annotate flys-backend/src/main/java/de/intevation/flys/importer/parsers/InfoGewParser.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
rev   line source
1211
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
1 package de.intevation.flys.importer.parsers;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
2
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
3 import java.io.File;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
4
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
5 import java.util.List;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
6 import java.util.ArrayList;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
7
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
8 import java.util.regex.Pattern;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
9 import java.util.regex.Matcher;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
10
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
11 import java.io.IOException;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
12 import java.io.LineNumberReader;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
13 import java.io.FileInputStream;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
14 import java.io.InputStreamReader;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
15
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
16 import org.apache.log4j.Logger;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
17
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
18 import de.intevation.flys.utils.FileTools;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
19
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
20 import de.intevation.flys.importer.ImportRiver;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
21
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
22 public class InfoGewParser
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
23 {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
24 private static Logger log = Logger.getLogger(InfoGewParser.class);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
25
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
26 public static final String ENCODING = "ISO-8859-1";
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
27
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
28 public static final Pattern GEWAESSER =
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
29 Pattern.compile("^\\s*Gew\u00e4sser\\s*:\\s*(.+)");
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
30
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
31 public static final Pattern WST_DATEI =
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
32 Pattern.compile("^\\s*WSTDatei\\s*:\\s*(.+)");
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
33
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
34 public static final Pattern BB_INFO =
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
35 Pattern.compile("^\\s*B\\+B-Info\\s*:\\s*(.+)");
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
36
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
37 protected ArrayList<ImportRiver> rivers;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
38
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
39 protected AnnotationClassifier annotationClassifier;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
40
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
41 public InfoGewParser() {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
42 this(null);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
43 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
44
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
45 public InfoGewParser(AnnotationClassifier annotationClassifier) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
46 rivers = new ArrayList<ImportRiver>();
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
47 this.annotationClassifier = annotationClassifier;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
48 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
49
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
50 public List<ImportRiver> getRivers() {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
51 return rivers;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
52 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
53
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
54 public static final String normalize(String f) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
55 return f.replace("\\", "/").replace("/", File.separator);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
56 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
57
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
58 public void parse(File file) throws IOException {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
59
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
60 LineNumberReader in = null;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
61
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
62 File root = file.getParentFile();
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
63
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
64 try {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
65 in =
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
66 new LineNumberReader(
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
67 new InputStreamReader(
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
68 new FileInputStream(file), ENCODING));
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
69
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
70 String line = null;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
71
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
72 String riverName = null;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
73 File wstFile = null;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
74 File bbInfoFile = null;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
75
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
76 while ((line = in.readLine()) != null) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
77 if ((line = line.trim()).length() == 0) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
78 continue;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
79 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
80 Matcher m = GEWAESSER.matcher(line);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
81
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
82 if (m.matches()) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
83 String river = m.group(1);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
84 log.info("Found river '" + river + "'");
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
85 if (riverName != null) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
86 rivers.add(new ImportRiver(
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
87 riverName,
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
88 wstFile,
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
89 bbInfoFile,
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
90 annotationClassifier));
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
91 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
92 riverName = river;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
93 wstFile = null;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
94 bbInfoFile = null;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
95 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
96 else if ((m = WST_DATEI.matcher(line)).matches()) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
97 String wstFilename = m.group(1);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
98 File wst = new File(wstFilename = normalize(wstFilename));
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
99 if (!wst.isAbsolute()) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
100 wst = new File(root, wstFilename);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
101 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
102 wst = FileTools.repair(wst);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
103 log.info("Found wst file '" + wst + "'");
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
104 if (!wst.isFile() || !wst.canRead()) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
105 log.warn("cannot access WST file '" + wstFilename + "'");
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
106 continue;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
107 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
108 wstFile = wst;
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
109 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
110 else if ((m = BB_INFO.matcher(line)).matches()) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
111 //TODO: Make it relative to the wst file.
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
112 String bbInfo = m.group(1);
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
113 bbInfoFile = new File(normalize(bbInfo));
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
114 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
115 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
116 if (riverName != null) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
117 rivers.add(new ImportRiver(
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
118 riverName,
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
119 wstFile,
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
120 bbInfoFile,
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
121 annotationClassifier));
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
122 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
123 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
124 finally {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
125 if (in != null) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
126 in.close();
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
127 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
128 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
129
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
130 for (ImportRiver river: rivers) {
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
131 river.parseDependencies();
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
132 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
133 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
134 }
f08fe480092c Moved file parsers to separate package.
Sascha L. Teichmann <sascha.teichmann@intevation.de>
parents:
diff changeset
135 // vim:set ts=4 sw=4 si et sta sts=4 fenc=utf8 :

http://dive4elements.wald.intevation.org