/* Partial definitions for data structures and routines for the regular
   expression library, version 0.12.

   Only contains defines & information used by RISC OS module RegEx which
   contains regex 0.12. See elsewhere in the archive for the full header.

   RISC OS port by Neil Bird, 1999, mailto:regex@fnxweb.com.


   Copyright (C) 1985, 1989, 1990, 1991, 1992, 1993, 1999
      Free Software Foundation, Inc.

   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2, or (at your option)
   any later version.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program; if not, write to the Free Software
   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */

#ifndef __REGEXMOD_H__
#define __REGEXMOD_H__




/* The following bits are used to determine the regexp syntax we
   recognize.  The set/not-set meanings are chosen so that Emacs syntax
   remains the value 0.  The bits are given in alphabetical order, and
   the definitions shifted by one from the previous bit; thus, when we
   add or remove a bit, only one other definition need change.  */


/* If this bit is not set, then \ inside a bracket expression is literal.
   If set, then such a \ quotes the following character.  */
#define RE_BACKSLASH_ESCAPE_IN_LISTS (1)

/* If this bit is not set, then + and ? are operators, and \+ and \? are
     literals.
   If set, then \+ and \? are operators and + and ? are literals.  */
#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)

/* If this bit is set, then character classes are supported.  They are:
     [:alpha:], [:upper:], [:lower:],  [:digit:], [:alnum:], [:xdigit:],
     [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
   If not set, then character classes are not supported.  */
#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)

/* If this bit is set, then ^ and $ are always anchors (outside bracket
     expressions, of course).
   If this bit is not set, then it depends:
        ^  is an anchor if it is at the beginning of a regular
           expression or after an open-group or an alternation operator;
        $  is an anchor if it is at the end of a regular expression, or
           before a close-group or an alternation operator.

   This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
   POSIX draft 11.2 says that * etc. in leading positions is undefined.
   We already implemented a previous draft which made those constructs
   invalid, though, so we haven't changed the code back.  */
#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)

/* If this bit is set, then special characters are always special
     regardless of where they are in the pattern.
   If this bit is not set, then special characters are special only in
     some contexts; otherwise they are ordinary.  Specifically,
     * + ? and intervals are only special when not after the beginning,
     open-group, or alternation operator.  */
#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)

/* If this bit is set, then *, +, ?, and { cannot be first in an re or
     immediately after an alternation or begin-group operator.  */
#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)

/* If this bit is set, then . matches newline.
   If not set, then it doesn't.  */
#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)

/* If this bit is set, then . doesn't match NUL.
   If not set, then it does.  */
#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)

/* If this bit is set, nonmatching lists [^...] do not match newline.
   If not set, they do.  */
#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)

/* If this bit is set, either \{...\} or {...} defines an
     interval, depending on RE_NO_BK_BRACES.
   If not set, \{, \}, {, and } are literals.  */
#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)

/* If this bit is set, +, ? and | aren't recognized as operators.
   If not set, they are.  */
#define RE_LIMITED_OPS (RE_INTERVALS << 1)

/* If this bit is set, newline is an alternation operator.
   If not set, newline is literal.  */
#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)

/* If this bit is set, then `{...}' defines an interval, and \{ and \}
     are literals.
  If not set, then `\{...\}' defines an interval.  */
#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)

/* If this bit is set, (...) defines a group, and \( and \) are literals.
   If not set, \(...\) defines a group, and ( and ) are literals.  */
#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)

/* If this bit is set, then \<digit> matches <digit>.
   If not set, then \<digit> is a back-reference.  */
#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)

/* If this bit is set, then | is an alternation operator, and \| is literal.
   If not set, then \| is an alternation operator, and | is literal.  */
#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)

/* If this bit is set, then an ending range point collating higher
     than the starting range point, as in [z-a], is invalid.
   If not set, then when ending range point collates higher than the
     starting range point, the range is ignored.  */
#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)

/* If this bit is set, then an unmatched ) is ordinary.
   If not set, then an unmatched ) is invalid.  */
#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)


/* If this bit is set, then the expression will be converted in it's
   entirety to be case-insensitive - for SWI RegEx_CompileExtendedPattern
   ONLY. An error will be generated if this bit is set in CompilePattern.
   When this bit is set, the extensions \i & \I are ignored.  */
#define RE_CASE_INSENSITIVE (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)



/* Define combinations of the above bits for the standard possibilities. */
#define RE_SYNTAX_DEFAULT -1  /* RegEx module specific */

#define RE_SYNTAX_EMACS 0

#define RE_SYNTAX_AWK                                                   \
  (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL                       \
   | RE_NO_BK_PARENS            | RE_NO_BK_REFS                         \
   | RE_NO_BK_VBAR               | RE_NO_EMPTY_RANGES                   \
   | RE_UNMATCHED_RIGHT_PAREN_ORD)

