ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/JSON-XS/XS.xs
(Generate patch)

Comparing JSON-XS/XS.xs (file contents):
Revision 1.5 by root, Thu Mar 22 23:24:18 2007 UTC vs.
Revision 1.8 by root, Fri Mar 23 16:13:59 2007 UTC

… …
12#define F_CANONICAL 0x00000008 12#define F_CANONICAL 0x00000008
13#define F_SPACE_BEFORE 0x00000010 13#define F_SPACE_BEFORE 0x00000010
14#define F_SPACE_AFTER 0x00000020 14#define F_SPACE_AFTER 0x00000020
15#define F_JSON_RPC 0x00000040 15#define F_JSON_RPC 0x00000040
16#define F_ALLOW_NONREF 0x00000080 16#define F_ALLOW_NONREF 0x00000080
17#define F_SHRINK 0x00000100
17 18
18#define F_PRETTY F_INDENT | F_SPACE_BEFORE | F_SPACE_AFTER 19#define F_PRETTY F_INDENT | F_SPACE_BEFORE | F_SPACE_AFTER
19#define F_DEFAULT 0 20#define F_DEFAULT 0
20 21
21#define INIT_SIZE 32 // initial scalar size to be allocated 22#define INIT_SIZE 32 // initial scalar size to be allocated
… …
53 croak ("object is not of type JSON::XS"); 54 croak ("object is not of type JSON::XS");
54 55
55 return &SvUVX (SvRV (sv)); 56 return &SvUVX (SvRV (sv));
56} 57}
57 58
59static void
60shrink (SV *sv)
61{
62 sv_utf8_downgrade (sv, 1);
63#ifdef SvPV_shrink_to_cur
64 SvPV_shrink_to_cur (sv);
65#endif
66}
67
58///////////////////////////////////////////////////////////////////////////// 68/////////////////////////////////////////////////////////////////////////////
59 69
60static void 70static void
61need (enc_t *enc, STRLEN len) 71need (enc_t *enc, STRLEN len)
62{ 72{
… …
85 95
86 while (str < end) 96 while (str < end)
87 { 97 {
88 unsigned char ch = *(unsigned char *)str; 98 unsigned char ch = *(unsigned char *)str;
89 99
90 if (ch == '"') 100 if (ch >= 0x20 && ch < 0x80) // most common case
91 { 101 {
102 if (ch == '"') // but with slow exceptions
103 {
92 need (enc, len += 1); 104 need (enc, len += 1);
93 *enc->cur++ = '\\'; 105 *enc->cur++ = '\\';
94 *enc->cur++ = '"'; 106 *enc->cur++ = '"';
95 ++str;
96 } 107 }
97 else if (ch == '\\') 108 else if (ch == '\\')
98 { 109 {
99 need (enc, len += 1); 110 need (enc, len += 1);
100 *enc->cur++ = '\\'; 111 *enc->cur++ = '\\';
101 *enc->cur++ = '\\'; 112 *enc->cur++ = '\\';
102 ++str;
103 } 113 }
104 else if (ch >= 0x20 && ch < 0x80) // most common case 114 else
105 {
106 *enc->cur++ = ch; 115 *enc->cur++ = ch;
107 ++str; 116
108 }
109 else if (ch == '\015')
110 {
111 need (enc, len += 1);
112 *enc->cur++ = '\\';
113 *enc->cur++ = 'r';
114 ++str;
115 }
116 else if (ch == '\012')
117 {
118 need (enc, len += 1);
119 *enc->cur++ = '\\';
120 *enc->cur++ = 'n';
121 ++str; 117 ++str;
122 } 118 }
123 else 119 else
124 { 120 {
125 STRLEN clen; 121 switch (ch)
126 UV uch;
127
128 if (is_utf8)
129 { 122 {
130 uch = utf8n_to_uvuni (str, end - str, &clen, UTF8_CHECK_ONLY); 123 case '\010': need (enc, len += 1); *enc->cur++ = '\\'; *enc->cur++ = 'b'; ++str; break;
131 if (clen == (STRLEN)-1) 124 case '\011': need (enc, len += 1); *enc->cur++ = '\\'; *enc->cur++ = 't'; ++str; break;
132 croak ("malformed UTF-8 character in string, cannot convert to JSON"); 125 case '\012': need (enc, len += 1); *enc->cur++ = '\\'; *enc->cur++ = 'n'; ++str; break;
133 } 126 case '\014': need (enc, len += 1); *enc->cur++ = '\\'; *enc->cur++ = 'f'; ++str; break;
134 else 127 case '\015': need (enc, len += 1); *enc->cur++ = '\\'; *enc->cur++ = 'r'; ++str; break;
135 {
136 uch = ch;
137 clen = 1;
138 }
139 128
140 if (uch < 0x80 || enc->flags & F_ASCII) 129 default:
