---

InvDocList.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.lemurproject.org/license.html
 *
 *==========================================================================
*/


#ifndef _INVDOCLIST_HPP
#define _INVDOCLIST_HPP

/*
 * NAME DATE - COMMENTS
 * tnt 03/2001 - created
 *
 *========================================================================*/
#include <cmath>
#include "InvFPTypes.hpp"
#include "common_headers.hpp"
#include "DocInfoList.hpp"
#include "MemCache.hpp"
#include "RVLCompress.hpp"

extern "C" {
  #include <cstdio>
}

#define DEFAULT 9

class InvDocList: public DocInfoList {
public:
  InvDocList();

  InvDocList(TERMID_T id, int len);
  InvDocList(MemCache* mc, TERMID_T id, int len);  
  InvDocList(MemCache* mc, TERMID_T id, int len, DOCID_T docid, LOC_T location);
  InvDocList(TERMID_T id, int listlen, LOC_T* list, int fr, DOCID_T* ldocid, int len);
  ~InvDocList();

  void setList(TERMID_T id, int listlen, LOC_T* list, int fr, DOCID_T* ldocid=NULL, int len=0);

  void setListSafe(TERMID_T id, int listlen, LOC_T* list, int fr, DOCID_T* ldocid, int len);

  void reset();

  void resetFree();

  bool allocMem();
  bool hasNoMem();

  virtual bool addTerm(DOCID_T docid);

  virtual bool append(InvDocList* tail);

  virtual void startIteration() const;
  virtual bool hasMore() const;
  virtual DocInfo* nextEntry() const;
  virtual void nextEntry(DocInfo* info) const;

  DOCID_T curDocID() const{ if (lastid == NULL) return -1; return *lastid; };
  COUNT_T docFreq() const{ return df; };
  int length() const{ return end-begin; };
  TERMID_T termID() const{ return uid; };
  int termLen() const{ return strlength; };
  virtual COUNT_T termCTF() const;
  int curDocIDdiff() const{ return lastid-begin; };
  int curDocIDtf() const{ return *(lastid+1); };
  int memorySize() const{ return size; };

  void binWrite(ofstream& of);

  bool binRead(ifstream& inf);

  void binWriteC(ofstream& of);

  bool binReadC(ifstream& inf);

protected:
  // Helper functions for iterator, subclasses should override
  virtual DocInfo* getElement(DocInfo* elem, POS_T position) const;
  virtual POS_T beginPosition() const { return (POS_T) begin; }
  virtual POS_T endPosition() const { return (POS_T) end; }
  virtual POS_T nextPosition(POS_T position) const;

  bool getMoreMem();
  int logb2(int num);

  virtual void deltaEncode();

  virtual void deltaDecode();

  // Use LOC_T* for TERMID/DOCID/COUNT/LOC.
  LOC_T* begin;         // pointer to the beginning of this list
  LOC_T* lastid;        // pointer to the most recent DocID added
  LOC_T* freq;          // pointer to the frequency of the last DocID
  LOC_T * end;            // pointer to the next free memory
  mutable LOC_T* iter;    // pointer tells us where we are in iteration
  int  size;            // how big are we, increment in powers of 2, start at 16K
  int  LOC_Tsize;       // sizeof(LOC_T) value
  int  strlength;       // the character length of our corresponding string
  TERMID_T  uid;                          // a unique ID for our string
  COUNT_T  df;                    // the document frequency for current term
  MemCache* cache;      // the cache to get memory from
  bool hascache;        // remember if we have our own cache

  bool READ_ONLY;    // flag for whether this list can be added
private:
  mutable DocInfo entry;
};

#endif

---