File: CDB_File.xs

package info (click to toggle)
libcdb-file-perl 0.84-2.1
  • links: PTS
  • area: main
  • in suites: sarge
  • size: 92 kB
  • ctags: 14
  • sloc: perl: 164; makefile: 48; sh: 12
file content (657 lines) | stat: -rw-r--r-- 14,557 bytes parent folder | download | duplicates (2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
/*

Most of this is reasonably straightforward.  The complications arise
when we are "iterating" over the CDB file, that is to say, using `keys'
or `values' or `each' to retrieve all the data in the file in order.
This interface stores extra data to allow us to track iterations: end
is a pointer to the end of data in the CDB file, and also a flag which
indicates whether we are iterating or not (note that the end of data
occurs at a position >= 2048); curkey is a copy of the current key;
curpos is the file offset of curkey; and fetch_advance is 0 for

    FIRSTKEY, fetch, NEXTKEY, fetch, NEXTKEY, fetch, ...

but 1 for

    FIRSTKEY, NEXTKEY, NEXTKEY, ..., fetch, fetch, fetch, ...

Don't tell the OO Police, but there are actually two different objects
called CDB_File.  One is created by TIEHASH, and accessed by the usual
tied hash methods (FETCH, FIRSTKEY, etc.).  The other is created by new,
and accessed by insert and finish.

In both cases, the object is a blessed reference to a scalar.  The
scalar contains either a struct cdbobj or a struct cdbmakeobj.

It gets a little messy in DESTROY: since this method will automatically
be called for both sorts of object, it distinguishes them by their
different sizes.

*/

#ifdef __cplusplus
extern "C" {
#endif

#include "EXTERN.h"
#include "perl.h"
#include "XSUB.h"

#include <sys/stat.h>
#include <sys/types.h>
#include <errno.h>
#include <fcntl.h>
#include <stdio.h>
#include <unistd.h>

/* We need to whistle up an error number for a file that is not a CDB
file.  The BSDish EFTYPE probably gives the most useful error message;
failing that we'll settle for the Single Unix Specification v2 EPROTO;
and finally the rather inappropriate, but universally(?) implented,
EINVAL. */
#ifdef EFTYPE
#else
#ifdef EPROTO
#define EFTYPE EPROTO
#else
#define EFTYPE EINVAL
#endif
#endif

/* These two provide backwards compatibility with perl 5.005. */
#ifndef WARN_UNINITIALIZED
#define ckWARN(x) dowarn
#define report_uninit() warn(warn_uninit)
#endif

#ifdef __cplusplus
}
#endif

struct cdb {
	PerlIO *f;  /* The file descriptor. */
	U32 end;    /* If non zero, the file offset of the first byte of hash tables. */
	SV *curkey; /* While iterating: a copy of the current key; */
	U32 curpos; /*                  the file offset of the current record. */
	int fetch_advance; /* the kludge */
	U32 size; /* initialized if map is nonzero */
	U32 loop; /* number of hash slots searched under this key */
	U32 khash; /* initialized if loop is nonzero */
	U32 kpos; /* initialized if loop is nonzero */
	U32 hpos; /* initialized if loop is nonzero */
	U32 hslots; /* initialized if loop is nonzero */
	U32 dpos; /* initialized if cdb_findnext() returns 1 */
	U32 dlen; /* initialized if cdb_findnext() returns 1 */
} ;

#define CDB_HPLIST 1000

struct cdb_hp { U32 h; U32 p; } ;

struct cdb_hplist {
	struct cdb_hp hp[CDB_HPLIST];
	struct cdb_hplist *next;
	int num;
} ;

struct cdb_make {
	PerlIO *f;            /* Handle of file being created. */
	char *fn;             /* Final name of file. */
	char *fntemp;         /* Temporary name of file. */
	char final[2048];
	char bspace[1024];
	U32 count[256];
	U32 start[256];
	struct cdb_hplist *head;
	struct cdb_hp *split; /* includes space for hash */
	struct cdb_hp *hash;
	U32 numentries;
	U32 pos;
	int fd;
} ;

