xapian-core/weight/weightinternal.h

   1 /** @file weightinternal.h
   2  * @brief Xapian::Weight::Internal class, holding database and term statistics.
   3  */
   4 /* Copyright (C) 2007 Lemur Consulting Ltd
   5  * Copyright (C) 2009,2010,2011,2013,2014,2015 Olly Betts
   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_WEIGHTINTERNAL_H
  23 #define XAPIAN_INCLUDED_WEIGHTINTERNAL_H
  24
  25 #include "xapian/weight.h"
  26
  27 #include "xapian/database.h"
  28 #include "xapian/error.h"
  29 #include "xapian/query.h"
  30
  31 #include "backends/database.h"
  32 #include "internaltypes.h"
  33 #include "omassert.h"
  34
  35 #include <cerrno>
  36 #include <cstdlib>
  37 #include <map>
  38 #include <string>
  39
  40 /// The frequencies for a term.
  41 struct TermFreqs {
  42     Xapian::doccount termfreq;
  43     Xapian::doccount reltermfreq;
  44     Xapian::termcount collfreq;
  45     double max_part;
  46
  47     TermFreqs() : termfreq(0), reltermfreq(0), collfreq(0), max_part(0.0) {}
  48     TermFreqs(Xapian::doccount termfreq_,
  49               Xapian::doccount reltermfreq_,
  50               Xapian::termcount collfreq_,
  51               double max_part_ = 0.0)
  52         : termfreq(termfreq_),
  53           reltermfreq(reltermfreq_),
  54           collfreq(collfreq_),
  55           max_part(max_part_) {}
  56
  57     void operator+=(const TermFreqs & other) {
  58         termfreq += other.termfreq;
  59         reltermfreq += other.reltermfreq;
  60         collfreq += other.collfreq;
  61         max_part += other.max_part;
  62     }
  63
  64     /// Return a std::string describing this object.
  65     std::string get_description() const;
  66 };
  67
  68 namespace Xapian {
  69
  70 class RSet;
  71
  72 /** Class to hold statistics for a given collection. */
  73 class Weight::Internal {
  74 #ifdef XAPIAN_ASSERTIONS
  75     /** Number of sub-databases. */
  76     size_t subdbs;
  77
  78     /** True if we've finalised the stats.
  79      *
  80      *  Used for assertions.
  81      */
  82     mutable bool finalised;
  83 #endif
  84
  85   public:
  86     /** Total length of all documents in the collection. */
  87     totlen_t total_length;
  88
  89     /** Number of documents in the collection. */
  90     Xapian::doccount collection_size;
  91
  92     /** Number of relevant documents in the collection. */
  93     Xapian::doccount rset_size;
  94
  95     /** Number of terms in the collection. */
  96     Xapian::termcount total_term_count;
  97
  98     /** Has max_part been set for any term?
  99      *
 100      *  If not, we can avoid having to serialise max_part.
 101      */
 102     bool have_max_part;
 103
 104     /** Database to get the bounds on doclength and wdf from. */
 105     Xapian::Database db;
 106
 107     /** The query. */
 108     Xapian::Query query;
 109
 110     /** Map of term frequencies and relevant term frequencies for the
 111      *  collection. */
 112     std::map<std::string, TermFreqs> termfreqs;
 113
 114     Internal()
 115         :
 116 #ifdef XAPIAN_ASSERTIONS
 117           subdbs(0), finalised(false),
 118 #endif
 119           total_length(0), collection_size(0), rset_size(0),
 120           total_term_count(0), have_max_part(false) { }
 121
 122     /** Add in the supplied statistics from a sub-database.
 123      *
 124      *  Used for remote databases, where we pass across a serialised stats
 125      *  object, unserialise it, and add it to our total.
 126      */
 127     Internal & operator+=(const Internal & inc);
 128
 129     void set_query(const Xapian::Query &query_) {
 130         AssertEq(subdbs, 0);
 131         query = query_;
 132     }
 133
 134     /// Accumulate the rtermfreqs for terms in the query.
 135     void accumulate_stats(const Xapian::Database::Internal &sub_db,
 136                           const Xapian::RSet &rset);
 137
 138     /** Get the frequencies for the given term.
 139      *
 140      *  termfreq is "n_t", the number of documents in the collection indexed by
 141      *  the given term.
