1 |
1831
|
tao
|
/**
|
2 |
|
|
* '$RCSfile$'
|
3 |
|
|
* Purpose: A Class that represents a structured query, and can be
|
4 |
|
|
* constructed from an XML serialization conforming to
|
5 |
|
|
* pathquery.dtd. The printSQL() method can be used to print
|
6 |
|
|
* a SQL serialization of the query.
|
7 |
|
|
* Copyright: 2000 Regents of the University of California and the
|
8 |
|
|
* National Center for Ecological Analysis and Synthesis
|
9 |
|
|
* Authors: Matt Jones
|
10 |
|
|
* Release: @release@
|
11 |
|
|
*
|
12 |
|
|
* '$Author$'
|
13 |
|
|
* '$Date$'
|
14 |
|
|
* '$Revision$'
|
15 |
|
|
*
|
16 |
|
|
* This program is free software; you can redistribute it and/or modify
|
17 |
|
|
* it under the terms of the GNU General Public License as published by
|
18 |
|
|
* the Free Software Foundation; either version 2 of the License, or
|
19 |
|
|
* (at your option) any later version.
|
20 |
|
|
*
|
21 |
|
|
* This program is distributed in the hope that it will be useful,
|
22 |
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
23 |
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
24 |
|
|
* GNU General Public License for more details.
|
25 |
|
|
*
|
26 |
|
|
* You should have received a copy of the GNU General Public License
|
27 |
|
|
* along with this program; if not, write to the Free Software
|
28 |
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
29 |
|
|
*/
|
30 |
|
|
|
31 |
|
|
package edu.ucsb.nceas.metacat;
|
32 |
|
|
|
33 |
|
|
import java.util.Vector;
|
34 |
|
|
|
35 |
|
|
/** a utility class that represents a single term in a query */
|
36 |
2068
|
jones
|
public class QueryTerm
|
37 |
|
|
{
|
38 |
|
|
|
39 |
1831
|
tao
|
private boolean casesensitive = false;
|
40 |
2068
|
jones
|
|
41 |
1831
|
tao
|
private String searchmode = null;
|
42 |
2068
|
jones
|
|
43 |
1831
|
tao
|
private String value = null;
|
44 |
2068
|
jones
|
|
45 |
1831
|
tao
|
private String pathexpr = null;
|
46 |
2068
|
jones
|
|
47 |
1831
|
tao
|
private boolean percentageSymbol = false;
|
48 |
2068
|
jones
|
|
49 |
1831
|
tao
|
private int countPercentageSearchItem = 0;
|
50 |
|
|
|
51 |
|
|
/**
|
52 |
2068
|
jones
|
* Construct a new instance of a query term for a free text search (using
|
53 |
|
|
* the value only)
|
54 |
|
|
*
|
55 |
|
|
* @param casesensitive
|
56 |
|
|
* flag indicating whether case is used to match
|
57 |
|
|
* @param searchmode
|
58 |
|
|
* determines what kind of substring match is performed (one of
|
59 |
|
|
* starts-with|ends-with|contains|matches-exactly)
|
60 |
|
|
* @param value
|
61 |
|
|
* the text value to match
|
62 |
1831
|
tao
|
*/
|
63 |
2068
|
jones
|
public QueryTerm(boolean casesensitive, String searchmode, String value)
|
64 |
|
|
{
|
65 |
|
|
this.casesensitive = casesensitive;
|
66 |
|
|
this.searchmode = searchmode;
|
67 |
|
|
this.value = value;
|
68 |
1831
|
tao
|
}
|
69 |
|
|
|
70 |
|
|
/**
|
71 |
|
|
* Construct a new instance of a query term for a structured search
|
72 |
|
|
* (matching the value only for those nodes in the pathexpr)
|
73 |
2068
|
jones
|
*
|
74 |
|
|
* @param casesensitive
|
75 |
|
|
* flag indicating whether case is used to match
|
76 |
|
|
* @param searchmode
|
77 |
|
|
* determines what kind of substring match is performed (one of
|
78 |
|
|
* starts-with|ends-with|contains|matches-exactly)
|
79 |
|
|
* @param value
|
80 |
|
|
* the text value to match
|
81 |
|
|
* @param pathexpr
|
82 |
|
|
* the hierarchical path to the nodes to be searched
