0
|
97 |
offset = Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
98 |
}
|
|
sl@0
|
99 |
end = format + offset;
|
|
sl@0
|
100 |
|
|
sl@0
|
101 |
/*
|
|
sl@0
|
102 |
* Find the close bracket so we can overallocate the set.
|
|
sl@0
|
103 |
*/
|
|
sl@0
|
104 |
|
|
sl@0
|
105 |
if (ch == ']') {
|
|
sl@0
|
106 |
end += Tcl_UtfToUniChar(end, &ch);
|
|
sl@0
|
107 |
}
|
|
sl@0
|
108 |
nranges = 0;
|
|
sl@0
|
109 |
while (ch != ']') {
|
|
sl@0
|
110 |
if (ch == '-') {
|
|
sl@0
|
111 |
nranges++;
|
|
sl@0
|
112 |
}
|
|
sl@0
|
113 |
end += Tcl_UtfToUniChar(end, &ch);
|
|
sl@0
|
114 |
}
|
|
sl@0
|
115 |
|
|
sl@0
|
116 |
cset->chars = (Tcl_UniChar *) ckalloc(sizeof(Tcl_UniChar)
|
|
sl@0
|
117 |
* (end - format - 1));
|
|
sl@0
|
118 |
if (nranges > 0) {
|
|
sl@0
|
119 |
cset->ranges = (struct Range *) ckalloc(sizeof(struct Range)*nranges);
|
|
sl@0
|
120 |
} else {
|
|
sl@0
|
121 |
cset->ranges = NULL;
|
|
sl@0
|
122 |
}
|
|
sl@0
|
123 |
|
|
sl@0
|
124 |
/*
|
|
sl@0
|
125 |
* Now build the character set.
|
|
sl@0
|
126 |
*/
|
|
sl@0
|
127 |
|
|
sl@0
|
128 |
cset->nchars = cset->nranges = 0;
|
|
sl@0
|
129 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
130 |
start = ch;
|
|
sl@0
|
131 |
if (ch == ']' || ch == '-') {
|
|
sl@0
|
132 |
cset->chars[cset->nchars++] = ch;
|
|
sl@0
|
133 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
134 |
}
|
|
sl@0
|
135 |
while (ch != ']') {
|
|
sl@0
|
136 |
if (*format == '-') {
|
|
sl@0
|
137 |
/*
|
|
sl@0
|
138 |
* This may be the first character of a range, so don't add
|
|
sl@0
|
139 |
* it yet.
|
|
sl@0
|
140 |
*/
|
|
sl@0
|
141 |
|
|
sl@0
|
142 |
start = ch;
|
|
sl@0
|
143 |
} else if (ch == '-') {
|
|
sl@0
|
144 |
/*
|
|
sl@0
|
145 |
* Check to see if this is the last character in the set, in which
|
|
sl@0
|
146 |
* case it is not a range and we should add the previous character
|
|
sl@0
|
147 |
* as well as the dash.
|
|
sl@0
|
148 |
*/
|
|
sl@0
|
149 |
|
|
sl@0
|
150 |
if (*format == ']') {
|
|
sl@0
|
151 |
cset->chars[cset->nchars++] = start;
|
|
sl@0
|
152 |
cset->chars[cset->nchars++] = ch;
|
|
sl@0
|
153 |
} else {
|
|
sl@0
|
154 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
155 |
|
|
sl@0
|
156 |
/*
|
|
sl@0
|
157 |
* Check to see if the range is in reverse order.
|
|
sl@0
|
158 |
*/
|
|
sl@0
|
159 |
|
|
sl@0
|
160 |
if (start < ch) {
|
|
sl@0
|
161 |
cset->ranges[cset->nranges].start = start;
|
|
sl@0
|
162 |
cset->ranges[cset->nranges].end = ch;
|
|
sl@0
|
163 |
} else {
|
|
sl@0
|
164 |
cset->ranges[cset->nranges].start = ch;
|
|
sl@0
|
165 |
cset->ranges[cset->nranges].end = start;
|
|
sl@0
|
166 |
}
|
|
sl@0
|
167 |
cset->nranges++;
|
|
sl@0
|
168 |
}
|
|
sl@0
|
169 |
} else {
|
|
sl@0
|
170 |
cset->chars[cset->nchars++] = ch;
|
|
sl@0
|
171 |
}
|
|
sl@0
|
172 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
173 |
}
|
|
sl@0
|
174 |
return format;
|
|
sl@0
|
175 |
}
|
|
sl@0
|
176 |
|
|
sl@0
|
177 |
/*
|
|
sl@0
|
178 |
*----------------------------------------------------------------------
|
|
sl@0
|
179 |
*
|
|
sl@0
|
180 |
* CharInSet --
|
|
sl@0
|
181 |
*
|
|
sl@0
|
182 |
* Check to see if a character matches the given set.
|
|
sl@0
|
183 |
*
|
|
sl@0
|
184 |
* Results:
|
|
sl@0
|
185 |
* Returns non-zero if the character matches the given set.
|
|
sl@0
|
186 |
*
|
|
sl@0
|
187 |
* Side effects:
|
|
sl@0
|
188 |
* None.
|
|
sl@0
|
189 |
*
|
|
sl@0
|
190 |
*----------------------------------------------------------------------
|
|
sl@0
|
191 |
*/
|
|
sl@0
|
192 |
|
|
sl@0
|
193 |
static int
|
|
sl@0
|
194 |
CharInSet(cset, c)
|
|
sl@0
|
195 |
CharSet *cset;
|
|
sl@0
|
196 |
int c; /* Character to test, passed as int because
|
|
sl@0
|
197 |
* of non-ANSI prototypes. */
|
|
sl@0
|
198 |
{
|
|
sl@0
|
199 |
Tcl_UniChar ch = (Tcl_UniChar) c;
|
|
sl@0
|
200 |
int i, match = 0;
|
|
sl@0
|
201 |
for (i = 0; i < cset->nchars; i++) {
|
|
sl@0
|
202 |
if (cset->chars[i] == ch) {
|
|
sl@0
|
203 |
match = 1;
|
|
sl@0
|
204 |
break;
|
|
sl@0
|
205 |
}
|
|
sl@0
|
206 |
}
|
|
sl@0
|
207 |
if (!match) {
|
|
sl@0
|
208 |
for (i = 0; i < cset->nranges; i++) {
|
|
sl@0
|
209 |
if ((cset->ranges[i].start <= ch)
|
|
sl@0
|
210 |
&& (ch <= cset->ranges[i].end)) {
|
|
sl@0
|
211 |
match = 1;
|
|
sl@0
|
212 |
break;
|
|
sl@0
|
213 |
}
|
|
sl@0
|
214 |
}
|
|
sl@0
|
215 |
}
|
|
sl@0
|
216 |
return (cset->exclude ? !match : match);
|
|
sl@0
|
217 |
}
|
|
sl@0
|
218 |
|
|
sl@0
|
219 |
/*
|
|
sl@0
|
220 |
*----------------------------------------------------------------------
|
|
sl@0
|
221 |
*
|
|
sl@0
|
222 |
* ReleaseCharSet --
|
|
sl@0
|
223 |
*
|
|
sl@0
|
224 |
* Free the storage associated with a character set.
|
|
sl@0
|
225 |
*
|
|
sl@0
|
226 |
* Results:
|
|
sl@0
|
227 |
* None.
|
|
sl@0
|
228 |
*
|
|
sl@0
|
229 |
* Side effects:
|
|
sl@0
|
230 |
* None.
|
|
sl@0
|
231 |
*
|
|
sl@0
|
232 |
*----------------------------------------------------------------------
|
|
sl@0
|
233 |
*/
|
|
sl@0
|
234 |
|
|
sl@0
|
235 |
static void
|
|
sl@0
|
236 |
ReleaseCharSet(cset)
|
|
sl@0
|
237 |
CharSet *cset;
|
|
sl@0
|
238 |
{
|
|
sl@0
|
239 |
ckfree((char *)cset->chars);
|
|
sl@0
|
240 |
if (cset->ranges) {
|
|
sl@0
|
241 |
ckfree((char *)cset->ranges);
|
|
sl@0
|
242 |
}
|
|
sl@0
|
243 |
}
|
|
sl@0
|
244 |
|
|
sl@0
|
245 |
/*
|
|
sl@0
|
246 |
*----------------------------------------------------------------------
|
|
sl@0
|
247 |
*
|
|
sl@0
|
248 |
* ValidateFormat --
|
|
sl@0
|
249 |
*
|
|
sl@0
|
250 |
* Parse the format string and verify that it is properly formed
|
|
sl@0
|
251 |
* and that there are exactly enough variables on the command line.
