0
|
1 /* Compare strings while treating digits characters numerically.
|
|
2 Copyright (C) 1997, 2002, 2005 Free Software Foundation, Inc.
|
|
3 This file is part of the libiberty library.
|
|
4 Contributed by Jean-François Bignolles <bignolle@ecoledoc.ibp.fr>, 1997.
|
|
5
|
|
6 Libiberty is free software; you can redistribute it and/or
|
|
7 modify it under the terms of the GNU Lesser General Public
|
|
8 License as published by the Free Software Foundation; either
|
|
9 version 2.1 of the License, or (at your option) any later version.
|
|
10
|
|
11 Libiberty is distributed in the hope that it will be useful,
|
|
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
14 Lesser General Public License for more details.
|
|
15
|
|
16 You should have received a copy of the GNU Lesser General Public
|
|
17 License along with the GNU C Library; if not, write to the Free
|
|
18 Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
|
|
19 02110-1301 USA. */
|
|
20
|
|
21 #include "libiberty.h"
|
|
22 #include "safe-ctype.h"
|
|
23
|
|
24 /*
|
|
25 @deftypefun int strverscmp (const char *@var{s1}, const char *@var{s2})
|
|
26 The @code{strverscmp} function compares the string @var{s1} against
|
|
27 @var{s2}, considering them as holding indices/version numbers. Return
|
|
28 value follows the same conventions as found in the @code{strverscmp}
|
|
29 function. In fact, if @var{s1} and @var{s2} contain no digits,
|
|
30 @code{strverscmp} behaves like @code{strcmp}.
|
|
31
|
|
32 Basically, we compare strings normally (character by character), until
|
|
33 we find a digit in each string - then we enter a special comparison
|
|
34 mode, where each sequence of digits is taken as a whole. If we reach the
|
|
35 end of these two parts without noticing a difference, we return to the
|
|
36 standard comparison mode. There are two types of numeric parts:
|
|
37 "integral" and "fractional" (those begin with a '0'). The types
|
|
38 of the numeric parts affect the way we sort them:
|
|
39
|
|
40 @itemize @bullet
|
|
41 @item
|
|
42 integral/integral: we compare values as you would expect.
|
|
43
|
|
44 @item
|
|
45 fractional/integral: the fractional part is less than the integral one.
|
|
46 Again, no surprise.
|
|
47
|
|
48 @item
|
|
49 fractional/fractional: the things become a bit more complex.
|
|
50 If the common prefix contains only leading zeroes, the longest part is less
|
|
51 than the other one; else the comparison behaves normally.
|
|
52 @end itemize
|
|
53
|
|
54 @smallexample
|
|
55 strverscmp ("no digit", "no digit")
|
|
56 @result{} 0 // @r{same behavior as strcmp.}
|
|
57 strverscmp ("item#99", "item#100")
|
|
58 @result{} <0 // @r{same prefix, but 99 < 100.}
|
|
59 strverscmp ("alpha1", "alpha001")
|
|
60 @result{} >0 // @r{fractional part inferior to integral one.}
|
|
61 strverscmp ("part1_f012", "part1_f01")
|
|
62 @result{} >0 // @r{two fractional parts.}
|
|
63 strverscmp ("foo.009", "foo.0")
|
|
64 @result{} <0 // @r{idem, but with leading zeroes only.}
|
|
65 @end smallexample
|
|
66
|
|
67 This function is especially useful when dealing with filename sorting,
|
|
68 because filenames frequently hold indices/version numbers.
|
|
69 @end deftypefun
|
|
70
|
|
71 */
|
|
72
|
|
73 /* states: S_N: normal, S_I: comparing integral part, S_F: comparing
|
|
74 fractional parts, S_Z: idem but with leading Zeroes only */
|
|
75 #define S_N 0x0
|
|
76 #define S_I 0x4
|
|
77 #define S_F 0x8
|
|
78 #define S_Z 0xC
|
|
79
|
|
80 /* result_type: CMP: return diff; LEN: compare using len_diff/diff */
|
|
81 #define CMP 2
|
|
82 #define LEN 3
|
|
83
|
|
84
|
|
85 /* Compare S1 and S2 as strings holding indices/version numbers,
|
|
86 returning less than, equal to or greater than zero if S1 is less than,
|
|
87 equal to or greater than S2 (for more info, see the Glibc texinfo doc). */
|
|
88
|
|
89 int
|
|
90 strverscmp (const char *s1, const char *s2)
|
|
91 {
|
|
92 const unsigned char *p1 = (const unsigned char *) s1;
|
|
93 const unsigned char *p2 = (const unsigned char *) s2;
|
|
94 unsigned char c1, c2;
|
|
95 int state;
|
|
96 int diff;
|
|
97
|
|
98 /* Symbol(s) 0 [1-9] others (padding)
|
|
99 Transition (10) 0 (01) d (00) x (11) - */
|
|
100 static const unsigned int next_state[] =
|
|
101 {
|
|
102 /* state x d 0 - */
|
|
103 /* S_N */ S_N, S_I, S_Z, S_N,
|
|
104 /* S_I */ S_N, S_I, S_I, S_I,
|
|
105 /* S_F */ S_N, S_F, S_F, S_F,
|
|
106 /* S_Z */ S_N, S_F, S_Z, S_Z
|
|
107 };
|
|
108
|
|
109 static const int result_type[] =
|
|
110 {
|
|
111 /* state x/x x/d x/0 x/- d/x d/d d/0 d/-
|
|
112 0/x 0/d 0/0 0/- -/x -/d -/0 -/- */
|
|
113
|
|
114 /* S_N */ CMP, CMP, CMP, CMP, CMP, LEN, CMP, CMP,
|
|
115 CMP, CMP, CMP, CMP, CMP, CMP, CMP, CMP,
|
|
116 /* S_I */ CMP, -1, -1, CMP, +1, LEN, LEN, CMP,
|
|
117 +1, LEN, LEN, CMP, CMP, CMP, CMP, CMP,
|
|
118 /* S_F */ CMP, CMP, CMP, CMP, CMP, LEN, CMP, CMP,
|
|
119 CMP, CMP, CMP, CMP, CMP, CMP, CMP, CMP,
|
|
120 /* S_Z */ CMP, +1, +1, CMP, -1, CMP, CMP, CMP,
|
|
121 -1, CMP, CMP, CMP
|
|
122 };
|
|
123
|
|
124 if (p1 == p2)
|
|
125 return 0;
|
|
126
|
|
127 c1 = *p1++;
|
|
128 c2 = *p2++;
|
|
129 /* Hint: '0' is a digit too. */
|
|
130 state = S_N | ((c1 == '0') + (ISDIGIT (c1) != 0));
|
|
131
|
|
132 while ((diff = c1 - c2) == 0 && c1 != '\0')
|
|
133 {
|
|
134 state = next_state[state];
|
|
135 c1 = *p1++;
|
|
136 c2 = *p2++;
|
|
137 state |= (c1 == '0') + (ISDIGIT (c1) != 0);
|
|
138 }
|
|
139
|
|
140 state = result_type[state << 2 | (((c2 == '0') + (ISDIGIT (c2) != 0)))];
|
|
141
|
|
142 switch (state)
|
|
143 {
|
|
144 case CMP:
|
|
145 return diff;
|
|
146
|
|
147 case LEN:
|
|
148 while (ISDIGIT (*p1++))
|
|
149 if (!ISDIGIT (*p2++))
|
|
150 return 1;
|
|
151
|
|
152 return ISDIGIT (*p2) ? -1 : diff;
|
|
153
|
|
154 default:
|
|
155 return state;
|
|
156 }
|
|
157 }
|