match.go

     1  // Copyright 2010 The Go Authors. All rights reserved.
     2  // Use of this source code is governed by a BSD-style
     3  // license that can be found in the LICENSE file.
     4  
     5  package path
     6  
     7  import (
     8  	"errors"
     9  	"internal/bytealg"
    10  	"unicode/utf8"
    11  )
    12  
    13  // ErrBadPattern indicates a pattern was malformed.
    14  var ErrBadPattern = errors.New("syntax error in pattern")
    15  
    16  // Match reports whether name matches the shell pattern.
    17  // The pattern syntax is:
    18  //
    19  //	pattern:
    20  //		{ term }
    21  //	term:
    22  //		'*'         matches any sequence of non-/ characters
    23  //		'?'         matches any single non-/ character
    24  //		'[' [ '^' ] { character-range } ']'
    25  //		            character class (must be non-empty)
    26  //		c           matches character c (c != '*', '?', '\\', '[')
    27  //		'\\' c      matches character c
    28  //
    29  //	character-range:
    30  //		c           matches character c (c != '\\', '-', ']')
    31  //		'\\' c      matches character c
    32  //		lo '-' hi   matches character c for lo <= c <= hi
    33  //
    34  // Match requires pattern to match all of name, not just a substring.
    35  // The only possible returned error is [ErrBadPattern], when pattern
    36  // is malformed.
    37  func Match(pattern, name string) (matched bool, err error) {
    38  Pattern:
    39  	for len(pattern) > 0 {
    40  		var star bool
    41  		var chunk string
    42  		star, chunk, pattern = scanChunk(pattern)
    43  		if star && chunk == "" {
    44  			// Trailing * matches rest of string unless it has a /.
    45  			return bytealg.IndexByteString(name, '/') < 0, nil
    46  		}
    47  		// Look for match at current position.
    48  		t, ok, err := matchChunk(chunk, name)
    49  		// if we're the last chunk, make sure we've exhausted the name
    50  		// otherwise we'll give a false result even if we could still match
    51  		// using the star
    52  		if ok && (len(t) == 0 || len(pattern) > 0) {
    53  			name = t
    54  			continue
    55  		}
    56  		if err != nil {
    57  			return false, err
    58  		}
    59  		if star {
    60  			// Look for match skipping i+1 bytes.
    61  			// Cannot skip /.
    62  			for i := 0; i < len(name) && name[i] != '/'; i++ {
    63  				t, ok, err := matchChunk(chunk, name[i+1:])
    64  				if ok {
    65  					// if we're the last chunk, make sure we exhausted the name
    66  					if len(pattern) == 0 && len(t) > 0 {
    67  						continue
    68  					}
    69  					name = t
    70  					continue Pattern
    71  				}
    72  				if err != nil {
    73  					return false, err
    74  				}
    75  			}
    76  		}
    77  		// Before returning false with no error,
    78  		// check that the remainder of the pattern is syntactically valid.
    79  		for len(pattern) > 0 {
    80  			_, chunk, pattern = scanChunk(pattern)
    81  			if _, _, err := matchChunk(chunk, ""); err != nil {
    82  				return false, err
    83  			}
    84  		}
    85  		return false, nil
    86  	}
    87  	return len(name) == 0, nil
    88  }
    89  
    90  // scanChunk gets the next segment of pattern, which is a non-star string
    91  // possibly preceded by a star.
    92  func scanChunk(pattern string) (star bool, chunk, rest string) {
    93  	for len(pattern) > 0 && pattern[0] == '*' {
    94  		pattern = pattern[1:]
    95  		star = true
    96  	}
    97  	inrange := false
    98  	for i := 0; i < len(pattern); i++ {
    99  		switch pattern[i] {
   100  		case '\\':
   101  			// error check handled in matchChunk: bad pattern.
   102  			if i+1 < len(pattern) {
   103  				i++
   104  			}
   105  		case '[':
   106  			inrange = true
   107  		case ']':
   108  			inrange = false
   109  		case '*':
   110  			if !inrange {
   111  				return star, pattern[:i], pattern[i:]
   112  			}
   113  		}
   114  	}
   115  	return star, pattern, ""
   116  }
   117  
   118  // matchChunk checks whether chunk matches the beginning of s.
   119  // If so, it returns the remainder of s (after the match).
   120  // Chunk is all single-character operators: literals, char classes, and ?.
   121  func matchChunk(chunk, s string) (rest string, ok bool, err error) {
   122  	// failed records whether the match has failed.
   123  	// After the match fails, the loop continues on processing chunk,
   124  	// checking that the pattern is well-formed but no longer reading s.
   125  	failed := false
   126  	for len(chunk) > 0 {
   127  		failed = failed || len(s) == 0
   128  		switch chunk[0] {
   129  		case '[':
   130  			// character class
   131  			var r rune
   132  			if !failed {
   133  				var n int
   134  				r, n = utf8.DecodeRuneInString(s)
   135  				s = s[n:]
   136  			}
   137  			chunk = chunk[1:]
   138  			// possibly negated
   139  			negated := false
   140  			if len(chunk) > 0 && chunk[0] == '^' {
   141  				negated = true
   142  				chunk = chunk[1:]
   143  			}
   144  			// parse all ranges
   145  			match := false
   146  			nrange := 0
   147  			for {
   148  				if len(chunk) > 0 && chunk[0] == ']' && nrange > 0 {
   149  					chunk = chunk[1:]
   150  					break
   151  				}
   152  				var lo, hi rune
   153  				if lo, chunk, err = getEsc(chunk); err != nil {
   154  					return "", false, err
   155  				}
   156  				hi = lo
   157  				if chunk[0] == '-' {
   158  					if hi, chunk, err = getEsc(chunk[1:]); err != nil {
   159  						return "", false, err
   160  					}
   161  				}
   162  				match = match || lo <= r && r <= hi
   163  				nrange++
   164  			}
   165  			failed = failed || match == negated
   166  
   167  		case '?':
   168  			if !failed {
   169  				failed = s[0] == '/'
   170  				_, n := utf8.DecodeRuneInString(s)
   171  				s = s[n:]
   172  			}
   173  			chunk = chunk[1:]
   174  
   175  		case '\\':
   176  			chunk = chunk[1:]
   177  			if len(chunk) == 0 {
   178  				return "", false, ErrBadPattern
   179  			}
   180  			fallthrough
   181  
   182  		default:
   183  			if !failed {
   184  				failed = chunk[0] != s[0]
   185  				s = s[1:]
   186  			}
   187  			chunk = chunk[1:]
   188  		}
   189  	}
   190  	if failed {
   191  		return "", false, nil
   192  	}
   193  	return s, true, nil
   194  }
   195  
   196  // getEsc gets a possibly-escaped character from chunk, for a character class.
   197  func getEsc(chunk string) (r rune, nchunk string, err error) {
   198  	if len(chunk) == 0 || chunk[0] == '-' || chunk[0] == ']' {
   199  		err = ErrBadPattern
   200  		return
   201  	}
   202  	if chunk[0] == '\\' {
   203  		chunk = chunk[1:]
   204  		if len(chunk) == 0 {
   205  			err = ErrBadPattern
   206  			return
   207  		}
   208  	}
   209  	r, n := utf8.DecodeRuneInString(chunk)
   210  	if r == utf8.RuneError && n == 1 {
   211  		err = ErrBadPattern
   212  	}
   213  	nchunk = chunk[n:]
   214  	if len(nchunk) == 0 {
   215  		err = ErrBadPattern
   216  	}
   217  	return
   218  }
   219
View as plain text