|
|
sl@0
|
252 |
*
|
|
sl@0
|
253 |
* Results:
|
|
sl@0
|
254 |
* A standard Tcl result.
|
|
sl@0
|
255 |
*
|
|
sl@0
|
256 |
* Side effects:
|
|
sl@0
|
257 |
* May place an error in the interpreter result.
|
|
sl@0
|
258 |
*
|
|
sl@0
|
259 |
*----------------------------------------------------------------------
|
|
sl@0
|
260 |
*/
|
|
sl@0
|
261 |
|
|
sl@0
|
262 |
static int
|
|
sl@0
|
263 |
ValidateFormat(interp, format, numVars, totalSubs)
|
|
sl@0
|
264 |
Tcl_Interp *interp; /* Current interpreter. */
|
|
sl@0
|
265 |
char *format; /* The format string. */
|
|
sl@0
|
266 |
int numVars; /* The number of variables passed to the
|
|
sl@0
|
267 |
* scan command. */
|
|
sl@0
|
268 |
int *totalSubs; /* The number of variables that will be
|
|
sl@0
|
269 |
* required. */
|
|
sl@0
|
270 |
{
|
|
sl@0
|
271 |
#define STATIC_LIST_SIZE 16
|
|
sl@0
|
272 |
int gotXpg, gotSequential, value, i, flags;
|
|
sl@0
|
273 |
char *end;
|
|
sl@0
|
274 |
Tcl_UniChar ch;
|
|
sl@0
|
275 |
int staticAssign[STATIC_LIST_SIZE];
|
|
sl@0
|
276 |
int *nassign = staticAssign;
|
|
sl@0
|
277 |
int objIndex, xpgSize, nspace = STATIC_LIST_SIZE;
|
|
sl@0
|
278 |
char buf[TCL_UTF_MAX+1];
|
|
sl@0
|
279 |
|
|
sl@0
|
280 |
/*
|
|
sl@0
|
281 |
* Initialize an array that records the number of times a variable
|
|
sl@0
|
282 |
* is assigned to by the format string. We use this to detect if
|
|
sl@0
|
283 |
* a variable is multiply assigned or left unassigned.
|
|
sl@0
|
284 |
*/
|
|
sl@0
|
285 |
|
|
sl@0
|
286 |
if (numVars > nspace) {
|
|
sl@0
|
287 |
nassign = (int*)ckalloc(sizeof(int) * numVars);
|
|
sl@0
|
288 |
nspace = numVars;
|
|
sl@0
|
289 |
}
|
|
sl@0
|
290 |
for (i = 0; i < nspace; i++) {
|
|
sl@0
|
291 |
nassign[i] = 0;
|
|
sl@0
|
292 |
}
|
|
sl@0
|
293 |
|
|
sl@0
|
294 |
xpgSize = objIndex = gotXpg = gotSequential = 0;
|
|
sl@0
|
295 |
|
|
sl@0
|
296 |
while (*format != '\0') {
|
|
sl@0
|
297 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
298 |
|
|
sl@0
|
299 |
flags = 0;
|
|
sl@0
|
300 |
|
|
sl@0
|
301 |
if (ch != '%') {
|
|
sl@0
|
302 |
continue;
|
|
sl@0
|
303 |
}
|
|
sl@0
|
304 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
305 |
if (ch == '%') {
|
|
sl@0
|
306 |
continue;
|
|
sl@0
|
307 |
}
|
|
sl@0
|
308 |
if (ch == '*') {
|
|
sl@0
|
309 |
flags |= SCAN_SUPPRESS;
|
|
sl@0
|
310 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
311 |
goto xpgCheckDone;
|
|
sl@0
|
312 |
}
|
|
sl@0
|
313 |
|
|
sl@0
|
314 |
if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
|
|
sl@0
|
315 |
/*
|
|
sl@0
|
316 |
* Check for an XPG3-style %n$ specification. Note: there
|
|
sl@0
|
317 |
* must not be a mixture of XPG3 specs and non-XPG3 specs
|
|
sl@0
|
318 |
* in the same format string.
|
|
sl@0
|
319 |
*/
|
|
sl@0
|
320 |
|
|
sl@0
|
321 |
value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */
|
|
sl@0
|
322 |
if (*end != '$') {
|
|
sl@0
|
323 |
goto notXpg;
|
|
sl@0
|
324 |
}
|
|
sl@0
|
325 |
format = end+1;
|
|
sl@0
|
326 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
327 |
gotXpg = 1;
|
|
sl@0
|
328 |
if (gotSequential) {
|
|
sl@0
|
329 |
goto mixedXPG;
|
|
sl@0
|
330 |
}
|
|
sl@0
|
331 |
objIndex = value - 1;
|
|
sl@0
|
332 |
if ((objIndex < 0) || (numVars && (objIndex >= numVars))) {
|
|
sl@0
|
333 |
goto badIndex;
|
|
sl@0
|
334 |
} else if (numVars == 0) {
|
|
sl@0
|
335 |
/*
|
|
sl@0
|
336 |
* In the case where no vars are specified, the user can
|
|
sl@0
|
337 |
* specify %9999$ legally, so we have to consider special
|
|
sl@0
|
338 |
* rules for growing the assign array. 'value' is
|
|
sl@0
|
339 |
* guaranteed to be > 0.
|
|
sl@0
|
340 |
*/
|
|
sl@0
|
341 |
xpgSize = (xpgSize > value) ? xpgSize : value;
|
|
sl@0
|
342 |
}
|
|
sl@0
|
343 |
goto xpgCheckDone;
|
|
sl@0
|
344 |
}
|
|
sl@0
|
345 |
|
|
sl@0
|
346 |
notXpg:
|
|
sl@0
|
347 |
gotSequential = 1;
|
|
sl@0
|
348 |
if (gotXpg) {
|
|
sl@0
|
349 |
mixedXPG:
|
|
sl@0
|
350 |
Tcl_SetResult(interp,
|
|
sl@0
|
351 |
"cannot mix \"%\" and \"%n$\" conversion specifiers",
|
|
sl@0
|
352 |
TCL_STATIC);
|
|
sl@0
|
353 |
goto error;
|
|
sl@0
|
354 |
}
|
|
sl@0
|
355 |
|
|
sl@0
|
356 |
xpgCheckDone:
|
|
sl@0
|
357 |
/*
|
|
sl@0
|
358 |
* Parse any width specifier.
|
|
sl@0
|
359 |
*/
|
|
sl@0
|
360 |
|
|
sl@0
|
361 |
if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
|
|
sl@0
|
362 |
value = strtoul(format-1, &format, 10); /* INTL: "C" locale. */
|
|
sl@0
|
363 |
flags |= SCAN_WIDTH;
|
|
sl@0
|
364 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
365 |
}
|
|
sl@0
|
366 |
|
|
sl@0
|
367 |
/*
|
|
sl@0
|
368 |
* Handle any size specifier.
|
|
sl@0
|
369 |
*/
|
|
sl@0
|
370 |
|
|
sl@0
|
371 |
switch (ch) {
|
|
sl@0
|
372 |
case 'l':
|
|
sl@0
|
373 |
case 'L':
|
|
sl@0
|
374 |
flags |= SCAN_LONGER;
|
|
sl@0
|
375 |
case 'h':
|
|
sl@0
|
376 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
377 |
}
|
|
sl@0
|
378 |
|
|
sl@0
|
379 |
if (!(flags & SCAN_SUPPRESS) && numVars && (objIndex >= numVars)) {
|
|
sl@0
|
380 |
goto badIndex;
|
|
sl@0
|
381 |
}
|
|
sl@0
|
382 |
|
|
sl@0
|
383 |
/*
|
|
sl@0
|
384 |
* Handle the various field types.
