[Encode] 1.77 Released
[p5sagit/p5-mst-13.2.git] / hv.h
CommitLineData
a0d0e21e 1/* hv.h
79072805 2 *
be3c0a43 3 * Copyright (c) 1991-2002, Larry Wall
79072805 4 *
5 * You may distribute under the terms of either the GNU General Public
6 * License or the Artistic License, as specified in the README file.
7 *
79072805 8 */
9
5cbe4eec 10/* typedefs to eliminate some typing */
79072805 11typedef struct he HE;
ff68c719 12typedef struct hek HEK;
79072805 13
5cbe4eec 14/* entry in hash value chain */
79072805 15struct he {
5cbe4eec 16 HE *hent_next; /* next entry in chain */
17 HEK *hent_hek; /* hash key */
18 SV *hent_val; /* scalar value that was hashed */
bbce6d69 19};
20
5cbe4eec 21/* hash key -- defined separately for use as shared pointer */
ff68c719 22struct hek {
5cbe4eec 23 U32 hek_hash; /* hash of key */
24 I32 hek_len; /* length of hash key */
25 char hek_key[1]; /* variable-length hash key */
e05949c7 26 /* the hash-key is \0-terminated */
ddac11c4 27 /* after the \0 there is a byte for flags, such as whether the key is
19692e8d 28 UTF8 */
79072805 29};
30
5cbe4eec 31/* hash structure: */
6ee623d5 32/* This structure must match the beginning of struct xpvmg in sv.h. */
79072805 33struct xpvhv {
463ee0b2 34 char * xhv_array; /* pointer to malloced string */
35 STRLEN xhv_fill; /* how full xhv_array currently is */
36 STRLEN xhv_max; /* subscript of last element of xhv_array */
6ee623d5 37 IV xhv_keys; /* how many elements in the array */
65202027 38 NV xnv_nv; /* numeric value, if any */
8aacddc1 39#define xhv_placeholders xnv_nv
79072805 40 MAGIC* xmg_magic; /* magic for scalar array */
41 HV* xmg_stash; /* class package */
42
79072805 43 I32 xhv_riter; /* current root of iterator */
44 HE *xhv_eiter; /* current entry of iterator */
45 PMOP *xhv_pmroot; /* list of pm's for this package */
46 char *xhv_name; /* name, if a symbol table */
79072805 47};
48
5cbe4eec 49/* hash a key */
5afd6d42 50/* FYI: This is the "One-at-a-Time" algorithm by Bob Jenkins
51 * from requirements by Colin Plumb.
52 * (http://burtleburtle.net/bob/hash/doobs.html) */
53/* The use of a temporary pointer and the casting games
54 * is needed to serve the dual purposes of
55 * (a) the hashed data being interpreted as "unsigned char" (new since 5.8,
56 * a "char" can be either signed or signed, depending on the compiler)
57 * (b) catering for old code that uses a "char"
58 */
bf6bd887 59#define PERL_HASH(hash,str,len) \
60 STMT_START { \
5afd6d42 61 register const char *s_PeRlHaSh_tmp = str; \
62 register const unsigned char *s_PeRlHaSh = (const unsigned char *)s_PeRlHaSh_tmp; \
bf6bd887 63 register I32 i_PeRlHaSh = len; \
64 register U32 hash_PeRlHaSh = 0; \
a6fe520e 65 while (i_PeRlHaSh--) { \
66 hash_PeRlHaSh += *s_PeRlHaSh++; \
67 hash_PeRlHaSh += (hash_PeRlHaSh << 10); \
68 hash_PeRlHaSh ^= (hash_PeRlHaSh >> 6); \
69 } \
70 hash_PeRlHaSh += (hash_PeRlHaSh << 3); \
71 hash_PeRlHaSh ^= (hash_PeRlHaSh >> 11); \
4fee5c71 72 (hash) = (hash_PeRlHaSh + (hash_PeRlHaSh << 15)); \
bf6bd887 73 } STMT_END
74
954c1994 75/*
ccfc67b7 76=head1 Hash Manipulation Functions
77
954c1994 78=for apidoc AmU||HEf_SVKEY
79This flag, used in the length slot of hash entries and magic structures,
d1be9408 80specifies the structure contains an C<SV*> pointer where a C<char*> pointer
954c1994 81is to be expected. (For information only--not to be used).
