LoadDbGapDownloads.java

package loaders;

import java.io.File;
import java.io.FileFilter;
import java.io.FilenameFilter;
import java.util.ArrayList;
import java.util.List;

import org.apache.log4j.BasicConfigurator;
import org.molgenis.MolgenisOptions;
import org.molgenis.framework.db.Database;
import org.molgenis.framework.db.Database.DatabaseAction;
import org.molgenis.util.SimpleTuple;

import app.CsvImport;
import app.DatabaseFactory;
import app.JDBCDatabase;

public class LoadDbGapDownloads
{
	public static void main(String[] args) throws Exception
	{
		// start logging
		// BasicConfigurator.configure();

		Database db = new JDBCDatabase("apps/phenoflow/phenoflow.properties");

		loadDbGaPData(db);
	}

	public static void loadDbGaPData(Database db) throws Exception
	{
		// it is assumed that DbGapToPheno has put its results here
		// This will need updating if run on a different machine
		// File rootDir = new File("D:\\Data\\dbgap");
		File rootDir = new File("../pheno_data/dbgap/");

		for (File investigationDir : rootDir.listFiles(new FileFilter()
		{
			public boolean accept(File file)
			{
				return file.isDirectory() && file.getName().startsWith("phs");
			}
		}))
		{
			System.out.println("Loading " + investigationDir);

			// need to preload measurements first, otherwise protocol
			// will complain it's missing them due to incorrect
			// autogenerated order
			// List<String> list = new ArrayList<String>();
			// list.add("investigation");
			// list.add("ontologyterm");
			// list.add("category");
			// list.add("measurement");
			// list.add("measurement");
			//
			// CsvImport.importAll(investigationDir, db, new SimpleTuple(),
			// list,
			// DatabaseAction.ADD_IGNORE_EXISTING, "");
			// measurements already in, proceed loading a complete set
			CsvImport.importAll(investigationDir, db, new SimpleTuple(), null, DatabaseAction.ADD_IGNORE_EXISTING, "");
		}

	}
}