--- /dev/null
+package combio.statistic;
+
+import java.text.ParseException;
+import java.text.SimpleDateFormat;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collections;
+import java.util.Date;
+import java.util.Iterator;
+import java.util.List;
+
+import me.prettyprint.cassandra.serializers.LongSerializer;
+import me.prettyprint.cassandra.serializers.StringSerializer;
+import me.prettyprint.hector.api.beans.ColumnSlice;
+import me.prettyprint.hector.api.beans.HColumn;
+import me.prettyprint.hector.api.beans.OrderedRows;
+import me.prettyprint.hector.api.beans.Row;
+import me.prettyprint.hector.api.factory.HFactory;
+import me.prettyprint.hector.api.query.QueryResult;
+import me.prettyprint.hector.api.query.RangeSlicesQuery;
+import me.prettyprint.hector.api.query.SliceQuery;
+import combio.cassandra.CassandraCreate;
+import combio.cassandra.DataBase;
+
+public class StatisticsProt {
+ private final static long MILLISECONDS_PER_DAY = 1000L * 60 * 60 * 24;
+ private CassandraCreate cc = new CassandraCreate();
+ private ArrayList<DataBase> query;
+
+ // query for the period from date1 till date2
+ public List<DataBase> readDetail(String dateInStringSt, String dateInStringEnd) {
+ long dateWorkSt = DateParsing(dateInStringSt);
+ long dateWorkEnd = DateParsing(dateInStringEnd);
+ query = new ArrayList<DataBase>();
+ while (dateWorkSt <= dateWorkEnd) {
+ SliceQuery<Long, String, String> result = HFactory.createSliceQuery(cc.GetKeyspace(), LongSerializer.get(), StringSerializer.get(), StringSerializer.get());
+ result.setColumnFamily("ProteinData");
+ result.setKey(dateWorkSt);
+ result.setRange(null, null, false, Integer.MAX_VALUE);
+ QueryResult <ColumnSlice<String, String>> columnSlice = result.execute();
+ DataBase db = new DataBase(DateFormat(dateWorkSt), columnSlice.get().getColumns().size());
+ query.add(db);
+ dateWorkSt += MILLISECONDS_PER_DAY ;
+ }
+ return query;
+ }
+
+ // query jobs for the period from dateInStringSt till dateInStringEnd
+ public List<DataBase> readLength(String dateInStringSt, String dateInStringEnd) {
+ query = new ArrayList<DataBase>();
+ long dateWorkSt = DateParsing(dateInStringSt);
+ long dateWorkEnd = DateParsing(dateInStringEnd);
+ while (dateWorkSt <= dateWorkEnd) {
+ List<Integer> timeResult = new ArrayList<Integer>();
+ SliceQuery<Long, String, String> result = HFactory.createSliceQuery(cc.GetKeyspace(), LongSerializer.get(), StringSerializer.get(), StringSerializer.get());
+ result.setColumnFamily("ProteinData");
+ result.setKey(dateWorkSt);
+ result.setRange(null, null, false, Integer.MAX_VALUE);
+ QueryResult <ColumnSlice<String, String>> columnSlice = result.execute();
+ List<HColumn<String, String>> col = columnSlice.get().getColumns();
+ Iterator<HColumn<String, String>> itCol = col.iterator();
+ for (int i = 0; i < 4; i++)
+ timeResult.add(i, 0);
+ while (itCol.hasNext()) {
+ String id = itCol.next().getName();
+ long lenResult = CountID(id);
+ if (lenResult <= 1)
+ timeResult.set(0, timeResult.get(0) + 1);
+ else if (lenResult > 1 && lenResult <= 10)
+ timeResult.set(1, timeResult.get(1) + 1);
+ else if (lenResult > 10 && lenResult <= 20)
+ timeResult.set(2, timeResult.get(2) + 1);
+ else
+ timeResult.set(3, timeResult.get(3) + 1);
+ }
+ DataBase db = new DataBase();
+ db.setTimeRez(timeResult);
+ db.setDate(DateFormat(dateWorkSt));
+ query.add(db);
+ List<Integer> test = query.get(0).getTimeRez();
+ dateWorkSt += MILLISECONDS_PER_DAY ;
+ }
+ return query;
+ }
+
+ //query by a protein name
+ public List<DataBase> readProt(String protIn) {
+ query = new ArrayList<DataBase>();
+ SliceQuery<String, String, String> result = HFactory.createSliceQuery(cc.GetKeyspace(), StringSerializer.get(), StringSerializer.get(), StringSerializer.get());
+ result.setColumnFamily("ProteinRow");
+ result.setKey(protIn);
+ result.setRange(null, null, false, Integer.MAX_VALUE);
+ QueryResult <ColumnSlice<String, String>> columnSlice = result.execute();
