sorting on the basis of datetime

I have a file that has duplicates and I have to select the one with the earliest time

Code:

---

0:1:OrderReceive,01.07.2010 10:38:09.984,0G508Q090,0000F9F5,A,15,0,15,0,115
0:1:OrderReceive,01.07.2010 10:30:36.903,0G508Q090,0000F9F5,A,15,0,15,0,115

0:2:OrderReceive,01.07.2010 10:23:58.390,0G508Q086,0000F9DA,A,15,0,15,0,70
0:2:OrderReceive,01.07.2010 10:16:25.340,0G508Q086,0000F9DA,A,15,0,15,0,70

---

where first two lines are identical except the time part and the last two lines are identical but again except for its time part. I need to select the second line of each pair (because they have the earliest time) and the final output should be

Code:

---

0:1:OrderReceive,01.07.2010 10:30:36.903,0G508Q090,0000F9F5,A,15,0,15,0,115
0:2:OrderReceive,01.07.2010 10:16:25.340,0G508Q086,0000F9DA,A,15,0,15,0,70

---

could anyone please help?

Have you thought of using the string comparison operators?

You have asked a number of questions concerning this data. I don't
know what the current status of your code is. But the solution I would use:

1) create a class with ALL of the diffrent fields in the record asmember variables.

2) read the file into a vector

Then sorting / removing duplicates come down to supply a function which
compares the fields you are interested in.

So the first question I would have ... what are the fields in the line:

0:1:OrderReceive,01.07.2010 10:38:09.984,0G508Q090,0000F9F5,A,15,0,15,0,115

yes that's exactly the line and the code I have so far for this task is

Code:

---


struct ordRecvblock
{
        std::string lineID;        // 0:1:OrderReceive
            std::string dateTime;        // 01.07.2010 10:38:09.984
        std::string remSTr;        // 0G508Q090,0000F9F5,A,15,0,15,0,115

        ordRecvblock()
        {}

        ordRecvblock(const std::string& _lineID, const std::string& _dateTime, const std::string& _remSTr):
        lineID(_lineID), dateTime(_dateTime), remSTr(_remSTr)
        {}

        bool operator==(const ordRecvblock& p) const
        {
                return lineID == p.lineID && remSTr == p.remSTr;
        }

        bool operator<(const ordRecvblock& p) const
        {
                if(remSTr < p.remSTr) return true;
                if(remSTr > p.remSTr) return false;

                return false;
        }
};

//.................................

size_t pos1 = line.find("OrderReceive");

if (pos1 != string::npos)
{
        istringstream ss(line);
        getline(ss,ordb.lineID,',');
        getline(ss,ordb.dateTime,',');
        getline(ss,ordb.remSTr);
        dupOrdTime.push_back(ordb);
        continue;
}
else
{
        std::sort(dupOrdTime.begin(), dupOrdTime.end());
       
        std::pair<std::vector<ordRecvblock>::iterator, std::vector<ordRecvblock>::iterator> ret;
        for(std::vector<ordRecvblock>::iterator i = dupOrdTime.begin(); i != dupOrdTime.end(); i = ret.second)
        {
                std::vector<ordRecvblock>* dupVec = new std::vector<ordRecvblock>; // stores values that were duplicated
                //cout << i->lineID << ',' << i->dateTime << ',' << i->remSTr << " =>\n";
                ret = std::equal_range(i, dupOrdTime.end(), *i);
                for(std::vector<ordRecvblock>::iterator j = ret.first; j != ret.second; ++j)
                {
                        //cout << j->lineID << ',' << j->dateTime << ',' << j->remSTr << endl;
                        dupVec->push_back(*j);
                }
       
                for(std::vector<ordRecvblock>::iterator i = dupVec.begin(); i != dupVec.end(); ++i)
                {
                        cout << i->lineID << ',' << i->dateTime << ',' << i->remSTr << endl;
                }
        }
}

---

Thanks Philips I have been able to do it :). Let me test it with different and many lines

1) I'm not sure what the :"else" section is doing. In it, you create a vector using
new, but do not do a delete, causing a memory leak. If you did not add anything
to the vector, whyb are you sorting it ? I really can't figure out that section at all.