static void writeerror() { croak("Write to CDB_File failed: %s", Strerror(errno)); }

static void readerror() { croak("Read of CDB_File failed: %s", Strerror(errno)); }

static void seekerror() { croak("Seek in CDB_File failed: %s", Strerror(errno)); }

static void nomem() { croak("Out of memory!"); }

static int cdb_make_start(struct cdb_make *c) {
	c->head = 0;
	c->split = 0;
	c->hash = 0;
	c->numentries = 0;
	c->pos = sizeof c->final;
	return PerlIO_seek(c->f, c->pos, SEEK_SET);
}

static int posplus(struct cdb_make *c, U32 len) {
	U32 newpos = c->pos + len;
	if (newpos < len) { errno = ENOMEM; return -1; }
	c->pos = newpos;
	return 0;
}

static int cdb_make_addend(struct cdb_make *c, unsigned int keylen, unsigned int datalen, U32 h) {
	struct cdb_hplist *head;

	head = c->head;
	if (!head || (head->num >= CDB_HPLIST)) {
		New(0xCDB, head, 1, struct cdb_hplist);
		head->num = 0;
		head->next = c->head;
		c->head = head;
	}
	head->hp[head->num].h = h;
	head->hp[head->num].p = c->pos;
	++head->num;
	++c->numentries;
	if (posplus(c, 8) == -1) return -1;
	if (posplus(c, keylen) == -1) return -1;
	if (posplus(c, datalen) == -1) return -1;
	return 0;
}

#define CDB_HASHSTART 5381

static U32 cdb_hashadd(U32 h, unsigned char c) {
	h += (h << 5);
	return h ^ c;
}

static U32 cdb_hash(char *buf, unsigned int len) {
	U32 h;

	h = CDB_HASHSTART;
	while (len) {
		h = cdb_hashadd(h,*buf++);
		--len;
	}
	return h;
}

static void uint32_pack(char s[4], U32 u) {
	s[0] = u & 255;
	u >>= 8;
	s[1] = u & 255;
	u >>= 8;
	s[2] = u & 255;
	s[3] = u >> 8;
}

static void uint32_unpack(char s[4], U32 *u) {
	U32 result;

	result = (unsigned char) s[3];
	result <<= 8;
	result += (unsigned char) s[2];
	result <<= 8;
	result += (unsigned char) s[1];
	result <<= 8;
	result += (unsigned char) s[0];

	*u = result;
}

static void cdb_findstart(struct cdb *c) {
	c->loop = 0;
}

static int cdb_read(struct cdb *c, char *buf, unsigned int len, U32 pos) {
	if (PerlIO_seek(c->f, pos, SEEK_SET) == -1) return -1;
	while (len > 0) {
		int r;
		do
			r = PerlIO_read(c->f, buf, len);
		while ((r == -1) && (errno == EINTR));
		if (r == -1) return -1;
		if (r == 0) {
			errno = EFTYPE;
			return -1;
		}
		buf += r;
		len -= r;
	}
	return 0;
}

static int match(struct cdb *c,char *key,unsigned int len, U32 pos) {
	char buf[32];
	int n;

	while (len > 0) {
		n = sizeof buf;
		if (n > len) n = len;
		if (cdb_read(c, buf, n, pos) == -1) return -1;
		if (memcmp(buf, key, n)) return 0;
		pos += n;
		key += n;
		len -= n;
	}
	return 1;
}

int cdb_findnext(struct cdb *c,char *key,unsigned int len) {
	char buf[8];
	U32 pos;
	U32 u;

  if (!c->loop) {
    u = cdb_hash(key,len);
    if (cdb_read(c,buf,8,(u << 3) & 2047) == -1) return -1;
    uint32_unpack(buf + 4,&c->hslots);
    if (!c->hslots) return 0;
    uint32_unpack(buf,&c->hpos);
    c->khash = u;
    u >>= 8;
    u %= c->hslots;
    u <<= 3;
    c->kpos = c->hpos + u;
  }