|
|
sl@0
|
385 |
*/
|
|
sl@0
|
386 |
|
|
sl@0
|
387 |
switch (ch) {
|
|
sl@0
|
388 |
case 'c':
|
|
sl@0
|
389 |
if (flags & SCAN_WIDTH) {
|
|
sl@0
|
390 |
Tcl_SetResult(interp,
|
|
sl@0
|
391 |
"field width may not be specified in %c conversion",
|
|
sl@0
|
392 |
TCL_STATIC);
|
|
sl@0
|
393 |
goto error;
|
|
sl@0
|
394 |
}
|
|
sl@0
|
395 |
/*
|
|
sl@0
|
396 |
* Fall through!
|
|
sl@0
|
397 |
*/
|
|
sl@0
|
398 |
case 'n':
|
|
sl@0
|
399 |
case 's':
|
|
sl@0
|
400 |
if (flags & SCAN_LONGER) {
|
|
sl@0
|
401 |
invalidLonger:
|
|
sl@0
|
402 |
buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
|
|
sl@0
|
403 |
Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
|
|
sl@0
|
404 |
"'l' modifier may not be specified in %", buf,
|
|
sl@0
|
405 |
" conversion", NULL);
|
|
sl@0
|
406 |
goto error;
|
|
sl@0
|
407 |
}
|
|
sl@0
|
408 |
/*
|
|
sl@0
|
409 |
* Fall through!
|
|
sl@0
|
410 |
*/
|
|
sl@0
|
411 |
case 'd':
|
|
sl@0
|
412 |
case 'e':
|
|
sl@0
|
413 |
case 'f':
|
|
sl@0
|
414 |
case 'g':
|
|
sl@0
|
415 |
case 'i':
|
|
sl@0
|
416 |
case 'o':
|
|
sl@0
|
417 |
case 'u':
|
|
sl@0
|
418 |
case 'x':
|
|
sl@0
|
419 |
break;
|
|
sl@0
|
420 |
/*
|
|
sl@0
|
421 |
* Bracket terms need special checking
|
|
sl@0
|
422 |
*/
|
|
sl@0
|
423 |
case '[':
|
|
sl@0
|
424 |
if (flags & SCAN_LONGER) {
|
|
sl@0
|
425 |
goto invalidLonger;
|
|
sl@0
|
426 |
}
|
|
sl@0
|
427 |
if (*format == '\0') {
|
|
sl@0
|
428 |
goto badSet;
|
|
sl@0
|
429 |
}
|
|
sl@0
|
430 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
431 |
if (ch == '^') {
|
|
sl@0
|
432 |
if (*format == '\0') {
|
|
sl@0
|
433 |
goto badSet;
|
|
sl@0
|
434 |
}
|
|
sl@0
|
435 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
436 |
}
|
|
sl@0
|
437 |
if (ch == ']') {
|
|
sl@0
|
438 |
if (*format == '\0') {
|
|
sl@0
|
439 |
goto badSet;
|
|
sl@0
|
440 |
}
|
|
sl@0
|
441 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
442 |
}
|
|
sl@0
|
443 |
while (ch != ']') {
|
|
sl@0
|
444 |
if (*format == '\0') {
|
|
sl@0
|
445 |
goto badSet;
|
|
sl@0
|
446 |
}
|
|
sl@0
|
447 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
448 |
}
|
|
sl@0
|
449 |
break;
|
|
sl@0
|
450 |
badSet:
|
|
sl@0
|
451 |
Tcl_SetResult(interp, "unmatched [ in format string",
|
|
sl@0
|
452 |
TCL_STATIC);
|
|
sl@0
|
453 |
goto error;
|
|
sl@0
|
454 |
default:
|
|
sl@0
|
455 |
{
|
|
sl@0
|
456 |
char buf[TCL_UTF_MAX+1];
|
|
sl@0
|
457 |
|
|
sl@0
|
458 |
buf[Tcl_UniCharToUtf(ch, buf)] = '\0';
|
|
sl@0
|
459 |
Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
|
|
sl@0
|
460 |
"bad scan conversion character \"", buf, "\"", NULL);
|
|
sl@0
|
461 |
goto error;
|
|
sl@0
|
462 |
}
|
|
sl@0
|
463 |
}
|
|
sl@0
|
464 |
if (!(flags & SCAN_SUPPRESS)) {
|
|
sl@0
|
465 |
if (objIndex >= nspace) {
|
|
sl@0
|
466 |
/*
|
|
sl@0
|
467 |
* Expand the nassign buffer. If we are using XPG specifiers,
|
|
sl@0
|
468 |
* make sure that we grow to a large enough size. xpgSize is
|
|
sl@0
|
469 |
* guaranteed to be at least one larger than objIndex.
|
|
sl@0
|
470 |
*/
|
|
sl@0
|
471 |
value = nspace;
|
|
sl@0
|
472 |
if (xpgSize) {
|
|
sl@0
|
473 |
nspace = xpgSize;
|
|
sl@0
|
474 |
} else {
|
|
sl@0
|
475 |
nspace += STATIC_LIST_SIZE;
|
|
sl@0
|
476 |
}
|
|
sl@0
|
477 |
if (nassign == staticAssign) {
|
|
sl@0
|
478 |
nassign = (void *)ckalloc(nspace * sizeof(int));
|
|
sl@0
|
479 |
for (i = 0; i < STATIC_LIST_SIZE; ++i) {
|
|
sl@0
|
480 |
nassign[i] = staticAssign[i];
|
|
sl@0
|
481 |
}
|
|
sl@0
|
482 |
} else {
|
|
sl@0
|
483 |
nassign = (void *)ckrealloc((void *)nassign,
|
|
sl@0
|
484 |
nspace * sizeof(int));
|
|
sl@0
|
485 |
}
|
|
sl@0
|
486 |
for (i = value; i < nspace; i++) {
|
|
sl@0
|
487 |
nassign[i] = 0;
|
|
sl@0
|
488 |
}
|
|
sl@0
|
489 |
}
|
|
sl@0
|
490 |
nassign[objIndex]++;
|
|
sl@0
|
491 |
objIndex++;
|
|
sl@0
|
492 |
}
|
|
sl@0
|
493 |
}
|
|
sl@0
|
494 |
|
|
sl@0
|
495 |
/*
|
|
sl@0
|
496 |
* Verify that all of the variable were assigned exactly once.
|
|
sl@0
|
497 |
*/
|
|
sl@0
|
498 |
|
|
sl@0
|
499 |
if (numVars == 0) {
|
|
sl@0
|
500 |
if (xpgSize) {
|
|
sl@0
|
501 |
numVars = xpgSize;
|
|
sl@0
|
502 |
} else {
|
|
sl@0
|
503 |
numVars = objIndex;
|
|
sl@0
|
504 |
}
|
|
sl@0
|
505 |
}
|
|
sl@0
|
506 |
if (totalSubs) {
|
|
sl@0
|
507 |
*totalSubs = numVars;
|
|
sl@0
|
508 |
}
|
|
sl@0
|
509 |
for (i = 0; i < numVars; i++) {
|
|
sl@0
|
510 |
if (nassign[i] > 1) {
|
|
sl@0
|
511 |
Tcl_SetResult(interp, "variable is assigned by multiple \"%n$\" conversion specifiers", TCL_STATIC);
|
|
sl@0
|
512 |
goto error;
|
|
sl@0
|
513 |
} else if (!xpgSize && (nassign[i] == 0)) {
|
|
sl@0
|
514 |
/*
|
|
sl@0
|
515 |
* If the space is empty, and xpgSize is 0 (means XPG wasn't
|
|
sl@0
|
516 |
* used, and/or numVars != 0), then too many vars were given
|
|
sl@0
|
517 |
*/
|
|
sl@0
|
518 |
Tcl_SetResult(interp, "variable is not assigned by any conversion specifiers", TCL_STATIC);
|
|
sl@0
|
519 |
goto error;
|
|
sl@0
|
520 |
}
|
|
sl@0
|
521 |
}
|
|
sl@0
|
522 |
|
|
sl@0
|
523 |
if (nassign != staticAssign) {
|
|
sl@0
|
524 |
ckfree((char *)nassign);
|
|
sl@0
|
525 |
}
|
|
sl@0
|
526 |
return TCL_OK;
|
|
sl@0
|
527 |
|
|
sl@0
|
528 |
badIndex:
|
|
sl@0
|
529 |
if (gotXpg) {
|
|
sl@0
|
530 |
Tcl_SetResult(interp, "\"%n$\" argument index out of range",
|
|
sl@0
|
531 |
TCL_STATIC);
|
|
sl@0
|
532 |
} else {
|
|
sl@0
|
533 |
Tcl_SetResult(interp,
|
|
sl@0
|
534 |
"different numbers of variable names and field specifiers",
|
|
sl@0
|
535 |
TCL_STATIC);
|
|
sl@0
|
536 |
}
|
|
sl@0
|
537 |
|
|
sl@0
|
538 |
error:
|
|
sl@0
|
539 |
if (nassign != staticAssign) {
|
|
sl@0
|
540 |
ckfree((char *)nassign);
|
|
sl@0
|
541 |
}
|
|
sl@0
|
542 |
return TCL_ERROR;
|
|
sl@0
|
543 |
#undef STATIC_LIST_SIZE
|
|
sl@0
|
544 |
}
|
|
sl@0
|
545 |
|
|
sl@0
|
546 |
/*
|
|
sl@0
|
547 |
*----------------------------------------------------------------------
|
|
sl@0
|
548 |
*
|
|
sl@0
|
549 |
* Tcl_ScanObjCmd --
|
|
sl@0
|
550 |
*
|
|
sl@0
|
551 |
* This procedure is invoked to process the "scan" Tcl command.
