[ViewVC] Diff of: cvs/CBOR-XS/XS.xs

Comparing CBOR-XS/XS.xs (file contents):
Revision 1.27 by root, Fri Nov 22 15:28:38 2013 UTC vs.
Revision 1.52 by root, Sun Apr 24 19:31:55 2016 UTC

…		…
7	#include <stdlib.h>	7	#include <stdlib.h>
8	#include <stdio.h>	8	#include <stdio.h>
9	#include <limits.h>	9	#include <limits.h>
10	#include <float.h>	10	#include <float.h>
11		11
		12	#define ECB_NO_THREADS 1
12	#include "ecb.h"	13	#include "ecb.h"
13		14
14	// compatibility with perl <5.18	15	// compatibility with perl <5.18
15	#ifndef HvNAMELEN_get	16	#ifndef HvNAMELEN_get
16	# define HvNAMELEN_get(hv) strlen (HvNAME (hv))	17	# define HvNAMELEN_get(hv) strlen (HvNAME (hv))
…		…
19	# define HvNAMELEN(hv) HvNAMELEN_get (hv)	20	# define HvNAMELEN(hv) HvNAMELEN_get (hv)
20	#endif	21	#endif
21	#ifndef HvNAMEUTF8	22	#ifndef HvNAMEUTF8
22	# define HvNAMEUTF8(hv) 0	23	# define HvNAMEUTF8(hv) 0
23	#endif	24	#endif
		25	#ifndef SvREFCNT_dec_NN
		26	# define SvREFCNT_dec_NN(sv) SvREFCNT_dec (sv)
		27	#endif
		28
		29	// known major and minor types
		30	enum cbor_type
		31	{
		32	MAJOR_SHIFT = 5,
		33	MINOR_MASK = 0x1f,
		34
		35	MAJOR_POS_INT = 0 << MAJOR_SHIFT,
		36	MAJOR_NEG_INT = 1 << MAJOR_SHIFT,
		37	MAJOR_BYTES = 2 << MAJOR_SHIFT,
		38	MAJOR_TEXT = 3 << MAJOR_SHIFT,
		39	MAJOR_ARRAY = 4 << MAJOR_SHIFT,
		40	MAJOR_MAP = 5 << MAJOR_SHIFT,
		41	MAJOR_TAG = 6 << MAJOR_SHIFT,
		42	MAJOR_MISC = 7 << MAJOR_SHIFT,
		43
		44	// INT/STRING/ARRAY/MAP subtypes
		45	LENGTH_EXT1 = 24,
		46	LENGTH_EXT2 = 25,
		47	LENGTH_EXT4 = 26,
		48	LENGTH_EXT8 = 27,
		49
		50	// SIMPLE types (effectively MISC subtypes)
		51	SIMPLE_FALSE = 20,
		52	SIMPLE_TRUE = 21,
		53	SIMPLE_NULL = 22,
		54	SIMPLE_UNDEF = 23,
		55
		56	// MISC subtype (unused)
		57	MISC_EXT1 = 24,
		58	MISC_FLOAT16 = 25,
		59	MISC_FLOAT32 = 26,
		60	MISC_FLOAT64 = 27,
		61
		62	// BYTES/TEXT/ARRAY/MAP
		63	MINOR_INDEF = 31,
		64	};
24		65
25	// known tags	66	// known tags
26	enum cbor_tag	67	enum cbor_tag
27	{	68	{
28	// inofficial extensions (pending iana registration)	69	// extensions
		70	CBOR_TAG_STRINGREF = 25, // http://cbor.schmorp.de/stringref
29	CBOR_TAG_PERL_OBJECT = 24, // http://cbor.schmorp.de/perl-object	71	CBOR_TAG_PERL_OBJECT = 26, // http://cbor.schmorp.de/perl-object
30	CBOR_TAG_GENERIC_OBJECT = 25, // http://cbor.schmorp.de/generic-object	72	CBOR_TAG_GENERIC_OBJECT = 27, // http://cbor.schmorp.de/generic-object
31	CBOR_TAG_VALUE_SHAREABLE = 26, // http://cbor.schmorp.de/value-sharing	73	CBOR_TAG_VALUE_SHAREABLE = 28, // http://cbor.schmorp.de/value-sharing
32	CBOR_TAG_VALUE_SHAREDREF = 27, // http://cbor.schmorp.de/value-sharing	74	CBOR_TAG_VALUE_SHAREDREF = 29, // http://cbor.schmorp.de/value-sharing
33	CBOR_TAG_STRINGREF_NAMESPACE = 65537, // http://cbor.schmorp.de/stringref	75	CBOR_TAG_STRINGREF_NAMESPACE = 256, // http://cbor.schmorp.de/stringref
34	CBOR_TAG_STRINGREF = 28, // http://cbor.schmorp.de/stringref
35	CBOR_TAG_INDIRECTION = 22098, // http://cbor.schmorp.de/indirection	76	CBOR_TAG_INDIRECTION = 22098, // http://cbor.schmorp.de/indirection
36		77
37	// rfc7049	78	// rfc7049
38	CBOR_TAG_DATETIME = 0, // rfc4287, utf-8	79	CBOR_TAG_DATETIME = 0, // rfc4287, utf-8
39	CBOR_TAG_TIMESTAMP = 1, // unix timestamp, any	80	CBOR_TAG_TIMESTAMP = 1, // unix timestamp, any
40	CBOR_TAG_POS_BIGNUM = 2, // byte string	81	CBOR_TAG_POS_BIGNUM = 2, // byte string
41	CBOR_TAG_NEG_BIGNUM = 3, // byte string	82	CBOR_TAG_NEG_BIGNUM = 3, // byte string
42	CBOR_TAG_DECIMAL = 4, // decimal fraction, array	83	CBOR_TAG_DECIMAL = 4, // decimal fraction, array
43	CBOR_TAG_BIGFLOAT = 5, // array	84	CBOR_TAG_BIGFLOAT = 5, // array
44		85
45	CBOR_TAG_CONV_B64U = 21, // base64url, any	86	CBOR_TAG_CONV_B64U = 21, // base64url, any
46	CBOR_TAG_CONV_B64 = 22, // base64, any	87	CBOR_TAG_CONV_B64 = 22, // base64, any
47	CBOR_TAG_CONV_HEX = 23, // base16, any	88	CBOR_TAG_CONV_HEX = 23, // base16, any
48	CBOR_TAG_CBOR = 24, // embedded cbor, byte string	89	CBOR_TAG_CBOR = 24, // embedded cbor, byte string
49		90
50	CBOR_TAG_URI = 32, // URI rfc3986, utf-8	91	CBOR_TAG_URI = 32, // URI rfc3986, utf-8
51	CBOR_TAG_B64U = 33, // base64url rfc4648, utf-8	92	CBOR_TAG_B64U = 33, // base64url rfc4648, utf-8
52	CBOR_TAG_B64 = 34, // base6 rfc46484, utf-8	93	CBOR_TAG_B64 = 34, // base6 rfc46484, utf-8
53	CBOR_TAG_REGEX = 35, // regex pcre/ecma262, utf-8	94	CBOR_TAG_REGEX = 35, // regex pcre/ecma262, utf-8
54	CBOR_TAG_MIME = 36, // mime message rfc2045, utf-8	95	CBOR_TAG_MIME = 36, // mime message rfc2045, utf-8
55		96
56	CBOR_TAG_MAGIC = 55799 // self-describe cbor	97	CBOR_TAG_MAGIC = 55799, // self-describe cbor
57	};	98	};
58		99
59	#define F_SHRINK 0x00000001UL	100	#define F_SHRINK 0x00000001UL
60	#define F_ALLOW_UNKNOWN 0x00000002UL	101	#define F_ALLOW_UNKNOWN 0x00000002UL
61	#define F_ALLOW_SHARING 0x00000004UL //TODO	102	#define F_ALLOW_SHARING 0x00000004UL
62	#define F_ALLOW_STRINGREF 0x00000008UL //TODO	103	#define F_ALLOW_CYCLES 0x00000008UL
		104	#define F_PACK_STRINGS 0x00000010UL
		105	#define F_UTF8_STRINGS 0x00000020UL
		106	#define F_VALIDATE_UTF8 0x00000040UL
63		107
64	#define INIT_SIZE 32 // initial scalar size to be allocated	108	#define INIT_SIZE 32 // initial scalar size to be allocated
65		109
66	#define SB do {	110	#define SB do {
67	#define SE } while (0)	111	#define SE } while (0)
