2
* Copyright (c) 2009 Kov Chai <tchaikov@gmail.com>
4
* The contents of this file are subject to the terms of either the GNU Lesser
5
* General Public License Version 2.1 only ("LGPL") or the Common Development and
6
* Distribution License ("CDDL")(collectively, the "License"). You may not use this
7
* file except in compliance with the License. You can obtain a copy of the CDDL at
8
* http://www.opensource.org/licenses/cddl1.php and a copy of the LGPLv2.1 at
9
* http://www.opensource.org/licenses/lgpl-license.php. See the License for the
10
* specific language governing permissions and limitations under the License. When
11
* distributing the software, include this License Header Notice in each file and
12
* include the full text of the License in the License file as well as the
15
* NOTICE PURSUANT TO SECTION 9 OF THE COMMON DEVELOPMENT AND DISTRIBUTION LICENSE
17
* For Covered Software in this distribution, this License shall be governed by the
18
* laws of the State of California (excluding conflict-of-law provisions).
19
* Any litigation relating to this License shall be subject to the jurisdiction of
20
* the Federal Courts of the Northern District of California and the state courts
21
* of the State of California, with venue lying in Santa Clara County, California.
25
* If you wish your version of this file to be governed by only the CDDL or only
26
* the LGPL Version 2.1, indicate your decision by adding "[Contributor]" elects to
27
* include this software in this distribution under the [CDDL or LGPL Version 2.1]
28
* license." If you don't indicate a single choice of license, a recipient has the
29
* option to distribute your version of this file under either the CDDL or the LGPL
30
* Version 2.1, or to extend the choice of license to its licensees as provided
31
* above. However, if you add LGPL Version 2.1 code and therefore, elected the LGPL
32
* Version 2 license, then the option applies only if the new code is made subject
33
* to such option by the copyright holder.
46
/* the ARPA style representation of sunpinyin's SLM */
50
TSIMWordId hw[N_GRAM];
56
void load(istream&, const TLexicon&);
57
int load_words(char* buf, const TLexicon& lexicon);
58
TLeaf() : wid(0), pr(.0), ch(0), bon(0), bol(0) {}
61
struct TNode : public TLeaf {
63
void load(istream&, const TLexicon&);
64
void load_level0(istream&);
67
typedef std::vector<TNode> TNodeLevel;
68
typedef std::vector<TLeaf> TLeafLevel;
71
TNodeLevel m_levels[N_GRAM+1]; /* [0..N_GRAM] */
72
TLeafLevel m_lastLevel;
73
const bool m_usingLogPr;
77
/* XXX, ARPA file does not provide these information.
78
so we assume this SLM is trigram, and does not use LogPr */
79
CArpaSlm() : m_usingLogPr(false), m_N(N_GRAM) {}
80
bool good() const { return m_levels[0].size() != 0; }
81
unsigned getN() const { return m_N; }
82
bool usingLogPr() const { return m_usingLogPr; }
83
const TNodeLevel& getLevel(unsigned lvl) const { return m_levels[lvl]; }
84
const TLeafLevel& getLastLevel() const { return m_lastLevel; }
85
unsigned getLevelSize(unsigned lvl) const {
88
return m_levels[lvl].size();
90
return m_lastLevel.size();
94
* initialize the `ch' and `wid' fields of each node in levels
97
void load(const char* filename, const TLexicon& lexicon);
101
* find out the first child of a given node in its next level
102
* @param lvl the level where node belongs to
103
* @param node the node
104
* @param last_child the child index of previous node
105
* @return the index of the found child
107
unsigned find_1st_child(unsigned lvl, const TNode& node, int last_child);