|
|
sl@0
|
552 |
* See the user documentation for details on what it does.
|
|
sl@0
|
553 |
*
|
|
sl@0
|
554 |
* Results:
|
|
sl@0
|
555 |
* A standard Tcl result.
|
|
sl@0
|
556 |
*
|
|
sl@0
|
557 |
* Side effects:
|
|
sl@0
|
558 |
* See the user documentation.
|
|
sl@0
|
559 |
*
|
|
sl@0
|
560 |
*----------------------------------------------------------------------
|
|
sl@0
|
561 |
*/
|
|
sl@0
|
562 |
|
|
sl@0
|
563 |
/* ARGSUSED */
|
|
sl@0
|
564 |
int
|
|
sl@0
|
565 |
Tcl_ScanObjCmd(dummy, interp, objc, objv)
|
|
sl@0
|
566 |
ClientData dummy; /* Not used. */
|
|
sl@0
|
567 |
Tcl_Interp *interp; /* Current interpreter. */
|
|
sl@0
|
568 |
int objc; /* Number of arguments. */
|
|
sl@0
|
569 |
Tcl_Obj *CONST objv[]; /* Argument objects. */
|
|
sl@0
|
570 |
{
|
|
sl@0
|
571 |
char *format;
|
|
sl@0
|
572 |
int numVars, nconversions, totalVars = -1;
|
|
sl@0
|
573 |
int objIndex, offset, i, result, code;
|
|
sl@0
|
574 |
long value;
|
|
sl@0
|
575 |
char *string, *end, *baseString;
|
|
sl@0
|
576 |
char op = 0;
|
|
sl@0
|
577 |
int base = 0;
|
|
sl@0
|
578 |
int underflow = 0;
|
|
sl@0
|
579 |
size_t width;
|
|
sl@0
|
580 |
long (*fn)() = NULL;
|
|
sl@0
|
581 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
582 |
Tcl_WideInt (*lfn)() = NULL;
|
|
sl@0
|
583 |
Tcl_WideInt wideValue;
|
|
sl@0
|
584 |
#endif
|
|
sl@0
|
585 |
Tcl_UniChar ch, sch;
|
|
sl@0
|
586 |
Tcl_Obj **objs = NULL, *objPtr = NULL;
|
|
sl@0
|
587 |
int flags;
|
|
sl@0
|
588 |
char buf[513]; /* Temporary buffer to hold scanned
|
|
sl@0
|
589 |
* number strings before they are
|
|
sl@0
|
590 |
* passed to strtoul. */
|
|
sl@0
|
591 |
|
|
sl@0
|
592 |
if (objc < 3) {
|
|
sl@0
|
593 |
Tcl_WrongNumArgs(interp, 1, objv,
|
|
sl@0
|
594 |
"string format ?varName varName ...?");
|
|
sl@0
|
595 |
return TCL_ERROR;
|
|
sl@0
|
596 |
}
|
|
sl@0
|
597 |
|
|
sl@0
|
598 |
format = Tcl_GetStringFromObj(objv[2], NULL);
|
|
sl@0
|
599 |
numVars = objc-3;
|
|
sl@0
|
600 |
|
|
sl@0
|
601 |
/*
|
|
sl@0
|
602 |
* Check for errors in the format string.
|
|
sl@0
|
603 |
*/
|
|
sl@0
|
604 |
|
|
sl@0
|
605 |
if (ValidateFormat(interp, format, numVars, &totalVars) == TCL_ERROR) {
|
|
sl@0
|
606 |
return TCL_ERROR;
|
|
sl@0
|
607 |
}
|
|
sl@0
|
608 |
|
|
sl@0
|
609 |
/*
|
|
sl@0
|
610 |
* Allocate space for the result objects.
|
|
sl@0
|
611 |
*/
|
|
sl@0
|
612 |
|
|
sl@0
|
613 |
if (totalVars > 0) {
|
|
sl@0
|
614 |
objs = (Tcl_Obj **) ckalloc(sizeof(Tcl_Obj*) * totalVars);
|
|
sl@0
|
615 |
for (i = 0; i < totalVars; i++) {
|
|
sl@0
|
616 |
objs[i] = NULL;
|
|
sl@0
|
617 |
}
|
|
sl@0
|
618 |
}
|
|
sl@0
|
619 |
|
|
sl@0
|
620 |
string = Tcl_GetStringFromObj(objv[1], NULL);
|
|
sl@0
|
621 |
baseString = string;
|
|
sl@0
|
622 |
|
|
sl@0
|
623 |
/*
|
|
sl@0
|
624 |
* Iterate over the format string filling in the result objects until
|
|
sl@0
|
625 |
* we reach the end of input, the end of the format string, or there
|
|
sl@0
|
626 |
* is a mismatch.
|
|
sl@0
|
627 |
*/
|
|
sl@0
|
628 |
|
|
sl@0
|
629 |
objIndex = 0;
|
|
sl@0
|
630 |
nconversions = 0;
|
|
sl@0
|
631 |
while (*format != '\0') {
|
|
sl@0
|
632 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
633 |
|
|
sl@0
|
634 |
flags = 0;
|
|
sl@0
|
635 |
|
|
sl@0
|
636 |
/*
|
|
sl@0
|
637 |
* If we see whitespace in the format, skip whitespace in the string.
|
|
sl@0
|
638 |
*/
|
|
sl@0
|
639 |
|
|
sl@0
|
640 |
if (Tcl_UniCharIsSpace(ch)) {
|
|
sl@0
|
641 |
offset = Tcl_UtfToUniChar(string, &sch);
|
|
sl@0
|
642 |
while (Tcl_UniCharIsSpace(sch)) {
|
|
sl@0
|
643 |
if (*string == '\0') {
|
|
sl@0
|
644 |
goto done;
|
|
sl@0
|
645 |
}
|
|
sl@0
|
646 |
string += offset;
|
|
sl@0
|
647 |
offset = Tcl_UtfToUniChar(string, &sch);
|
|
sl@0
|
648 |
}
|
|
sl@0
|
649 |
continue;
|
|
sl@0
|
650 |
}
|
|
sl@0
|
651 |
|
|
sl@0
|
652 |
if (ch != '%') {
|
|
sl@0
|
653 |
literal:
|
|
sl@0
|
654 |
if (*string == '\0') {
|
|
sl@0
|
655 |
underflow = 1;
|
|
sl@0
|
656 |
goto done;
|
|
sl@0
|
657 |
}
|
|
sl@0
|
658 |
string += Tcl_UtfToUniChar(string, &sch);
|
|
sl@0
|
659 |
if (ch != sch) {
|
|
sl@0
|
660 |
goto done;
|
|
sl@0
|
661 |
}
|
|
sl@0
|
662 |
continue;
|
|
sl@0
|
663 |
}
|
|
sl@0
|
664 |
|
|
sl@0
|
665 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
666 |
if (ch == '%') {
|
|
sl@0
|
667 |
goto literal;
|
|
sl@0
|
668 |
}
|
|
sl@0
|
669 |
|
|
sl@0
|
670 |
/*
|
|
sl@0
|
671 |
* Check for assignment suppression ('*') or an XPG3-style
|
|
sl@0
|
672 |
* assignment ('%n$').