|
83 |
1831
|
tao
|
*/
|
84 |
2068
|
jones
|
public QueryTerm(boolean casesensitive, String searchmode, String value,
|
85 |
|
|
String pathexpr)
|
86 |
|
|
{
|
87 |
|
|
this(casesensitive, searchmode, value);
|
88 |
|
|
this.pathexpr = pathexpr;
|
89 |
1831
|
tao
|
}
|
90 |
|
|
|
91 |
|
|
/** determine if the QueryTerm is case sensitive */
|
92 |
2068
|
jones
|
public boolean isCaseSensitive()
|
93 |
|
|
{
|
94 |
|
|
return casesensitive;
|
95 |
1831
|
tao
|
}
|
96 |
|
|
|
97 |
|
|
/** get the searchmode parameter */
|
98 |
2068
|
jones
|
public String getSearchMode()
|
99 |
|
|
{
|
100 |
|
|
return searchmode;
|
101 |
1831
|
tao
|
}
|
102 |
2068
|
jones
|
|
103 |
1831
|
tao
|
/** get the Value parameter */
|
104 |
2068
|
jones
|
public String getValue()
|
105 |
|
|
{
|
106 |
|
|
return value;
|
107 |
1831
|
tao
|
}
|
108 |
|
|
|
109 |
|
|
/** get the path expression parameter */
|
110 |
2068
|
jones
|
public String getPathExpression()
|
111 |
|
|
{
|
112 |
|
|
return pathexpr;
|
113 |
1831
|
tao
|
}
|
114 |
2068
|
jones
|
|
115 |
|
|
/** get the percentage count for one query term */
|
116 |
1831
|
tao
|
public int getPercentageSymbolCount()
|
117 |
|
|
{
|
118 |
2068
|
jones
|
return countPercentageSearchItem;
|
119 |
1831
|
tao
|
}
|
120 |
|
|
|
121 |
|
|
/**
|
122 |
|
|
* create a SQL serialization of the query that this instance represents
|
123 |
|
|
*/
|
124 |
2068
|
jones
|
public String printSQL(boolean useXMLIndex)
|
125 |
|
|
{
|
126 |
|
|
StringBuffer self = new StringBuffer();
|
127 |
1831
|
tao
|
|
128 |
2068
|
jones
|
// Uppercase the search string if case match is not important
|
129 |
|
|
String casevalue = null;
|
130 |
|
|
String nodedataterm = null;
|
131 |
1831
|
tao
|
|
132 |
2068
|
jones
|
if (casesensitive) {
|
133 |
|
|
nodedataterm = "nodedata";
|
134 |
|
|
casevalue = value;
|
135 |
|
|
} else {
|
136 |
|
|
nodedataterm = "UPPER(nodedata)";
|
137 |
|
|
casevalue = value.toUpperCase();
|
138 |
|
|
}
|
139 |
1831
|
tao
|
|
140 |
2068
|
jones
|
// Add appropriate wildcards to search string
|
141 |
|
|
String searchexpr = null;
|
142 |
|
|
if (searchmode.equals("starts-with")) {
|
143 |
|
|
searchexpr = nodedataterm + " LIKE '" + casevalue + "%' ";
|
144 |
|
|
} else if (searchmode.equals("ends-with")) {
|
145 |
|
|
searchexpr = nodedataterm + " LIKE '%" + casevalue + "' ";
|
146 |
|
|
} else if (searchmode.equals("contains")) {
|
147 |
|
|
if (!casevalue.equals("%")) {
|
148 |
|
|
searchexpr = nodedataterm + " LIKE '%" + casevalue + "%' ";
|
149 |
|
|
} else {
|
150 |
|
|
searchexpr = nodedataterm + " LIKE '" + casevalue + "' ";
|
151 |
|
|
// find percentage symbol
|
152 |
|
|
percentageSymbol = true;
|
153 |
|
|
}
|
154 |
|
|
} else if (searchmode.equals("not-contains")) {
|
155 |
|
|
searchexpr = nodedataterm + " NOT LIKE '%" + casevalue + "%' ";
|
156 |
|
|
} else if (searchmode.equals("equals")) {
|
157 |
|
|
searchexpr = nodedataterm + " = '" + casevalue + "' ";
|
158 |
|
|
} else if (searchmode.equals("isnot-equal")) {
|
159 |
|
|
searchexpr = nodedataterm + " != '" + casevalue + "' ";
|
160 |
1831
|
tao
|
} else {
|
161 |
2068
|
jones
|
String oper = null;
|
162 |
|
|
if (searchmode.equals("greater-than")) {
|
163 |
|
|
oper = ">";
|
164 |
|
|
nodedataterm = "nodedata";
|
165 |
|
|
} else if (searchmode.equals("greater-than-equals")) {
|
166 |
|
|
oper = ">=";
|
167 |
|
|
nodedataterm = "nodedata";
|
168 |
|
|
} else if (searchmode.equals("less-than")) {