…		…
86	typedef struct {	130	typedef struct {
87	U32 flags;	131	U32 flags;
88	U32 max_depth;	132	U32 max_depth;
89	STRLEN max_size;	133	STRLEN max_size;
90	SV *filter;	134	SV *filter;
		135
		136	// for the incremental parser
		137	STRLEN incr_pos; // the current offset into the text
		138	STRLEN incr_need; // minimum bytes needed to decode
		139	AV *incr_count; // for every nesting level, the number of outstanding values, or -1 for indef.
91	} CBOR;	140	} CBOR;
92		141
93	ecb_inline void	142	ecb_inline void
94	cbor_init (CBOR *cbor)	143	cbor_init (CBOR *cbor)
95	{	144	{
…		…
99		148
100	ecb_inline void	149	ecb_inline void
101	cbor_free (CBOR *cbor)	150	cbor_free (CBOR *cbor)
102	{	151	{
103	SvREFCNT_dec (cbor->filter);	152	SvREFCNT_dec (cbor->filter);
		153	SvREFCNT_dec (cbor->incr_count);
104	}	154	}
105		155
106	/////////////////////////////////////////////////////////////////////////////	156	/////////////////////////////////////////////////////////////////////////////
107	// utility functions	157	// utility functions
108		158
…		…
138	{	188	{
139	return idx > 23	189	return idx > 23
140	? idx > 0xffU	190	? idx > 0xffU
141	? idx > 0xffffU	191	? idx > 0xffffU
142	? idx > 0xffffffffU	192	? idx > 0xffffffffU
		193	? 11
143	? 7	194	: 7
144	: 6
145	: 5	195	: 5
146	: 4	196	: 4
147	: 3;	197	: 3;
148	}	198	}
149		199
…		…
186	static void	236	static void
187	encode_uint (enc_t *enc, int major, UV len)	237	encode_uint (enc_t *enc, int major, UV len)
188	{	238	{
189	need (enc, 9);	239	need (enc, 9);
190		240
191	if (len < 24)	241	if (ecb_expect_true (len < LENGTH_EXT1))
192	*enc->cur++ = major \| len;	242	*enc->cur++ = major \| len;
193	else if (len <= 0xff)	243	else if (ecb_expect_true (len <= 0xffU))
194	{	244	{
195	*enc->cur++ = major \| 24;	245	*enc->cur++ = major \| LENGTH_EXT1;
196	*enc->cur++ = len;	246	*enc->cur++ = len;
197	}	247	}
198	else if (len <= 0xffff)	248	else if (len <= 0xffffU)
199	{	249	{
200	*enc->cur++ = major \| 25;	250	*enc->cur++ = major \| LENGTH_EXT2;
201	*enc->cur++ = len >> 8;	251	*enc->cur++ = len >> 8;
202	*enc->cur++ = len;	252	*enc->cur++ = len;
203	}	253	}
204	else if (len <= 0xffffffff)	254	else if (len <= 0xffffffffU)
205	{	255	{
206	*enc->cur++ = major \| 26;	256	*enc->cur++ = major \| LENGTH_EXT4;
207	*enc->cur++ = len >> 24;	257	*enc->cur++ = len >> 24;
208	*enc->cur++ = len >> 16;	258	*enc->cur++ = len >> 16;
209	*enc->cur++ = len >> 8;	259	*enc->cur++ = len >> 8;
210	*enc->cur++ = len;	260	*enc->cur++ = len;
211	}	261	}
212	else	262	else
213	{	263	{
214	*enc->cur++ = major \| 27;	264	*enc->cur++ = major \| LENGTH_EXT8;
215	*enc->cur++ = len >> 56;	265	*enc->cur++ = len >> 56;
216	*enc->cur++ = len >> 48;	266	*enc->cur++ = len >> 48;
217	*enc->cur++ = len >> 40;	267	*enc->cur++ = len >> 40;
218	*enc->cur++ = len >> 32;	268	*enc->cur++ = len >> 32;
219	*enc->cur++ = len >> 24;	269	*enc->cur++ = len >> 24;
…		…
224	}	274	}
225		275
226	ecb_inline void	276	ecb_inline void
227	encode_tag (enc_t *enc, UV tag)	277	encode_tag (enc_t *enc, UV tag)
228	{	278	{
229	encode_uint (enc, 0xc0, tag);	279	encode_uint (enc, MAJOR_TAG, tag);
		280	}
		281
		282	ecb_inline void
		283	encode_str (enc_t enc, int utf8, char str, STRLEN len)
		284	{
		285	if (ecb_expect_false (enc->cbor.flags & F_UTF8_STRINGS))
		286	if (!utf8)
		287	{
		288	// exceptional path for bytze strings that need to be utf8-encoded
		289	STRLEN ulen = len;
		290	U8 p, pend = (U8 *)str + len;
		291
		292	for (p = (U8 *)str; p < pend; ++p)
		293	ulen += *p >> 7; // count set high bits
		294
		295	encode_uint (enc, MAJOR_TEXT, ulen);
		296
		297	need (enc, ulen);
		298	for (p = (U8 *)str; p < pend; ++p)
		299	if (*p < 0x80)
		300	enc->cur++ = p;
		301	else
		302	{
		303	enc->cur++ = 0xc0 + (p >> 6);
		304	enc->cur++ = 0x80 + (p & 63);
		305	}
		306
		307	return;
		308	}
		309
		310	encode_uint (enc, utf8 ? MAJOR_TEXT : MAJOR_BYTES, len);
		311	need (enc, len);
		312	memcpy (enc->cur, str, len);
		313	enc->cur += len;
230	}	314	}
231		315
232	static void	316	static void
233	encode_str (enc_t enc, int utf8, char str, STRLEN len)	317	encode_strref (enc_t enc, int utf8, char str, STRLEN len)
234	{	318	{
235	if (ecb_expect_false (enc->cbor.flags & F_ALLOW_STRINGREF))	319	if (ecb_expect_false (enc->cbor.flags & F_PACK_STRINGS))
236	{	320	{
237	SV **svp = hv_fetch (enc->stringref[!!utf8], str, len, 1);	321	SV **svp = hv_fetch (enc->stringref[!!utf8], str, len, 1);
238		322
239	if (SvOK (*svp))	323	if (SvOK (*svp))
240	{	324	{
241	// already registered, use stringref	325	// already registered, use stringref
242	encode_tag (enc, CBOR_TAG_STRINGREF);	326	encode_tag (enc, CBOR_TAG_STRINGREF);
243	encode_uint (enc, 0x00, SvUV (*svp));	327	encode_uint (enc, MAJOR_POS_INT, SvUV (*svp));
244	return;	328	return;
245	}	329	}
246	else if (len >= minimum_string_length (enc->stringref_idx))	330	else if (len >= minimum_string_length (enc->stringref_idx))
247	{	331	{
248	// register only	332	// register only
249	sv_setuv (*svp, enc->stringref_idx);	333	sv_setuv (*svp, enc->stringref_idx);
250	++enc->stringref_idx;	334	++enc->stringref_idx;
251	}	335	}
252	}	336	}
253		337
254	encode_uint (enc, utf8 ? 0x60 : 0x40, len);	338	encode_str (enc, utf8, str, len);
255	need (enc, len);
256	memcpy (enc->cur, str, len);
257	enc->cur += len;
258	}	339	}
259		340
260	static void encode_sv (enc_t enc, SV sv);	341	static void encode_sv (enc_t enc, SV sv);
261		342
262	static void	343	static void
…		…
267	if (enc->depth >= enc->cbor.max_depth)	348	if (enc->depth >= enc->cbor.max_depth)
268	croak (ERR_NESTING_EXCEEDED);	349	croak (ERR_NESTING_EXCEEDED);
269		350
270	++enc->depth;	351	++enc->depth;
271		352
272	encode_uint (enc, 0x80, len + 1);	353	encode_uint (enc, MAJOR_ARRAY, len + 1);
273		354
		355	if (SvMAGICAL (av))
274	for (i = 0; i <= len; ++i)	356	for (i = 0; i <= len; ++i)
275	{	357	{
276	SV **svp = av_fetch (av, i, 0);	358	SV **svp = av_fetch (av, i, 0);
277	encode_sv (enc, svp ? *svp : &PL_sv_undef);	359	encode_sv (enc, svp ? *svp : &PL_sv_undef);
278	}	360	}
		361	else
		362	for (i = 0; i <= len; ++i)
		363	{
		364	SV *sv = AvARRAY (av)[i];
		365	encode_sv (enc, sv ? sv : &PL_sv_undef);
		366	}
279		367
280	--enc->depth;	368	--enc->depth;
281	}	369	}
282		370
283	static void	371	static void
…		…
292		380
293	int pairs = hv_iterinit (hv);	381	int pairs = hv_iterinit (hv);
294	int mg = SvMAGICAL (hv);	382	int mg = SvMAGICAL (hv);
295		383
296	if (mg)	384	if (mg)
297	encode_ch (enc, 0xa0 \| 31);	385	encode_ch (enc, MAJOR_MAP \| MINOR_INDEF);
298	else	386	else
299	encode_uint (enc, 0xa0, pairs);	387	encode_uint (enc, MAJOR_MAP, pairs);
300		388
301	while ((he = hv_iternext (hv)))	389	while ((he = hv_iternext (hv)))
302	{	390	{
303	if (HeKLEN (he) == HEf_SVKEY)	391	if (HeKLEN (he) == HEf_SVKEY)
304	encode_sv (enc, HeSVKEY (he));	392	encode_sv (enc, HeSVKEY (he));
305	else	393	else
306	encode_str (enc, HeKUTF8 (he), HeKEY (he), HeKLEN (he));	394	encode_strref (enc, HeKUTF8 (he), HeKEY (he), HeKLEN (he));
307		395
308	encode_sv (enc, ecb_expect_false (mg) ? hv_iterval (hv, he) : HeVAL (he));	396	encode_sv (enc, ecb_expect_false (mg) ? hv_iterval (hv, he) : HeVAL (he));
309	}	397	}
310		398
311	if (mg)	399	if (mg)
312	encode_ch (enc, 0xe0 \| 31);	400	encode_ch (enc, MAJOR_MISC \| MINOR_INDEF);
313		401
314	--enc->depth;	402	--enc->depth;
315	}	403	}
316		404
317	// encode objects, arrays and special \0=false and \1=true values.	405	// encode objects, arrays and special \0=false and \1=true values.
