src/regex/regex2.h

   1 /*
   2  * First, the stuff that ends up in the outside-world include file
   3  = typedef off_t regoff_t;
   4  = typedef struct {
   5  =      int re_magic;
   6  =      size_t re_nsub;         // number of parenthesized subexpressions
   7  =      const char *re_endp;    // end pointer for REG_PEND
   8  =      struct re_guts *re_g;   // none of your business :-)
   9  = } regex_t;
  10  = typedef struct {
  11  =      regoff_t rm_so;         // start of match
  12  =      regoff_t rm_eo;         // end of match
  13  = } regmatch_t;
  14  */
  15 /*
  16  * internals of regex_t
  17  */
  18 #define MAGIC1  ((('r'^0200)<<8) | 'e')
  19
  20 /*
  21  * The internal representation is a *strip*, a sequence of
  22  * operators ending with an endmarker.  (Some terminology etc. is a
  23  * historical relic of earlier versions which used multiple strips.)
  24  * Certain oddities in the representation are there to permit running
  25  * the machinery backwards; in particular, any deviation from sequential
  26  * flow must be marked at both its source and its destination.  Some
  27  * fine points:
  28  *
  29  * - OPLUS_ and O_PLUS are *inside* the loop they create.
  30  * - OQUEST_ and O_QUEST are *outside* the bypass they create.
  31  * - OCH_ and O_CH are *outside* the multi-way branch they create, while
  32  *   OOR1 and OOR2 are respectively the end and the beginning of one of
  33  *   the branches.  Note that there is an implicit OOR2 following OCH_
  34  *   and an implicit OOR1 preceding O_CH.
  35  *
  36  * In state representations, an operator's bit is on to signify a state
  37  * immediately *preceding* "execution" of that operator.
  38  */
  39 typedef long sop;               /* strip operator */
  40 typedef long sopno;
  41 #define OPRMASK 0x7c000000
  42 #define OPDMASK 0x03ffffff
  43 #define OPSHIFT (26)
  44 #define OP(n)   ((n)&OPRMASK)
  45 #define OPND(n) ((n)&OPDMASK)
  46 #define SOP(op, opnd)   ((op)|(opnd))
  47 /* operators                       meaning      operand                 */
  48 /*                                              (back, fwd are offsets) */
  49 #define OEND    (1<<OPSHIFT)    /* endmarker    -                       */
  50 #define OCHAR   (2<<OPSHIFT)    /* character    unsigned char           */
  51 #define OBOL    (3<<OPSHIFT)    /* left anchor  -                       */
  52 #define OEOL    (4<<OPSHIFT)    /* right anchor -                       */
  53 #define OANY    (5<<OPSHIFT)    /* .            -                       */
  54 #define OANYOF  (6<<OPSHIFT)    /* [...]        set number              */
  55 #define OBACK_  (7<<OPSHIFT)    /* begin \d     paren number            */
  56 #define O_BACK  (8<<OPSHIFT)    /* end \d       paren number            */
  57 #define OPLUS_  (9<<OPSHIFT)    /* + prefix     fwd to suffix           */
  58 #define O_PLUS  (10<<OPSHIFT)   /* + suffix     back to prefix          */
  59 #define OQUEST_ (11<<OPSHIFT)   /* ? prefix     fwd to suffix           */
  60 #define O_QUEST (12<<OPSHIFT)   /* ? suffix     back to prefix          */
  61 #define OLPAREN (13<<OPSHIFT)   /* (            fwd to )                */
  62 #define ORPAREN (14<<OPSHIFT)   /* )            back to (               */
  63 #define OCH_    (15<<OPSHIFT)   /* begin choice fwd to OOR2             */
  64 #define OOR1    (16<<OPSHIFT)   /* | pt. 1      back to OOR1 or OCH_    */
  65 #define OOR2    (17<<OPSHIFT)   /* | pt. 2      fwd to OOR2 or O_CH     */
  66 #define O_CH    (18<<OPSHIFT)   /* end choice   back to OOR1            */
  67 #define OBOW    (19<<OPSHIFT)   /* begin word   -                       */
  68 #define OEOW    (20<<OPSHIFT)   /* end word     -                       */
  69
  70 /*
  71  * Structure for [] character-set representation.  Character sets are
  72  * done as bit vectors, grouped 8 to a byte vector for compactness.
  73  * The individual set therefore has both a pointer to the byte vector
  74  * and a mask to pick out the relevant bit of each byte.  A hash code
  75  * simplifies testing whether two sets could be identical.
  76  *
  77  * This will get trickier for multicharacter collating elements.  As
  78  * preliminary hooks for dealing with such things, we also carry along
  79  * a string of multi-character elements, and decide the size of the
  80  * vectors at run time.
  81  */
  82 typedef struct {
  83         uch *ptr;               /* -> uch [csetsize] */
  84         uch mask;               /* bit within array */
  85         uch hash;               /* hash code */
  86         size_t smultis;
  87         char *multis;           /* -> char[smulti]  ab\0cd\0ef\0\0 */
  88 } cset;
  89 /* note that CHadd and CHsub are unsafe, and CHIN doesn't yield 0/1 */
  90 #define CHadd(cs, c)    ((cs)->ptr[(uch)(c)] |= (cs)->mask, (cs)->hash += (c))
  91 #define CHsub(cs, c)    ((cs)->ptr[(uch)(c)] &= ~(cs)->mask, (cs)->hash -= (c))
  92 #define CHIN(cs, c)     ((cs)->ptr[(uch)(c)] & (cs)->mask)
  93 #define MCadd(p, cs, cp)        mcadd(p, cs, cp)        /* regcomp() internal fns */
  94 #define MCsub(p, cs, cp)        mcsub(p, cs, cp)
  95 #define MCin(p, cs, cp) mcin(p, cs, cp)
  96
  97 /* stuff for character categories */
  98 typedef unsigned char cat_t;
  99
 100 /*
 101  * main compiled-expression structure
 102  */
 103 struct re_guts {
 104         int magic;
 105 #               define  MAGIC2  ((('R'^0200)<<8)|'E')
 106         sop *strip;             /* malloced area for strip */
 107         int csetsize;           /* number of bits in a cset vector */
 108         int ncsets;             /* number of csets in use */
 109         cset *sets;             /* -> cset [ncsets] */
 110         uch *setbits;           /* -> uch[csetsize][ncsets/CHAR_BIT] */
 111         int cflags;             /* copy of regcomp() cflags argument */
 112         sopno nstates;          /* = number of sops */
 113         sopno firststate;       /* the initial OEND (normally 0) */
 114         sopno laststate;        /* the final OEND */
 115         int iflags;             /* internal flags */
 116 #               define  USEBOL  01      /* used ^ */
 117 #               define  USEEOL  02      /* used $ */
 118 #               define  BAD     04      /* something wrong */
 119         int nbol;               /* number of ^ used */
 120         int neol;               /* number of $ used */
 121         int ncategories;        /* how many character categories */
 122         cat_t *categories;      /* ->catspace[-CHAR_MIN] */
 123         char *must;             /* match must contain this string */
 124         int mlen;               /* length of must */
 125         size_t nsub;            /* copy of re_nsub */
 126         int backrefs;           /* does it use back references? */
 127         sopno nplus;            /* how deep does it nest +s? */
 128         /* catspace must be last */
 129         cat_t catspace[1];      /* actually [NC] */
 130 };
 131
 132 /* misc utilities */
 133 #define OUT     (CHAR_MAX+1)    /* a non-character value */
 134 #define ISWORD(c)       (isalnum(c) || (c) == '_')