  while (c->loop < c->hslots) {
    if (cdb_read(c,buf,8,c->kpos) == -1) return -1;
    uint32_unpack(buf + 4,&pos);
    if (!pos) return 0;
    c->loop += 1;
    c->kpos += 8;
    if (c->kpos == c->hpos + (c->hslots << 3)) c->kpos = c->hpos;
    uint32_unpack(buf,&u);
    if (u == c->khash) {
      if (cdb_read(c,buf,8,pos) == -1) return -1;
      uint32_unpack(buf,&u);
      if (u == len)
	switch(match(c,key,len,pos + 8)) {
	  case -1:
	    return -1;
	  case 1:
	    uint32_unpack(buf + 4,&c->dlen);
	    c->dpos = pos + 8 + len;
	    return 1;
	}
    }
  }

  return 0;
}

static int cdb_find(struct cdb *c, char *key, unsigned int len) {
  cdb_findstart(c);
  return cdb_findnext(c,key,len);
}

static void iter_start(struct cdb *c) {
	char buf[4];

	c->curpos = 2048;
	if (cdb_read(c, buf, 4, 0) == -1) readerror();
	uint32_unpack(buf, &c->end);
	c->curkey = NEWSV(0xcdb, 1);
	c->fetch_advance = 0;
}

static int iter_key(struct cdb *c) {
	char buf[8];
	U32 klen;

	if (c->curpos < c->end) {
		if (cdb_read(c, buf, 8, c->curpos) == -1) readerror();
		uint32_unpack(buf, &klen);
		(void)SvPOK_only(c->curkey);
		SvGROW(c->curkey, klen); SvCUR_set(c->curkey, klen);
		if (cdb_read(c, SvPVX(c->curkey), klen, c->curpos + 8) == -1) readerror();
		return 1;
	}
	return 0;
}

static void iter_advance(struct cdb *c) {
	char buf[8];
	U32 klen, dlen;

	if (cdb_read(c, buf, 8, c->curpos) == -1) readerror();
	uint32_unpack(buf, &klen);
	uint32_unpack(buf + 4, &dlen);
	c->curpos += 8 + klen + dlen;
}

static void iter_end(struct cdb *c) {
	if (c->end != 0) {
		c->end = 0;
		SvREFCNT_dec(c->curkey);
	}
}

#define cdb_datapos(c) ((c)->dpos)
#define cdb_datalen(c) ((c)->dlen)

MODULE = CDB_File		PACKAGE = CDB_File	PREFIX = cdb_

SV *
cdb_TIEHASH(dbtype, filename)
	char *		dbtype
	char *		filename

	PROTOTYPE: $$

	CODE:
	struct cdb cdb;
	SV *cdbp;

	cdb.f = PerlIO_open(filename, "r");
	if (!cdb.f) XSRETURN_NO;
	cdb.end = 0;
	cdbp = newSVpv((char *)&cdb, sizeof(struct cdb));
	RETVAL = newRV_noinc(cdbp);
	sv_bless(RETVAL, gv_stashpv(dbtype, 0));
	/* Prevent the user stomping on the cdb. */
	SvREADONLY_on(cdbp);

	OUTPUT:
		RETVAL

SV *
cdb_FETCH(db, k, n = 0)
	SV *		db
	SV *		k
	unsigned int	n
	
	PROTOTYPE: $$;$

	PREINIT:
	struct cdb *this;
	PerlIO *f;
	char buf[8];
	int found;
	off_t pos;
	STRLEN klen, x;
	U32 klen0;
	char *kp;

	CODE:
	if (!SvOK(k)) {
		if (ckWARN(WARN_UNINITIALIZED)) report_uninit();
		XSRETURN_UNDEF;
	}
	this = (struct cdb *)SvPV(SvRV(db), PL_na);
	kp = SvPV(k, klen);
	if (this->end && sv_eq(this->curkey, k)) {
		if (cdb_read(this, buf, 8, this->curpos) == -1) readerror();
		uint32_unpack(buf + 4, &this->dlen);
		this->dpos = this->curpos + 8 + klen;
		if (this->fetch_advance) {
			iter_advance(this);
			if (!iter_key(this)) iter_end(this);
		}
		found = 1;
	} else {
		cdb_findstart(this);
		do {
			found = cdb_findnext(this, kp, klen);
			if ((found != 0) && (found != 1)) readerror();
		} while (found && n--);
	}
	ST(0) = sv_newmortal();
	if (found && sv_upgrade(ST(0), SVt_PV)) {
		U32 dlen = cdb_datalen(this);