318	static void	406	static void
319	encode_rv (enc_t enc, SV sv)	407	encode_rv (enc_t enc, SV sv)
320	{	408	{
321	SvGETMAGIC (sv);	409	SvGETMAGIC (sv);
322
323	if (ecb_expect_false (enc->cbor.flags & F_ALLOW_SHARING)
324	&& ecb_expect_false (SvREFCNT (sv) > 1))
325	{
326	if (!enc->shareable)
327	enc->shareable = (HV )sv_2mortal ((SV )newHV ());
328
329	SV *svp = hv_fetch (enc->shareable, (char )&sv, sizeof (sv), 1);
330
331	if (SvOK (*svp))
332	{
333	encode_tag (enc, CBOR_TAG_VALUE_SHAREDREF);
334	encode_uint (enc, 0x00, SvUV (*svp));
335	return;
336	}
337	else
338	{
339	sv_setuv (*svp, enc->shareable_idx);
340	++enc->shareable_idx;
341	encode_tag (enc, CBOR_TAG_VALUE_SHAREABLE);
342	}
343	}
344		410
345	svtype svt = SvTYPE (sv);	411	svtype svt = SvTYPE (sv);
346		412
347	if (ecb_expect_false (SvOBJECT (sv)))	413	if (ecb_expect_false (SvOBJECT (sv)))
348	{	414	{
…		…
355	HV *tagged_stash = !CBOR_SLOW \|\| cbor_tagged_stash	421	HV *tagged_stash = !CBOR_SLOW \|\| cbor_tagged_stash
356	? cbor_tagged_stash	422	? cbor_tagged_stash
357	: gv_stashpv ("CBOR::XS::Tagged" , 1);	423	: gv_stashpv ("CBOR::XS::Tagged" , 1);
358		424
359	HV *stash = SvSTASH (sv);	425	HV *stash = SvSTASH (sv);
360	GV *method;
361		426
362	if (stash == boolean_stash)	427	if (stash == boolean_stash)
363	encode_ch (enc, SvIV (sv) ? 0xe0 \| 21 : 0xe0 \| 20);	428	{
		429	encode_ch (enc, SvIV (sv) ? MAJOR_MISC \| SIMPLE_TRUE : MAJOR_MISC \| SIMPLE_FALSE);
		430	return;
		431	}
364	else if (stash == error_stash)	432	else if (stash == error_stash)
365	encode_ch (enc, 0xe0 \| 23);	433	{
		434	encode_ch (enc, MAJOR_MISC \| SIMPLE_UNDEF);
		435	return;
		436	}
366	else if (stash == tagged_stash)	437	else if (stash == tagged_stash)
367	{	438	{
368	if (svt != SVt_PVAV)	439	if (svt != SVt_PVAV)
369	croak ("encountered CBOR::XS::Tagged object that isn't an array");	440	croak ("encountered CBOR::XS::Tagged object that isn't an array");
370		441
371	encode_uint (enc, 0xc0, SvUV (av_fetch ((AV )sv, 0, 1)));	442	encode_uint (enc, MAJOR_TAG, SvUV (av_fetch ((AV )sv, 0, 1)));
372	encode_sv (enc, av_fetch ((AV )sv, 1, 1));	443	encode_sv (enc, av_fetch ((AV )sv, 1, 1));
		444
		445	return;
		446	}
		447	}
		448
		449	if (ecb_expect_false (SvREFCNT (sv) > 1)
		450	&& ecb_expect_false (enc->cbor.flags & F_ALLOW_SHARING))
		451	{
		452	if (!enc->shareable)
		453	enc->shareable = (HV )sv_2mortal ((SV )newHV ());
		454
		455	SV *svp = hv_fetch (enc->shareable, (char )&sv, sizeof (sv), 1);
		456
		457	if (SvOK (*svp))
373	}	458	{
		459	encode_tag (enc, CBOR_TAG_VALUE_SHAREDREF);
		460	encode_uint (enc, MAJOR_POS_INT, SvUV (*svp));
		461	return;
		462	}
		463	else
		464	{
		465	sv_setuv (*svp, enc->shareable_idx);
		466	++enc->shareable_idx;
		467	encode_tag (enc, CBOR_TAG_VALUE_SHAREABLE);
		468	}
		469	}
		470
		471	if (ecb_expect_false (SvOBJECT (sv)))
		472	{
		473	HV *stash = SvSTASH (sv);
		474	GV *method;
		475
374	else if ((method = gv_fetchmethod_autoload (stash, "TO_CBOR", 0)))	476	if ((method = gv_fetchmethod_autoload (stash, "TO_CBOR", 0)))
375	{	477	{
376	dSP;	478	dSP;
377		479
378	ENTER; SAVETMPS; PUSHMARK (SP);	480	ENTER; SAVETMPS;
		481	PUSHMARK (SP);
379	// we re-bless the reference to get overload and other niceties right	482	// we re-bless the reference to get overload and other niceties right
380	XPUSHs (sv_bless (sv_2mortal (newRV_inc (sv)), stash));	483	XPUSHs (sv_bless (sv_2mortal (newRV_inc (sv)), stash));
381		484
382	PUTBACK;	485	PUTBACK;
383	// G_SCALAR ensures that return value is 1	486	// G_SCALAR ensures that return value is 1
…		…
396	}	499	}
397	else if ((method = gv_fetchmethod_autoload (stash, "FREEZE", 0)) != 0)	500	else if ((method = gv_fetchmethod_autoload (stash, "FREEZE", 0)) != 0)
398	{	501	{
399	dSP;	502	dSP;
400		503
401	ENTER; SAVETMPS; PUSHMARK (SP);	504	ENTER; SAVETMPS;
		505	SAVESTACK_POS ();
		506	PUSHMARK (SP);
402	EXTEND (SP, 2);	507	EXTEND (SP, 2);
403	// we re-bless the reference to get overload and other niceties right	508	// we re-bless the reference to get overload and other niceties right
404	PUSHs (sv_bless (sv_2mortal (newRV_inc (sv)), stash));	509	PUSHs (sv_bless (sv_2mortal (newRV_inc (sv)), stash));
405	PUSHs (sv_cbor);	510	PUSHs (sv_cbor);
406		511
…		…
411	// catch this surprisingly common error	516	// catch this surprisingly common error
412	if (count == 1 && SvROK (TOPs) && SvRV (TOPs) == sv)	517	if (count == 1 && SvROK (TOPs) && SvRV (TOPs) == sv)
413	croak ("%s::FREEZE(CBOR) method returned same object as was passed instead of a new one", HvNAME (stash));	518	croak ("%s::FREEZE(CBOR) method returned same object as was passed instead of a new one", HvNAME (stash));
414		519
415	encode_tag (enc, CBOR_TAG_PERL_OBJECT);	520	encode_tag (enc, CBOR_TAG_PERL_OBJECT);
416	encode_uint (enc, 0x80, count + 1);	521	encode_uint (enc, MAJOR_ARRAY, count + 1);
417	encode_str (enc, HvNAMEUTF8 (stash), HvNAME (stash), HvNAMELEN (stash));	522	encode_strref (enc, HvNAMEUTF8 (stash), HvNAME (stash), HvNAMELEN (stash));
418		523
419	while (count)	524	while (count)
420	encode_sv (enc, SP[1 - count--]);	525	encode_sv (enc, SP[1 - count--]);
421		526
422	PUTBACK;	527	PUTBACK;
…		…
443	{	548	{
444	double nv = SvNVX (sv);	549	double nv = SvNVX (sv);
445		550
446	need (enc, 9);	551	need (enc, 9);
447		552
448	if (ecb_expect_false (nv == (U32)nv))	553	if (ecb_expect_false (nv == (NV)(U32)nv))
449	encode_uint (enc, 0x00, (U32)nv);	554	encode_uint (enc, MAJOR_POS_INT, (U32)nv);
450	//TODO: maybe I32?	555	//TODO: maybe I32?
