1 /******************************************************************************
2 * Copyright (C) 2008 by Niko Valimaki <nvalimak@cs.helsinki.fi> *
3 * Text collection interface for an in-memory XQuery/XPath engine *
5 * This program is free software; you can redistribute it and/or modify *
6 * it under the terms of the GNU Lesser General Public License as published *
7 * by the Free Software Foundation; either version 2 of the License, or *
8 * (at your option) any later version. *
10 * This program is distributed in the hope that it will be useful, *
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of *
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
13 * GNU Lesser General Public License for more details. *
15 * You should have received a copy of the GNU Lesser General Public License *
16 * along with this program; if not, write to the *
17 * Free Software Foundation, Inc., *
18 * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. *
19 ******************************************************************************/
21 #ifndef _SXSI_TextCollection_h_
22 #define _SXSI_TextCollection_h_
24 #include "Tools.h" // Defines ulong and uchar.
26 #include <utility> // Defines std::pair.
32 * General interface for a text collection
34 * Class is virtual, make objects by calling
35 * the static method InitTextCollection().
40 // Type of document identifier
42 // Type for text position (FIXME ulong or long?)
43 typedef ulong TextPosition;
48 * New samplerate can be given, otherwise will use the one specified in the save file!
49 * Note: This is not a static method; call InitTextCollection() first to get the object handle.
51 * Throws an exception if std::fread() fails.
54 static TextCollection* Load(FILE *, unsigned samplerate = 0);
57 * Save data structure into a file
59 * Throws an exception if std::fwrite() fails.
61 virtual void Save(FILE *) const = 0;
66 virtual ~TextCollection() { };
69 * Tests if the string pointed to by DocId is empty
71 virtual bool EmptyText(DocId) const = 0;
76 * Returns the i'th text in the collection.
77 * The numbering starts from 0.
79 virtual uchar* GetText(DocId) const = 0;
81 * Returns substring [i, j] of k'th text
83 * Note: Parameters i and j are text positions inside the k'th text.
85 // virtual uchar* GetText(DocId, TextPosition, TextPosition) const = 0;
87 * Returns backwards (reverse) iterator to the end of i'th text
89 * Note: Do we need this?
90 * Forward iterator would be really in-efficient compared to
93 * TODO Define and implement const_reverse_iterator.
95 //const_reverse_iterator rend(DocId) const;
100 // Is there a text prefixed by given string?
101 virtual bool IsPrefix(uchar const *) const = 0;
102 // Is there a text having given string as a suffix?
103 virtual bool IsSuffix(uchar const *) const = 0;
104 // Is there a text that equals given string?
105 virtual bool IsEqual(uchar const *) const = 0;
106 // Does a text contain given string?
107 virtual bool IsContains(uchar const *) const = 0;
108 // Is there a text that is lexicographically less than given string?
109 virtual bool IsLessThan(uchar const *) const = 0;
112 * Existential queries for given DocId interval.
114 virtual bool IsPrefix(uchar const *, DocId, DocId) const = 0;
115 virtual bool IsSuffix(uchar const *, DocId, DocId) const = 0;
116 virtual bool IsEqual(uchar const *, DocId, DocId) const = 0;
117 virtual bool IsContains(uchar const *, DocId, DocId) const = 0;
118 virtual bool IsLessThan(uchar const *, DocId, DocId) const = 0;
122 * Result is the number of occurrences.
124 virtual ulong Count(uchar const *) const = 0;
126 * More counting queries
127 * Result is the number of documents.
129 virtual unsigned CountPrefix(uchar const *) const = 0;
130 virtual unsigned CountSuffix(uchar const *) const = 0;
131 virtual unsigned CountEqual(uchar const *) const = 0;
132 virtual unsigned CountContains(uchar const *) const = 0;
133 virtual unsigned CountLessThan(uchar const *) const = 0;
136 * Counting queries for given DocId interval
138 virtual unsigned CountPrefix(uchar const *, DocId, DocId) const = 0;
139 virtual unsigned CountSuffix(uchar const *, DocId, DocId) const = 0;
140 virtual unsigned CountEqual(uchar const *, DocId, DocId) const = 0;
141 virtual unsigned CountContains(uchar const *, DocId, DocId) const = 0;
142 virtual unsigned CountLessThan(uchar const *, DocId, DocId) const = 0;
145 * Document reporting queries
147 * Result is a vector of document id's in some undefined order.
149 // Data type for results
150 typedef std::vector<DocId> document_result;
151 virtual document_result Prefix(uchar const *) const = 0;
152 virtual document_result Suffix(uchar const *) const = 0;
153 virtual document_result Equal(uchar const *) const = 0;
154 virtual document_result Contains(uchar const *) const = 0;
155 virtual document_result LessThan(uchar const *) const = 0;
158 * Document reporting queries for given DocId interval.
160 virtual document_result Prefix(uchar const *, DocId, DocId) const = 0;
161 virtual document_result Suffix(uchar const *, DocId, DocId) const = 0;
162 virtual document_result Equal(uchar const *, DocId, DocId) const = 0;
163 virtual document_result Contains(uchar const *, DocId, DocId) const = 0;
164 virtual document_result LessThan(uchar const *, DocId, DocId) const = 0;
167 * Full reporting queries
169 * Result is a vector of pairs <doc id, offset> in some undefined order.
171 // Data type for results
172 typedef std::vector<std::pair<DocId, TextPosition> > full_result;
173 virtual full_result FullContains(uchar const *) const = 0;
174 // Full reporting query for given DocId interval
175 virtual full_result FullContains(uchar const *, DocId, DocId) const = 0;
178 // Protected constructor; use TextCollectionBuilder
179 TextCollection() { };
181 // No copy constructor or assignment
182 TextCollection(TextCollection const&);
183 TextCollection& operator = (TextCollection const&);