|
|
sl@0
|
673 |
*/
|
|
sl@0
|
674 |
|
|
sl@0
|
675 |
if (ch == '*') {
|
|
sl@0
|
676 |
flags |= SCAN_SUPPRESS;
|
|
sl@0
|
677 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
678 |
} else if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
|
|
sl@0
|
679 |
value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */
|
|
sl@0
|
680 |
if (*end == '$') {
|
|
sl@0
|
681 |
format = end+1;
|
|
sl@0
|
682 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
683 |
objIndex = (int) value - 1;
|
|
sl@0
|
684 |
}
|
|
sl@0
|
685 |
}
|
|
sl@0
|
686 |
|
|
sl@0
|
687 |
/*
|
|
sl@0
|
688 |
* Parse any width specifier.
|
|
sl@0
|
689 |
*/
|
|
sl@0
|
690 |
|
|
sl@0
|
691 |
if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */
|
|
sl@0
|
692 |
width = strtoul(format-1, &format, 10); /* INTL: "C" locale. */
|
|
sl@0
|
693 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
694 |
} else {
|
|
sl@0
|
695 |
width = 0;
|
|
sl@0
|
696 |
}
|
|
sl@0
|
697 |
|
|
sl@0
|
698 |
/*
|
|
sl@0
|
699 |
* Handle any size specifier.
|
|
sl@0
|
700 |
*/
|
|
sl@0
|
701 |
|
|
sl@0
|
702 |
switch (ch) {
|
|
sl@0
|
703 |
case 'l':
|
|
sl@0
|
704 |
case 'L':
|
|
sl@0
|
705 |
flags |= SCAN_LONGER;
|
|
sl@0
|
706 |
/*
|
|
sl@0
|
707 |
* Fall through so we skip to the next character.
|
|
sl@0
|
708 |
*/
|
|
sl@0
|
709 |
case 'h':
|
|
sl@0
|
710 |
format += Tcl_UtfToUniChar(format, &ch);
|
|
sl@0
|
711 |
}
|
|
sl@0
|
712 |
|
|
sl@0
|
713 |
/*
|
|
sl@0
|
714 |
* Handle the various field types.
|
|
sl@0
|
715 |
*/
|
|
sl@0
|
716 |
|
|
sl@0
|
717 |
switch (ch) {
|
|
sl@0
|
718 |
case 'n':
|
|
sl@0
|
719 |
if (!(flags & SCAN_SUPPRESS)) {
|
|
sl@0
|
720 |
objPtr = Tcl_NewIntObj(string - baseString);
|
|
sl@0
|
721 |
Tcl_IncrRefCount(objPtr);
|
|
sl@0
|
722 |
objs[objIndex++] = objPtr;
|
|
sl@0
|
723 |
}
|
|
sl@0
|
724 |
nconversions++;
|
|
sl@0
|
725 |
continue;
|
|
sl@0
|
726 |
|
|
sl@0
|
727 |
case 'd':
|
|
sl@0
|
728 |
op = 'i';
|
|
sl@0
|
729 |
base = 10;
|
|
sl@0
|
730 |
fn = (long (*)())strtol;
|
|
sl@0
|
731 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
732 |
lfn = (Tcl_WideInt (*)())strtoll;
|
|
sl@0
|
733 |
#endif
|
|
sl@0
|
734 |
break;
|
|
sl@0
|
735 |
case 'i':
|
|
sl@0
|
736 |
op = 'i';
|
|
sl@0
|
737 |
base = 0;
|
|
sl@0
|
738 |
fn = (long (*)())strtol;
|
|
sl@0
|
739 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
740 |
lfn = (Tcl_WideInt (*)())strtoll;
|
|
sl@0
|
741 |
#endif
|
|
sl@0
|
742 |
break;
|
|
sl@0
|
743 |
case 'o':
|
|
sl@0
|
744 |
op = 'i';
|
|
sl@0
|
745 |
base = 8;
|
|
sl@0
|
746 |
fn = (long (*)())strtoul;
|
|
sl@0
|
747 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
748 |
lfn = (Tcl_WideInt (*)())strtoull;
|
|
sl@0
|
749 |
#endif
|
|
sl@0
|
750 |
break;
|
|
sl@0
|
751 |
case 'x':
|
|
sl@0
|
752 |
op = 'i';
|
|
sl@0
|
753 |
base = 16;
|
|
sl@0
|
754 |
fn = (long (*)())strtoul;
|
|
sl@0
|
755 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
756 |
lfn = (Tcl_WideInt (*)())strtoull;
|
|
sl@0
|
757 |
#endif
|
|
sl@0
|
758 |
break;
|
|
sl@0
|
759 |
case 'u':
|
|
sl@0
|
760 |
op = 'i';
|
|
sl@0
|
761 |
base = 10;
|
|
sl@0
|
762 |
flags |= SCAN_UNSIGNED;
|
|
sl@0
|
763 |
fn = (long (*)())strtoul;
|
|
sl@0
|
764 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
765 |
lfn = (Tcl_WideInt (*)())strtoull;
|
|
sl@0
|
766 |
#endif
|
|
sl@0
|
767 |
break;
|
|
sl@0
|
768 |
|
|
sl@0
|
769 |
case 'f':
|
|
sl@0
|
770 |
case 'e':
|
|
sl@0
|
771 |
case 'g':
|
|
sl@0
|
772 |
op = 'f';
|
|
sl@0
|
773 |
break;
|
|
sl@0
|
774 |
|
|
sl@0
|
775 |
case 's':
|
|
sl@0
|
776 |
op = 's';
|
|
sl@0
|
777 |
break;
|
|
sl@0
|
778 |
|
|
sl@0
|
779 |
case 'c':
|
|
sl@0
|
780 |
op = 'c';
|
|
sl@0
|
781 |
flags |= SCAN_NOSKIP;
|
|
sl@0
|
782 |
break;
|
|
sl@0
|
783 |
case '[':
|
|
sl@0
|
784 |
op = '[';
|
|
sl@0
|
785 |
flags |= SCAN_NOSKIP;
|
|
sl@0
|
786 |
break;
|
|
sl@0
|
787 |
}
|
|
sl@0
|
788 |
|
|
sl@0
|
789 |
/*
|
|
sl@0
|
790 |
* At this point, we will need additional characters from the
|
|
sl@0
|
791 |
* string to proceed.
|
|
sl@0
|
792 |
*/
|
|
sl@0
|
793 |
|
|
sl@0
|
794 |
if (*string == '\0') {
|
|
sl@0
|
795 |
underflow = 1;
|
|
sl@0
|
796 |
goto done;
|
|
sl@0
|
797 |
}
|
|
sl@0
|
798 |
|
|
sl@0
|
799 |
/*
|
|
sl@0
|
800 |
* Skip any leading whitespace at the beginning of a field unless
|
|
sl@0
|
801 |
* the format suppresses this behavior.
|
|
sl@0
|
802 |
*/
|
|
sl@0
|
803 |
|
|
sl@0
|
804 |
if (!(flags & SCAN_NOSKIP)) {
|
|
sl@0
|
805 |
while (*string != '\0') {
|
|
sl@0
|
806 |
offset = Tcl_UtfToUniChar(string, &sch);
|
|
sl@0
|
807 |
if (!Tcl_UniCharIsSpace(sch)) {
|
|
sl@0
|
808 |
break;
|
|
sl@0
|
809 |
}
|
|
sl@0
|
810 |
string += offset;
|
|
sl@0
|
811 |
}
|
|
sl@0
|
812 |
if (*string == '\0') {
|
|
sl@0
|
813 |
underflow = 1;
|
|
sl@0
|
814 |
goto done;
|
|
sl@0
|
815 |
}
|
|
sl@0
|
816 |
}
|
|
sl@0
|
817 |
|
|
sl@0
|
818 |
/*
|
|
sl@0
|
819 |
* Perform the requested scanning operation.
