match.c

国外网站上的一些精典的C程序

/*
 EPSHeader

   File: match.c
   Author: J. Kercheval
   Created: Sat, 01/05/1991  22:21:49
*/
/*
 EPSRevision History

   J. Kercheval  Wed, 02/20/1991  22:29:01  Released to Public Domain
   J. Kercheval  Fri, 02/22/1991  15:29:01  fix '\' bugs (two :( of them)
   J. Kercheval  Sun, 03/10/1991  19:31:29  add error return to matche()
   J. Kercheval  Sun, 03/10/1991  20:11:11  add is_valid_pattern code
   J. Kercheval  Sun, 03/10/1991  20:37:11  beef up main()
   J. Kercheval  Tue, 03/12/1991  22:25:10  Released as V1.1 to Public Domain
*/

/*
   Wildcard Pattern Matching
*/


#include "match.h"

int matche_after_star (register char *pattern, register char *text);
int fast_match_after_star (register char *pattern, register char *text);

/*----------------------------------------------------------------------------
*
* Return TRUE if PATTERN has any special wildcard characters
*
----------------------------------------------------------------------------*/

BOOLEAN is_pattern (char *p)
{
      while (*p)
      {
            switch (*p++)
            {
            case '?':
            case '*':
            case '[':
            case '\\':
                  return TRUE;
            }
      }
      return FALSE;
}


/*----------------------------------------------------------------------------
*
* Return TRUE if PATTERN has is a well formed regular expression according
* to the above syntax
*
* error_type is a return code based on the type of pattern error.  Zero is
* returned in error_type if the pattern is a valid one.  error_type return
* values are as follows:
*
*   PATTERN_VALID - pattern is well formed
*   PATTERN_ESC   - pattern has invalid escape ('\' at end of pattern)
*   PATTERN_RANGE - [..] construct has a no end range in a '-' pair (ie [a-])
*   PATTERN_CLOSE - [..] construct has no end bracket (ie [abc-g )
*   PATTERN_EMPTY - [..] construct is empty (ie [])
*
----------------------------------------------------------------------------*/

BOOLEAN is_valid_pattern (char *p, int *error_type)
{
      /* init error_type */
      *error_type = PATTERN_VALID;

      /* loop through pattern to EOS */
      while (*p)
      {
            /* determine pattern type */
            switch (*p)
            {
            /* check literal escape, it cannot be at end of pattern */
            case '\\':
                  if (!*++p)
                  {
                        *error_type = PATTERN_ESC;
                        return FALSE;
                  }
                  p++;
                  break;
                  
                  /* the [..] construct must be well formed */
            case '[':
                  p++;

                  /* if the next character is ']' then bad pattern */
                  if (*p == ']')
                  {
                        *error_type = PATTERN_EMPTY;
                        return FALSE;
                  }
                
                  /* if end of pattern here then bad pattern */
                  if (!*p)
                  {
                        *error_type = PATTERN_CLOSE;
                        return FALSE;
                  }

                  /* loop to end of [..] construct */
                  while (*p != ']')
                  {
                        /* check for literal escape */
                        if (*p == '\\')
                        {
                              p++;

                              /* if end of pattern here then bad pattern */
                              if (!*p++)
                              {
                                    *error_type = PATTERN_ESC;
                                    return FALSE;
                              }
                        }
                        else  p++;

                        /* if end of pattern here then bad pattern */
                        if (!*p)
                        {
                              *error_type = PATTERN_CLOSE;
                              return FALSE;
                        }

                        /* if this a range */
                        if (*p == '-')
                        {
                              /* we must have an end of range */
                              if (!*++p || *p == ']')
                              {
                                    *error_type = PATTERN_RANGE;
                                    return FALSE;
                              }
                              else
                              {

                                    /* check for literal escape */
                                    if (*p == '\\')
                                          p++;

                                    /* if end of pattern here
                                       then bad pattern           */
                                    if (!*p++)
                                    {
                                          *error_type = PATTERN_ESC;
                                          return FALSE;
                                    }
                              }
                        }
                  }
                  break;

                  /* all other characters are valid pattern elements */
            case '*':
            case '?':
            default:
                  p++;                              /* "normal" character */
                  break;
            }
      }
      return TRUE;
}


/*----------------------------------------------------------------------------
*
*  Match the pattern PATTERN against the string TEXT;
*
*  returns MATCH_VALID if pattern matches, or an errorcode as follows
*  otherwise:
*
*            MATCH_PATTERN  - bad pattern
*            MATCH_LITERAL  - match failure on literal mismatch
*            MATCH_RANGE    - match failure on [..] construct
*            MATCH_ABORT    - premature end of text string
*            MATCH_END      - premature end of pattern string
*            MATCH_VALID    - valid match
*
*
*  A match means the entire string TEXT is used up in matching.
*
*  In the pattern string:
*       `*' matches any sequence of characters (zero or more)
*       `?' matches any character
*       [SET] matches any character in the specified set,
*       [!SET] or [^SET] matches any character not in the specified set.
*
*  A set is composed of characters or ranges; a range looks like
*  character hyphen character (as in 0-9 or A-Z).  [0-9a-zA-Z_] is the
*  minimal set of characters allowed in the [..] pattern construct.
*  Other characters are allowed (ie. 8 bit characters) if your system
*  will support them.
*
*  To suppress the special syntactic significance of any of `[]*?!^-\',
*  and match the character exactly, precede it with a `\'.
*
----------------------------------------------------------------------------*/

int matche (register char *p, register char *t)
{
      register char range_start, range_end;  /* start and end in range */

      BOOLEAN invert;             /* is this [..] or [!..] */
      BOOLEAN member_match;       /* have I matched the [..] construct? */
      BOOLEAN loop;               /* should I terminate? */

      for ( ; *p; p++, t++)
      {
            /* if this is the end of the text
               then this is the end of the match */

            if (!*t)
            {
                  return ( *p == '*' && *++p == '\0' ) ?
                        MATCH_VALID : MATCH_ABORT;
            }

            /* determine and react to pattern type */

            switch (*p)
            {
            case '?':                     /* single any character match */
                  break;

            case '*':                     /* multiple any character match */
                  return matche_after_star (p, t);

            /* [..] construct, single member/exclusion character match */
            case '[':
            {
                  /* move to beginning of range */

                  p++;

                  /* check if this is a member match or exclusion match */

                  invert = FALSE;
                  if (*p == '!' || *p == '^')
                  {
                        invert = TRUE;
                        p++;
                  }

                  /* if closing bracket here or at range start then we have a
                        malformed pattern */

                  if (*p == ']')
                  {
                        return MATCH_PATTERN;
                  }

                  member_match = FALSE;
                  loop = TRUE;

                  while (loop)
                  {
                        /* if end of construct then loop is done */

                        if (*p == ']')
                        {
                              loop = FALSE;
                              continue;
                        }

                        /* matching a '!', '^', '-', '\' or a ']' */

                        if (*p == '\\')
                        {
                              range_start = range_end = *++p;
                        }
                        else  range_start = range_end = *p;

                        /* if end of pattern then bad pattern (Missing ']') */

                        if (!*p)
                              return MATCH_PATTERN;

                        /* check for range bar */
                        if (*++p == '-')
                        {
                              /* get the range end */

                              range_end = *++p;
                              
                              /* if end of pattern or construct
                                 then bad pattern */