---

InvFPIndex.hpp

Go to the documentation of this file.

/*==========================================================================
 * Copyright (c) 2001 Carnegie Mellon University.  All Rights Reserved.
 *
 * Use of the Lemur Toolkit for Language Modeling and Information Retrieval
 * is subject to the terms of the software license set forth in the LICENSE
 * file included with this software, and also available at
 * http://www.cs.cmu.edu/~lemur/license.html
 *
 *==========================================================================
*/


#ifndef _INVFPINDEX_HPP
#define _INVFPINDEX_HPP

#include "common_headers.hpp"
#include "Index.hpp"
#include "InvFPDocList.hpp"
#include "InvFPTermList.hpp"
#include "InvFPTypes.hpp"


// for counts array
#define UNIQUE_TERMS 0
#define TOTAL_TERMS  1
#define DOCS         2
#define DT_FILES     3
#define INV_FILES    4

// for names array
#define DOC_INDEX    0
#define DOC_LOOKUP   1
#define TERM_INDEX   2
#define TERM_LOOKUP  3
#define TERM_IDS     4
#define DOC_IDS      5

class InvFPIndex : public Index {
public:
   InvFPIndex();
   InvFPIndex(const char* indexName);
  ~InvFPIndex(); 



  bool open(const char* indexName);

  bool openName(char* filename);



  int term(const char* word);

  const char* term(int termID);

  int document(const char* docIDStr);

  const char* document(int docID); 




  int docCount() { return counts[DOCS]; };

  int termCountUnique() { return counts[UNIQUE_TERMS]; };

  int termCount(int termID) const;

  int termCount() const { return counts[TOTAL_TERMS]; };

  float docLengthAvg();

  int docCount(int termID);

  int docLength(DOCID_T docID) const;

  int docLengthCounted(int docID);




  DocInfoList* docInfoList(int termID);

  TermInfoList* termInfoList(int docID);

  TermInfoList* termInfoListSeq(int docID);

private:
  bool fullToc(const char* fileName);
  bool mainToc(char* fileName);
  bool indexLookup();
  bool invFileIDs();
  bool dtLookup();
  bool dtFileIDs();
  bool termIDs();
  bool docIDs();

  int* counts;    // array to hold all the overall count stats of this db
  char** names;   // array to hold all the names for files we need for this db
  float aveDocLen; // the average document length in this index
  inv_entry* lookup;  // the array holding entries (index is termid)
  dt_entry* dtlookup; // the array holding entries to dt index (index of array is docid)
  TERM_T* terms;   // array of the term spellings (index is termid)
  EXDOCID_T* docnames; // array of the external docids (index is docid)
  char** dtfiles; // array of dt index filenames
  char** invfiles; // array of inv index filenames
  map<TERM_T, TERMID_T, ltstr> termtable; // table of terms to termid
  map<EXDOCID_T, DOCID_T, ltstr> doctable; // table of exdocids to docid
};

#endif

---