I have deleted the vector in my code. Sorry I missed it here :P

Code:

---

if (pos1 != string::npos)
{
        istringstream ss(line);
        getline(ss,ordb.lineID,',');
        getline(ss,ordb.dateTime,',');
        getline(ss,ordb.remSTr);

        dupOrdTime.push_back(ordb); // I am populating the vector here

        continue;
}

---

and finally trying to find duplicates

Code:

---

for(std::vector<ordRecvblock>::iterator i = dupOrdTime.begin(); i != dupOrdTime.end(); i = ret.second)
{
        std::vector<ordRecvblock>* dupVec = new std::vector<ordRecvblock>; // stores values that were duplicated
               
                ret = std::equal_range(i, dupOrdTime.end(), *i);
                for(std::vector<ordRecvblock>::iterator j = ret.first; j != ret.second; ++j)
                {
                        dupVec->push_back(*j);
                }
       
                for(std::vector<ordRecvblock>::iterator i = dupVec.begin(); i != dupVec.end(); ++i)
                {
                        cout << i->lineID << ',' << i->dateTime << ',' << i->remSTr << endl; // printing the duplicates
                }
        }

---

Do I necessarily have to convert the date/time part of the string to proper date time (time_t&)? because the simple string comparison is working and is picking up the correct time

You almost have it ...

Code:

---

        istringstream ss(line);
        getline(ss,ordb.lineID,',');
        getline(ss,ordb.dateTime,',');
        getline(ss,ordb.remSTr);

        // Check if the "lineID + remSTr" combination is already in the vector.
        // If it is, compare dates and use the lesser one

        vector<ordRecvblock>::iterator it = find(dupOrdTime.begin(),dupOrdTime.end(),ordb);

        if (it != dupOrdTime.end())
        {
            if (ordb.dateTime < it->dateTime)
            {
                cout << "removing : " << it->lineID << ',' << it->dateTime << ',' << it->remSTr << endl;

                *it = ordb;
            }
            else
            {
                cout << "not adding: " << ordb.lineID << ',' << ordb.dateTime << ',' << ordb.remSTr << endl;
            }
        }
        else
        {
            dupOrdTime.push_back(ordb);
        }

---

Oh thanks you sooo much Philip :)

Hello Philips. Could you please tell me at which point do I include/print those strings which do not have any matches at all while deleting/excluding the ones which do have one/more matches?

Thanks

Code:

---

istringstream ss(line);

getline(ss,ordb.lineID,',');
getline(ss,ordb.dateTime,',');
getline(ss,ordb.remSTr);

// Check if the "lineID + remSTr" combination is already in the vector.
// If it is, compare dates and use the lesser one

vector<ordRecvblock>::iterator it = find(OrdTime->begin(),OrdTime->end(),ordb);

if (it != OrdTime->end())
{
                    if (ordb.dateTime < it->dateTime)
                    {
                                cout << "    adding : " << ordb.lineID << ',' << ordb.dateTime << ',' << ordb.remSTr << endl;

                                *it = ordb;
                    }
}
else
{
                    cout << "NOT adding : " << ordb.lineID << ',' << ordb.dateTime << ',' << ordb.remSTr << endl;
                    OrdTime->push_back(ordb);
}

---

and also, there can be more than one similar string but with different time and I have to select the one with the minimum time out of many. Could you please tell me how do I go about that?

1) in your last "cout" statement , you say "NOT adding" , but you are adding

2) I am guessing that your operator < is not what you want. But you need to tell us exactly
what it means for 2 records to be duplicates. I think that it should probably be:

Code:

---

        bool operator<(const ordRecvblock& p) const
        {
                if(remSTr < p.remSTr) return true;
                if(remSTr > p.remSTr) return false;

                return lineID < p.lineID;
        }

---

2) " and also, there can be more than one similar string but with different time and I have to select the one with the minimum time out of many. Could you please tell me how do I go about that?"

Doesn't the code do that ? If not, post an example.

3) What do you mean by : "Could you please tell me at which point do I include/print those strings which do not have any matches at all while deleting/excluding the ones which do have one/more matches?"

Do you mean records that do not have any duplicates ? If so, you have to do that
after reading in all the data. You can create a map to count the number of times
a parttcular string occurs in the file (ignoring the dateTime):

Code:

---

map<ordRecvblock,int> record_count;

---

While reading ...

Code:

---

getline(ss,ordb.remSTr);
       
++record_count[ordb]; // keeps track of how many time "ordb" is in the file

---

And when completely done reading ...

Code:

---

  map<ordRecvblock,int>::iterator it = record_count.begin();
 
  while (it != record_count.end())
  {
    cout << it->first.lineID << "|" << it->first.remSTr << " : count = " << it->second << "\n";
    ++it;
  }

---

oh thanks Philip. That's soo soo soooo nice of you :). Let me make changes to my code according to your suggestions :)