usr.bin/sed/regex.h

   1 /* Definitions for data structures callers pass the regex library.
   2
   3    Copyright (C) 1985, 1989-90 Free Software Foundation, Inc.
   4
   5    This program is free software; you can redistribute it and/or modify
   6    it under the terms of the GNU General Public License as published by
   7    the Free Software Foundation; either version 2, or (at your option)
   8    any later version.
   9
  10    This program is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13    GNU General Public License for more details.
  14
  15    You should have received a copy of the GNU General Public License
  16    along with this program; if not, write to the Free Software
  17    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
  18
  19
  20 #ifndef __REGEXP_LIBRARY
  21 #define __REGEXP_LIBRARY
  22
  23 /* Define number of parens for which we record the beginnings and ends.
  24    This affects how much space the `struct re_registers' type takes up.  */
  25 #ifndef RE_NREGS
  26 #define RE_NREGS 10
  27 #endif
  28
  29 #define BYTEWIDTH 8
  30
  31
  32 /* Maximum number of duplicates an interval can allow.  */
  33 #define RE_DUP_MAX  ((1 << 15) - 1)
  34
  35
  36 /* This defines the various regexp syntaxes.  */
  37 extern int obscure_syntax;
  38
  39
  40 /* The following bits are used in the obscure_syntax variable to choose among
  41    alternative regexp syntaxes.  */
  42
  43 /* If this bit is set, plain parentheses serve as grouping, and backslash
  44      parentheses are needed for literal searching.
  45    If not set, backslash-parentheses are grouping, and plain parentheses
  46      are for literal searching.  */
  47 #define RE_NO_BK_PARENS 1
  48
  49 /* If this bit is set, plain | serves as the `or'-operator, and \| is a
  50      literal.
  51    If not set, \| serves as the `or'-operator, and | is a literal.  */
  52 #define RE_NO_BK_VBAR (1 << 1)
  53
  54 /* If this bit is not set, plain + or ? serves as an operator, and \+, \? are
  55      literals.
  56    If set, \+, \? are operators and plain +, ? are literals.  */
  57 #define RE_BK_PLUS_QM (1 << 2)
  58
  59 /* If this bit is set, | binds tighter than ^ or $.
  60    If not set, the contrary.  */
  61 #define RE_TIGHT_VBAR (1 << 3)
  62
  63 /* If this bit is set, then treat newline as an OR operator.
  64    If not set, treat it as a normal character.  */
  65 #define RE_NEWLINE_OR (1 << 4)
  66
  67 /* If this bit is set, then special characters may act as normal
  68    characters in some contexts. Specifically, this applies to:
  69         ^ -- only special at the beginning, or after ( or |;
  70         $ -- only special at the end, or before ) or |;
  71         *, +, ? -- only special when not after the beginning, (, or |.
  72    If this bit is not set, special characters (such as *, ^, and $)
  73    always have their special meaning regardless of the surrounding
  74    context.  */
  75 #define RE_CONTEXT_INDEP_OPS (1 << 5)
  76
  77 /* If this bit is not set, then \ before anything inside [ and ] is taken as
  78      a real \.
  79    If set, then such a \ escapes the following character.  This is a
  80      special case for awk.  */
  81 #define RE_AWK_CLASS_HACK (1 << 6)
  82
  83 /* If this bit is set, then \{ and \} or { and } serve as interval operators.
  84    If not set, then \{ and \} and { and } are treated as literals.  */
  85 #define RE_INTERVALS (1 << 7)
  86
  87 /* If this bit is not set, then \{ and \} serve as interval operators and
  88      { and } are literals.
  89    If set, then { and } serve as interval operators and \{ and \} are
  90      literals.  */
  91 #define RE_NO_BK_CURLY_BRACES (1 << 8)
  92
  93 /* If this bit is set, then character classes are supported; they are:
  94      [:alpha:], [:upper:], [:lower:],  [:digit:], [:alnum:], [:xdigit:],
  95      [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
  96    If not set, then character classes are not supported.  */
  97 #define RE_CHAR_CLASSES (1 << 9)
  98
  99 /* If this bit is set, then the dot re doesn't match a null byte.