 142      *
 143      *  reltermfreq is "r_t", the number of relevant documents in the
 144      *  collection indexed by the given term.
 145      *
 146      *  collfreq is the total number of occurrences of the term in all
 147      *  documents.
 148      */
 149     bool get_stats(const std::string & term,
 150                    Xapian::doccount & termfreq,
 151                    Xapian::doccount & reltermfreq,
 152                    Xapian::termcount & collfreq) const {
 153 #ifdef XAPIAN_ASSERTIONS
 154         finalised = true;
 155 #endif
 156         // We pass an empty std::string for term when calculating the extra
 157         // weight.
 158         if (term.empty()) {
 159             termfreq = collection_size;
 160             collfreq = collection_size;
 161             reltermfreq = rset_size;
 162             return true;
 163         }
 164
 165         auto i = termfreqs.find(term);
 166         if (i == termfreqs.end()) {
 167             termfreq = reltermfreq = collfreq = 0;
 168             return false;
 169         }
 170
 171         termfreq = i->second.termfreq;
 172         reltermfreq = i->second.reltermfreq;
 173         collfreq = i->second.collfreq;
 174         return true;
 175     }
 176
 177     /// Get just the termfreq.
 178     bool get_stats(const std::string & term,
 179                    Xapian::doccount & termfreq) const {
 180         Xapian::doccount dummy1;
 181         Xapian::termcount dummy2;
 182         return get_stats(term, termfreq, dummy1, dummy2);
 183     }
 184
 185     /// Get the termweight.
 186     bool get_termweight(const std::string & term, double & termweight) const {
 187 #ifdef XAPIAN_ASSERTIONS
 188         finalised = true;
 189 #endif
 190         termweight = 0.0;
 191         if (term.empty()) {
 192             return false;
 193         }
 194
 195         auto i = termfreqs.find(term);
 196         if (i == termfreqs.end()) {
 197             return false;
 198         }
 199
 200         termweight = i->second.max_part;
 201         return true;
 202     }
 203
 204     /** Get the minimum and maximum termweights.
 205      *
 206      *  Used by the snippet code.
 207      */
 208     void get_max_termweight(double & min_tw, double & max_tw) {
 209         auto i = termfreqs.begin();
 210         while (i != termfreqs.end() && i->second.max_part == 0.0) ++i;
 211         if (rare(i == termfreqs.end())) {
 212             min_tw = max_tw = 0.0;
 213             return;
 214         }
 215         min_tw = max_tw = i->second.max_part;
 216         while (++i != termfreqs.end()) {
 217             double max_part = i->second.max_part;
 218             if (max_part > max_tw) {
 219                 max_tw = max_part;
 220             } else if (max_part < min_tw && max_part != 0.0) {
 221                 min_tw = max_part;
 222             }
 223         }
 224     }
 225
 226     /// Set max_part for a term.
 227     void set_max_part(const std::string & term, double max_part) {
 228         have_max_part = true;
 229         Assert(!term.empty());
 230         auto i = termfreqs.find(term);
 231         if (i != termfreqs.end())
 232             i->second.max_part += max_part;
 233     }
 234
 235     Xapian::doclength get_average_length() const {
 236 #ifdef XAPIAN_ASSERTIONS
 237         finalised = true;
 238 #endif
 239         if (rare(collection_size == 0)) return 0;
 240         return Xapian::doclength(total_length) / collection_size;
 241     }
 242
 243     /** Set the "bounds" stats from Database @a db. */
 244     void set_bounds_from_db(const Xapian::Database &db_) {
 245         Assert(!finalised);
 246         db = db_;
 247     }
 248
 249     /// Return a std::string describing this object.
 250     std::string get_description() const;
 251
 252     static bool double_param(const char ** p, double * ptr_val) {
 253         char *end;
 254         errno = 0;
 255         double v = strtod(*p, &end);
 256         if (*p == end || errno) return false;
 257         *p = end;
 258         *ptr_val = v;
 259         return true;
 260     }
 261
 262     static void parameter_error(const char * msg,
 263                                 const std::string & scheme) {
 264         string m(msg);
 265         m += ": '";
 266         m += scheme;
 267         m += "'";
 268         throw InvalidArgumentError(m);
 269     }
 270 };
 271
 272 }
 273
 274 #endif // XAPIAN_INCLUDED_WEIGHTINTERNAL_H