7017
|
1 ## Copyright (C) 1996, 1999, 2000, 2002, 2003, 2005, 2006, 2007 |
|
2 ## Kurt Hornik |
2325
|
3 ## |
2313
|
4 ## This file is part of Octave. |
|
5 ## |
|
6 ## Octave is free software; you can redistribute it and/or modify it |
|
7 ## under the terms of the GNU General Public License as published by |
7016
|
8 ## the Free Software Foundation; either version 3 of the License, or (at |
|
9 ## your option) any later version. |
2313
|
10 ## |
|
11 ## Octave is distributed in the hope that it will be useful, but |
|
12 ## WITHOUT ANY WARRANTY; without even the implied warranty of |
|
13 ## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
|
14 ## General Public License for more details. |
|
15 ## |
|
16 ## You should have received a copy of the GNU General Public License |
7016
|
17 ## along with Octave; see the file COPYING. If not, see |
|
18 ## <http://www.gnu.org/licenses/>. |
2272
|
19 |
3361
|
20 ## -*- texinfo -*- |
|
21 ## @deftypefn {Function File} {} findstr (@var{s}, @var{t}, @var{overlap}) |
|
22 ## Return the vector of all positions in the longer of the two strings |
|
23 ## @var{s} and @var{t} where an occurrence of the shorter of the two starts. |
|
24 ## If the optional argument @var{overlap} is nonzero, the returned vector |
|
25 ## can include overlapping positions (this is the default). For example, |
3426
|
26 ## |
3361
|
27 ## @example |
|
28 ## findstr ("ababab", "a") |
|
29 ## @result{} [ 1, 3, 5 ] |
|
30 ## findstr ("abababa", "aba", 0) |
|
31 ## @result{} [ 1, 5 ] |
|
32 ## @end example |
|
33 ## @end deftypefn |
2272
|
34 |
3891
|
35 ## Note that this implementation swaps the strings if second one is longer |
|
36 ## than the first, so try to put the longer one first. |
|
37 ## |
5428
|
38 ## Author: Kurt Hornik <Kurt.Hornik@wu-wien.ac.at> |
2355
|
39 ## Adapted-By: jwe |
2314
|
40 |
2311
|
41 function v = findstr (s, t, overlap) |
2275
|
42 |
|
43 if (nargin < 2 || nargin > 3) |
6046
|
44 print_usage (); |
2275
|
45 endif |
|
46 |
5348
|
47 if (all (size (s) > 1) || all (size (t) > 1)) |
|
48 error ("findstr: arguments must have only one non-singleton dimension"); |
3891
|
49 endif |
|
50 |
2275
|
51 if (nargin == 2) |
|
52 overlap = 1; |
2272
|
53 endif |
|
54 |
3891
|
55 ## Make S be the longer string. |
|
56 if (length (s) < length (t)) |
|
57 tmp = s; |
|
58 s = t; |
|
59 t = tmp; |
|
60 endif |
|
61 |
|
62 l_s = length (s); |
|
63 l_t = length (t); |
|
64 |
|
65 if (l_t == 0) |
4321
|
66 ## zero length target: return empty set |
|
67 v = []; |
3891
|
68 |
|
69 elseif (l_t == 1) |
|
70 ## length one target: simple find |
|
71 v = find (s == t); |
|
72 |
|
73 elseif (l_t == 2) |
|
74 ## length two target: find first at i and second at i+1 |
|
75 v = find (s(1:l_s-1) == t(1) & s(2:l_s) == t(2)); |
|
76 |
|
77 else |
|
78 ## length three or more: match the first three by find then go through |
|
79 ## the much smaller list to determine which of them are real matches |
|
80 limit = l_s - l_t + 1; |
|
81 v = find (s(1:limit) == t(1) |
|
82 & s(2:limit+1) == t(2) |
|
83 & s (3:limit+2) == t(3)); |
|
84 endif |
3759
|
85 |
3891
|
86 ## Need to search the index vector if our find was too short |
|
87 ## (target length > 3), or if we don't allow overlaps. Note though |
|
88 ## that there cannot be any overlaps if the first character in the |
|
89 ## target is different from the remaining characters in the target, |
|
90 ## so a single character, two different characters, or first character |
|
91 ## different from the second two don't need to be searched. |
|
92 if (l_t >= 3 || (! overlap && l_t > 1 && any (t(1) == t(2:l_t)))) |
|
93 ## force strings to be both row vectors or both column vectors |
|
94 if (all (size (s) != size (t))) |
|
95 t = t.'; |
|
96 endif |
|
97 |
|
98 ## determine which ones to keep |
|
99 keep = zeros (size (v)); |
|
100 ind = 0:l_t-1; |
|
101 if (overlap) |
|
102 for idx = 1:length (v) |
|
103 keep(idx) = all (s(v(idx) + ind) == t); |
|
104 endfor |
2272
|
105 else |
3891
|
106 next = 1; # first possible position for next non-overlapping match |
|
107 for idx = 1:length (v) |
|
108 if (v(idx) >= next && s(v(idx) + ind) == t) |
|
109 keep(idx) = 1; |
|
110 next = v(idx) + l_t; # skip to the next possible match position |
|
111 else |
|
112 keep(idx) = 0; |
|
113 endif |
|
114 endfor |
2272
|
115 endif |
3891
|
116 if (! isempty (v)) |
|
117 v = v(find (keep)); |
|
118 endif |
|
119 endif |
5400
|
120 |
|
121 if (isempty (v)) |
|
122 v = []; |
|
123 endif |
|
124 |
3891
|
125 ## Always return a column vector, because that's what the old one did |
|
126 if (rows (v) > 1) |
|
127 v = v.'; |
2272
|
128 endif |
|
129 |
|
130 endfunction |
7411
|
131 |
|
132 %!assert((findstr ("abababa", "a") == [1, 3, 5, 7] |
|
133 %! && findstr ("abababa", "aba") == [1, 3, 5] |
|
134 %! && findstr ("abababa", "aba", 0) == [1, 5])); |
|
135 |
|
136 %!error findstr (); |
|
137 |
|
138 %!error findstr ("foo", "bar", 3, 4); |
|
139 |