|
|
sl@0
|
820 |
*/
|
|
sl@0
|
821 |
|
|
sl@0
|
822 |
switch (op) {
|
|
sl@0
|
823 |
case 's':
|
|
sl@0
|
824 |
/*
|
|
sl@0
|
825 |
* Scan a string up to width characters or whitespace.
|
|
sl@0
|
826 |
*/
|
|
sl@0
|
827 |
|
|
sl@0
|
828 |
if (width == 0) {
|
|
sl@0
|
829 |
width = (size_t) ~0;
|
|
sl@0
|
830 |
}
|
|
sl@0
|
831 |
end = string;
|
|
sl@0
|
832 |
while (*end != '\0') {
|
|
sl@0
|
833 |
offset = Tcl_UtfToUniChar(end, &sch);
|
|
sl@0
|
834 |
if (Tcl_UniCharIsSpace(sch)) {
|
|
sl@0
|
835 |
break;
|
|
sl@0
|
836 |
}
|
|
sl@0
|
837 |
end += offset;
|
|
sl@0
|
838 |
if (--width == 0) {
|
|
sl@0
|
839 |
break;
|
|
sl@0
|
840 |
}
|
|
sl@0
|
841 |
}
|
|
sl@0
|
842 |
if (!(flags & SCAN_SUPPRESS)) {
|
|
sl@0
|
843 |
objPtr = Tcl_NewStringObj(string, end-string);
|
|
sl@0
|
844 |
Tcl_IncrRefCount(objPtr);
|
|
sl@0
|
845 |
objs[objIndex++] = objPtr;
|
|
sl@0
|
846 |
}
|
|
sl@0
|
847 |
string = end;
|
|
sl@0
|
848 |
break;
|
|
sl@0
|
849 |
|
|
sl@0
|
850 |
case '[': {
|
|
sl@0
|
851 |
CharSet cset;
|
|
sl@0
|
852 |
|
|
sl@0
|
853 |
if (width == 0) {
|
|
sl@0
|
854 |
width = (size_t) ~0;
|
|
sl@0
|
855 |
}
|
|
sl@0
|
856 |
end = string;
|
|
sl@0
|
857 |
|
|
sl@0
|
858 |
format = BuildCharSet(&cset, format);
|
|
sl@0
|
859 |
while (*end != '\0') {
|
|
sl@0
|
860 |
offset = Tcl_UtfToUniChar(end, &sch);
|
|
sl@0
|
861 |
if (!CharInSet(&cset, (int)sch)) {
|
|
sl@0
|
862 |
break;
|
|
sl@0
|
863 |
}
|
|
sl@0
|
864 |
end += offset;
|
|
sl@0
|
865 |
if (--width == 0) {
|
|
sl@0
|
866 |
break;
|
|
sl@0
|
867 |
}
|
|
sl@0
|
868 |
}
|
|
sl@0
|
869 |
ReleaseCharSet(&cset);
|
|
sl@0
|
870 |
|
|
sl@0
|
871 |
if (string == end) {
|
|
sl@0
|
872 |
/*
|
|
sl@0
|
873 |
* Nothing matched the range, stop processing
|
|
sl@0
|
874 |
*/
|
|
sl@0
|
875 |
goto done;
|
|
sl@0
|
876 |
}
|
|
sl@0
|
877 |
if (!(flags & SCAN_SUPPRESS)) {
|
|
sl@0
|
878 |
objPtr = Tcl_NewStringObj(string, end-string);
|
|
sl@0
|
879 |
Tcl_IncrRefCount(objPtr);
|
|
sl@0
|
880 |
objs[objIndex++] = objPtr;
|
|
sl@0
|
881 |
}
|
|
sl@0
|
882 |
string = end;
|
|
sl@0
|
883 |
|
|
sl@0
|
884 |
break;
|
|
sl@0
|
885 |
}
|
|
sl@0
|
886 |
case 'c':
|
|
sl@0
|
887 |
/*
|
|
sl@0
|
888 |
* Scan a single Unicode character.
|
|
sl@0
|
889 |
*/
|
|
sl@0
|
890 |
|
|
sl@0
|
891 |
string += Tcl_UtfToUniChar(string, &sch);
|
|
sl@0
|
892 |
if (!(flags & SCAN_SUPPRESS)) {
|
|
sl@0
|
893 |
objPtr = Tcl_NewIntObj((int)sch);
|
|
sl@0
|
894 |
Tcl_IncrRefCount(objPtr);
|
|
sl@0
|
895 |
objs[objIndex++] = objPtr;
|
|
sl@0
|
896 |
}
|
|
sl@0
|
897 |
break;
|
|
sl@0
|
898 |
|
|
sl@0
|
899 |
case 'i':
|
|
sl@0
|
900 |
/*
|
|
sl@0
|
901 |
* Scan an unsigned or signed integer.
|
|
sl@0
|
902 |
*/
|
|
sl@0
|
903 |
|
|
sl@0
|
904 |
if ((width == 0) || (width > sizeof(buf) - 1)) {
|
|
sl@0
|
905 |
width = sizeof(buf) - 1;
|
|
sl@0
|
906 |
}
|
|
sl@0
|
907 |
flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_NOZERO;
|
|
sl@0
|
908 |
for (end = buf; width > 0; width--) {
|
|
sl@0
|
909 |
switch (*string) {
|
|
sl@0
|
910 |
/*
|
|
sl@0
|
911 |
* The 0 digit has special meaning at the beginning of
|
|
sl@0
|
912 |
* a number. If we are unsure of the base, it
|
|
sl@0
|
913 |
* indicates that we are in base 8 or base 16 (if it is
|
|
sl@0
|
914 |
* followed by an 'x').
|
|
sl@0
|
915 |
*
|
|
sl@0
|
916 |
* 8.1 - 8.3.4 incorrectly handled 0x... base-16
|
|
sl@0
|
917 |
* cases for %x by not reading the 0x as the
|
|
sl@0
|
918 |
* auto-prelude for base-16. [Bug #495213]
|
|
sl@0
|
919 |
*/
|
|
sl@0
|
920 |
case '0':
|
|
sl@0
|
921 |
if (base == 0) {
|
|
sl@0
|
922 |
base = 8;
|
|
sl@0
|
923 |
flags |= SCAN_XOK;
|
|
sl@0
|
924 |
}
|
|
sl@0
|
925 |
if (base == 16) {
|
|
sl@0
|
926 |
flags |= SCAN_XOK;
|
|
sl@0
|
927 |
}
|
|
sl@0
|
928 |
if (flags & SCAN_NOZERO) {
|
|
sl@0
|
929 |
flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS
|
|
sl@0
|
930 |
| SCAN_NOZERO);
|
|
sl@0
|
931 |
} else {
|
|
sl@0
|
932 |
flags &= ~(SCAN_SIGNOK | SCAN_XOK
|
|
sl@0
|
933 |
| SCAN_NODIGITS);
|
|
sl@0
|
934 |
}
|
|
sl@0
|
935 |
goto addToInt;
|
|
sl@0
|
936 |
|
|
sl@0
|
937 |
case '1': case '2': case '3': case '4':
|
|
sl@0
|
938 |
case '5': case '6': case '7':
|
|
sl@0
|
939 |
if (base == 0) {
|
|
sl@0
|
940 |
base = 10;
|
|
sl@0
|
941 |
}
|
|
sl@0
|
942 |
flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
|
|
sl@0
|
943 |
goto addToInt;
|
|
sl@0
|
944 |
|
|
sl@0
|
945 |
case '8': case '9':
|
|
sl@0
|
946 |
if (base == 0) {
|
|
sl@0
|
947 |
base = 10;
|
|
sl@0
|
948 |
}
|
|
sl@0
|
949 |
if (base <= 8) {
|
|
sl@0
|
950 |
break;
|
|
sl@0
|
951 |
}
|
|
sl@0
|
952 |
flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
|
|
sl@0
|
953 |
goto addToInt;
|
|
sl@0
|
954 |
|
|
sl@0
|
955 |
case 'A': case 'B': case 'C':
|
|
sl@0
|
956 |
case 'D': case 'E': case 'F':
|
|
sl@0
|
957 |
case 'a': case 'b': case 'c':
|
|
sl@0
|
958 |
case 'd': case 'e': case 'f':
|
|
sl@0
|
959 |
if (base <= 10) {
|
|
sl@0
|
960 |
break;
|
|
sl@0
|
961 |
}
|
|
sl@0
|
962 |
flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS);
|
|
sl@0
|
963 |
goto addToInt;
|
|
sl@0
|
964 |
|
|
sl@0
|
965 |
case '+': case '-':
|
|
sl@0
|
966 |
if (flags & SCAN_SIGNOK) {
|
|
sl@0
|
967 |
flags &= ~SCAN_SIGNOK;
|
|
sl@0
|
968 |
goto addToInt;
|
|
sl@0
|
969 |
}
|
|
sl@0
|
970 |
break;
|
|
sl@0
|
971 |
|
|
sl@0
|
972 |
case 'x': case 'X':
|
|
sl@0
|
973 |
if ((flags & SCAN_XOK) && (end == buf+1)) {
|
|
sl@0
|
974 |
base = 16;
|
|
sl@0
|
975 |
flags &= ~SCAN_XOK;
|
|
sl@0
|
976 |
goto addToInt;
|
|
sl@0
|
977 |
}
|
|
sl@0
|
978 |
break;
|
|
sl@0
|
979 |
}
|
|
sl@0
|
980 |
|
|
sl@0
|
981 |
/*
|
|
sl@0
|
982 |
* We got an illegal character so we are done accumulating.
