1#include <stdio.h>
2#include <stdlib.h>
3#include <stdarg.h>
4#include <ctype.h>
5#include <wchar.h>
6#include <wctype.h>
7#include <limits.h>
8#include <string.h>
9
10#include "stdio_impl.h"
11#include "shgetc.h"
12#include "intscan.h"
13#include "floatscan.h"
14#include "libc.h"
15
16#define SIZE_hh -2
17#define SIZE_h  -1
18#define SIZE_def 0
19#define SIZE_l   1
20#define SIZE_L   2
21#define SIZE_ll  3
22
23static void store_int(void *dest, int size, unsigned long long i)
24{
25	if (!dest) return;
26	switch (size) {
27	case SIZE_hh:
28		*(char *)dest = i;
29		break;
30	case SIZE_h:
31		*(short *)dest = i;
32		break;
33	case SIZE_def:
34		*(int *)dest = i;
35		break;
36	case SIZE_l:
37		*(long *)dest = i;
38		break;
39	case SIZE_ll:
40		*(long long *)dest = i;
41		break;
42	}
43}
44
45static void *arg_n(va_list ap, unsigned int n)
46{
47	void *p;
48	unsigned int i;
49	va_list ap2;
50	va_copy(ap2, ap);
51	for (i=n; i>1; i--) va_arg(ap2, void *);
52	p = va_arg(ap2, void *);
53	va_end(ap2);
54	return p;
55}
56
57static int in_set(const wchar_t *set, int c)
58{
59	int j;
60	const wchar_t *p = set;
61	if (*p == '-') {
62		if (c=='-') return 1;
63		p++;
64	} else if (*p == ']') {
65		if (c==']') return 1;
66		p++;
67	}
68	for (; *p && *p != ']'; p++) {
69		if (*p=='-' && p[1] && p[1] != ']')
70			for (j=p++[-1]; j<*p; j++)
71				if (c==j) return 1;
72		if (c==*p) return 1;
73	}
74	return 0;
75}
76
77#if 1
78#undef getwc
79#define getwc(f) \
80	((f)->rpos < (f)->rend && *(f)->rpos < 128 ? *(f)->rpos++ : (getwc)(f))
81
82#undef ungetwc
83#define ungetwc(c,f) \
84	((f)->rend && (c)<128U ? *--(f)->rpos : ungetwc((c),(f)))
85#endif
86
87int vfwscanf(FILE *restrict f, const wchar_t *restrict fmt, va_list ap)
88{
89	int width;
90	int size;
91	int alloc;
92	const wchar_t *p;
93	int c, t;
94	char *s;
95	wchar_t *wcs;
96	void *dest=NULL;
97	int invert;
98	int matches=0;
99	off_t pos = 0, cnt;
100	static const char size_pfx[][3] = { "hh", "h", "", "l", "L", "ll" };
101	char tmp[3*sizeof(int)+10];
102	const wchar_t *set;
103	size_t i, k;
104
105	FLOCK(f);
106
107	fwide(f, 1);
108
109	for (p=fmt; *p; p++) {
110
111		alloc = 0;
112
113		if (iswspace(*p)) {
114			while (iswspace(p[1])) p++;
115			while (iswspace((c=getwc(f)))) pos++;
116			ungetwc(c, f);
117			continue;
118		}
119		if (*p != '%' || p[1] == '%') {
120			p += *p=='%';
121			c = getwc(f);
122			if (c!=*p) {
123				ungetwc(c, f);
124				if (c<0) goto input_fail;
125				goto match_fail;
126			}
127			pos++;
128			continue;
129		}
130
131		p++;
132		if (*p=='*') {
133			dest = 0; p++;
134		} else if (iswdigit(*p) && p[1]=='$') {
135			dest = arg_n(ap, *p-'0'); p+=2;
136		} else {
137			dest = va_arg(ap, void *);
138		}
139
140		for (width=0; iswdigit(*p); p++) {
141			width = 10*width + *p - '0';
142		}
143
144		if (*p=='m') {
145			wcs = 0;
146			s = 0;
147			alloc = !!dest;
148			p++;
149		} else {
150			alloc = 0;
151		}
152
153		size = SIZE_def;
154		switch (*p++) {
155		case 'h':
156			if (*p == 'h') p++, size = SIZE_hh;
157			else size = SIZE_h;
158			break;
159		case 'l':
160			if (*p == 'l') p++, size = SIZE_ll;
161			else size = SIZE_l;
162			break;
163		case 'j':
164			size = SIZE_ll;
165			break;
166		case 'z':
167		case 't':
168			size = SIZE_l;
169			break;
170		case 'L':
171			size = SIZE_L;
172			break;
173		case 'd': case 'i': case 'o': case 'u': case 'x':
174		case 'a': case 'e': case 'f': case 'g':
175		case 'A': case 'E': case 'F': case 'G': case 'X':