141 {
142 if (uch > 0xFFFFUL)
143 { 130 {
131 STRLEN clen;
132 UV uch;
133
134 if (is_utf8)
135 {
136 uch = utf8n_to_uvuni (str, end - str, &clen, UTF8_CHECK_ONLY);
137 if (clen == (STRLEN)-1)
138 croak ("malformed UTF-8 character in string, cannot convert to JSON");
139 }
140 else
141 {
142 uch = ch;
143 clen = 1;
144 }
145
146 if (uch < 0x80 || enc->flags & F_ASCII)
147 {
148 if (uch > 0xFFFFUL)
149 {
144 need (enc, len += 11); 150 need (enc, len += 11);
145 sprintf (enc->cur, "\\u%04x\\u%04x", 151 sprintf (enc->cur, "\\u%04x\\u%04x",
146 (uch - 0x10000) / 0x400 + 0xD800, 152 (uch - 0x10000) / 0x400 + 0xD800,
147 (uch - 0x10000) % 0x400 + 0xDC00); 153 (uch - 0x10000) % 0x400 + 0xDC00);
148 enc->cur += 12; 154 enc->cur += 12;
155 }
156 else
157 {
158 static char hexdigit [16] = "0123456789abcdef";
159 need (enc, len += 5);
160 *enc->cur++ = '\\';
161 *enc->cur++ = 'u';
162 *enc->cur++ = hexdigit [ uch >> 12 ];
163 *enc->cur++ = hexdigit [(uch >> 8) & 15];
164 *enc->cur++ = hexdigit [(uch >> 4) & 15];
165 *enc->cur++ = hexdigit [(uch >> 0) & 15];
166 }
167
168 str += clen;
169 }
170 else if (is_utf8)
171 {
172 need (enc, len += clen);
173 do
174 {
175 *enc->cur++ = *str++;
176 }
177 while (--clen);
178 }
179 else
180 {
181 need (enc, len += 10); // never more than 11 bytes needed
182 enc->cur = uvuni_to_utf8_flags (enc->cur, uch, 0);
183 ++str;
184 }
149 } 185 }
150 else
151 {
152 static char hexdigit [16] = "0123456789abcdef";
153 need (enc, len += 5);
154 *enc->cur++ = '\\';
155 *enc->cur++ = 'u';
156 *enc->cur++ = hexdigit [ uch >> 12 ];
157 *enc->cur++ = hexdigit [(uch >> 8) & 15];
158 *enc->cur++ = hexdigit [(uch >> 4) & 15];
159 *enc->cur++ = hexdigit [(uch >> 0) & 15];
160 }
161
162 str += clen;
163 }
164 else if (is_utf8)
165 {
166 need (enc, len += clen);
167 do
168 {
169 *enc->cur++ = *str++;
170 }
171 while (--clen);
172 }
173 else
174 {
175 need (enc, 10); // never more than 11 bytes needed
176 enc->cur = uvuni_to_utf8_flags (enc->cur, uch, 0);
177 ++str;
178 } 186 }
179 } 187 }
180 188
181 --len; 189 --len;
182 } 190 }
… …
309 317
310 if (fast) 318 if (fast)
311 qsort (hes, count, sizeof (HE *), he_cmp_fast); 319 qsort (hes, count, sizeof (HE *), he_cmp_fast);
312 else 320 else
313 { 321 {
314 // hack to disable "use bytes" 322 // hack to forcefully disable "use bytes"
315 COP *oldcop = PL_curcop, cop; 323 COP cop = *PL_curcop;
316 cop.op_private = 0; 324 cop.op_private = 0;
325
326 ENTER;
327 SAVETMPS;
328
329 SAVEVPTR (PL_curcop);
317 PL_curcop = &cop; 330 PL_curcop = &cop;
318 331
319 SAVETMPS;
320 qsort (hes, count, sizeof (HE *), he_cmp_slow); 332 qsort (hes, count, sizeof (HE *), he_cmp_slow);
333
321 FREETMPS; 334 FREETMPS;
322 335 LEAVE;
323 PL_curcop = oldcop;
324 } 336 }
325 337
326 for (i = 0; i < count; ++i) 338 for (i = 0; i < count; ++i)
327 { 339 {
328 INDENT; 340 INDENT;
… …
425 437
426 if (!(flags & (F_ASCII | F_UTF8))) 438 if (!(flags & (F_ASCII | F_UTF8)))
427 SvUTF8_on (enc.sv); 439 SvUTF8_on (enc.sv);
428 440
429 SvCUR_set (enc.sv, enc.cur - SvPVX (enc.sv)); 441 SvCUR_set (enc.sv, enc.cur - SvPVX (enc.sv));
442
443 if (enc.flags & F_SHRINK)
444 shrink (enc.sv);
445
430 return enc.sv; 446 return enc.sv;
431} 447}
432 448
433///////////////////////////////////////////////////////////////////////////// 449/////////////////////////////////////////////////////////////////////////////
434 450
… …
572 else if (ch >= 0x80) 588 else if (ch >= 0x80)