 100    If not set, it does.  */
 101 #define RE_DOT_NOT_NULL (1 << 10)
 102
 103 /* If this bit is set, then [^...] doesn't match a newline.
 104    If not set, it does.  */
 105 #define RE_HAT_NOT_NEWLINE (1 << 11)
 106
 107 /* If this bit is set, back references are recognized.
 108    If not set, they aren't.  */
 109 #define RE_NO_BK_REFS (1 << 12)
 110
 111 /* If this bit is set, back references must refer to a preceding
 112    subexpression.  If not set, a back reference to a nonexistent
 113    subexpression is treated as literal characters.  */
 114 #define RE_NO_EMPTY_BK_REF (1 << 13)
 115
 116 /* If this bit is set, bracket expressions can't be empty.
 117    If it is set, they can be empty.  */
 118 #define RE_NO_EMPTY_BRACKETS (1 << 14)
 119
 120 /* If this bit is set, then *, +, ? and { cannot be first in an re or
 121    immediately after a |, or a (.  Furthermore, a | cannot be first or
 122    last in an re, or immediately follow another | or a (.  Also, a ^
 123    cannot appear in a nonleading position and a $ cannot appear in a
 124    nontrailing position (outside of bracket expressions, that is).  */
 125 #define RE_CONTEXTUAL_INVALID_OPS (1 << 15)
 126
 127 /* If this bit is set, then +, ? and | aren't recognized as operators.
 128    If it's not, they are.  */
 129 #define RE_LIMITED_OPS (1 << 16)
 130
 131 /* If this bit is set, then an ending range point has to collate higher
 132      or equal to the starting range point.
 133    If it's not set, then when the ending range point collates higher
 134      than the starting range point, the range is just considered empty.  */
 135 #define RE_NO_EMPTY_RANGES (1 << 17)
 136
 137 /* If this bit is set, then a hyphen (-) can't be an ending range point.
 138    If it isn't, then it can.  */
 139 #define RE_NO_HYPHEN_RANGE_END (1 << 18)
 140
 141
 142 /* Define combinations of bits for the standard possibilities.  */
 143 #define RE_SYNTAX_POSIX_AWK (RE_NO_BK_PARENS | RE_NO_BK_VBAR \
 144                         | RE_CONTEXT_INDEP_OPS)
 145 #define RE_SYNTAX_AWK (RE_NO_BK_PARENS | RE_NO_BK_VBAR \
 146                         | RE_CONTEXT_INDEP_OPS | RE_AWK_CLASS_HACK)
 147 #define RE_SYNTAX_EGREP (RE_NO_BK_PARENS | RE_NO_BK_VBAR \
 148                         | RE_CONTEXT_INDEP_OPS | RE_NEWLINE_OR)
 149 #define RE_SYNTAX_GREP (RE_BK_PLUS_QM | RE_NEWLINE_OR)
 150 #define RE_SYNTAX_EMACS 0
 151 #define RE_SYNTAX_POSIX_BASIC (RE_INTERVALS | RE_BK_PLUS_QM             \
 152                         | RE_CHAR_CLASSES | RE_DOT_NOT_NULL             \
 153                         | RE_HAT_NOT_NEWLINE | RE_NO_EMPTY_BK_REF       \
 154                         | RE_NO_EMPTY_BRACKETS | RE_LIMITED_OPS         \
 155                         | RE_NO_EMPTY_RANGES | RE_NO_HYPHEN_RANGE_END)
 156
 157 #define RE_SYNTAX_POSIX_EXTENDED (RE_INTERVALS | RE_NO_BK_CURLY_BRACES     \
 158                         | RE_NO_BK_VBAR | RE_NO_BK_PARENS                  \
 159                         | RE_HAT_NOT_NEWLINE | RE_CHAR_CLASSES             \
 160                         | RE_NO_EMPTY_BRACKETS | RE_CONTEXTUAL_INVALID_OPS \
 161                         | RE_NO_BK_REFS | RE_NO_EMPTY_RANGES               \
 162                         | RE_NO_HYPHEN_RANGE_END)
 163
 164
 165 /* This data structure is used to represent a compiled pattern.  */
 166
 167 struct re_pattern_buffer
 168   {
 169     char *buffer;       /* Space holding the compiled pattern commands.  */
 170     long allocated;     /* Size of space that `buffer' points to. */
 171     long used;          /* Length of portion of buffer actually occupied  */
 172     char *fastmap;      /* Pointer to fastmap, if any, or zero if none.  */
 173                         /* re_search uses the fastmap, if there is one,
 174                            to skip over totally implausible characters.  */
 175     char *translate;    /* Translate table to apply to all characters before
 176                            comparing, or zero for no translation.