+ Iterator <HColumn<String, String>> it = columnSlice.get().getColumns().iterator();
+ while (it.hasNext()) {
+ HColumn<String, String> col = it.next();
+ DataBase db = new DataBase();
+ db.setProt(protIn);
+ db.setId(col.getName());
+ db.setJpred(col.getValue());
+ query.add(db);
+ }
+ return query;
+ }
+
+
+ //query by a sequence (whether exists proteins contains this sequence in the name)
+ public List<DataBase> readPart(String protIn) {
+ int row_count = 10000;
+ query = new ArrayList<DataBase>();
+ RangeSlicesQuery<String, String, String> result = HFactory.createRangeSlicesQuery(cc.GetKeyspace(), StringSerializer.get(), StringSerializer.get(), StringSerializer.get());
+ result.setColumnFamily("ProteinRow");
+ result.setRange(null, null, false, Integer.MAX_VALUE);
+ result.setRowCount(row_count);
+ String last_key = null;
+ while (true) {
+ result.setKeys(last_key, null);
+ QueryResult <OrderedRows<String,String, String>> columnSlice = result.execute();
+ OrderedRows<String, String, String> rows = columnSlice.get();
+ Iterator<Row<String, String, String>> rowsIterator = rows.iterator();
+ while (rowsIterator.hasNext()) {
+ Row<String, String, String> row = rowsIterator.next();
+ last_key = row.getKey();
+ if (last_key.matches("(.*)" + protIn + "(.*)")) {
+ Iterator <HColumn<String, String>> it = row.getColumnSlice().getColumns().iterator();
+ while (it.hasNext()) {
+ HColumn<String, String> col = it.next();
+ List<String> subProt = new ArrayList<String>();
+ String subStr = last_key;
+ while (subStr.length() > 0 && subStr.contains(protIn)) {
+ String first = subStr.substring(0, subStr.indexOf(protIn));
+ if (first.length() > 0)
+ subProt.add(first);
+ subProt.add(protIn);
+ subStr = subStr.substring(subStr.indexOf(protIn) + protIn.length(), subStr.length());
+ }
+ if (subStr.length() > 0)
+ subProt.add(subStr);
+ DataBase db = new DataBase();
+ db.setProt(last_key);
+ db.setId(col.getName());
+ db.setJpred(col.getValue());
+ db.setSubProt(subProt);
+ query.add(db);
+ }
+ }
+ }
+ if (rows.getCount() < row_count)
+ break;
+ }
+ return query;
+ }
+
+ // convert String to Date
+ private static long DateParsing(String datInput) {
+ long dateWorkSt = 0;
+ SimpleDateFormat formatter = new SimpleDateFormat("yyyy/MM/dd");
+ try {
+ dateWorkSt = formatter.parse(datInput).getTime();
+ } catch (ParseException e) {
+ e.printStackTrace();
+ }
+ return dateWorkSt;
+ }
+
+ // convert String to Date
+ private static long TimeConvert(String datInput) {
+ long dateWorkSt = 0;
+ SimpleDateFormat formatter = new SimpleDateFormat("yyyy/MM/dd:hh:mm:ss");
+ try {
+ dateWorkSt = formatter.parse(datInput).getTime();
+ } catch (ParseException e) {
+ e.printStackTrace();
+ }
+ // System.out.println("start reverce" + DateFormat1(dateWorkSt));
+ return dateWorkSt;
+ }
+
+ // convert long to date in string format
+ private static String DateFormat(long inDate){
+ SimpleDateFormat datformat = new SimpleDateFormat("dd/MM/yyyy");
+ String dateString = datformat.format(new Date(inDate));
+ return dateString;
+ }
+
+ private static String DateFormat1(long inDate){
+ SimpleDateFormat datformat = new SimpleDateFormat("dd/MM/yyyy:hh:mm:ss");
+ String dateString = datformat.format(new Date(inDate));
+ return dateString;
+ }
+
+ public long CountID(String id) {
+ SliceQuery<String, String, String> sliceQuery = HFactory.createSliceQuery(cc.GetKeyspace(), StringSerializer.get(), StringSerializer.get(), StringSerializer.get());
+ sliceQuery.setColumnFamily("ProteinLog").setKey(id).setRange("", "", false, 100);
+ QueryResult<ColumnSlice<String, String>> result = sliceQuery.execute();
+ String datBegin = result.get().getColumnByName("DataBegin").getValue();
+ String datEnd = result.get().getColumnByName("DataEnd").getValue();
+ long datBeginLong = TimeConvert(datBegin);
+ long datEndLong = TimeConvert(datEnd);
+ return datEndLong-datBeginLong;
+
+ }
+
+}
git://source.jalview.org / jabaws.git / blobdiff