|
|
sl@0
|
983 |
*/
|
|
sl@0
|
984 |
|
|
sl@0
|
985 |
break;
|
|
sl@0
|
986 |
|
|
sl@0
|
987 |
addToInt:
|
|
sl@0
|
988 |
/*
|
|
sl@0
|
989 |
* Add the character to the temporary buffer.
|
|
sl@0
|
990 |
*/
|
|
sl@0
|
991 |
|
|
sl@0
|
992 |
*end++ = *string++;
|
|
sl@0
|
993 |
if (*string == '\0') {
|
|
sl@0
|
994 |
break;
|
|
sl@0
|
995 |
}
|
|
sl@0
|
996 |
}
|
|
sl@0
|
997 |
|
|
sl@0
|
998 |
/*
|
|
sl@0
|
999 |
* Check to see if we need to back up because we only got a
|
|
sl@0
|
1000 |
* sign or a trailing x after a 0.
|
|
sl@0
|
1001 |
*/
|
|
sl@0
|
1002 |
|
|
sl@0
|
1003 |
if (flags & SCAN_NODIGITS) {
|
|
sl@0
|
1004 |
if (*string == '\0') {
|
|
sl@0
|
1005 |
underflow = 1;
|
|
sl@0
|
1006 |
}
|
|
sl@0
|
1007 |
goto done;
|
|
sl@0
|
1008 |
} else if (end[-1] == 'x' || end[-1] == 'X') {
|
|
sl@0
|
1009 |
end--;
|
|
sl@0
|
1010 |
string--;
|
|
sl@0
|
1011 |
}
|
|
sl@0
|
1012 |
|
|
sl@0
|
1013 |
|
|
sl@0
|
1014 |
/*
|
|
sl@0
|
1015 |
* Scan the value from the temporary buffer. If we are
|
|
sl@0
|
1016 |
* returning a large unsigned value, we have to convert it back
|
|
sl@0
|
1017 |
* to a string since Tcl only supports signed values.
|
|
sl@0
|
1018 |
*/
|
|
sl@0
|
1019 |
|
|
sl@0
|
1020 |
if (!(flags & SCAN_SUPPRESS)) {
|
|
sl@0
|
1021 |
*end = '\0';
|
|
sl@0
|
1022 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
1023 |
if (flags & SCAN_LONGER) {
|
|
sl@0
|
1024 |
wideValue = (Tcl_WideInt) (*lfn)(buf, NULL, base);
|
|
sl@0
|
1025 |
if ((flags & SCAN_UNSIGNED) && (wideValue < 0)) {
|
|
sl@0
|
1026 |
/* INTL: ISO digit */
|
|
sl@0
|
1027 |
sprintf(buf, "%" TCL_LL_MODIFIER "u",
|
|
sl@0
|
1028 |
(Tcl_WideUInt)wideValue);
|
|
sl@0
|
1029 |
objPtr = Tcl_NewStringObj(buf, -1);
|
|
sl@0
|
1030 |
} else {
|
|
sl@0
|
1031 |
objPtr = Tcl_NewWideIntObj(wideValue);
|
|
sl@0
|
1032 |
}
|
|
sl@0
|
1033 |
} else {
|
|
sl@0
|
1034 |
#endif /* !TCL_WIDE_INT_IS_LONG */
|
|
sl@0
|
1035 |
value = (long) (*fn)(buf, NULL, base);
|
|
sl@0
|
1036 |
if ((flags & SCAN_UNSIGNED) && (value < 0)) {
|
|
sl@0
|
1037 |
sprintf(buf, "%lu", value); /* INTL: ISO digit */
|
|
sl@0
|
1038 |
objPtr = Tcl_NewStringObj(buf, -1);
|
|
sl@0
|
1039 |
} else if ((flags & SCAN_LONGER)
|
|
sl@0
|
1040 |
|| (unsigned long) value > UINT_MAX) {
|
|
sl@0
|
1041 |
objPtr = Tcl_NewLongObj(value);
|
|
sl@0
|
1042 |
} else {
|
|
sl@0
|
1043 |
objPtr = Tcl_NewIntObj(value);
|
|
sl@0
|
1044 |
}
|
|
sl@0
|
1045 |
#ifndef TCL_WIDE_INT_IS_LONG
|
|
sl@0
|
1046 |
}
|
|
sl@0
|
1047 |
#endif
|
|
sl@0
|
1048 |
Tcl_IncrRefCount(objPtr);
|
|
sl@0
|
1049 |
objs[objIndex++] = objPtr;
|
|
sl@0
|
1050 |
}
|
|
sl@0
|
1051 |
|
|
sl@0
|
1052 |
break;
|
|
sl@0
|
1053 |
|
|
sl@0
|
1054 |
case 'f':
|
|
sl@0
|
1055 |
/*
|
|
sl@0
|
1056 |
* Scan a floating point number
|
|
sl@0
|
1057 |
*/
|
|
sl@0
|
1058 |
|
|
sl@0
|
1059 |
if ((width == 0) || (width > sizeof(buf) - 1)) {
|
|
sl@0
|
1060 |
width = sizeof(buf) - 1;
|
|
sl@0
|
1061 |
}
|
|
sl@0
|
1062 |
flags &= ~SCAN_LONGER;
|
|
sl@0
|
1063 |
flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_PTOK | SCAN_EXPOK;
|
|
sl@0
|
1064 |
for (end = buf; width > 0; width--) {
|
|
sl@0
|
1065 |
switch (*string) {
|
|
sl@0
|
1066 |
case '0': case '1': case '2': case '3':
|
|
sl@0
|
1067 |
case '4': case '5': case '6': case '7':
|
|
sl@0
|
1068 |
case '8': case '9':
|
|
sl@0
|
1069 |
flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS);
|
|
sl@0
|
1070 |
goto addToFloat;
|
|
sl@0
|
1071 |
case '+': case '-':
|
|
sl@0
|
1072 |
if (flags & SCAN_SIGNOK) {
|
|
sl@0
|
1073 |
flags &= ~SCAN_SIGNOK;
|
|
sl@0
|
1074 |
goto addToFloat;
|
|
sl@0
|
1075 |
}
|
|
sl@0
|
1076 |
break;
|
|
sl@0
|
1077 |
case '.':
|
|
sl@0
|
1078 |
if (flags & SCAN_PTOK) {
|
|
sl@0
|
1079 |
flags &= ~(SCAN_SIGNOK | SCAN_PTOK);
|
|
sl@0
|
1080 |
goto addToFloat;
|
|
sl@0
|
1081 |
}
|
|
sl@0
|
1082 |
break;
|
|
sl@0
|
1083 |
case 'e': case 'E':
|
|
sl@0
|
1084 |
/*
|
|
sl@0
|
1085 |
* An exponent is not allowed until there has
|
|
sl@0
|
1086 |
* been at least one digit.