451	else if (ecb_expect_false (nv == (float)nv))	556	else if (ecb_expect_false (nv == (float)nv))
452	{	557	{
453	uint32_t fp = ecb_float_to_binary32 (nv);	558	uint32_t fp = ecb_float_to_binary32 (nv);
454		559
455	*enc->cur++ = 0xe0 \| 26;	560	*enc->cur++ = MAJOR_MISC \| MISC_FLOAT32;
456		561
457	if (!ecb_big_endian ())	562	if (!ecb_big_endian ())
458	fp = ecb_bswap32 (fp);	563	fp = ecb_bswap32 (fp);
459		564
460	memcpy (enc->cur, &fp, 4);	565	memcpy (enc->cur, &fp, 4);
…		…
462	}	567	}
463	else	568	else
464	{	569	{
465	uint64_t fp = ecb_double_to_binary64 (nv);	570	uint64_t fp = ecb_double_to_binary64 (nv);
466		571
467	*enc->cur++ = 0xe0 \| 27;	572	*enc->cur++ = MAJOR_MISC \| MISC_FLOAT64;
468		573
469	if (!ecb_big_endian ())	574	if (!ecb_big_endian ())
470	fp = ecb_bswap64 (fp);	575	fp = ecb_bswap64 (fp);
471		576
472	memcpy (enc->cur, &fp, 8);	577	memcpy (enc->cur, &fp, 8);
…		…
481		586
482	if (SvPOKp (sv))	587	if (SvPOKp (sv))
483	{	588	{
484	STRLEN len;	589	STRLEN len;
485	char *str = SvPV (sv, len);	590	char *str = SvPV (sv, len);
486	encode_str (enc, SvUTF8 (sv), str, len);	591	encode_strref (enc, SvUTF8 (sv), str, len);
487	}	592	}
488	else if (SvNOKp (sv))	593	else if (SvNOKp (sv))
489	encode_nv (enc, sv);	594	encode_nv (enc, sv);
490	else if (SvIOKp (sv))	595	else if (SvIOKp (sv))
491	{	596	{
492	if (SvIsUV (sv))	597	if (SvIsUV (sv))
493	encode_uint (enc, 0x00, SvUVX (sv));	598	encode_uint (enc, MAJOR_POS_INT, SvUVX (sv));
494	else if (SvIVX (sv) >= 0)	599	else if (SvIVX (sv) >= 0)
495	encode_uint (enc, 0x00, SvIVX (sv));	600	encode_uint (enc, MAJOR_POS_INT, SvIVX (sv));
496	else	601	else
497	encode_uint (enc, 0x20, -(SvIVX (sv) + 1));	602	encode_uint (enc, MAJOR_NEG_INT, -(SvIVX (sv) + 1));
498	}	603	}
499	else if (SvROK (sv))	604	else if (SvROK (sv))
500	encode_rv (enc, SvRV (sv));	605	encode_rv (enc, SvRV (sv));
501	else if (!SvOK (sv))	606	else if (!SvOK (sv))
502	encode_ch (enc, 0xe0 \| 22);	607	encode_ch (enc, MAJOR_MISC \| SIMPLE_NULL);
503	else if (enc->cbor.flags & F_ALLOW_UNKNOWN)	608	else if (enc->cbor.flags & F_ALLOW_UNKNOWN)
504	encode_ch (enc, 0xe0 \| 23);	609	encode_ch (enc, MAJOR_MISC \| SIMPLE_UNDEF);
505	else	610	else
506	croak ("encountered perl type (%s,0x%x) that CBOR cannot handle, check your input data",	611	croak ("encountered perl type (%s,0x%x) that CBOR cannot handle, check your input data",
507	SvPV_nolen (sv), (unsigned int)SvFLAGS (sv));	612	SvPV_nolen (sv), (unsigned int)SvFLAGS (sv));
508	}	613	}
509		614
510	static SV *	615	static SV *
511	encode_cbor (SV scalar, CBOR cbor)	616	encode_cbor (SV scalar, CBOR cbor)
512	{	617	{
513	enc_t enc = { };	618	enc_t enc = { 0 };
514		619
515	enc.cbor = *cbor;	620	enc.cbor = *cbor;
516	enc.sv = sv_2mortal (NEWSV (0, INIT_SIZE));	621	enc.sv = sv_2mortal (NEWSV (0, INIT_SIZE));
517	enc.cur = SvPVX (enc.sv);	622	enc.cur = SvPVX (enc.sv);
518	enc.end = SvEND (enc.sv);	623	enc.end = SvEND (enc.sv);
519		624
520	SvPOK_only (enc.sv);	625	SvPOK_only (enc.sv);
521		626
522	if (cbor->flags & F_ALLOW_STRINGREF)	627	if (cbor->flags & F_PACK_STRINGS)
523	{	628	{
524	encode_tag (&enc, CBOR_TAG_STRINGREF_NAMESPACE);	629	encode_tag (&enc, CBOR_TAG_STRINGREF_NAMESPACE);
525	enc.stringref[0]= (HV )sv_2mortal ((SV )newHV ());	630	enc.stringref[0]= (HV )sv_2mortal ((SV )newHV ());
526	enc.stringref[1]= (HV )sv_2mortal ((SV )newHV ());	631	enc.stringref[1]= (HV )sv_2mortal ((SV )newHV ());
527	}	632	}
…		…
562	#define DEC_DEC_DEPTH --dec->depth	667	#define DEC_DEC_DEPTH --dec->depth
563		668
564	static UV	669	static UV
565	decode_uint (dec_t *dec)	670	decode_uint (dec_t *dec)
566	{	671	{
567	switch (*dec->cur & 31)	672	U8 m = *dec->cur & MINOR_MASK;
568	{	673	++dec->cur;
569	case 0: case 1: case 2: case 3: case 4: case 5: case 6: case 7:
570	case 8: case 9: case 10: case 11: case 12: case 13: case 14: case 15:
571	case 16: case 17: case 18: case 19: case 20: case 21: case 22: case 23:
572	return *dec->cur++ & 31;
573		674
574	case 24:	675	if (ecb_expect_true (m < LENGTH_EXT1))
		676	return m;
		677	else if (ecb_expect_true (m == LENGTH_EXT1))
		678	{
575	WANT (2);	679	WANT (1);
576	dec->cur += 2;	680	dec->cur += 1;
577	return dec->cur[-1];	681	return dec->cur[-1];
578		682	}
579	case 25:	683	else if (ecb_expect_true (m == LENGTH_EXT2))
		684	{
580	WANT (3);	685	WANT (2);
581	dec->cur += 3;	686	dec->cur += 2;
582	return (((UV)dec->cur[-2]) << 8)	687	return (((UV)dec->cur[-2]) << 8)
583	\| ((UV)dec->cur[-1]);	688	\| ((UV)dec->cur[-1]);
584		689	}
585	case 26:	690	else if (ecb_expect_true (m == LENGTH_EXT4))
		691	{
586	WANT (5);	692	WANT (4);
587	dec->cur += 5;	693	dec->cur += 4;
588	return (((UV)dec->cur[-4]) << 24)	694	return (((UV)dec->cur[-4]) << 24)
589	\| (((UV)dec->cur[-3]) << 16)	695	\| (((UV)dec->cur[-3]) << 16)
590	\| (((UV)dec->cur[-2]) << 8)	696	\| (((UV)dec->cur[-2]) << 8)
591	\| ((UV)dec->cur[-1]);	697	\| ((UV)dec->cur[-1]);
592		698	}
593	case 27:	699	else if (ecb_expect_true (m == LENGTH_EXT8))
		700	{
594	WANT (9);	701	WANT (8);
595	dec->cur += 9;	702	dec->cur += 8;
		703
		704	return
		705	#if UVSIZE < 8
		706	0
		707	#else
596	return (((UV)dec->cur[-8]) << 56)	708	(((UV)dec->cur[-8]) << 56)
597	\| (((UV)dec->cur[-7]) << 48)	709	\| (((UV)dec->cur[-7]) << 48)
598	\| (((UV)dec->cur[-6]) << 40)	710	\| (((UV)dec->cur[-6]) << 40)
599	\| (((UV)dec->cur[-5]) << 32)	711	\| (((UV)dec->cur[-5]) << 32)
		712	#endif
600	\| (((UV)dec->cur[-4]) << 24)	713	\| (((UV)dec->cur[-4]) << 24)
601	\| (((UV)dec->cur[-3]) << 16)	714	\| (((UV)dec->cur[-3]) << 16)