573 { 589 {
574 STRLEN clen; 590 STRLEN clen;
575 UV uch = utf8n_to_uvuni (dec->cur, dec->end - dec->cur, &clen, UTF8_CHECK_ONLY); 591 UV uch = utf8n_to_uvuni (dec->cur, dec->end - dec->cur, &clen, UTF8_CHECK_ONLY);
576 if (clen == (STRLEN)-1) 592 if (clen == (STRLEN)-1)
577 ERR ("malformed UTF-8 character in string, cannot convert to JSON"); 593 ERR ("malformed UTF-8 character in JSON string");
578 594
579 APPEND_GROW (clen); 595 APPEND_GROW (clen);
580 do 596 do
581 { 597 {
582 *cur++ = *dec->cur++; 598 *cur++ = *dec->cur++;
… …
598 SvPOK_only (sv); 614 SvPOK_only (sv);
599 *SvEND (sv) = 0; 615 *SvEND (sv) = 0;
600 616
601 if (utf8) 617 if (utf8)
602 SvUTF8_on (sv); 618 SvUTF8_on (sv);
619
620 if (dec->flags & F_SHRINK)
621 shrink (sv);
603 622
604 return sv; 623 return sv;
605 624
606fail: 625fail:
607 SvREFCNT_dec (sv); 626 SvREFCNT_dec (sv);
… …
829 ERR ("'null' expected"); 848 ERR ("'null' expected");
830 849
831 break; 850 break;
832 851
833 default: 852 default:
834 ERR ("malformed json string"); 853 ERR ("malformed json string, neither array, object, number, string or atom");
835 break; 854 break;
836 } 855 }
837 856
838fail: 857fail:
839 return 0; 858 return 0;
… …
859 878
860 sv = decode_sv (&dec); 879 sv = decode_sv (&dec);
861 880
862 if (!sv) 881 if (!sv)
863 { 882 {
883 IV offset = dec.flags & F_UTF8
884 ? dec.cur - SvPVX (string)
864 IV offset = utf8_distance (dec.cur, SvPVX (string)); 885 : utf8_distance (dec.cur, SvPVX (string));
865 SV *uni = sv_newmortal (); 886 SV *uni = sv_newmortal ();
887
866 // horrible hack to silence warning inside pv_uni_display 888 // horrible hack to silence warning inside pv_uni_display
867 COP cop; 889 COP cop = *PL_curcop;
868 memset (&cop, 0, sizeof (cop));
869 cop.cop_warnings = pWARN_NONE; 890 cop.cop_warnings = pWARN_NONE;
891 ENTER;
870 SAVEVPTR (PL_curcop); 892 SAVEVPTR (PL_curcop);
871 PL_curcop = &cop; 893 PL_curcop = &cop;
872
873 pv_uni_display (uni, dec.cur, dec.end - dec.cur, 20, UNI_DISPLAY_QQ); 894 pv_uni_display (uni, dec.cur, dec.end - dec.cur, 20, UNI_DISPLAY_QQ);
895 LEAVE;
896
874 croak ("%s, at character offset %d (%s)", 897 croak ("%s, at character offset %d (%s)",
875 dec.err, 898 dec.err,
876 (int)offset, 899 (int)offset,
877 dec.cur != dec.end ? SvPV_nolen (uni) : "(end of string)"); 900 dec.cur != dec.end ? SvPV_nolen (uni) : "(end of string)");
878 } 901 }
… …
910 CODE: 933 CODE:
911 RETVAL = sv_bless (newRV_noinc (newSVuv (F_DEFAULT)), json_stash); 934 RETVAL = sv_bless (newRV_noinc (newSVuv (F_DEFAULT)), json_stash);
912 OUTPUT: 935 OUTPUT:
913 RETVAL 936 RETVAL
914 937
915SV *ascii (SV *self, int enable) 938SV *ascii (SV *self, int enable = 1)
916 ALIAS: 939 ALIAS:
917 ascii = F_ASCII 940 ascii = F_ASCII
918 utf8 = F_UTF8 941 utf8 = F_UTF8
919 indent = F_INDENT 942 indent = F_INDENT
920 canonical = F_CANONICAL 943 canonical = F_CANONICAL
921 space_before = F_SPACE_BEFORE 944 space_before = F_SPACE_BEFORE
922 space_after = F_SPACE_AFTER 945 space_after = F_SPACE_AFTER
923 json_rpc = F_JSON_RPC 946 json_rpc = F_JSON_RPC
924 pretty = F_PRETTY 947 pretty = F_PRETTY
925 allow_nonref = F_ALLOW_NONREF 948 allow_nonref = F_ALLOW_NONREF
949 shrink = F_SHRINK
926 CODE: 950 CODE:
927{ 951{
928 UV *uv = SvJSON (self); 952 UV *uv = SvJSON (self);
929 if (enable) 953 if (enable)
930 *uv |= ix; 954 *uv |= ix;

Diff Legend

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