C++ THash::Save方法代码示例

void SaveAll()
{
	printf("\n<<< SAVING STARTS (PLEASE BE PATIENT!!!) >>> .......\n");
	TZipOut resOut("CascadesFullUrlsOnTwitterData.rar");
	cascadesInUrlsOnTwitter.Save(resOut);
	printf("The size of CascadesFullUrlsOnTwitterData was: %d\n",cascadesInUrlsOnTwitter.Len());

	printf("\n<<<<<<<< SAVING DONE >>>>>>>>\n\n");
}

//.........这里部分代码省略.........
			{
				if(fileExists(qContentFname))
				{
					// removing the quotes' content file
					system(TStr::Fmt("rm %s",qContentFname.CStr()).CStr());
				}
			}
			else
			{
				if(fileExists(qContentFname))
				{
					THash<TStr,TInt> quotesContent;
					THash<TInt,TSecTmV> CascadesOnTwitter;

					TZipIn ZquotesIn(qContentFname);
					quotesContent.Load(ZquotesIn);
					printf("Q%d loading done, it contains %d quotes.\n",indx-1,quotesContent.Len());

					conn = PQconnectdb("dbname=twitter host=postgresql01.mpi-sws.org user=twitter [email protected]");

					if (PQstatus(conn) == CONNECTION_BAD)
					{
						printf("We were unable to connect to the database");
						return 1;
					}

					// we use cursors/fetch to speed up the process; batch of 10000 tweets
					PQexec(conn, "begin work");
					PQexec(conn,TStr::Fmt("declare mycursor cursor for select tweettext, extract(epoch from tweettime) from tweets where tweettime >= timestamp '%s' and tweettime < timestamp '%s'", StartDate.CStr(), EndDate.CStr()).CStr());

					do
					{
						res = PQexec(conn, "FETCH 1000000 IN mycursor");    // all of them are: 1675401026
						if (PQresultStatus(res) == PGRES_TUPLES_OK)
						{
							rec_count = PQntuples(res);
							total_number_tweets += rec_count;
							printf("Adding %d tweets... (total: %d)\n", rec_count, total_number_tweets);

							for (row=0; row<rec_count; row++)
							{
								TweetStr = PQgetvalue(res, row, 0);
								tweet_date = TStr(PQgetvalue(res, row, 1)).GetFlt();
								TweetStrLc = TweetStr.ToLc();
								for(q=0;q<quotesContent.Len();q++)
								{
									if (TweetStrLc.SearchStr(quotesContent.GetKey(q)) > -1)
									{
										TSecTm td(tweet_date);
										id = CascadesOnTwitter.GetKeyId(quotesContent[q]);
										if(id == -1)
										{
											CascadesOnTwitter.AddDat(quotesContent[q]).Add(td);
										}
										else
										{
											CascadesOnTwitter.GetDat(quotesContent[q]).AddSorted(td);
										}
									}
								}
							}

							PQclear(res);
						}
						else
						{
							rec_count = 0;
						}
					}
					while (rec_count);

					PQexec(conn, "close mycursor");
					PQexec(conn, "commit work");
					PQfinish(conn);


					// Save the results
					TZipOut zout(resultFname);
					CascadesOnTwitter.Save(zout);

					// Remove the qoutes' content file
					system(TStr::Fmt("rm %s",qContentFname.CStr()).CStr());
				}
			}
		}

		printf("\n\nD O N E\n\n");
	}
	catch(exception& ex)
	{
		printf("\nError1 happened, it was: %s\n\n",ex.what());
	}
	catch(TPt<TExcept>& ex)
	{
		printf("\nError2 happened: %s\n\n",ex[0].GetStr().CStr());
	}

	printf("\nrun time: %s (%s)\n", ExeTm.GetTmStr(), TSecTm::GetCurTm().GetTmStr().CStr());
	return 0;
}

int main(int argc, char* argv[])
{
	TExeTm ExeTm;
	THash< TStr , CascadeElementV > quotesFiltered;
	double* vol_me;
	uint period = 9 * 3600;   // 9 days because of NIFTY paper

	printf("((((( Starting The Filtering Cascades CODE )))))\n");
	try
	{
		Env = TEnv(argc, argv, TNotify::StdNotify);
		Env.PrepArgs(TStr::Fmt("\nFiltering Memes Cascades. build: %s, %s. Time: %s", __TIME__, __DATE__, TExeTm::GetCurTm()));

		// ---== Loading Data ==---
		TZipIn ZquotesIn("QuotesPreprocessedData_NIFTY.rar");		///("/agbs/cluster/oaskaris/Data_Preparing_Codes/RESULTS/QuotesPreprocessedData_NIFTY.rar");
		THash< TStr , CascadeElementV > quotes;
		quotes.Load(ZquotesIn);
		printf("Loaded QuotesPreprocessedData_NIFTY has instances: %d\n\n\n",quotes.Len());

		// NIFTY Method for Filtering by Peaks
		uint begin = TSecTm(2008,7,31,0,0,0).GetAbsSecs();
		uint end = TSecTm(2009,10,1,0,0,0).GetAbsSecs();
		TSecTmV memesTimes;
		int bins = (end - begin) / period;
		for(int c=0;c<quotes.Len();c++)
		{
			memesTimes.Clr();
			for(int i=0;i<quotes[c].Len();i++)
			{
				memesTimes.Add(quotes[c][i].time);
			}
			vol_me = Tools::calculateHistOfCascade(memesTimes,begin,period,false);

			// calculating mean and standard deviation
			double mean = 0;
			for(int i=0;i<bins;i++)
			{
				mean += vol_me[i];
			}
			mean /= bins;

			double std = 0;
			for(int i=0;i<bins;i++)
			{
				std += pow(vol_me[i]-mean , 2);
			}
			std = sqrt(std / (bins-1));

			// peak definition by NIFTY: a point is a peak if its volume in 9 days binning is 1 standard deviation higher than the average frequency
			double maxVolume = mean + std;
			int peakCnt = 0;
			for(int i=0;i<bins;i++)
			{
				if(vol_me[i] > maxVolume)
				{
					peakCnt++;
				}
			}
			// if there is more than 5 peaks ignore this quote, since it is not a meme
			if(peakCnt > 5)
			{
				delete[] vol_me;
				continue;
			}

			quotesFiltered.AddDat(quotes.GetKey(c),quotes[c]);
			delete[] vol_me;
		}

		TZipOut mout("QuotesPreprocessedData_NIFTY_FINALFILTERED.rar");
		quotesFiltered.Save(mout);
		printf("Saved QuotesPreprocessedData_NIFTY_FINALFILTERED has instances: %d\n\n\n",quotesFiltered.Len());

		printf("\nThe Meme Filter for plotting had been done successfully.\n");
	}
	catch(exception& ex)
	{
		printf("\nError1 happened, it was: %s\n\n",ex.what());
	}
	catch(TPt<TExcept>& ex)
	{
		printf("\nError2 happened: %s\n\n",ex[0].GetStr().CStr());
	}

	printf("\nrun time: %s (%s)\n", ExeTm.GetTmStr(), TSecTm::GetCurTm().GetTmStr().CStr());
	return 0;
}