602	\| (((UV)dec->cur[-2]) << 8)	715	\| (((UV)dec->cur[-2]) << 8)
603	\| ((UV)dec->cur[-1]);	716	\| ((UV)dec->cur[-1]);
604		717	}
605	default:	718	else
606	ERR ("corrupted CBOR data (unsupported integer minor encoding)");	719	ERR ("corrupted CBOR data (unsupported integer minor encoding)");
607	}
608		720
609	fail:	721	fail:
610	return 0;	722	return 0;
611	}	723	}
612		724
…		…
617	{	729	{
618	AV *av = newAV ();	730	AV *av = newAV ();
619		731
620	DEC_INC_DEPTH;	732	DEC_INC_DEPTH;
621		733
622	if ((*dec->cur & 31) == 31)	734	if (*dec->cur == (MAJOR_ARRAY \| MINOR_INDEF))
623	{	735	{
624	++dec->cur;	736	++dec->cur;
625		737
626	for (;;)	738	for (;;)
627	{	739	{
628	WANT (1);	740	WANT (1);
629		741
630	if (*dec->cur == (0xe0 \| 31))	742	if (*dec->cur == (MAJOR_MISC \| MINOR_INDEF))
631	{	743	{
632	++dec->cur;	744	++dec->cur;
633	break;	745	break;
634	}	746	}
635		747
…		…
638	}	750	}
639	else	751	else
640	{	752	{
641	int i, len = decode_uint (dec);	753	int i, len = decode_uint (dec);
642		754
		755	WANT (len); // complexity check for av_fill - need at least one byte per value, do not allow supersize arrays
643	av_fill (av, len - 1);	756	av_fill (av, len - 1);
644		757
645	for (i = 0; i < len; ++i)	758	for (i = 0; i < len; ++i)
646	AvARRAY (av)[i] = decode_sv (dec);	759	AvARRAY (av)[i] = decode_sv (dec);
647	}	760	}
…		…
660	{	773	{
661	// for speed reasons, we specialcase single-string	774	// for speed reasons, we specialcase single-string
662	// byte or utf-8 strings as keys, but only when !stringref	775	// byte or utf-8 strings as keys, but only when !stringref
663		776
664	if (ecb_expect_true (!dec->stringref))	777	if (ecb_expect_true (!dec->stringref))
665	if (dec->cur >= 0x40 && dec->cur <= 0x40 + 27)	778	if (ecb_expect_true ((U8)(*dec->cur - MAJOR_BYTES) <= LENGTH_EXT8))
666	{	779	{
667	I32 len = decode_uint (dec);	780	I32 len = decode_uint (dec);
668	char key = (char )dec->cur;	781	char key = (char )dec->cur;
669		782
		783	WANT (len);
670	dec->cur += len;	784	dec->cur += len;
671		785
672	if (ecb_expect_false (dec->stringref))
673	av_push (dec->stringref, newSVpvn (key, len));
674
675	hv_store (hv, key, len, decode_sv (dec), 0);	786	hv_store (hv, key, len, decode_sv (dec), 0);
676		787
677	return;	788	return;
678	}	789	}
679	else if (dec->cur >= 0x60 && dec->cur <= 0x60 + 27)	790	else if (ecb_expect_true ((U8)(*dec->cur - MAJOR_TEXT) <= LENGTH_EXT8))
680	{	791	{
681	I32 len = decode_uint (dec);	792	I32 len = decode_uint (dec);
682	char key = (char )dec->cur;	793	char key = (char )dec->cur;
683		794
		795	WANT (len);
684	dec->cur += len;	796	dec->cur += len;
685		797
686	if (ecb_expect_false (dec->stringref))	798	if (ecb_expect_false (dec->cbor.flags & F_VALIDATE_UTF8))
687	av_push (dec->stringref, newSVpvn_utf8 (key, len, 1));	799	if (!is_utf8_string (key, len))
		800	ERR ("corrupted CBOR data (invalid UTF-8 in map key)");
688		801
689	hv_store (hv, key, -len, decode_sv (dec), 0);	802	hv_store (hv, key, -len, decode_sv (dec), 0);
690		803
691	return;	804	return;
692	}	805	}
…		…
694	SV *k = decode_sv (dec);	807	SV *k = decode_sv (dec);
695	SV *v = decode_sv (dec);	808	SV *v = decode_sv (dec);
696		809
697	hv_store_ent (hv, k, v, 0);	810	hv_store_ent (hv, k, v, 0);
698	SvREFCNT_dec (k);	811	SvREFCNT_dec (k);
		812
		813	fail:
		814	;
699	}	815	}
700		816
701	static SV *	817	static SV *
702	decode_hv (dec_t *dec)	818	decode_hv (dec_t *dec)
703	{	819	{
704	HV *hv = newHV ();	820	HV *hv = newHV ();
705		821
706	DEC_INC_DEPTH;	822	DEC_INC_DEPTH;
707		823
708	if ((*dec->cur & 31) == 31)	824	if (*dec->cur == (MAJOR_MAP \| MINOR_INDEF))
709	{	825	{
710	++dec->cur;	826	++dec->cur;
711		827
712	for (;;)	828	for (;;)
713	{	829	{
714	WANT (1);	830	WANT (1);
715		831
716	if (*dec->cur == (0xe0 \| 31))	832	if (*dec->cur == (MAJOR_MISC \| MINOR_INDEF))
717	{	833	{
718	++dec->cur;	834	++dec->cur;
719	break;	835	break;
720	}	836	}
721		837
…		…
742	static SV *	858	static SV *
743	decode_str (dec_t *dec, int utf8)	859	decode_str (dec_t *dec, int utf8)
744	{	860	{
745	SV *sv = 0;	861	SV *sv = 0;
746		862
747	if ((*dec->cur & 31) == 31)	863	if ((*dec->cur & MINOR_MASK) == MINOR_INDEF)
748	{	864	{
		865	// indefinite length strings
749	++dec->cur;	866	++dec->cur;
750		867
		868	U8 major = *dec->cur & MAJOR_MISC;
		869
751	sv = newSVpvn ("", 0);	870	sv = newSVpvn ("", 0);
752		871
753	// not very fast, and certainly not robust against illegal input
754	for (;;)	872	for (;;)
755	{	873	{
756	WANT (1);	874	WANT (1);
757		875
758	if (*dec->cur == (0xe0 \| 31))	876	if ((*dec->cur - major) > LENGTH_EXT8)
		877	if (*dec->cur == (MAJOR_MISC \| MINOR_INDEF))
759	{	878	{
760	++dec->cur;	879	++dec->cur;
761	break;	880	break;
762	}	881	}
		882	else
		883	ERR ("corrupted CBOR data (invalid chunks in indefinite length string)");
763		884
764	sv_catsv (sv, decode_sv (dec));	885	STRLEN len = decode_uint (dec);
		886
		887	WANT (len);
		888	sv_catpvn (sv, dec->cur, len);
		889	dec->cur += len;
765	}	890	}
766	}	891	}
767	else	892	else
768	{	893	{
769	STRLEN len = decode_uint (dec);	894	STRLEN len = decode_uint (dec);
…		…
776	&& SvCUR (sv) >= minimum_string_length (AvFILLp (dec->stringref) + 1))	901	&& SvCUR (sv) >= minimum_string_length (AvFILLp (dec->stringref) + 1))
777	av_push (dec->stringref, SvREFCNT_inc_NN (sv));	902	av_push (dec->stringref, SvREFCNT_inc_NN (sv));
778	}	903	}
779		904
780	if (utf8)	905	if (utf8)
		906	{
		907	if (ecb_expect_false (dec->cbor.flags & F_VALIDATE_UTF8))
		908	if (!is_utf8_string (SvPVX (sv), SvCUR (sv)))
