the-tree/lib/glob.c

0001 #include <linux/module.h>
0002 #include <linux/glob.h>
0003
0004 /*
0005  * The only reason this code can be compiled as a module is because the
0006  * ATA code that depends on it can be as well.  In practice, they're
0007  * both usually compiled in and the module overhead goes away.
0008  */
0009 MODULE_DESCRIPTION("glob(7) matching");
0010 MODULE_LICENSE("Dual MIT/GPL");
0011
0012 /**
0013  * glob_match - Shell-style pattern matching, like !fnmatch(pat, str, 0)
0014  * @pat: Shell-style pattern to match, e.g. "*.[ch]".
0015  * @str: String to match.  The pattern must match the entire string.
0016  *
0017  * Perform shell-style glob matching, returning true (1) if the match
0018  * succeeds, or false (0) if it fails.  Equivalent to !fnmatch(@pat, @str, 0).
0019  *
0020  * Pattern metacharacters are ?, *, [ and \.
0021  * (And, inside character classes, !, - and ].)
0022  *
0023  * This is small and simple implementation intended for device blacklists
0024  * where a string is matched against a number of patterns.  Thus, it
0025  * does not preprocess the patterns.  It is non-recursive, and run-time
0026  * is at most quadratic: strlen(@str)*strlen(@pat).
0027  *
0028  * An example of the worst case is glob_match("*aaaaa", "aaaaaaaaaa");
0029  * it takes 6 passes over the pattern before matching the string.
0030  *
0031  * Like !fnmatch(@pat, @str, 0) and unlike the shell, this does NOT
0032  * treat / or leading . specially; it isn't actually used for pathnames.
0033  *
0034  * Note that according to glob(7) (and unlike bash), character classes
0035  * are complemented by a leading !; this does not support the regex-style
0036  * [^a-z] syntax.
0037  *
0038  * An opening bracket without a matching close is matched literally.
0039  */
0040 bool __pure glob_match(char const *pat, char const *str)
0041 {
0042     /*
0043      * Backtrack to previous * on mismatch and retry starting one
0044      * character later in the string.  Because * matches all characters
0045      * (no exception for /), it can be easily proved that there's
0046      * never a need to backtrack multiple levels.
0047      */
0048     char const *back_pat = NULL, *back_str;
0049
0050     /*
0051      * Loop over each token (character or class) in pat, matching
0052      * it against the remaining unmatched tail of str.  Return false
0053      * on mismatch, or true after matching the trailing nul bytes.
0054      */
0055     for (;;) {
0056         unsigned char c = *str++;
0057         unsigned char d = *pat++;
0058
0059         switch (d) {
0060         case '?':   /* Wildcard: anything but nul */
0061             if (c == '\0')
0062                 return false;
0063             break;
0064         case '*':   /* Any-length wildcard */
0065             if (*pat == '\0')   /* Optimize trailing * case */
0066                 return true;
0067             back_pat = pat;
0068             back_str = --str;   /* Allow zero-length match */
0069             break;
0070         case '[': { /* Character class */
0071             bool match = false, inverted = (*pat == '!');
0072             char const *class = pat + inverted;
0073             unsigned char a = *class++;
0074
0075             /*
0076              * Iterate over each span in the character class.
0077              * A span is either a single character a, or a
0078              * range a-b.  The first span may begin with ']'.
0079              */
0080             do {
0081                 unsigned char b = a;
0082
0083                 if (a == '\0')  /* Malformed */
0084                     goto literal;
0085
0086                 if (class[0] == '-' && class[1] != ']') {
0087                     b = class[1];
0088
0089                     if (b == '\0')
0090                         goto literal;
0091
0092                     class += 2;
0093                     /* Any special action if a > b? */
0094                 }
0095                 match |= (a <= c && c <= b);
0096             } while ((a = *class++) != ']');
0097
0098             if (match == inverted)
0099                 goto backtrack;
0100             pat = class;
0101             }
0102             break;
0103         case '\\':
0104             d = *pat++;
0105             fallthrough;
0106         default:    /* Literal character */
0107 literal:
0108             if (c == d) {
0109                 if (d == '\0')
0110                     return true;
0111                 break;
0112             }
0113 backtrack:
0114             if (c == '\0' || !back_pat)
0115                 return false;   /* No point continuing */
0116             /* Try again from last *, one character later in str. */
0117             pat = back_pat;
0118             str = ++back_str;
0119             break;
0120         }
0121     }
0122 }
0123 EXPORT_SYMBOL(glob_match);