view flys-backend/src/main/java/de/intevation/flys/importer/parsers/SedimentDensityParser.java @ 4524:be9e28cff0c4

Parse and store year in sediment densities. * Year is parsed ('guessed') from description column in .csv file.
author Raimund Renkert <rrenkert@intevation.de>
date Wed, 14 Nov 2012 17:24:55 +0100
parents f63b39799d2d
children 3694d8f48e16
line wrap: on
line source
package de.intevation.flys.importer.parsers;

import java.io.File;
import java.io.IOException;

import java.math.BigDecimal;

import java.text.DateFormat;
import java.text.NumberFormat;
import java.text.ParseException;

import java.util.ArrayList;
import java.util.Calendar;
import java.util.Date;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import org.apache.log4j.Logger;

import de.intevation.flys.importer.ImportDepth;
import de.intevation.flys.importer.ImportSedimentDensity;
import de.intevation.flys.importer.ImportSedimentDensityValue;
import de.intevation.flys.importer.ImportUnit;
import de.intevation.flys.utils.DateGuesser;


public class SedimentDensityParser extends LineParser {

    private static final Logger log =
        Logger.getLogger(SedimentDensityParser.class);

    private int densitsyColumn = 1;

    public static final NumberFormat nf = NumberFormat.getInstance(DEFAULT_LOCALE);


    public static final Pattern META_UNIT =
        Pattern.compile("^Einheit: \\[(.*)\\].*");

    public static final Pattern META_DEPTH =
        Pattern.compile("^Tiefe: (\\w++)-(\\w++)( (\\w++))?.*");

    protected List<ImportSedimentDensity> sedimentDensities;

    protected ImportSedimentDensity current;

    protected String currentDescription;


    public SedimentDensityParser() {
        sedimentDensities = new ArrayList<ImportSedimentDensity>();
    }


    @Override
    public void parse(File file) throws IOException {
        currentDescription = file.getName();

        super.parse(file);
    }


    @Override
    protected void reset() {
        current = new ImportSedimentDensity(currentDescription);
    }


    @Override
    protected void finish() {
        if (current != null) {
            sedimentDensities.add(current);
        }
    }


    @Override
    protected void handleLine(int lineNum, String line) {
        if (line.startsWith(START_META_CHAR)) {
            handleMetaLine(stripMetaLine(line));
        }
        else {
            handleDataLine(line);
        }
    }


    protected void handleMetaLine(String line) {
        if (handleMetaUnit(line)) {
            return;
        }
        else if (handleMetaDepth(line)) {
            return;
        }
        else if (handleMetaColumns(line)) {
            return;
        }
        else {
            log.warn("Unknown meta line: '" + line + "'");
        }
    }


    private boolean handleMetaColumns(String line) {
        String[] columns = line.split(";");
        for (int i = 0; i < columns.length; i++) {
            if (columns[i].contains("Sedimentdichte")) {
                this.densitsyColumn = i;
                return true;
            }
        }
        return false;
    }


    protected boolean handleMetaUnit(String line) {
        Matcher m = META_UNIT.matcher(line);

        if (m.matches()) {
            String unit = m.group(1);

            current.setUnit(new ImportUnit(unit));

            return true;
        }

        return false;
    }


    protected boolean handleMetaDepth(String line) {
        Matcher m = META_DEPTH.matcher(line);

        if (m.matches()) {
            String lo   = m.group(1);
            String up   = m.group(2);
            String unit = m.group(4);

            log.info("Found sediment density depth: " + lo + " - " + up + " " + unit);

            try {
                ImportDepth depth = new ImportDepth(
                    new BigDecimal(nf.parse(lo).doubleValue()),
                    new BigDecimal(nf.parse(up).doubleValue()),
                    new ImportUnit(unit)
                );

                current.setDepth(depth);

                return true;
            }
            catch (ParseException pe) {
                log.warn("Error while parsing numbers in: '" + line + "'");
            }
        }
        else {
            log.debug("Meta line doesn't contain depth information: " + line);
        }

        return false;
    }


    protected void handleDataLine(String line) {
        String[] vals = line.split(SEPERATOR_CHAR);

        if (vals == null || vals.length < 3) {
            log.warn("skip invalid data line: '" + line + "'");
            return;
        }

        BigDecimal km;
        BigDecimal density;
        try {
            km      = new BigDecimal(nf.parse(vals[0]).doubleValue());
            density = new BigDecimal(nf.parse(vals[this.densitsyColumn]).doubleValue());

        }
        catch (ParseException pe) {
            log.warn("Error while parsing numbers in '" + line + "'");
            return;
        }

        BigDecimal year = null;
        try {
            year =
                new BigDecimal(nf.parse(vals[vals.length - 1]).doubleValue());
        }
        catch(ParseException pe) {
            try {
                Date d = DateGuesser.guessDate(vals[vals.length - 1]);
                Calendar c = Calendar.getInstance();
                c.setTime(d);
                year = new BigDecimal(c.get(Calendar.YEAR));
            }
            catch (IllegalArgumentException iae) {
                log.warn("Error while parsing date in '" + line + "'");
                return;
            }
        }

        current.addValue(new ImportSedimentDensityValue(
            km,
            density,
            year,
            vals[vals.length - 1])
        );
    }


    public List<ImportSedimentDensity> getSedimentDensities() {
        return sedimentDensities;
    }
}
// vim:set ts=4 sw=4 si et sta sts=4 fenc=utf8 :

http://dive4elements.wald.intevation.org