176		case 's': case 'c': case '[':
177		case 'S': case 'C':
178		case 'p': case 'n':
179			p--;
180			break;
181		default:
182			goto fmt_fail;
183		}
184
185		t = *p;
186
187		/* Transform S,C -> ls,lc */
188		if ((t&0x2f)==3) {
189			size = SIZE_l;
190			t |= 32;
191		}
192
193		if (t != 'n') {
194			if (t != '[' && (t|32) != 'c')
195				while (iswspace((c=getwc(f)))) pos++;
196			else
197				c=getwc(f);
198			if (c < 0) goto input_fail;
199			ungetwc(c, f);
200		}
201
202		switch (t) {
203		case 'n':
204			store_int(dest, size, pos);
205			/* do not increment match count, etc! */
206			continue;
207
208		case 's':
209		case 'c':
210		case '[':
211			if (t == 'c') {
212				if (width<1) width = 1;
213				invert = 1;
214				set = L"";
215			} else if (t == 's') {
216				invert = 1;
217				set = (const wchar_t[]){
218					' ', '\t', '\n', '\r', 11, 12,  0x0085,
219					0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005,
220					0x2006, 0x2008, 0x2009, 0x200a,
221					0x2028, 0x2029, 0x205f, 0x3000, 0 };
222			} else {
223				if (*++p == '^') p++, invert = 1;
224				else invert = 0;
225				set = p;
226				if (*p==']') p++;
227				while (*p!=']') {
228					if (!*p) goto fmt_fail;
229					p++;
230				}
231			}
232
233			s = (size == SIZE_def) ? dest : 0;
234			wcs = (size == SIZE_l) ? dest : 0;
235
236			int gotmatch = 0;
237
238			if (width < 1) width = -1;
239
240			i = 0;
241			if (alloc) {
242				k = t=='c' ? width+1U : 31;
243				if (size == SIZE_l) {
244					wcs = malloc(k*sizeof(wchar_t));
245					if (!wcs) goto alloc_fail;
246				} else {
247					s = malloc(k);
248					if (!s) goto alloc_fail;
249				}
250			}
251			while (width) {
252				if ((c=getwc(f))<0) break;
253				if (in_set(set, c) == invert)
254					break;
255				if (wcs) {
256					wcs[i++] = c;
257					if (alloc && i==k) {
258						k += k+1;
259						wchar_t *tmp = realloc(wcs, k*sizeof(wchar_t));
260						if (!tmp) goto alloc_fail;
261						wcs = tmp;
262					}
263				} else if (size != SIZE_l) {
264					int l = wctomb(s?s+i:tmp, c);
265					if (l<0) goto input_fail;
266					i += l;
267					if (alloc && i > k-4) {
268						k += k+1;
269						char *tmp = realloc(s, k);
270						if (!tmp) goto alloc_fail;
271						s = tmp;
272					}
273				}
274				pos++;
275				width-=(width>0);
276				gotmatch=1;
277			}
278			if (width) {
279				ungetwc(c, f);
280				if (t == 'c' || !gotmatch) goto match_fail;
281			}
282
283			if (alloc) {
284				if (size == SIZE_l) *(wchar_t **)dest = wcs;
285				else *(char **)dest = s;
286			}
287			if (t != 'c') {
288				if (wcs) wcs[i] = 0;
289				if (s) s[i] = 0;
290			}
291			break;
292
293		case 'd': case 'i': case 'o': case 'u': case 'x':
294		case 'a': case 'e': case 'f': case 'g':
295		case 'A': case 'E': case 'F': case 'G': case 'X':
296		case 'p':
297			if (width < 1) width = 0;
298			snprintf(tmp, sizeof tmp, "%.*s%.0d%s%c%%lln",
299				1+!dest, "%*", width, size_pfx[size+2], t);
300			cnt = 0;
301			if (fscanf(f, tmp, dest?dest:&cnt, &cnt) == -1)
302				goto input_fail;
303			else if (!cnt)
304				goto match_fail;
305			pos += cnt;
306			break;
307		default:
308			goto fmt_fail;
309		}
310
311		if (dest) matches++;
312	}
313	if (0) {
314fmt_fail:
315alloc_fail:
316input_fail:
317		if (!matches) matches--;
318match_fail:
319		if (alloc) {
320			free(s);
321			free(wcs);
322		}
323	}
324	FUNLOCK(f);
325	return matches;
326}
327
328weak_alias(vfwscanf,__isoc99_vfwscanf);
329