		909	ERR ("corrupted CBOR data (invalid UTF-8 in text string)");
		910
781	SvUTF8_on (sv);	911	SvUTF8_on (sv);
		912	}
782		913
783	return sv;	914	return sv;
784		915
785	fail:	916	fail:
786	SvREFCNT_dec (sv);	917	SvREFCNT_dec (sv);
…		…
818	}	949	}
819	break;	950	break;
820		951
821	case CBOR_TAG_STRINGREF:	952	case CBOR_TAG_STRINGREF:
822	{	953	{
823	if ((*dec->cur >> 5) != 0)	954	if ((*dec->cur >> MAJOR_SHIFT) != (MAJOR_POS_INT >> MAJOR_SHIFT))
824	ERR ("corrupted CBOR data (stringref index not an unsigned integer)");	955	ERR ("corrupted CBOR data (stringref index not an unsigned integer)");
825		956
826	UV idx = decode_uint (dec);	957	UV idx = decode_uint (dec);
827		958
828	if (!dec->stringref \|\| (int)idx > AvFILLp (dec->stringref))	959	if (!dec->stringref \|\| (int)idx > AvFILLp (dec->stringref))
…		…
835	case CBOR_TAG_VALUE_SHAREABLE:	966	case CBOR_TAG_VALUE_SHAREABLE:
836	{	967	{
837	if (ecb_expect_false (!dec->shareable))	968	if (ecb_expect_false (!dec->shareable))
838	dec->shareable = (AV )sv_2mortal ((SV )newAV ());	969	dec->shareable = (AV )sv_2mortal ((SV )newAV ());
839		970
		971	if (dec->cbor.flags & F_ALLOW_CYCLES)
		972	{
840	sv = newSV (0);	973	sv = newSV (0);
841	av_push (dec->shareable, SvREFCNT_inc_NN (sv));	974	av_push (dec->shareable, SvREFCNT_inc_NN (sv));
842		975
843	SV *osv = decode_sv (dec);	976	SV *osv = decode_sv (dec);
844	sv_setsv (sv, osv);	977	sv_setsv (sv, osv);
845	SvREFCNT_dec_NN (osv);	978	SvREFCNT_dec_NN (osv);
		979	}
		980	else
		981	{
		982	av_push (dec->shareable, &PL_sv_undef);
		983	int idx = AvFILLp (dec->shareable);
		984	sv = decode_sv (dec);
		985	av_store (dec->shareable, idx, SvREFCNT_inc_NN (sv));
		986	}
846	}	987	}
847	break;	988	break;
848		989
849	case CBOR_TAG_VALUE_SHAREDREF:	990	case CBOR_TAG_VALUE_SHAREDREF:
850	{	991	{
851	if ((*dec->cur >> 5) != 0)	992	if ((*dec->cur >> MAJOR_SHIFT) != (MAJOR_POS_INT >> MAJOR_SHIFT))
852	ERR ("corrupted CBOR data (sharedref index not an unsigned integer)");	993	ERR ("corrupted CBOR data (sharedref index not an unsigned integer)");
853		994
854	UV idx = decode_uint (dec);	995	UV idx = decode_uint (dec);
855		996
856	if (!dec->shareable \|\| (int)idx > AvFILLp (dec->shareable))	997	if (!dec->shareable \|\| (int)idx > AvFILLp (dec->shareable))
857	ERR ("corrupted CBOR data (sharedref index out of bounds)");	998	ERR ("corrupted CBOR data (sharedref index out of bounds)");
858		999
859	sv = SvREFCNT_inc_NN (AvARRAY (dec->shareable)[idx]);	1000	sv = SvREFCNT_inc_NN (AvARRAY (dec->shareable)[idx]);
		1001
		1002	if (sv == &PL_sv_undef)
		1003	ERR ("cyclic CBOR data structure found, but allow_cycles is not enabled");
860	}	1004	}
861	break;	1005	break;
862		1006
863	case CBOR_TAG_PERL_OBJECT:	1007	case CBOR_TAG_PERL_OBJECT:
864	{	1008	{
…		…
879	if (!method)	1023	if (!method)
880	ERR ("cannot decode perl-object (package does not have a THAW method)");	1024	ERR ("cannot decode perl-object (package does not have a THAW method)");
881		1025
882	dSP;	1026	dSP;
883		1027
884	ENTER; SAVETMPS; PUSHMARK (SP);	1028	ENTER; SAVETMPS;
		1029	PUSHMARK (SP);
885	EXTEND (SP, len + 1);	1030	EXTEND (SP, len + 1);
886	// we re-bless the reference to get overload and other niceties right	1031	// we re-bless the reference to get overload and other niceties right
887	PUSHs (*av_fetch (av, 0, 1));	1032	PUSHs (*av_fetch (av, 0, 1));
888	PUSHs (sv_cbor);	1033	PUSHs (sv_cbor);
889		1034
…		…
914	default:	1059	default:
915	{	1060	{
916	sv = decode_sv (dec);	1061	sv = decode_sv (dec);
917		1062
918	dSP;	1063	dSP;
919	ENTER; SAVETMPS; PUSHMARK (SP);	1064	ENTER; SAVETMPS;
		1065	SAVESTACK_POS ();
		1066	PUSHMARK (SP);
920	EXTEND (SP, 2);	1067	EXTEND (SP, 2);
921	PUSHs (newSVuv (tag));	1068	PUSHs (newSVuv (tag));
922	PUSHs (sv);	1069	PUSHs (sv);
923		1070
924	PUTBACK;	1071	PUTBACK;
…		…
965	static SV *	1112	static SV *
966	decode_sv (dec_t *dec)	1113	decode_sv (dec_t *dec)
967	{	1114	{
968	WANT (1);	1115	WANT (1);
969		1116
970	switch (*dec->cur >> 5)	1117	switch (*dec->cur >> MAJOR_SHIFT)
971	{	1118	{
972	case 0: // unsigned int	1119	case MAJOR_POS_INT >> MAJOR_SHIFT: return newSVuv (decode_uint (dec));
973	return newSVuv (decode_uint (dec));	1120	case MAJOR_NEG_INT >> MAJOR_SHIFT: return newSViv (-1 - (IV)decode_uint (dec));
974	case 1: // negative int	1121	case MAJOR_BYTES >> MAJOR_SHIFT: return decode_str (dec, 0);
975	return newSViv (-1 - (IV)decode_uint (dec));	1122	case MAJOR_TEXT >> MAJOR_SHIFT: return decode_str (dec, 1);
976	case 2: // octet string	1123	case MAJOR_ARRAY >> MAJOR_SHIFT: return decode_av (dec);
977	return decode_str (dec, 0);	1124	case MAJOR_MAP >> MAJOR_SHIFT: return decode_hv (dec);
978	case 3: // utf-8 string	1125	case MAJOR_TAG >> MAJOR_SHIFT: return decode_tagged (dec);
979	return decode_str (dec, 1);	1126
980	case 4: // array	1127	case MAJOR_MISC >> MAJOR_SHIFT:
981	return decode_av (dec);
982	case 5: // map
983	return decode_hv (dec);
984	case 6: // tag
985	return decode_tagged (dec);
986	case 7: // misc
987	switch (*dec->cur++ & 31)	1128	switch (*dec->cur++ & MINOR_MASK)
988	{	1129	{
989	case 20:	1130	case SIMPLE_FALSE:
990	#if CBOR_SLOW	1131	#if CBOR_SLOW
991	types_false = get_bool ("Types::Serialiser::false");	1132	types_false = get_bool ("Types::Serialiser::false");
992	#endif	1133	#endif
993	return newSVsv (types_false);	1134	return newSVsv (types_false);
994	case 21:	1135	case SIMPLE_TRUE:
995	#if CBOR_SLOW	1136	#if CBOR_SLOW
996	types_true = get_bool ("Types::Serialiser::true");	1137	types_true = get_bool ("Types::Serialiser::true");