#define RE_SYNTAX_POSIX_AWK                                             \
  (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS)

#define RE_SYNTAX_GREP                                                  \
  (RE_BK_PLUS_QM              | RE_CHAR_CLASSES                         \
   | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS                            \
   | RE_NEWLINE_ALT)

#define RE_SYNTAX_EGREP                                                 \
  (RE_CHAR_CLASSES        | RE_CONTEXT_INDEP_ANCHORS                    \
   | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE                    \
   | RE_NEWLINE_ALT       | RE_NO_BK_PARENS                             \
   | RE_NO_BK_VBAR)

#define RE_SYNTAX_POSIX_EGREP                                           \
  (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES)

/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff.  */
#define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC

#define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC

/* Syntax bits common to both basic and extended POSIX regex syntax.  */
#define _RE_SYNTAX_POSIX_COMMON                                         \
  (RE_CHAR_CLASSES | RE_DOT_NEWLINE      | RE_DOT_NOT_NULL              \
   | RE_INTERVALS  | RE_NO_EMPTY_RANGES)

#define RE_SYNTAX_POSIX_BASIC                                           \
  (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM)

/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
   RE_LIMITED_OPS, i.e., \? \+ \| are not recognized.  Actually, this
   isn't minimal, since other operators, such as \`, aren't disabled.  */
#define RE_SYNTAX_POSIX_MINIMAL_BASIC                                   \
  (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)

#define RE_SYNTAX_POSIX_EXTENDED                                        \
  (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS                   \
   | RE_CONTEXT_INDEP_OPS  | RE_NO_BK_BRACES                            \
   | RE_NO_BK_PARENS       | RE_NO_BK_VBAR                              \
   | RE_UNMATCHED_RIGHT_PAREN_ORD)

/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS
   replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added.  */
#define RE_SYNTAX_POSIX_MINIMAL_EXTENDED                                \
  (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS                  \
   | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES                           \
   | RE_NO_BK_PARENS        | RE_NO_BK_REFS                             \
   | RE_NO_BK_VBAR          | RE_UNMATCHED_RIGHT_PAREN_ORD)





/* RegEx RISC OS error numbers will be listed in the final release */


/* If any error codes are removed, changed, or added, update the
   `re_error_msg' table in regex.c.  */
typedef enum
{
  REG_NOERROR = 0,      /* Success.  */
  REG_NOMATCH,          /* Didn't find a match (for regexec).  */

  /* POSIX regcomp return error codes.  (In the order listed in the
     standard.)  */
  REG_BADPAT,           /* Invalid pattern.  */
  REG_ECOLLATE,         /* Not implemented.  */
  REG_ECTYPE,           /* Invalid character class name.  */
  REG_EESCAPE,          /* Trailing backslash.  */
  REG_ESUBREG,          /* Invalid back reference.  */
  REG_EBRACK,           /* Unmatched left bracket.  */
  REG_EPAREN,           /* Parenthesis imbalance.  */
  REG_EBRACE,           /* Unmatched \{.  */
  REG_BADBR,            /* Invalid contents of \{\}.  */
  REG_ERANGE,           /* Invalid range end.  */
  REG_ESPACE,           /* Ran out of memory.  */
  REG_BADRPT,           /* No preceding re for repetition op.  */

  /* Error codes we've added.  */
  REG_EEND,             /* Premature end.  */
  REG_ESIZE,            /* Compiled pattern bigger than 2^16 bytes.  */
  REG_ERPAREN           /* Unmatched ) or \); not returned from regcomp.  */
} reg_errcode_t;




/* Type for byte offsets within the string.  POSIX mandates this.  */
typedef int regoff_t;


/* This is the structure we store register match data in.  See
   regex.texinfo for a full description of what registers match.  */
struct re_registers
{
  unsigned num_regs;
  regoff_t *start;
  regoff_t *end;
};

/* the 'reg_p' return data from the RegEx module is a pointer to this struct */



/* Declarations for routines.  I'll tidy this up soon. */


/* Compile the regular expression PATTERN, with length LENGTH
   and syntax given by SYNTAX.
   Return NULL if successful, and an error string if not.

SWI RegEx_CompilePattern:
      // R0 = handle to reuse, or 0 to get new one
      // R1 = pattern
      // R2 = length (strlen if -1)
      // R3 = syntax (previous/default if -1)
      //   returns:
      // R0 = allocated handle (if R0=0 on input)
      // R4 = error-code
*/



/* Compile a fastmap for the compiled pattern in HANDLE; used to
   accelerate searches.  Return 0 if successful and -2 if was an
   internal error.

