xapian-core/include/xapian/stem.h

   1 /** @file
   2  * @brief stemming algorithms
   3  */
   4 /* Copyright (C) 2005,2007,2010,2011,2013,2014,2015,2018,2019 Olly Betts
   5  * Copyright (C) 2010 Evgeny Sizikov
   6  *
   7  * This program is free software; you can redistribute it and/or
   8  * modify it under the terms of the GNU General Public License as
   9  * published by the Free Software Foundation; either version 2 of the
  10  * License, or (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License
  18  * along with this program; if not, write to the Free Software
  19  * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA
  20  */
  21
  22 #ifndef XAPIAN_INCLUDED_STEM_H
  23 #define XAPIAN_INCLUDED_STEM_H
  24
  25 #if !defined XAPIAN_IN_XAPIAN_H && !defined XAPIAN_LIB_BUILD
  26 # error Never use <xapian/stem.h> directly; include <xapian.h> instead.
  27 #endif
  28
  29 #include <xapian/constinfo.h>
  30 #include <xapian/intrusive_ptr.h>
  31 #include <xapian/visibility.h>
  32
  33 #include <string>
  34
  35 namespace Xapian {
  36
  37 /// Class representing a stemming algorithm implementation.
  38 class XAPIAN_VISIBILITY_DEFAULT StemImplementation
  39     : public Xapian::Internal::intrusive_base
  40 {
  41     /// Don't allow assignment.
  42     void operator=(const StemImplementation &);
  43
  44     /// Don't allow copying.
  45     StemImplementation(const StemImplementation &);
  46
  47   public:
  48     /// Default constructor.
  49     StemImplementation() { }
  50
  51     /// Virtual destructor.
  52     virtual ~StemImplementation();
  53
  54     /// Stem the specified word.
  55     virtual std::string operator()(const std::string & word) = 0;
  56
  57     /// Return a string describing this object.
  58     virtual std::string get_description() const = 0;
  59 };
  60
  61 /// Class representing a stemming algorithm.
  62 class XAPIAN_VISIBILITY_DEFAULT Stem {
  63   public:
  64     /// @private @internal Reference counted internals.
  65     Xapian::Internal::intrusive_ptr<StemImplementation> internal;
  66
  67     /// Copy constructor.
  68     Stem(const Stem & o);
  69
  70     /// Assignment.
  71     Stem & operator=(const Stem & o);
  72
  73 #ifdef XAPIAN_MOVE_SEMANTICS
  74     /// Move constructor.
  75     Stem(Stem && o);
  76
  77     /// Move assignment operator.
  78     Stem & operator=(Stem && o);
  79 #endif
  80
  81     /** Construct a Xapian::Stem object which doesn't change terms.
  82      *
  83      *  Equivalent to Stem("none").
  84      */
  85     Stem();
  86
  87     /** Construct a Xapian::Stem object for a particular language.
  88      *
  89      *  @param language Either the English name for the language
  90      *                  or the two letter ISO639 code.
  91      *
  92      *  The following language names are understood (aliases follow the
  93      *  name):
  94      *
  95      *  - none - don't stem terms
  96      *  - arabic (ar) - Since Xapian 1.3.5
  97      *  - armenian (hy) - Since Xapian 1.3.0
  98      *  - basque (eu) - Since Xapian 1.3.0
  99      *  - catalan (ca) - Since Xapian 1.3.0
 100      *  - danish (da)
 101      *  - dutch (nl)
 102      *  - english (en) - Martin Porter's 2002 revision of his stemmer
 103      *  - earlyenglish - Early English (e.g. Shakespeare, Dickens) stemmer
 104      *    (since Xapian 1.3.2)
 105      *  - english_lovins (lovins) - Lovin's stemmer
 106      *  - english_porter (porter) - Porter's stemmer as described in
 107      *                  his 1980 paper
 108      *  - finnish (fi)
 109      *  - french (fr)
 110      *  - german (de)
 111      *  - german2 - Normalises umlauts and &szlig;
 112      *  - hungarian (hu)
 113      *  - indonesian (id) - Since Xapian 1.4.6
 114      *  - irish (ga) - Since Xapian 1.4.7
 115      *  - italian (it)
 116      *  - kraaij_pohlmann - A different Dutch stemmer
 117      *  - lithuanian (lt) - Since Xapian 1.4.7
 118      *  - nepali (ne) - Since Xapian 1.4.7
 119      *  - norwegian (nb, nn, no)
 120      *  - portuguese (pt)
 121      *  - romanian (ro)
 122      *  - russian (ru)
 123      *  - spanish (es)
 124      *  - swedish (sv)
 125      *  - tamil (ta) - Since Xapian 1.4.7
 126      *  - turkish (tr)
 127      *
 128      *  @param fallback If true then treat unknown @a language as "none",
 129      *                  otherwise an exception is thrown (default: false).
 130      *                  Parameter added in Xapian 1.4.14 - older versions
 131      *                  always threw an exception.
 132      *
 133      *  @exception      Xapian::InvalidArgumentError is thrown if
 134      *                  @a language isn't recognised and @a fallback is false.
 135      *
 136      *  @{
 137      */
 138     explicit Stem(const std::string& language);
 139     Stem(const std::string& language, bool fallback);
 140     /** @} */
 141
 142     /** Construct a Xapian::Stem object with a user-provided stemming algorithm.
 143      *
 144      *  You can subclass Xapian::StemImplementation to implement your own
 145      *  stemming algorithm (or to wrap a third-party algorithm) and then wrap
 146      *  your implementation in a Xapian::Stem object to pass to the Xapian API.
 147      *
 148      *  @param p        The user-subclassed StemImplementation object.  This
 149      *                  is reference counted, and so will be automatically
 150      *                  deleted by the Xapian::Stem wrapper when no longer
 151      *                  required.
 152      */
 153     explicit Stem(StemImplementation * p);
 154
 155     /// Destructor.
 156     ~Stem();
 157
 158     /** Stem a word.
 159      *
 160      *  @param word             a word to stem.
 161      *  @return         the stem
 162      */
 163     std::string operator()(const std::string &word) const;
 164
 165     /// Return true if this is a no-op stemmer.
 166     bool is_none() const { return !internal.get(); }
 167
 168     /// Return a string describing this object.
 169     std::string get_description() const;
 170
 171     /** Return a list of available languages.
 172      *
 173      *  Each stemmer is only included once in the list (not once for
 174      *  each alias).  The name included is the English name of the
 175      *  language.
 176      *
 177      *  The list is returned as a string, with language names separated by
 178      *  spaces.  This is a static method, so a Xapian::Stem object is not
 179      *  required for this operation.
 180      */
 181     static std::string get_available_languages() {
 182         const struct Xapian::Internal::constinfo * info =
 183             Xapian::Internal::get_constinfo_();
 184         return std::string(info->stemmer_data, info->stemmer_name_len);
 185     }
 186 };
 187
 188 }
 189
 190 #endif // XAPIAN_INCLUDED_STEM_H