|
|
sl@0
|
1087 |
*/
|
|
sl@0
|
1088 |
|
|
sl@0
|
1089 |
if ((flags & (SCAN_NODIGITS | SCAN_EXPOK))
|
|
sl@0
|
1090 |
== SCAN_EXPOK) {
|
|
sl@0
|
1091 |
flags = (flags & ~(SCAN_EXPOK|SCAN_PTOK))
|
|
sl@0
|
1092 |
| SCAN_SIGNOK | SCAN_NODIGITS;
|
|
sl@0
|
1093 |
goto addToFloat;
|
|
sl@0
|
1094 |
}
|
|
sl@0
|
1095 |
break;
|
|
sl@0
|
1096 |
}
|
|
sl@0
|
1097 |
|
|
sl@0
|
1098 |
/*
|
|
sl@0
|
1099 |
* We got an illegal character so we are done accumulating.
|
|
sl@0
|
1100 |
*/
|
|
sl@0
|
1101 |
|
|
sl@0
|
1102 |
break;
|
|
sl@0
|
1103 |
|
|
sl@0
|
1104 |
addToFloat:
|
|
sl@0
|
1105 |
/*
|
|
sl@0
|
1106 |
* Add the character to the temporary buffer.
|
|
sl@0
|
1107 |
*/
|
|
sl@0
|
1108 |
|
|
sl@0
|
1109 |
*end++ = *string++;
|
|
sl@0
|
1110 |
if (*string == '\0') {
|
|
sl@0
|
1111 |
break;
|
|
sl@0
|
1112 |
}
|
|
sl@0
|
1113 |
}
|
|
sl@0
|
1114 |
|
|
sl@0
|
1115 |
/*
|
|
sl@0
|
1116 |
* Check to see if we need to back up because we saw a
|
|
sl@0
|
1117 |
* trailing 'e' or sign.
|
|
sl@0
|
1118 |
*/
|
|
sl@0
|
1119 |
|
|
sl@0
|
1120 |
if (flags & SCAN_NODIGITS) {
|
|
sl@0
|
1121 |
if (flags & SCAN_EXPOK) {
|
|
sl@0
|
1122 |
/*
|
|
sl@0
|
1123 |
* There were no digits at all so scanning has
|
|
sl@0
|
1124 |
* failed and we are done.
|
|
sl@0
|
1125 |
*/
|
|
sl@0
|
1126 |
if (*string == '\0') {
|
|
sl@0
|
1127 |
underflow = 1;
|
|
sl@0
|
1128 |
}
|
|
sl@0
|
1129 |
goto done;
|
|
sl@0
|
1130 |
}
|
|
sl@0
|
1131 |
|
|
sl@0
|
1132 |
/*
|
|
sl@0
|
1133 |
* We got a bad exponent ('e' and maybe a sign).
|
|
sl@0
|
1134 |
*/
|
|
sl@0
|
1135 |
|
|
sl@0
|
1136 |
end--;
|
|
sl@0
|
1137 |
string--;
|
|
sl@0
|
1138 |
if (*end != 'e' && *end != 'E') {
|
|
sl@0
|
1139 |
end--;
|
|
sl@0
|
1140 |
string--;
|
|
sl@0
|
1141 |
}
|
|
sl@0
|
1142 |
}
|
|
sl@0
|
1143 |
|
|
sl@0
|
1144 |
/*
|
|
sl@0
|
1145 |
* Scan the value from the temporary buffer.
|
|
sl@0
|
1146 |
*/
|
|
sl@0
|
1147 |
|
|
sl@0
|
1148 |
if (!(flags & SCAN_SUPPRESS)) {
|
|
sl@0
|
1149 |
double dvalue;
|
|
sl@0
|
1150 |
*end = '\0';
|
|
sl@0
|
1151 |
dvalue = strtod(buf, NULL);
|
|
sl@0
|
1152 |
objPtr = Tcl_NewDoubleObj(dvalue);
|
|
sl@0
|
1153 |
Tcl_IncrRefCount(objPtr);
|
|
sl@0
|
1154 |
objs[objIndex++] = objPtr;
|
|
sl@0
|
1155 |
}
|
|
sl@0
|
1156 |
break;
|
|
sl@0
|
1157 |
}
|
|
sl@0
|
1158 |
nconversions++;
|
|
sl@0
|
1159 |
}
|
|
sl@0
|
1160 |
|
|
sl@0
|
1161 |
done:
|
|
sl@0
|
1162 |
result = 0;
|
|
sl@0
|
1163 |
code = TCL_OK;
|
|
sl@0
|
1164 |
|
|
sl@0
|
1165 |
if (numVars) {
|
|
sl@0
|
1166 |
/*
|
|
sl@0
|
1167 |
* In this case, variables were specified (classic scan)
|
|
sl@0
|
1168 |
*/
|
|
sl@0
|
1169 |
for (i = 0; i < totalVars; i++) {
|
|
sl@0
|
1170 |
if (objs[i] != NULL) {
|
|
sl@0
|
1171 |
Tcl_Obj *tmpPtr;
|
|
sl@0
|
1172 |
|
|
sl@0
|
1173 |
result++;
|
|
sl@0
|
1174 |
tmpPtr = Tcl_ObjSetVar2(interp, objv[i+3], NULL, objs[i], 0);
|
|
sl@0
|
1175 |
Tcl_DecrRefCount(objs[i]);
|
|
sl@0
|
1176 |
if (tmpPtr == NULL) {
|
|
sl@0
|
1177 |
Tcl_AppendStringsToObj(Tcl_GetObjResult(interp),
|
|
sl@0
|
1178 |
"couldn't set variable \"",
|
|
sl@0
|
1179 |
Tcl_GetString(objv[i+3]), "\"", (char *) NULL);
|
|
sl@0
|
1180 |
code = TCL_ERROR;
|
|
sl@0
|
1181 |
}
|
|
sl@0
|
1182 |
}
|
|
sl@0
|
1183 |
}
|
|
sl@0
|
1184 |
} else {
|
|
sl@0
|
1185 |
/*
|
|
sl@0
|
1186 |
* Here no vars were specified, we want a list returned (inline scan)
|
|
sl@0
|
1187 |
*/
|
|
sl@0
|
1188 |
objPtr = Tcl_NewObj();
|
|
sl@0
|
1189 |
for (i = 0; i < totalVars; i++) {
|
|
sl@0
|
1190 |
if (objs[i] != NULL) {
|
|
sl@0
|
1191 |
Tcl_ListObjAppendElement(NULL, objPtr, objs[i]);
|
|
sl@0
|
1192 |
Tcl_DecrRefCount(objs[i]);
|
|
sl@0
|
1193 |
} else {
|
|
sl@0
|
1194 |
/*
|
|
sl@0
|
1195 |
* More %-specifiers than matching chars, so we
|
|
sl@0
|
1196 |
* just spit out empty strings for these
|
|
sl@0
|
1197 |
*/
|
|
sl@0
|
1198 |
Tcl_ListObjAppendElement(NULL, objPtr, Tcl_NewObj());
|
|
sl@0
|
1199 |
}
|
|
sl@0
|
1200 |
}
|
|
sl@0
|
1201 |
}
|
|
sl@0
|
1202 |
if (objs != NULL) {
|
|
sl@0
|
1203 |
ckfree((char*) objs);
|
|
sl@0
|
1204 |
}
|
|
sl@0
|
1205 |
if (code == TCL_OK) {
|
|
sl@0
|
1206 |
if (underflow && (nconversions == 0)) {
|
|
sl@0
|
1207 |
if (numVars) {
|
|
sl@0
|
1208 |
objPtr = Tcl_NewIntObj(-1);
|
|
sl@0
|
1209 |
} else {
|
|
sl@0
|
1210 |
if (objPtr) {
|
|
sl@0
|
1211 |
Tcl_SetListObj(objPtr, 0, NULL);
|
|
sl@0
|
1212 |
} else {
|
|
sl@0
|
1213 |
objPtr = Tcl_NewObj();
|
|
sl@0
|
1214 |
}
|
|
sl@0
|
1215 |
}
|
|
sl@0
|
1216 |
} else if (numVars) {
|
|
sl@0
|
1217 |
objPtr = Tcl_NewIntObj(result);
|
|
sl@0
|
1218 |
}
|
|
sl@0
|
1219 |
Tcl_SetObjResult(interp, objPtr);
|
|
sl@0
|
1220 |
}
|
|
sl@0
|
1221 |
return code;
|
|
sl@0
|
1222 |
}
|