| 1 | /* invlist.h |
| 2 | * |
| 3 | * Copyright (C) 2012 by Larry Wall and others |
| 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 | * |
| 8 | * This file is a home for static inline functions that cannot go in other |
| 9 | * headers files, because they depend on proto.h (included after most other |
| 10 | * headers) or struct definitions. |
| 11 | * |
| 12 | */ |
| 13 | |
| 14 | #if defined(PERL_IN_UTF8_C) || defined(PERL_IN_REGCOMP_C) || defined(PERL_IN_REGEXEC_C) |
| 15 | |
| 16 | #define INVLIST_LEN_OFFSET 0 /* Number of elements in the inversion list */ |
| 17 | #define INVLIST_ITER_OFFSET 1 /* Current iteration position */ |
| 18 | #define INVLIST_PREVIOUS_INDEX_OFFSET 2 /* Place to cache index of previous |
| 19 | result */ |
| 20 | |
| 21 | /* This is a combination of a version and data structure type, so that one |
| 22 | * being passed in can be validated to be an inversion list of the correct |
| 23 | * vintage. When the structure of the header is changed, a new random number |
| 24 | * in the range 2**31-1 should be generated and the new() method changed to |
| 25 | * insert that at this location. Then, if an auxiliary program doesn't change |
| 26 | * correspondingly, it will be discovered immediately */ |
| 27 | #define INVLIST_VERSION_ID_OFFSET 3 |
| 28 | #define INVLIST_VERSION_ID 290655244 |
| 29 | |
| 30 | /* For safety, when adding new elements, remember to #undef them at the end of |
| 31 | * the inversion list code section */ |
| 32 | |
| 33 | #define INVLIST_ZERO_OFFSET 4 /* 0 or 1; must be last element in header */ |
| 34 | /* The UV at position ZERO contains either 0 or 1. If 0, the inversion list |
| 35 | * contains the code point U+00000, and begins here. If 1, the inversion list |
| 36 | * doesn't contain U+0000, and it begins at the next UV in the array. |
| 37 | * Inverting an inversion list consists of adding or removing the 0 at the |
| 38 | * beginning of it. By reserving a space for that 0, inversion can be made |
| 39 | * very fast */ |
| 40 | |
| 41 | #define HEADER_LENGTH (INVLIST_ZERO_OFFSET + 1) |
| 42 | |
| 43 | /* An element is in an inversion list iff its index is even numbered: 0, 2, 4, |
| 44 | * etc */ |
| 45 | #define ELEMENT_RANGE_MATCHES_INVLIST(i) (! ((i) & 1)) |
| 46 | #define PREV_RANGE_MATCHES_INVLIST(i) (! ELEMENT_RANGE_MATCHES_INVLIST(i)) |
| 47 | |
| 48 | PERL_STATIC_INLINE UV* |
| 49 | S__get_invlist_len_addr(pTHX_ SV* invlist) |
| 50 | { |
| 51 | /* Return the address of the UV that contains the current number |
| 52 | * of used elements in the inversion list */ |
| 53 | |
| 54 | PERL_ARGS_ASSERT__GET_INVLIST_LEN_ADDR; |
| 55 | |
| 56 | return (UV *) (SvPVX(invlist) + (INVLIST_LEN_OFFSET * sizeof (UV))); |
| 57 | } |
| 58 | |
| 59 | PERL_STATIC_INLINE UV |
| 60 | S__invlist_len(pTHX_ SV* const invlist) |
| 61 | { |
| 62 | /* Returns the current number of elements stored in the inversion list's |
| 63 | * array */ |
| 64 | |
| 65 | PERL_ARGS_ASSERT__INVLIST_LEN; |
| 66 | |
| 67 | return *_get_invlist_len_addr(invlist); |
| 68 | } |
| 69 | |
| 70 | PERL_STATIC_INLINE bool |
| 71 | S__invlist_contains_cp(pTHX_ SV* const invlist, const UV cp) |
| 72 | { |
| 73 | /* Does <invlist> contain code point <cp> as part of the set? */ |
| 74 | |
| 75 | IV index = _invlist_search(invlist, cp); |
| 76 | |
| 77 | PERL_ARGS_ASSERT__INVLIST_CONTAINS_CP; |
| 78 | |
| 79 | return index >= 0 && ELEMENT_RANGE_MATCHES_INVLIST(index); |
| 80 | } |
| 81 | |
| 82 | #endif |