82
ccfc67b7 83=head1 Handy Values
84
954c1994 85=for apidoc AmU||Nullhv
86Null HV pointer.
87
ccfc67b7 88=head1 Hash Manipulation Functions
89
954c1994 90=for apidoc Am|char*|HvNAME|HV* stash
91Returns the package name of a stash. See C<SvSTASH>, C<CvSTASH>.
92
93=for apidoc Am|void*|HeKEY|HE* he
94Returns the actual pointer stored in the key slot of the hash entry. The
95pointer may be either C<char*> or C<SV*>, depending on the value of
96C<HeKLEN()>. Can be assigned to. The C<HePV()> or C<HeSVKEY()> macros are
97usually preferable for finding the value of a key.
98
99=for apidoc Am|STRLEN|HeKLEN|HE* he
100If this is negative, and amounts to C<HEf_SVKEY>, it indicates the entry
101holds an C<SV*> key. Otherwise, holds the actual length of the key. Can
102be assigned to. The C<HePV()> macro is usually preferable for finding key
103lengths.
104
105=for apidoc Am|SV*|HeVAL|HE* he
106Returns the value slot (type C<SV*>) stored in the hash entry.
107
108=for apidoc Am|U32|HeHASH|HE* he
109Returns the computed hash stored in the hash entry.
110
111=for apidoc Am|char*|HePV|HE* he|STRLEN len
112Returns the key slot of the hash entry as a C<char*> value, doing any
113necessary dereferencing of possibly C<SV*> keys. The length of the string
114is placed in C<len> (this is a macro, so do I<not> use C<&len>). If you do
115not care about what the length of the key is, you may use the global
116variable C<PL_na>, though this is rather less efficient than using a local
117variable. Remember though, that hash keys in perl are free to contain
118embedded nulls, so using C<strlen()> or similar is not a good way to find
119the length of hash keys. This is very similar to the C<SvPV()> macro
120described elsewhere in this document.
121
122=for apidoc Am|SV*|HeSVKEY|HE* he
123Returns the key as an C<SV*>, or C<Nullsv> if the hash entry does not
124contain an C<SV*> key.
125
126=for apidoc Am|SV*|HeSVKEY_force|HE* he
127Returns the key as an C<SV*>. Will create and return a temporary mortal
128C<SV*> if the hash entry contains only a C<char*> key.
129
130=for apidoc Am|SV*|HeSVKEY_set|HE* he|SV* sv
131Sets the key to a given C<SV*>, taking care to set the appropriate flags to
132indicate the presence of an C<SV*> key, and returns the same
133C<SV*>.
134
135=cut
136*/
bf6bd887 137
bf5b86ae 138/* these hash entry flags ride on hent_klen (for use only in magic/tied HVs) */
d1be9408 139#define HEf_SVKEY -2 /* hent_key is an SV* */
bf6bd887 140
141
79072805 142#define Nullhv Null(HV*)
4efcf9a2 143#define HvARRAY(hv) (*(HE***)&((XPVHV*) SvANY(hv))->xhv_array)
79072805 144#define HvFILL(hv) ((XPVHV*) SvANY(hv))->xhv_fill
463ee0b2 145#define HvMAX(hv) ((XPVHV*) SvANY(hv))->xhv_max
79072805 146#define HvRITER(hv) ((XPVHV*) SvANY(hv))->xhv_riter
