Mercurial > dive4elements > river

annotate flys-backend/src/main/java/de/intevation/flys/importer/InfoGewParser.java @ 485:6b231041dc18

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Importer: Try to extract time ranges from at files. flys-backend/trunk@1811 c6561f87-3c4e-4783-a992-168aeb5c3f6f
author Sascha L. Teichmann <sascha.teichmann@intevation.de>
date Tue, 03 May 2011 17:09:38 +0000
parents cf8cbcb6a10d
children 763c4137d6e1
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
180
3908bfb2aa43 Store imported rivers into database. Needs testing!
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 177
diff changeset
41 public List<ImportRiver> getRivers() {
3908bfb2aa43 Store imported rivers into database. Needs testing!
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 177
diff changeset
42 return rivers;
3908bfb2aa43 Store imported rivers into database. Needs testing!
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 177
diff changeset
43 }
3908bfb2aa43 Store imported rivers into database. Needs testing!
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 177
diff changeset
44
177
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
45 public static final String normalize(String f) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
46 return f.replace("\\", "/").replace("/", File.separator);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
47 }
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 public void parse(File file) throws IOException {
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 LineNumberReader in = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
52
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
53 File root = file.getParentFile();
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
54
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
55 ImportRiver importRiver = new ImportRiver();
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
56 try {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
57 in =
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
58 new LineNumberReader(
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
59 new InputStreamReader(
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
60 new FileInputStream(file), ENCODING));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
61
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
62 String line = 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 String riverName = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
65 File wstFile = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
66 File bbInfoFile = null;
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 while ((line = in.readLine()) != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
69 if ((line = line.trim()).length() == 0) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
70 continue;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
71 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
72 Matcher m = GEWAESSER.matcher(line);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
73
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
74 if (m.matches()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
75 String river = m.group(1);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
76 log.info("Found river '" + river + "'");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
77 if (riverName != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
78 rivers.add(new ImportRiver(riverName, wstFile, bbInfoFile));
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 riverName = river;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
81 wstFile = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
82 bbInfoFile = null;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
83 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
84 else if ((m = WST_DATEI.matcher(line)).matches()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
85 String wstFilename = m.group(1);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
86 File wst = new File(wstFilename = normalize(wstFilename));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
87 if (!wst.isAbsolute()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
88 wst = new File(root, wstFilename);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
89 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
90 wst = FileTools.repair(wst);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
91 log.info("Found wst file '" + wst + "'");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
92 if (!wst.isFile() || !wst.canRead()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
93 log.warn("cannot access WST file '" + wstFilename + "'");
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
94 continue;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
95 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
96 wstFile = wst;
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
97 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
98 else if ((m = BB_INFO.matcher(line)).matches()) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
99 //TODO: Make it relative to the wst file.
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
100 String bbInfo = m.group(1);
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
101 bbInfoFile = new File(normalize(bbInfo));
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 if (riverName != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
105 rivers.add(new ImportRiver(riverName, wstFile, bbInfoFile));
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
106 }
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 finally {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
109 if (in != null) {
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
110 in.close();
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
111 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
112 }
184
4ab2c3bd474c Added parsing of PEGEL.GLT files.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 180
diff changeset
113
4ab2c3bd474c Added parsing of PEGEL.GLT files.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 180
diff changeset
114 for (ImportRiver river: rivers) {
186
cf8cbcb6a10d Added parser to read *.KM files.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 184
diff changeset
115 river.parseDependencies();
184
4ab2c3bd474c Added parsing of PEGEL.GLT files.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents: 180
diff changeset
116 }
177
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
117 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
118 }
31895d24387e Importer: Added info gew parser.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
119 // vim:set ts=4 sw=4 si et sta sts=4 fenc=utf8 :
http://dive4elements.wald.intevation.org