Mercurial > dive4elements > river

comparison flys-backend/src/main/java/de/intevation/flys/importer/AtFileParser.java @ 508:a9c7f6ec3a5a 2.3.1

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
merged flys-backend/2.3.1
author Thomas Arendsen Hein <thomas@intevation.de>
date Fri, 28 Sep 2012 12:14:12 +0200
parents ee693b8fbb55
children 23fe9ac1c3b4
comparison
equal deleted inserted replaced
462:ebf049a1eb53 508:a9c7f6ec3a5a
1 package de.intevation.flys.importer;
2
3 import java.io.BufferedReader;
4 import java.io.File;
5 import java.io.FileInputStream;
6 import java.io.InputStreamReader;
7 import java.io.IOException;
8 import java.math.BigDecimal;
9
10 import org.apache.log4j.Logger;
11
12 import de.intevation.flys.importer.ImportDischargeTable;
13 import de.intevation.flys.importer.ImportDischargeTableValue;
14
15 import java.util.regex.Pattern;
16 import java.util.regex.Matcher;
17
18 import java.util.Date;
19 import java.util.Calendar;
20
21 public class AtFileParser {
22
23 public static final String ENCODING = "ISO-8859-1";
24
25 private static Logger logger = Logger.getLogger(AtFileParser.class);
26
27
28 // regular expression from hell to find out time range
29 public static final Pattern DATE_LINE = Pattern.compile(
30 "^\\*\\s*Abflu[^t]+tafel?\\s*([^\\d]+)" +
31 "(\\d{1,2})?\\.?(\\d{1,2})?\\.?(\\d{2,4})\\s*(?:(?:bis)|-)?\\s*" +
32 "(?:(\\d{1,2})?\\.?(\\d{1,2})?\\.?(\\d{2,4}))?\\s*.*$");
33
34 public AtFileParser() {
35 }
36
37
38 public ImportDischargeTable parse(File file) throws IOException {
39 return parse(file, "", 0);
40 }
41
42 public ImportDischargeTable parse(
43 File file,
44 String prefix,
45 int kind
46 )
47 throws IOException {
48
49 logger.info("parsing AT file: " + file);
50
51 BufferedReader br = null;
52
53 String line = null;
54
55 boolean beginning = true;
56
57 ImportDischargeTable dischargeTable =
58 new ImportDischargeTable(kind, prefix + file.getName());
59
60 Date from = null;
61 Date to = null;
62
63 try {
64 br = new BufferedReader(
65 new InputStreamReader(
66 new FileInputStream(file), ENCODING));
67
68 while ((line = br.readLine()) != null) {
69
70 String tmp = line.trim();
71
72 if (tmp.length() == 0) {
73 continue;
74 }
75
76 Matcher m = DATE_LINE.matcher(tmp);
77 if (m.matches()) {
78 from = guessDate(m.group(2), m.group(3), m.group(4));
79 to = guessDate(m.group(5), m.group(6), m.group(7));
80 if (from == null) {
81 Date t = from; from = to; to = t;
82 }
83 continue;
84 }
85
86 if (tmp.startsWith("#! name=")) {
87 // XXX Skip the name, because we don't know where to save
88 // it at the moment
89
90 //String name = tmp.substring(8);
91 continue;
92 }
93
94 if (tmp.startsWith("#") || tmp.startsWith("*")) {
95 continue;
96 }
97
98 String[] splits = tmp.replace(',', '.').split("\\s+");
99
100 if ((splits.length < 2) || (splits.length > 11)) {
101 logger.warn("Found an invalid row in the AT file.");
102 continue;
103 }
104
105 String strW = splits[0].trim();
106 double W = Double.parseDouble(strW);
107
108 /* shift is used to differenciate between lines with
109 * exactly 10 Qs and lines with less than 10 Qs. The shift
110 * is only modified when it is the first line.
111 */
112 int shift = 0;
113
114 if (splits.length != 11 && beginning) {
115 shift = 11 - splits.length;
116 }
117
118
119 for (int i = 1; i < splits.length; i++) {
120 double iW = W + shift + i;
121 double iQ = Double.parseDouble(splits[i].trim());
122
123 dischargeTable.addDischargeTableValue(
124 new ImportDischargeTableValue(
125 new BigDecimal(iQ/100.0),
126 new BigDecimal(iW/100.0)));
127 }
128
129 beginning = false;
130 }
131 }
132 catch (NumberFormatException pe) {
133 logger.warn(pe.getMessage());
134 }
135 finally {
136 if (br != null) {
137 br.close();
138 }
139 }
140
141 if (from != null) {
142 if (to != null && from.compareTo(to) > 0) {
143 Date t = from; from = to; to = t;
144 }
145 logger.info("from: " + from + " to: " + to);
146 ImportTimeInterval interval = new ImportTimeInterval(from, to);
147 dischargeTable.setTimeInterval(interval);
148 }
149
150 logger.info("Finished parsing AT file: " + file);
151
152 return dischargeTable;
153 }
154
155 public static Date guessDate(String day, String month, String year) {
156 if (day == null && month == null && year == null) {
157 return null;
158 }
159
160 logger.debug("day: " + day + " month: " + month + " year: " + year);
161
162 int dayI = 15;
163 if (day != null) {
164 try {
165 dayI = Integer.parseInt(day.trim());
166 }
167 catch (NumberFormatException nfe) {
168 }
169 }
170
171 int monthI = 6;
172 if (month != null) {
173 try {
174 monthI = Integer.parseInt(month.trim());
175 }
176 catch (NumberFormatException nfe) {
177 }
178 }
179
180 int yearI = 1900;
181 if (year != null) {
182 try {
183 yearI = Integer.parseInt(year.trim());
184 if (yearI < 100) {
185 if (yearI < 20) {
186 yearI += 2000;
187 }
188 else {
189 yearI += 1900;
190 }
191 }
192 }
193 catch (NumberFormatException nfe) {
194 }
195 }
196
197 Calendar cal = Calendar.getInstance();
198 cal.set(yearI, monthI-1, dayI, 12, 0, 0);
199 long ms = cal.getTimeInMillis();
200 cal.setTimeInMillis(ms - ms%1000);
201 return cal.getTime();
202 }
203 }
204 // vim:set ts=4 sw=4 si et sta sts=4 fenc=utf8 :
http://dive4elements.wald.intevation.org