Mercurial > dive4elements > river

annotate flys-backend/src/main/java/de/intevation/flys/importer/InfoGewParser.java @ 177:31895d24387e

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Importer: Added info gew parser. flys-backend/trunk@1485 c6561f87-3c4e-4783-a992-168aeb5c3f6f
author Sascha L. Teichmann <sascha.teichmann@intevation.de>
date Wed, 16 Mar 2011 17:01:55 +0000
parents
children 3908bfb2aa43
rev   line source
177
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
1 package de.intevation.flys.importer;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
2
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
3 import java.io.File;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
4
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
5 import java.util.List;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
6 import java.util.ArrayList;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
7
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
8 import java.util.regex.Pattern;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
9 import java.util.regex.Matcher;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
10
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
11 import java.io.IOException;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
12 import java.io.LineNumberReader;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
13 import java.io.FileInputStream;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
14 import java.io.InputStreamReader;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
15
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
16 import org.apache.log4j.Logger;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
17
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
18 import de.intevation.flys.utils.FileTools;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
19
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
20 public class InfoGewParser
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
21 {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
22 private static Logger log = Logger.getLogger(InfoGewParser.class);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
23
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
24 public static final String ENCODING = "ISO-8859-1";
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
25
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
26 public static final Pattern GEWAESSER =
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
27 Pattern.compile("^\\s*Gew\u00e4sser\\s*:\\s*(.+)");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
28
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
29 public static final Pattern WST_DATEI =
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
30 Pattern.compile("^\\s*WSTDatei\\s*:\\s*(.+)");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
31
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
32 public static final Pattern BB_INFO =
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
33 Pattern.compile("^\\s*B\\+B-Info\\s*:\\s*(.+)");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
34
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
35 protected ArrayList<ImportRiver> rivers;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
36
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
37 public InfoGewParser() {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
38 rivers = new ArrayList<ImportRiver>();
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
39 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
40
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
41 public static final String normalize(String f) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
42 return f.replace("\\", "/").replace("/", File.separator);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
43 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
44
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
45 public void parse(File file) throws IOException {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
46
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
47 LineNumberReader in = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
48
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
49 File root = file.getParentFile();
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
50
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
51 ImportRiver importRiver = new ImportRiver();
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
52 try {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
53 in =
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
54 new LineNumberReader(
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
55 new InputStreamReader(
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
56 new FileInputStream(file), ENCODING));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
57
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
58 String line = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
59
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
60 String riverName = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
61 File wstFile = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
62 File bbInfoFile = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
63
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
64 while ((line = in.readLine()) != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
65 if ((line = line.trim()).length() == 0) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
66 continue;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
67 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
68 Matcher m = GEWAESSER.matcher(line);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
69
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
70 if (m.matches()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
71 String river = m.group(1);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
72 log.info("Found river '" + river + "'");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
73 if (riverName != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
74 rivers.add(new ImportRiver(riverName, wstFile, bbInfoFile));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
75 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
76 riverName = river;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
77 wstFile = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
78 bbInfoFile = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
79 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
80 else if ((m = WST_DATEI.matcher(line)).matches()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
81 String wstFilename = m.group(1);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
82 File wst = new File(wstFilename = normalize(wstFilename));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
83 if (!wst.isAbsolute()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
84 wst = new File(root, wstFilename);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
85 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
86 wst = FileTools.repair(wst);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
87 log.info("Found wst file '" + wst + "'");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
88 if (!wst.isFile() || !wst.canRead()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
89 log.warn("cannot access WST file '" + wstFilename + "'");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
90 continue;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
91 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
92 wstFile = wst;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
93 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
94 else if ((m = BB_INFO.matcher(line)).matches()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
95 //TODO: Make it relative to the wst file.
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
96 String bbInfo = m.group(1);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
97 bbInfoFile = new File(normalize(bbInfo));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
98 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
99 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
100 if (riverName != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
101 rivers.add(new ImportRiver(riverName, wstFile, bbInfoFile));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
102 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
103 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
104 finally {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
105 if (in != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
106 in.close();
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
107 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
108 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
109 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
110 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
111 // vim:set ts=4 sw=4 si et sta sts=4 fenc=utf8 :
http://dive4elements.wald.intevation.org