		(void)SvPOK_only(ST(0));
		SvGROW(ST(0), dlen + 1); SvCUR_set(ST(0),  dlen);
		if (cdb_read(this, SvPVX(ST(0)), dlen, cdb_datapos(this)) == -1) readerror();
		SvPV(ST(0), PL_na)[dlen] = '\0';
	}

int
cdb_EXISTS(db, k)
	SV *		db
	SV *		k

	PROTOTYPE: $$

	CODE:
	struct cdb *this;
	STRLEN klen;
	char *kp;

	if (!SvOK(k)) {
		if (ckWARN(WARN_UNINITIALIZED)) report_uninit();
		XSRETURN_NO;
	}
	this = (struct cdb *)SvPV(SvRV(db), PL_na);
	kp = SvPV(k, klen);
	RETVAL = cdb_find(this, kp, klen);
	if (RETVAL != 0 && RETVAL != 1) readerror();

	OUTPUT:
		RETVAL

void
cdb_DESTROY(db)
	SV *		db

	PROTOTYPE: $

	CODE:
	struct cdb *this;

	if (SvCUR(SvRV(db)) == sizeof(struct cdb)) { /* It came from TIEHASH. */
		this = (struct cdb *)SvPV(SvRV(db), PL_na);
		iter_end(this);
		PerlIO_close(this->f); /* close() on O_RDONLY can't fail */
	}

SV *
cdb_FIRSTKEY(db)
	SV *		db

	PROTOTYPE: $

	CODE:
	struct cdb *this;
	char buf[8];
	U32 klen;

	this = (struct cdb *)SvPV(SvRV(db), PL_na);

	iter_start(this);
	if (iter_key(this))
		ST(0) = sv_mortalcopy(this->curkey);
	else
		XSRETURN_UNDEF; /* empty database */

SV *
cdb_NEXTKEY(db, k)
	SV *		db
	SV *		k

	PROTOTYPE: $$

	CODE:
	struct cdb *this;
	char buf[8], *kp;
	int found;
	off_t pos;
	U32 dlen, klen0;
	STRLEN klen1;

	if (!SvOK(k)) {
		if (ckWARN(WARN_UNINITIALIZED)) report_uninit();
		XSRETURN_UNDEF;
	}
	this = (struct cdb *)SvPV(SvRV(db), PL_na);
	if (this->end == 0 || !sv_eq(this->curkey, k))
		croak("Use CDB_File::FIRSTKEY before CDB_File::NEXTKEY");
	iter_advance(this);
	if (iter_key(this))
		ST(0) = sv_mortalcopy(this->curkey);
	else {
		iter_start(this);
		(void)iter_key(this); /* prepare curkey for FETCH */
		this->fetch_advance = 1;
		XSRETURN_UNDEF;
	}

SV *
cdb_new(this, fn, fntemp)
	char *		this
	char *		fn
	char *		fntemp

	PROTOTYPE: $$$

	CODE:
	SV *cdbmp;
	struct cdb_make cdbmake;
	int i;
	mode_t oldum;

	oldum = umask(0222);
	cdbmake.f = PerlIO_open(fntemp, "w");
	umask(oldum);
	if (!cdbmake.f) XSRETURN_UNDEF;

	if (cdb_make_start(&cdbmake) < 0) XSRETURN_UNDEF;

	/* Oh, for referential transparency. */
	New(0, cdbmake.fn, strlen(fn) + 1, char);
	New(0, cdbmake.fntemp, strlen(fntemp) + 1, char);
	strncpy(cdbmake.fn, fn, strlen(fn) + 1);
	strncpy(cdbmake.fntemp, fntemp, strlen(fntemp) + 1);

	cdbmp = newSVpv((char *)&cdbmake, sizeof(struct cdb_make));
	RETVAL = newRV_noinc(cdbmp);
	sv_bless(RETVAL, gv_stashpv(this, 0));