147#define HvEITER(hv) ((XPVHV*) SvANY(hv))->xhv_eiter
148#define HvPMROOT(hv) ((XPVHV*) SvANY(hv))->xhv_pmroot
149#define HvNAME(hv) ((XPVHV*) SvANY(hv))->xhv_name
a0d0e21e 150
8aacddc1 151/* the number of keys (including any placeholers) */
152#define XHvTOTALKEYS(xhv) ((xhv)->xhv_keys)
153
154/* The number of placeholders in the enumerated-keys hash */
205c8ad3 155#define XHvPLACEHOLDERS(xhv) ((xhv)->xhv_placeholders)
8aacddc1 156
dd2710a1 157/* the number of keys that exist() (i.e. excluding placeholders) */
158#define XHvUSEDKEYS(xhv) (XHvTOTALKEYS(xhv) - (IV)XHvPLACEHOLDERS(xhv))
8aacddc1 159
160/*
161 * HvKEYS gets the number of keys that actually exist(), and is provided
162 * for backwards compatibility with old XS code. The core uses HvUSEDKEYS
163 * (keys, excluding placeholdes) and HvTOTALKEYS (including placeholders)
164 */
165#define HvKEYS(hv) XHvUSEDKEYS((XPVHV*) SvANY(hv))
166#define HvUSEDKEYS(hv) XHvUSEDKEYS((XPVHV*) SvANY(hv))
167#define HvTOTALKEYS(hv) XHvTOTALKEYS((XPVHV*) SvANY(hv))
168#define HvPLACEHOLDERS(hv) XHvPLACEHOLDERS((XPVHV*) SvANY(hv))
169
bf6bd887 170#define HvSHAREKEYS(hv) (SvFLAGS(hv) & SVphv_SHAREKEYS)
171#define HvSHAREKEYS_on(hv) (SvFLAGS(hv) |= SVphv_SHAREKEYS)
172#define HvSHAREKEYS_off(hv) (SvFLAGS(hv) &= ~SVphv_SHAREKEYS)
173
19692e8d 174/* This is an optimisation flag. It won't be set if all hash keys have a 0
175 * flag. Currently the only flags relate to utf8.
176 * Hence it won't be set if all keys are 8 bit only. It will be set if any key
177 * is utf8 (including 8 bit keys that were entered as utf8, and need upgrading
178 * when retrieved during iteration. It may still be set when there are no longer
179 * any utf8 keys.
180 */
181#define HvHASKFLAGS(hv) (SvFLAGS(hv) & SVphv_HASKFLAGS)
182#define HvHASKFLAGS_on(hv) (SvFLAGS(hv) |= SVphv_HASKFLAGS)
183#define HvHASKFLAGS_off(hv) (SvFLAGS(hv) &= ~SVphv_HASKFLAGS)
574c8022 184
bf5b86ae 185#define HvLAZYDEL(hv) (SvFLAGS(hv) & SVphv_LAZYDEL)
186#define HvLAZYDEL_on(hv) (SvFLAGS(hv) |= SVphv_LAZYDEL)
187#define HvLAZYDEL_off(hv) (SvFLAGS(hv) &= ~SVphv_LAZYDEL)
188
a0d0e21e 189/* Maybe amagical: */
190/* #define HV_AMAGICmb(hv) (SvFLAGS(hv) & (SVpgv_badAM | SVpgv_AM)) */
191
192#define HV_AMAGIC(hv) (SvFLAGS(hv) & SVpgv_AM)
193#define HV_AMAGIC_on(hv) (SvFLAGS(hv) |= SVpgv_AM)
194#define HV_AMAGIC_off(hv) (SvFLAGS(hv) &= ~SVpgv_AM)
195
196/*
197#define HV_AMAGICbad(hv) (SvFLAGS(hv) & SVpgv_badAM)
198#define HV_badAMAGIC_on(hv) (SvFLAGS(hv) |= SVpgv_badAM)
199#define HV_badAMAGIC_off(hv) (SvFLAGS(hv) &= ~SVpgv_badAM)
200*/
bf6bd887 201
202#define Nullhe Null(HE*)
203#define HeNEXT(he) (he)->hent_next
ff68c719 204#define HeKEY_hek(he) (he)->hent_hek
205#define HeKEY(he) HEK_KEY(HeKEY_hek(he))
bbce6d69 206#define HeKEY_sv(he) (*(SV**)HeKEY(he))
