1 | /**********************************************************************
|
---|
2 | *
|
---|
3 | * phrase.h -- definition of the phrase object used by suffix.cpp
|
---|
4 | *
|
---|
5 | * Copyright 2000 Gordon W. Paynter
|
---|
6 | * Copyright 2000 The New Zealand Digital Library Project
|
---|
7 | *
|
---|
8 | * A component of the Greenstone digital library software
|
---|
9 | * from the New Zealand Digital Library Project at the
|
---|
10 | * University of Waikato, New Zealand.
|
---|
11 | *
|
---|
12 | * This program is free software; you can redistribute it and/or modify
|
---|
13 | * it under the terms of the GNU General Public License as published by
|
---|
14 | * the Free Software Foundation; either version 2 of the License, or
|
---|
15 | * (at your option) any later version.
|
---|
16 | *
|
---|
17 | * This program is distributed in the hope that it will be useful,
|
---|
18 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
19 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
20 | * GNU General Public License for more details.
|
---|
21 | *
|
---|
22 | * You should have received a copy of the GNU General Public License
|
---|
23 | * along with this program; if not, write to the Free Software
|
---|
24 | * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
25 | *
|
---|
26 | *********************************************************************/
|
---|
27 |
|
---|
28 | #ifndef PHRASE_H
|
---|
29 | #define PHRASE_H
|
---|
30 |
|
---|
31 | // use the standard namespace
|
---|
32 | #if !defined (GSDL_NAMESPACE_BROKEN)
|
---|
33 | #if defined(GSDL_USE_OBJECTSPACE)
|
---|
34 | using namespace ospace::std;
|
---|
35 | #else
|
---|
36 | using namespace std;
|
---|
37 | #endif
|
---|
38 | #endif
|
---|
39 |
|
---|
40 | #if defined(GSDL_USE_STL_H)
|
---|
41 | # include <vector.h>
|
---|
42 | #else
|
---|
43 | # include <vector>
|
---|
44 | #endif
|
---|
45 |
|
---|
46 | #include "suffix.h"
|
---|
47 |
|
---|
48 | class Phrase {
|
---|
49 |
|
---|
50 | public:
|
---|
51 |
|
---|
52 | // The phrase itself is stored with two pointers: forward
|
---|
53 | // points to its first cell, back points to its last.
|
---|
54 | // The length is always stored in length.
|
---|
55 | // If one of these is set, all must be set, and it must
|
---|
56 | // be true that (forward + length - 1) = back
|
---|
57 | symbol *forward;
|
---|
58 | symbol *back;
|
---|
59 | cellcount length;
|
---|
60 |
|
---|
61 | // Phrase location in the suffix array
|
---|
62 | int suffixFound;
|
---|
63 | symbol *firstSuffix;
|
---|
64 | symbol *lastSuffix;
|
---|
65 | cellindex firstSuffixIndex;
|
---|
66 | cellindex lastSuffixIndex;
|
---|
67 | cellcount suffixFrequency;
|
---|
68 |
|
---|
69 | // Phrase location in the prefix array
|
---|
70 | int prefixFound;
|
---|
71 | symbol* firstPrefix;
|
---|
72 | symbol* lastPrefix;
|
---|
73 | cellindex firstPrefixIndex;
|
---|
74 | cellindex lastPrefixIndex;
|
---|
75 | cellcount prefixFrequency;
|
---|
76 |
|
---|
77 | // Constructor function
|
---|
78 | // First argument is an array of words, second is the length of
|
---|
79 | // the phrase, third is the direction (SUFFIX of PREFIX) in
|
---|
80 | // which the words should be read (defaults to forwards).
|
---|
81 | Phrase(symbol *words, cellcount size, int direction);
|
---|
82 |
|
---|
83 | // An empty phrase can be created without arguments, but is
|
---|
84 | // good for nothing and may not be used with any public fuctions.
|
---|
85 | // We therefore only use it internally.
|
---|
86 | Phrase();
|
---|
87 |
|
---|
88 | // Represent the phrase as a string
|
---|
89 | char *toString();
|
---|
90 |
|
---|
91 | // Find an initial set of candidate phrases in the suffix/prefix array
|
---|
92 | int initialSuffixCandidates(vector<Phrase> &results);
|
---|
93 | int initialPrefixCandidates(vector<Phrase> &results);
|
---|
94 |
|
---|
95 | // Does the phrase have a unique extension?
|
---|
96 | int hasUniqueSuffixExtension();
|
---|
97 | int hasUniquePrefixExtension();
|
---|
98 |
|
---|
99 | // Extend a phrase by exactly one symbol
|
---|
100 | int expandUniquePrefixExtensionByOne();
|
---|
101 | int expandUniqueSuffixExtensionByOne();
|
---|
102 |
|
---|
103 | // Find the phrase in the suffix/prefix array
|
---|
104 | int findFirstAndLastSuffix();
|
---|
105 | int findFirstAndLastPrefix();
|
---|
106 | int findFirstAndLastSuffix(cellindex begin, cellindex end);
|
---|
107 | int findFirstAndLastPrefix(cellindex begin, cellindex end);
|
---|
108 |
|
---|
109 | // Make sure the phrase location in the suffix/prefix array is known
|
---|
110 | int ensureSuffixFound();
|
---|
111 | int ensurePrefixFound();
|
---|
112 |
|
---|
113 | private:
|
---|
114 |
|
---|
115 | // Does the phrase have a unique suffix/prefix extension?
|
---|
116 | // if yes, then 1; if no then 0; if unknown then -1;
|
---|
117 | int uniqueSuffixExtension;
|
---|
118 | int uniquePrefixExtension;
|
---|
119 |
|
---|
120 | // reset a phrase
|
---|
121 | int empty();
|
---|
122 |
|
---|
123 | // reset phrase information relating to location in suffix/prefix array
|
---|
124 | int clearSuffix();
|
---|
125 | int clearPrefix();
|
---|
126 |
|
---|
127 | // increase the length of a phrase "in place"
|
---|
128 | int Phrase::increaseSuffixLength(cellcount l);
|
---|
129 | int Phrase::increasePrefixLength(cellcount l);
|
---|
130 |
|
---|
131 | // Compare the phrase to a given array of symbols
|
---|
132 | int compareSuffix(symbol *words, cellcount length);
|
---|
133 | int comparePrefix(symbol *words, cellcount length);
|
---|
134 |
|
---|
135 | // Create a new phrase that is longer than this one, yet as short as possible.
|
---|
136 | Phrase Phrase::newPhraseShortestSuffixExpansion(cellindex i);
|
---|
137 | Phrase Phrase::newPhraseShortestPrefixExpansion(cellindex i);
|
---|
138 |
|
---|
139 | // Extend a phrase until it no longer has a unique extanesion
|
---|
140 | int expandWhileUniqueSuffixExtension();
|
---|
141 | int expandWhileUniquePrefixExtension();
|
---|
142 | };
|
---|
143 |
|
---|
144 |
|
---|
145 | bool isShorter(Phrase p1, Phrase p2);
|
---|
146 | bool isLonger(Phrase p1, Phrase p2);
|
---|
147 |
|
---|
148 | #endif
|
---|