SWI RegEx_CompileFastmap:
      // R0 = handle
      //   returns:
      // R1 = error-code
*/



/* Search in the string STRING (with length LENGTH) for the pattern
   compiled into HANDLE.  Start searching at position START, for RANGE
   characters.  Return the starting position of the match, -1 for no
   match, or -2 for an internal error.  Also return register
   information in REGS.

SWI RegEx_Search:
      // R0 = handle
      // R1 = string
      // R2 = length (strlen if -1)
      // R3 = start
      // R4 = range  (length-start if -1)
      //   returns:
      // R5 = position of match (or -1=none or -2=error)
      // R6 = pointer to regs
*/



/* Like `Search', but search in the concatenation of STRING1 and
   STRING2.  Also, stop searching at index START + STOP.

SWI RegEx_Search2:
      // R0 = handle
      // R1 = string1
      // R2 = length1 (strlen if -1)
      // R3 = string2
      // R4 = length2 (strlen if -1)
      // R5 = start
      // R6 = range  (length1+length2-start if -1)
      // R7 = stop   (range if -1)
      //   returns:
      // R8 = position of match (or -1=none or -2=error)
      // R9 = pointer to regs
*/



/* Like `Search', but return how many characters in STRING the regexp
   in BUFFER matched, starting at position START.

SWI RegEx_Match:
      // R0 = handle
      // R1 = string
      // R2 = length (strlen if -1)
      // R3 = start
      //   returns:
      // R4 = number of matching characters
      // R5 = pointer to regs
*/



/* Relates to `Match' as `Search2' relates to `Search'.

SWI RegEx_Match2:
      // R0 = handle
      // R1 = string1
      // R2 = length1 (strlen if -1)
      // R3 = string2
      // R4 = length2 (strlen if -1)
      // R5 = start
      // R6 = stop
      //   returns:
      // R7 = number of matching characters
      // R8 = pointer to regs
*/



/* Set HANDLE's regs to hold NUM_REGS registers, storing them in STARTS and
   ENDS.  Subsequent matches using BUFFER and regs will use this memory
   for recording register information.  STARTS and ENDS must be
   allocated with malloc, and must each be at least `NUM_REGS * sizeof
   (regoff_t)' bytes long.

   If NUM_REGS == 0, then subsequent matches should allocate their own
   register data.

   Unless this function is called, the first search or match using
   PATTERN_BUFFER will allocate & manage its own register data.

SWI RegEx_SetRegisters:
      // R0 = handle
      // R1 = count
      // R2 = starts
      // R3 = ends
*/



/* Frees up all memory associated with HANDLE, except in the case where
   HANDLE has had its regs. set with SetRegisters - users are expected
   to free such regs. themselves.

case RegEx_Free:
      // R0 = handle
      //   returns:
      // R0 = 0
*/


/* Compile the regular expression PATTERN with syntax given by SYNTAX.
   Allows additional extended escape sequences (backspace followed by letter)
   for shorthand inclusion of common matches, plus inline case-sensitivity
   switching. Also works with RE_CASE_INSENSITIVE global syntax flag.
   Return NULL if successful, and an error string if not.


SWI RegEx_CompileExtendedPattern:
      // R0 = handle to reuse, or 0 to get new one
      // R1 = pattern
      // R2 = syntax (previous/default if -1)
      //   returns:
      // R0 = allocated handle (if R0=0 on input)
      // R3 = error-code


   These are the currently allowed extended sequences (not valid within []) :

         \a  Match an alpha-numeric character
         \A  Match a non-alpha-numeric character
         \d  Match a digit
         \D  Match a non-digit
         \f  Match a float-digit (digit plus '.')
         \F  Match a non-float-digit
         \h  Match a hex-digit
         \H  Match a non-hex-digit
         \i  Start matching case-insensitively
         \I  Stop matching case-insensitively (if previous \i found)
         \l  Match lower-case word character
         \L  Match non-lower-case word character
         \n  Match new-line (for those not using C)
         \N  Match non-new-line
         \r  Match carriage-return (for those not using C)
         \R  Match non-carriage-return
         \s  Match a whitespace character
         \S  Match a non-whitespace character
         \t  Match tab (for those not using C)
         \T  Match non-tab
         \u  Match upper-case word character
         \U  Match non-upper-case word character
         \w  Match a "word" character (alphanumeric plus "_")
         \W  Match a non-word character

*/


/* Returns the module's default syntax (that initially used when -1 given).

case RegEx_GetDefaultSyntax:
      //   returns:
      // R0 = syntax
*/


/* Sets the module's default syntax (that initially used when -1 given).

case RegEx_SetDefaultSyntax:
      // R0 = syntax
*/



#endif /* not __REGEXMOD_H__ */
