Home | History | Annotate | Download | only in slm
      1 /*
      2  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS HEADER.
      3  *
      4  * Copyright (c) 2007 Sun Microsystems, Inc. All Rights Reserved.
      5  *
      6  * The contents of this file are subject to the terms of either the GNU Lesser
      7  * General Public License Version 2.1 only ("LGPL") or the Common Development and
      8  * Distribution License ("CDDL")(collectively, the "License"). You may not use this
      9  * file except in compliance with the License. You can obtain a copy of the CDDL at
     10  * http://www.opensource.org/licenses/cddl1.php and a copy of the LGPLv2.1 at
     11  * http://www.opensource.org/licenses/lgpl-license.php. See the License for the
     12  * specific language governing permissions and limitations under the License. When
     13  * distributing the software, include this License Header Notice in each file and
     14  * include the full text of the License in the License file as well as the
     15  * following notice:
     16  *
     17  * NOTICE PURSUANT TO SECTION 9 OF THE COMMON DEVELOPMENT AND DISTRIBUTION LICENSE
     18  * (CDDL)
     19  * For Covered Software in this distribution, this License shall be governed by the
     20  * laws of the State of California (excluding conflict-of-law provisions).
     21  * Any litigation relating to this License shall be subject to the jurisdiction of
     22  * the Federal Courts of the Northern District of California and the state courts
     23  * of the State of California, with venue lying in Santa Clara County, California.
     24  *
     25  * Contributor(s):
     26  *
     27  * If you wish your version of this file to be governed by only the CDDL or only
     28  * the LGPL Version 2.1, indicate your decision by adding "[Contributor]" elects to
     29  * include this software in this distribution under the [CDDL or LGPL Version 2.1]
     30  * license." If you don't indicate a single choice of license, a recipient has the
     31  * option to distribute your version of this file under either the CDDL or the LGPL
     32  * Version 2.1, or to extend the choice of license to its licensees as provided
     33  * above. However, if you add LGPL Version 2.1 code and therefore, elected the LGPL
     34  * Version 2 license, then the option applies only if the new code is made subject
     35  * to such option by the copyright holder.
     36  */
     37 
     38 #ifndef _SIM_FILE_MERGE_H
     39 #define _SIM_FILE_MERGE_H
     40 
     41 #include "../portability.h"
     42 
     43 #include <stdio.h>
     44 #include <deque>
     45 #include <vector>
     46 #include <algorithm>
     47 
     48 template<class unit_type>
     49 class TUnitAndParaInfo {
     50 public:
     51     typedef unit_type TUnit;
     52     TUnitAndParaInfo() : unit(), runOut(false) {}
     53 public:
     54     TUnit unit;
     55     bool runOut;
     56 };
     57 
     58 template<class unit_type>
     59 class file_para
     60 {
     61 public:
     62     typedef unit_type TUnit;
     63     typedef TUnitAndParaInfo<TUnit> UnitAndParaInfo;
     64     typedef std::deque<UnitAndParaInfo> TItemBuf;
     65     typedef typename TItemBuf::iterator TIBIterator;
     66     typedef typename TItemBuf::const_iterator TIBConstIterator;
     67 
     68     file_para(FILE* p_file, size_t start, size_t end)
     69         : fp(p_file), cur_offset(start), last_offset(end), runOut(false), buf() {}
     70 
     71     UnitAndParaInfo& operator*()
     72         {
     73             if (buf.size() == 0) {
     74                 for (int i=0; i < BUF_SIZE; ++i)
     75                 {
     76                     buf.push_back(UnitAndParaInfo());
     77                     UnitAndParaInfo& e = buf.back();
     78                     e.runOut = runOut = !(e.unit.read(fp, cur_offset, last_offset));
     79                     if (runOut) break;
     80                 }
     81             }
     82             return buf.front();
     83         }
     84 
     85     file_para& operator++()
     86         {
     87             if (buf.size() == 0)
     88                 this->operator*();
     89             buf.pop_front();
     90             return *this;
     91         }
     92 
     93     bool operator< (file_para& another)
     94         {
     95             const UnitAndParaInfo& me = this->operator*();
     96             const UnitAndParaInfo& you = *another;
     97             if (me.runOut){
     98                 if (you.runOut)
     99                     return ((char*)&me < (char*)&you);
    100                 else
    101                     return true;
    102             } else {
    103                 if (you.runOut)
    104                     return false;
    105                 else {
    106                     if (me.unit > you.unit) return true;
    107                     if (me.unit == you.unit) return ((char*)&me < (char*)&you);
    108                     return false;
    109                 }
    110             }
    111         }
    112 
    113 private:
    114     static const int BUF_SIZE=1024;
    115     FILE * fp;
    116     bool runOut;
    117     size_t cur_offset, last_offset;
    118     TItemBuf buf;
    119 };
    120 
    121 template <class PPara>
    122 class PtrCompare {
    123 public:
    124     bool operator()(const PPara& p1, const PPara& p2)
    125         { return (*p1 < *p2); }
    126 };
    127 
    128 template<class unit_type>
    129 class CMultiWayFileMerger
    130 {
    131 public:
    132     typedef file_para<unit_type> TPara;
    133     typedef std::vector<TPara*> TParaVec;
    134 
    135     void addPara(FILE *fp, size_t first_offset, size_t last_offset)
    136         {
    137             paras.push_back(new TPara(fp, first_offset, last_offset));
    138         }
    139     void start()
    140         {
    141             std::make_heap(paras.begin(), paras.end(), PtrCompare<TPara*>());
    142         }
    143     TPara* getBest() //You then have to deal same items form different part
    144         {
    145             std::pop_heap(paras.begin(), paras.end(), PtrCompare<TPara*>());
    146             return paras.back();
    147         }
    148     void next()
    149         {
    150             ++(*(paras.back()));
    151             std::push_heap(paras.begin(), paras.end(), PtrCompare<TPara*>());
    152         }
    153     ~CMultiWayFileMerger() { for (int i=0; i < paras.size(); ++i) delete paras[i]; }
    154 private:
    155     TParaVec paras;
    156 };
    157 
    158 #endif
    159