	OUTPUT:
		RETVAL

void
cdb_insert(cdbmake, k, v)
	SV *		cdbmake
	SV *		k
	SV *		v

	PROTOTYPE: $$$

	CODE:
	char *kp, *vp, packbuf[8];
	int c, i;
	STRLEN klen, vlen;
	struct cdb_make *this;
	U32 h;

	this = (struct cdb_make *)SvPV(SvRV(cdbmake), PL_na);
	kp = SvPV(k, klen); vp = SvPV(v, vlen);
	uint32_pack(packbuf, klen);
	uint32_pack(packbuf + 4, vlen);

	if (PerlIO_write(this->f, packbuf, 8) < 8) writeerror();

	h = cdb_hash(kp, klen);
	if (PerlIO_write(this->f, kp, klen) < klen) writeerror();
	if (PerlIO_write(this->f, vp, vlen) < vlen) writeerror();

	if (cdb_make_addend(this, klen, vlen, h) == -1) nomem();


int
cdb_finish(cdbmake)
	SV *		cdbmake;

	PROTOTYPE: $

	CODE:
	char buf[8];
	int i;
	struct cdb_make *this;
	U32 len, u;
	U32 count, memsize, where;
	struct cdb_hplist *x;
	struct cdb_hp *hp;

	this = (struct cdb_make *)SvPV(SvRV(cdbmake), PL_na);

	for (i = 0; i < 256; ++i)
		this->count[i] = 0;

	for (x = this->head; x; x = x->next) {
		i = x->num;
		while (i--)
			++this->count[255 & x->hp[i].h];
		}

	memsize = 1;
	for (i = 0; i < 256; ++i) {
		u = this->count[i] * 2;
		if (u > memsize)
			memsize = u;
	}

	memsize += this->numentries; /* no overflow possible up to now */
	u = (U32) 0 - (U32) 1;
	u /= sizeof(struct cdb_hp);
	if (memsize > u) { errno = ENOMEM; XSRETURN_UNDEF; }

	New(0xCDB, this->split, memsize, struct cdb_hp);

	this->hash = this->split + this->numentries;

	u = 0;
	for (i = 0; i < 256; ++i) {
		u += this->count[i]; /* bounded by numentries, so no overflow */
		this->start[i] = u;
	}

	for (x = this->head; x; x = x->next) {
		i = x->num;
		while (i--)
			this->split[--this->start[255 & x->hp[i].h]] = x->hp[i];
	}

	for (i = 0; i < 256; ++i) {
		count = this->count[i];

		len = count + count; /* no overflow possible */
		uint32_pack(this->final + 8 * i, this->pos);
		uint32_pack(this->final + 8 * i + 4, len);

		for (u = 0; u < len; ++u)
			this->hash[u].h = this->hash[u].p = 0;

		hp = this->split + this->start[i];
		for (u = 0; u < count; ++u) {
			where = (hp->h >> 8) % len;
			while (this->hash[where].p)
				if (++where == len)
					where = 0;
			this->hash[where] = *hp++;
		}

		for (u = 0; u < len; ++u) {
			uint32_pack(buf, this->hash[u].h);
			uint32_pack(buf + 4, this->hash[u].p);
			if (PerlIO_write(this->f, buf, 8) == -1) XSRETURN_UNDEF;
			if (posplus(this, 8) == -1) XSRETURN_UNDEF;
		}
	}

	if (PerlIO_flush(this->f) == EOF) writeerror();
	PerlIO_rewind(this->f);

	if (PerlIO_write(this->f, this->final, sizeof this->final) < sizeof this->final) writeerror();
	if (PerlIO_flush(this->f) == EOF) writeerror();

	if (fsync(PerlIO_fileno(this->f)) == -1) XSRETURN_NO;
	if (PerlIO_close(this->f) == EOF) XSRETURN_NO;

	if (rename(this->fntemp, this->fn)) XSRETURN_NO;
	
	Safefree(this->fn);
	Safefree(this->fntemp);

	RETVAL = 1;

	OUTPUT:
		RETVAL