ff68c719 207#define HeKLEN(he) HEK_LEN(HeKEY_hek(he))
da58a35d 208#define HeKUTF8(he) HEK_UTF8(HeKEY_hek(he))
19692e8d 209#define HeKWASUTF8(he) HEK_WASUTF8(HeKEY_hek(he))
da58a35d 210#define HeKLEN_UTF8(he) (HeKUTF8(he) ? -HeKLEN(he) : HeKLEN(he))
19692e8d 211#define HeKFLAGS(he) HEK_FLAGS(HeKEY_hek(he))
bf6bd887 212#define HeVAL(he) (he)->hent_val
ff68c719 213#define HeHASH(he) HEK_HASH(HeKEY_hek(he))
1e422769 214#define HePV(he,lp) ((HeKLEN(he) == HEf_SVKEY) ? \
215 SvPV(HeKEY_sv(he),lp) : \
216 (((lp = HeKLEN(he)) >= 0) ? \
217 HeKEY(he) : Nullch))
218
bbce6d69 219#define HeSVKEY(he) ((HeKEY(he) && \
220 HeKLEN(he) == HEf_SVKEY) ? \
221 HeKEY_sv(he) : Nullsv)
222
223#define HeSVKEY_force(he) (HeKEY(he) ? \
224 ((HeKLEN(he) == HEf_SVKEY) ? \
225 HeKEY_sv(he) : \
79cb57f6 226 sv_2mortal(newSVpvn(HeKEY(he), \
bbce6d69 227 HeKLEN(he)))) : \
3280af22 228 &PL_sv_undef)
1e422769 229#define HeSVKEY_set(he,sv) ((HeKLEN(he) = HEf_SVKEY), (HeKEY_sv(he) = sv))
bbce6d69 230
ff68c719 231#define Nullhek Null(HEK*)
71be2cbc 232#define HEK_BASESIZE STRUCT_OFFSET(HEK, hek_key[0])
ff68c719 233#define HEK_HASH(hek) (hek)->hek_hash
234#define HEK_LEN(hek) (hek)->hek_len
235#define HEK_KEY(hek) (hek)->hek_key
19692e8d 236#define HEK_FLAGS(hek) (*((unsigned char *)(HEK_KEY(hek))+HEK_LEN(hek)+1))
237
238#define HVhek_UTF8 0x01 /* Key is utf8 encoded. */
239#define HVhek_WASUTF8 0x02 /* Key is bytes here, but was supplied as utf8. */
240#define HVhek_FREEKEY 0x100 /* Internal flag to say key is malloc()ed. */
e16e2ff8 241#define HVhek_PLACEHOLD 0x200 /* Internal flag to create placeholder.
242 * (may change, but Storable is a core module) */
19692e8d 243#define HVhek_MASK 0xFF
244
245#define HEK_UTF8(hek) (HEK_FLAGS(hek) & HVhek_UTF8)
246#define HEK_UTF8_on(hek) (HEK_FLAGS(hek) |= HVhek_UTF8)
247#define HEK_UTF8_off(hek) (HEK_FLAGS(hek) &= ~HVhek_UTF8)
248#define HEK_WASUTF8(hek) (HEK_FLAGS(hek) & HVhek_WASUTF8)
249#define HEK_WASUTF8_on(hek) (HEK_FLAGS(hek) |= HVhek_WASUTF8)
250#define HEK_WASUTF8_off(hek) (HEK_FLAGS(hek) &= ~HVhek_WASUTF8)
d18c6117 251
5cbe4eec 252/* calculate HV array allocation */
d18c6117 253#if defined(STRANGE_MALLOC) || defined(MYMALLOC)
254# define PERL_HV_ARRAY_ALLOC_BYTES(size) ((size) * sizeof(HE*))
255#else
256# define MALLOC_OVERHEAD 16
257# define PERL_HV_ARRAY_ALLOC_BYTES(size) \
258 (((size) < 64) \
259 ? (size) * sizeof(HE*) \
260 : (size) * sizeof(HE*) * 2 - MALLOC_OVERHEAD)
261#endif
37d85e3a 262
e16e2ff8 263/* Flags for hv_iternext_flags. */
264#define HV_ITERNEXT_WANTPLACEHOLDERS 0x01 /* Don't skip placeholders. */
265
37d85e3a 266/* available as a function in hv.c */
267#define Perl_sharepvn(sv, len, hash) HEK_KEY(share_hek(sv, len, hash))
268#define sharepvn(sv, len, hash) Perl_sharepvn(sv, len, hash)