997	#endif	1138	#endif
998	return newSVsv (types_true);	1139	return newSVsv (types_true);
999	case 22:	1140	case SIMPLE_NULL:
1000	return newSVsv (&PL_sv_undef);	1141	return newSVsv (&PL_sv_undef);
1001	case 23:	1142	case SIMPLE_UNDEF:
1002	#if CBOR_SLOW	1143	#if CBOR_SLOW
1003	types_error = get_bool ("Types::Serialiser::error");	1144	types_error = get_bool ("Types::Serialiser::error");
1004	#endif	1145	#endif
1005	return newSVsv (types_error);	1146	return newSVsv (types_error);
1006		1147
1007	case 25:	1148	case MISC_FLOAT16:
1008	{	1149	{
1009	WANT (2);	1150	WANT (2);
1010		1151
1011	uint16_t fp = (dec->cur[0] << 8) \| dec->cur[1];	1152	uint16_t fp = (dec->cur[0] << 8) \| dec->cur[1];
1012	dec->cur += 2;	1153	dec->cur += 2;
1013		1154
1014	return newSVnv (ecb_binary16_to_float (fp));	1155	return newSVnv (ecb_binary16_to_float (fp));
1015	}	1156	}
1016		1157
1017	case 26:	1158	case MISC_FLOAT32:
1018	{	1159	{
1019	uint32_t fp;	1160	uint32_t fp;
1020	WANT (4);	1161	WANT (4);
1021	memcpy (&fp, dec->cur, 4);	1162	memcpy (&fp, dec->cur, 4);
1022	dec->cur += 4;	1163	dec->cur += 4;
…		…
1025	fp = ecb_bswap32 (fp);	1166	fp = ecb_bswap32 (fp);
1026		1167
1027	return newSVnv (ecb_binary32_to_float (fp));	1168	return newSVnv (ecb_binary32_to_float (fp));
1028	}	1169	}
1029		1170
1030	case 27:	1171	case MISC_FLOAT64:
1031	{	1172	{
1032	uint64_t fp;	1173	uint64_t fp;
1033	WANT (8);	1174	WANT (8);
1034	memcpy (&fp, dec->cur, 8);	1175	memcpy (&fp, dec->cur, 8);
1035	dec->cur += 8;	1176	dec->cur += 8;
…		…
1038	fp = ecb_bswap64 (fp);	1179	fp = ecb_bswap64 (fp);
1039		1180
1040	return newSVnv (ecb_binary64_to_double (fp));	1181	return newSVnv (ecb_binary64_to_double (fp));
1041	}	1182	}
1042		1183
1043	// 0..19 unassigned	1184	// 0..19 unassigned simple
1044	// 24 reserved + unassigned (reserved values are not encodable)	1185	// 24 reserved + unassigned simple (reserved values are not encodable)
		1186	// 28-30 unassigned misc
		1187	// 31 break code
1045	default:	1188	default:
1046	ERR ("corrupted CBOR data (reserved/unassigned major 7 value)");	1189	ERR ("corrupted CBOR data (reserved/unassigned/unexpected major 7 value)");
1047	}	1190	}
1048		1191
1049	break;	1192	break;
1050	}	1193	}
1051		1194
…		…
1054	}	1197	}
1055		1198
1056	static SV *	1199	static SV *
1057	decode_cbor (SV string, CBOR cbor, char **offset_return)	1200	decode_cbor (SV string, CBOR cbor, char **offset_return)
1058	{	1201	{
1059	dec_t dec = { };	1202	dec_t dec = { 0 };
1060	SV *sv;	1203	SV *sv;
1061	STRLEN len;	1204	STRLEN len;
1062	char *data = SvPVbyte (string, len);	1205	char *data = SvPVbyte (string, len);
1063		1206
1064	if (len > cbor->max_size && cbor->max_size)	1207	if (len > cbor->max_size && cbor->max_size)
…		…
1078	if (dec.cur != dec.end && !dec.err)	1221	if (dec.cur != dec.end && !dec.err)
1079	dec.err = "garbage after CBOR object";	1222	dec.err = "garbage after CBOR object";
1080		1223
1081	if (dec.err)	1224	if (dec.err)
1082	{	1225	{
		1226	if (dec.shareable)
		1227	{
		1228	// need to break cyclic links, which whould all be in shareable
		1229	int i;
		1230	SV **svp;
		1231
		1232	for (i = av_len (dec.shareable) + 1; i--; )
		1233	if ((svp = av_fetch (dec.shareable, i, 0)))
		1234	sv_setsv (*svp, &PL_sv_undef);
		1235	}
		1236
1083	SvREFCNT_dec (sv);	1237	SvREFCNT_dec (sv);
1084	croak ("%s, at offset %d (octet 0x%02x)", dec.err, dec.cur - (U8 )data, (int)(uint8_t)dec.cur);	1238	croak ("%s, at offset %d (octet 0x%02x)", dec.err, dec.cur - (U8 )data, (int)(uint8_t)dec.cur);
1085	}	1239	}
1086		1240
1087	sv = sv_2mortal (sv);	1241	sv = sv_2mortal (sv);
1088		1242
1089	return sv;	1243	return sv;
1090	}	1244	}
1091		1245
		1246	/////////////////////////////////////////////////////////////////////////////
		1247	// incremental parser
		1248
		1249	#define INCR_DONE(cbor) (AvFILLp (cbor->incr_count) < 0)
		1250
		1251	// returns 0 for notyet, 1 for success or error
		1252	static int
		1253	incr_parse (CBOR self, SV cborstr)
		1254	{
		1255	STRLEN cur;
		1256	SvPV (cborstr, cur);
		1257
		1258	while (ecb_expect_true (self->incr_need <= cur))
		1259	{
		1260	// table of integer count bytes
		1261	static I8 incr_len[MINOR_MASK + 1] = {
		1262	0, 0, 0, 0, 0, 0, 0, 0,
		1263	0, 0, 0, 0, 0, 0, 0, 0,
		1264	0, 0, 0, 0, 0, 0, 0, 0,
		1265	1, 2, 4, 8,-1,-1,-1,-2
		1266	};
		1267
		1268	const U8 *p = SvPVX (cborstr) + self->incr_pos;
		1269	U8 m = *p & MINOR_MASK;
		1270	IV count = SvIVX (AvARRAY (self->incr_count)[AvFILLp (self->incr_count)]);
		1271	I8 ilen = incr_len[m];
		1272
		1273	self->incr_need = self->incr_pos + 1;
		1274
		1275	if (ecb_expect_false (ilen < 0))
		1276	{
		1277	if (m != MINOR_INDEF)
		1278	return 1; // error
		1279
		1280	if (*p == (MAJOR_MISC \| MINOR_INDEF))
		1281	{
		1282	if (count >= 0)
		1283	return 1; // error
		1284
		1285	count = 1;
		1286	}
		1287	else
		1288	{
		1289	av_push (self->incr_count, newSViv (-1)); //TODO: nest
		1290	count = -1;
		1291	}
		1292	}
		1293	else
		1294	{
		1295	self->incr_need += ilen;
		1296	if (ecb_expect_false (self->incr_need > cur))
		1297	return 0;
		1298
		1299	int major = *p >> MAJOR_SHIFT;
		1300
		1301	switch (major)
		1302	{
		1303	case MAJOR_TAG >> MAJOR_SHIFT:
		1304	++count; // tags merely prefix another value
		1305	break;
		1306
		1307	case MAJOR_BYTES >> MAJOR_SHIFT:
		1308	case MAJOR_TEXT >> MAJOR_SHIFT:
		1309	case MAJOR_ARRAY >> MAJOR_SHIFT:
		1310	case MAJOR_MAP >> MAJOR_SHIFT:
		1311	{
		1312	UV len;
		1313
