Mercurial > dive4elements > river
diff flys-backend/src/main/java/de/intevation/flys/importer/parsers/AnnotationsParser.java @ 1211:f08fe480092c
Moved file parsers to separate package.
flys-backend/trunk@2337 c6561f87-3c4e-4783-a992-168aeb5c3f6f
author | Sascha L. Teichmann <sascha.teichmann@intevation.de> |
---|---|
date | Fri, 15 Jul 2011 13:07:45 +0000 |
parents | |
children | c5c48f52dc7b |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/flys-backend/src/main/java/de/intevation/flys/importer/parsers/AnnotationsParser.java Fri Jul 15 13:07:45 2011 +0000 @@ -0,0 +1,196 @@ +package de.intevation.flys.importer.parsers; + +import java.util.HashMap; +import java.util.TreeSet; +import java.util.List; +import java.util.ArrayList; + +import java.io.IOException; +import java.io.File; +import java.io.LineNumberReader; +import java.io.InputStreamReader; +import java.io.FileInputStream; + +import java.math.BigDecimal; + +import org.apache.log4j.Logger; + +import de.intevation.flys.utils.FileTools; + +import de.intevation.flys.importer.ImportAnnotation; +import de.intevation.flys.importer.ImportRange; +import de.intevation.flys.importer.ImportEdge; +import de.intevation.flys.importer.ImportAnnotationType; +import de.intevation.flys.importer.ImportAttribute; +import de.intevation.flys.importer.ImportPosition; + +public class AnnotationsParser +{ + private static Logger log = Logger.getLogger(AnnotationsParser.class); + + public static final String ENCODING = "ISO-8859-1"; + + public static final String [] TO_SCAN = { + "Basisdaten", + "Streckendaten" + }; + + protected HashMap<String, ImportAttribute> attributes; + protected HashMap<String, ImportPosition> positions; + protected TreeSet<ImportAnnotation> annotations; + protected AnnotationClassifier classifier; + + public AnnotationsParser() { + this(null); + } + + public AnnotationsParser(AnnotationClassifier classifier) { + attributes = new HashMap<String, ImportAttribute>(); + positions = new HashMap<String, ImportPosition>(); + annotations = new TreeSet<ImportAnnotation>(); + this.classifier = classifier; + } + + public void parseFile(File file) throws IOException { + log.info("parsing km file: '" + file + "'"); + + ImportAnnotationType defaultIAT = null; + + if (classifier != null) { + defaultIAT = classifier.classifyFile( + file.getName(), + classifier.getDefaultType()); + } + + LineNumberReader in = null; + try { + in = + new LineNumberReader( + new InputStreamReader( + new FileInputStream(file), ENCODING)); + + String line = null; + while ((line = in.readLine()) != null) { + if ((line = line.trim()).length() == 0 + || line.startsWith("*")) { + continue; + } + + String [] parts = line.split("\\s*;\\s*"); + + if (parts.length < 3) { + log.warn("not enough columns in line " + + in.getLineNumber()); + continue; + } + + ImportPosition position = positions.get(parts[0]); + if (position == null) { + position = new ImportPosition(parts[0]); + positions.put(parts[0], position); + } + + ImportAttribute attribute = attributes.get(parts[1]); + if (attribute == null) { + attribute = new ImportAttribute(parts[1]); + attributes.put(parts[1], attribute); + } + + String [] r = parts[2].replace(",", ".").split("\\s*#\\s*"); + + BigDecimal from, to; + + try { + from = new BigDecimal(r[0]); + to = r.length < 2 ? null : new BigDecimal(r[1]); + if (to != null && from.compareTo(to) > 0) { + BigDecimal t = from; from = to; to = t; + } + } + catch (NumberFormatException nfe) { + log.warn("invalid number in line " + in.getLineNumber()); + continue; + } + + ImportEdge edge = null; + + if (parts.length == 4) { // Only 'Unterkante' + try { + edge = new ImportEdge( + null, + new BigDecimal(parts[3].trim().replace(',', '.'))); + } + catch (NumberFormatException nfe) { + log.warn("cannot parse 'Unterkante' in line " + + in.getLineNumber()); + } + } + else if (parts.length > 4) { // 'Unterkante' and 'Oberkante' + String bottom = parts[3].trim().replace(',', '.'); + String top = parts[4].trim().replace(',', '.'); + try { + BigDecimal b = bottom.length() == 0 + ? null + : new BigDecimal(bottom); + BigDecimal t = top.length() == 0 + ? null + : new BigDecimal(top); + edge = new ImportEdge(t, b); + } + catch (NumberFormatException nfe) { + log.warn( + "cannot parse 'Unterkante' or 'Oberkante' in line " + + in.getLineNumber()); + } + } + + ImportRange range = new ImportRange(from, to); + + ImportAnnotationType type = classifier != null + ? classifier.classifyDescription(line, defaultIAT) + : null; + + ImportAnnotation annotation = new ImportAnnotation( + attribute, position, range, edge, type); + + if (!annotations.add(annotation)) { + log.warn("duplicated annotation '" + parts[0] + + "' in line " + in.getLineNumber()); + } + } + } + finally { + if (in != null) { + in.close(); + } + } + } + + public void parse(File root) throws IOException { + + for (String toScan: TO_SCAN) { + File directory = FileTools.repair(new File(root, toScan)); + if (!directory.isDirectory()) { + log.warn("'" + directory + "' is not a directory."); + continue; + } + File [] files = directory.listFiles(); + if (files == null) { + log.warn("cannot list directory '" + directory + "'"); + continue; + } + + for (File file: files) { + if (file.isFile() && file.canRead() + && file.getName().toLowerCase().endsWith(".km")) { + parseFile(file); + } + } + } // for all directories to scan + } + + public List<ImportAnnotation> getAnnotations() { + return new ArrayList<ImportAnnotation>(annotations); + } +} +// vim:set ts=4 sw=4 si et sta sts=4 fenc=utf8 :