 177                            The translation is applied to a pattern when it is
 178                            compiled and to data when it is matched.  */
 179     char fastmap_accurate;
 180                         /* Set to zero when a new pattern is stored,
 181                            set to one when the fastmap is updated from it.  */
 182     char can_be_null;   /* Set to one by compiling fastmap
 183                            if this pattern might match the null string.
 184                            It does not necessarily match the null string
 185                            in that case, but if this is zero, it cannot.
 186                            2 as value means can match null string
 187                            but at end of range or before a character
 188                            listed in the fastmap.  */
 189   };
 190
 191
 192 /* search.c (search_buffer) needs this one value.  It is defined both in
 193    regex.c and here.  */
 194 #define RE_EXACTN_VALUE 1
 195
 196
 197 /* Structure to store register contents data in.
 198
 199    Pass the address of such a structure as an argument to re_match, etc.,
 200    if you want this information back.
 201
 202    For i from 1 to RE_NREGS - 1, start[i] records the starting index in
 203    the string of where the ith subexpression matched, and end[i] records
 204    one after the ending index.  start[0] and end[0] are analogous, for
 205    the entire pattern.  */
 206
 207 struct re_registers
 208   {
 209     int start[RE_NREGS];
 210     int end[RE_NREGS];
 211   };
 212
 213
 214 \f
 215 #ifdef __STDC__
 216
 217 extern char *re_compile_pattern (char *, int, struct re_pattern_buffer *);
 218 /* Is this really advertised?  */
 219 extern void re_compile_fastmap (struct re_pattern_buffer *);
 220 extern int re_search (struct re_pattern_buffer *, char*, int, int, int,
 221                       struct re_registers *);
 222 extern int re_search_2 (struct re_pattern_buffer *, char *, int,
 223                         char *, int, int, int,
 224                         struct re_registers *, int);
 225 extern int re_match (struct re_pattern_buffer *, char *, int, int,
 226                      struct re_registers *);
 227 extern int re_match_2 (struct re_pattern_buffer *, char *, int,
 228                        char *, int, int, struct re_registers *, int);
 229
 230 /* 4.2 bsd compatibility.  */
 231 extern char *re_comp (char *);
 232 extern int re_exec (char *);
 233
 234 #else /* !__STDC__ */
 235
 236 extern char *re_compile_pattern ();
 237 /* Is this really advertised? */
 238 extern void re_compile_fastmap ();
 239 extern int re_search (), re_search_2 ();
 240 extern int re_match (), re_match_2 ();
 241
 242 /* 4.2 bsd compatibility.  */
 243 extern char *re_comp ();
 244 extern int re_exec ();
 245
 246 #endif /* __STDC__ */
 247
 248
 249 #ifdef SYNTAX_TABLE
 250 extern char *re_syntax_table;
 251 #endif
 252
 253 #endif /* !__REGEXP_LIBRARY */