		1314	if (ecb_expect_false (ilen))
		1315	{
		1316	len = 0;
		1317
		1318	do {
		1319	len = (len << 8) \| *++p;
		1320	} while (--ilen);
		1321	}
		1322	else
		1323	len = m;
		1324
		1325	switch (major)
		1326	{
		1327	case MAJOR_BYTES >> MAJOR_SHIFT:
		1328	case MAJOR_TEXT >> MAJOR_SHIFT:
		1329	self->incr_need += len;
		1330	if (ecb_expect_false (self->incr_need > cur))
		1331	return 0;
		1332
		1333	break;
		1334
		1335	case MAJOR_MAP >> MAJOR_SHIFT:
		1336	len <<= 1;
		1337	case MAJOR_ARRAY >> MAJOR_SHIFT:
		1338	if (len)
		1339	{
		1340	av_push (self->incr_count, newSViv (len + 1)); //TODO: nest
		1341	count = len + 1;
		1342	}
		1343	break;
		1344	}
		1345	}
		1346	}
		1347	}
		1348
		1349	self->incr_pos = self->incr_need;
		1350
		1351	if (count > 0)
		1352	{
		1353	while (!--count)
		1354	{
		1355	if (!AvFILLp (self->incr_count))
		1356	return 1; // done
		1357
		1358	SvREFCNT_dec_NN (av_pop (self->incr_count));
		1359	count = SvIVX (AvARRAY (self->incr_count)[AvFILLp (self->incr_count)]);
		1360	}
		1361
		1362	SvIVX (AvARRAY (self->incr_count)[AvFILLp (self->incr_count)]) = count;
		1363	}
		1364	}
		1365
		1366	return 0;
		1367	}
		1368
		1369
1092	/////////////////////////////////////////////////////////////////////////////	1370	/////////////////////////////////////////////////////////////////////////////
1093	// XS interface functions	1371	// XS interface functions
1094		1372
1095	MODULE = CBOR::XS PACKAGE = CBOR::XS	1373	MODULE = CBOR::XS PACKAGE = CBOR::XS
1096		1374
…		…
1136	void shrink (CBOR *self, int enable = 1)	1414	void shrink (CBOR *self, int enable = 1)
1137	ALIAS:	1415	ALIAS:
1138	shrink = F_SHRINK	1416	shrink = F_SHRINK
1139	allow_unknown = F_ALLOW_UNKNOWN	1417	allow_unknown = F_ALLOW_UNKNOWN
1140	allow_sharing = F_ALLOW_SHARING	1418	allow_sharing = F_ALLOW_SHARING
1141	allow_stringref = F_ALLOW_STRINGREF	1419	allow_cycles = F_ALLOW_CYCLES
		1420	pack_strings = F_PACK_STRINGS
		1421	utf8_strings = F_UTF8_STRINGS
		1422	validate_utf8 = F_VALIDATE_UTF8
1142	PPCODE:	1423	PPCODE:
1143	{	1424	{
1144	if (enable)	1425	if (enable)
1145	self->flags \|= ix;	1426	self->flags \|= ix;
1146	else	1427	else
…		…
1152	void get_shrink (CBOR *self)	1433	void get_shrink (CBOR *self)
1153	ALIAS:	1434	ALIAS:
1154	get_shrink = F_SHRINK	1435	get_shrink = F_SHRINK
1155	get_allow_unknown = F_ALLOW_UNKNOWN	1436	get_allow_unknown = F_ALLOW_UNKNOWN
1156	get_allow_sharing = F_ALLOW_SHARING	1437	get_allow_sharing = F_ALLOW_SHARING
1157	get_allow_stringref = F_ALLOW_STRINGREF	1438	get_allow_cycles = F_ALLOW_CYCLES
		1439	get_pack_strings = F_PACK_STRINGS
		1440	get_validate_utf8 = F_VALIDATE_UTF8
1158	PPCODE:	1441	PPCODE:
1159	XPUSHs (boolSV (self->flags & ix));	1442	XPUSHs (boolSV (self->flags & ix));
1160		1443
1161	void max_depth (CBOR *self, U32 max_depth = 0x80000000UL)	1444	void max_depth (CBOR *self, U32 max_depth = 0x80000000UL)
1162	PPCODE:	1445	PPCODE:
…		…
1211	EXTEND (SP, 2);	1494	EXTEND (SP, 2);
1212	PUSHs (sv);	1495	PUSHs (sv);
1213	PUSHs (sv_2mortal (newSVuv (offset - SvPVX (cborstr))));	1496	PUSHs (sv_2mortal (newSVuv (offset - SvPVX (cborstr))));
1214	}	1497	}
1215		1498
		1499	void incr_parse (CBOR self, SV cborstr)
		1500	ALIAS:
		1501	incr_parse_multiple = 1
		1502	PPCODE:
		1503	{
		1504	if (SvUTF8 (cborstr))
		1505	sv_utf8_downgrade (cborstr, 0);
		1506
		1507	if (!self->incr_count)
		1508	{
		1509	self->incr_count = newAV ();
		1510	self->incr_pos = 0;
		1511	self->incr_need = 1;
		1512
		1513	av_push (self->incr_count, newSViv (1));
		1514	}
		1515
		1516	do
		1517	{
		1518	if (!incr_parse (self, cborstr))
		1519	{
		1520	if (self->incr_need > self->max_size && self->max_size)
		1521	croak ("attempted decode of CBOR text of %lu bytes size, but max_size is set to %lu",
		1522	(unsigned long)self->incr_need, (unsigned long)self->max_size);
		1523
		1524	break;
		1525	}
		1526
		1527	SV *sv;
		1528	char *offset;
		1529
		1530	PUTBACK; sv = decode_cbor (cborstr, self, &offset); SPAGAIN;
		1531	XPUSHs (sv);
		1532
		1533	sv_chop (cborstr, offset);
		1534
		1535	av_clear (self->incr_count);
		1536	av_push (self->incr_count, newSViv (1));
		1537
		1538	self->incr_pos = 0;
		1539	self->incr_need = self->incr_pos + 1;
		1540	}
		1541	while (ix);
		1542	}
		1543
		1544	void incr_reset (CBOR *self)
		1545	CODE:
		1546	{
		1547	SvREFCNT_dec (self->incr_count);
		1548	self->incr_count = 0;
		1549	}
		1550
1216	void DESTROY (CBOR *self)	1551	void DESTROY (CBOR *self)
1217	PPCODE:	1552	PPCODE:
1218	cbor_free (self);	1553	cbor_free (self);
1219		1554
1220	PROTOTYPES: ENABLE	1555	PROTOTYPES: ENABLE
1221		1556
1222	void encode_cbor (SV *scalar)	1557	void encode_cbor (SV *scalar)
		1558	ALIAS:
		1559	encode_cbor = 0
		1560	encode_cbor_sharing = F_ALLOW_SHARING
1223	PPCODE:	1561	PPCODE:
1224	{	1562	{
1225	CBOR cbor;	1563	CBOR cbor;
1226	cbor_init (&cbor);	1564	cbor_init (&cbor);
		1565	cbor.flags \|= ix;
1227	PUTBACK; scalar = encode_cbor (scalar, &cbor); SPAGAIN;	1566	PUTBACK; scalar = encode_cbor (scalar, &cbor); SPAGAIN;
1228	XPUSHs (scalar);	1567	XPUSHs (scalar);
1229	}	1568	}
1230		1569
1231	void decode_cbor (SV *cborstr)	1570	void decode_cbor (SV *cborstr)

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing CBOR-XS/XS.xs (file contents): Revision 1.27 by root, Fri Nov 22 15:28:38 2013 UTC vs. Revision 1.52 by root, Sun Apr 24 19:31:55 2016 UTC

Diff Legend

Comparing CBOR-XS/XS.xs (file contents):
Revision 1.27 by root, Fri Nov 22 15:28:38 2013 UTC vs.
Revision 1.52 by root, Sun Apr 24 19:31:55 2016 UTC