|
169 |
|
|
oper = "<";
|
170 |
|
|
nodedataterm = "nodedata";
|
171 |
|
|
} else if (searchmode.equals("less-than-equals")) {
|
172 |
|
|
oper = "<=";
|
173 |
|
|
nodedataterm = "nodedata";
|
174 |
|
|
} else {
|
175 |
|
|
System.out
|
176 |
|
|
.println("NOT expected case. NOT recognized operator: "
|
177 |
|
|
+ searchmode);
|
178 |
|
|
return null;
|
179 |
|
|
}
|
180 |
|
|
try {
|
181 |
|
|
// it is number; numeric comparison
|
182 |
|
|
// but we need to make sure there is no string in node data
|
183 |
|
|
String getRidOfString = " AND UPPER(nodedata) = LOWER(nodedata)"
|
184 |
|
|
+ " AND LTRIM(nodedata) != ' ' "
|
185 |
|
|
+ " AND nodedata IS NOT NULL ";
|
186 |
|
|
searchexpr = nodedataterm + " " + oper + " "
|
187 |
|
|
+ new Double(casevalue) + " " + getRidOfString;
|
188 |
|
|
} catch (NumberFormatException nfe) {
|
189 |
|
|
// these are characters; character comparison
|
190 |
|
|
searchexpr = nodedataterm + " " + oper + " '" + casevalue
|
191 |
|
|
+ "' ";
|
192 |
|
|
}
|
193 |
1831
|
tao
|
}
|
194 |
|
|
|
195 |
2068
|
jones
|
self.append("SELECT DISTINCT docid FROM xml_nodes WHERE \n");
|
196 |
|
|
self.append(searchexpr);
|
197 |
|
|
if (pathexpr != null) {
|
198 |
|
|
|
199 |
|
|
// use XML Index
|
200 |
|
|
if (useXMLIndex) {
|
201 |
|
|
if (!hasAttributeInPath(pathexpr)) {
|
202 |
|
|
// without attributes in path
|
203 |
|
|
self.append("AND parentnodeid IN ");
|
204 |
|
|
} else {
|
205 |
|
|
// has a attribute in path
|
206 |
|
|
String attributeName = QuerySpecification
|
207 |
|
|
.getAttributeName(pathexpr);
|
208 |
|
|
self.append("AND nodetype LIKE 'ATTRIBUTE' AND nodename LIKE '"
|
209 |
|
|
+ attributeName + "' ");
|
210 |
|
|
self.append("AND parentnodeid IN ");
|
211 |
|
|
pathexpr = QuerySpecification
|
212 |
|
|
.newPathExpressionWithOutAttribute(pathexpr);
|
213 |
|
|
|
214 |
|
|
}
|
215 |
|
|
self.append("(SELECT nodeid FROM xml_index WHERE path LIKE "
|
216 |
|
|
+ "'" + pathexpr + "') ");
|
217 |
|
|
} else {
|
218 |
|
|
// without using XML Index; using nested statements instead
|
219 |
|
|
self.append("AND parentnodeid IN ");
|
220 |
|
|
self.append(useNestedStatements(pathexpr));
|
221 |
|
|
}
|
222 |
|
|
} else if ((value.trim()).equals("%")) {
|
223 |
|
|
//if pathexpr is null and search value is %, is a
|
224 |
|
|
// percentageSearchItem
|
225 |
|
|
// the count number will be increase one
|
226 |
|
|
countPercentageSearchItem++;
|
227 |
|
|
|
228 |
1831
|
tao
|
}
|
229 |
|
|
|
230 |
2068
|
jones
|
return self.toString();
|
231 |
1831
|
tao
|
}
|
232 |
2068
|
jones
|
|
233 |
|
|
/** A method to judge if a path have attribute */
|
234 |
1831
|
tao
|
private boolean hasAttributeInPath(String path)
|
235 |
|
|
{
|
236 |
2068
|
jones
|
if (path.indexOf(QuerySpecification.ATTRIBUTESYMBOL) != -1) {
|
237 |
|
|
return true;
|
238 |
|
|
} else {
|
239 |
|
|
return false;
|
240 |
|
|
}
|
241 |
1831
|
tao
|
}
|
242 |
2068
|
jones
|
|
243 |
|
|
/*
|
244 |
|
|
* Constraint the query with @pathexp without using the XML Index, but
|
245 |
|
|
* nested SQL statements instead. The query migth be slower.
|
246 |
1831
|
tao
|
*/
|
247 |
2069
|
jones
|
public static String useNestedStatements(String pathexpr)
|
248 |
1831
|
tao
|
{
|
249 |
2068
|
jones
|
StringBuffer nestedStmts = new StringBuffer();
|
250 |
|
|
Vector nodes = new Vector();
|
251 |
|
|
String path = pathexpr;
|
252 |
|
|
int inx = 0;
|
253 |
1831
|
tao
|
|
254 |
2068
|
jones
|
do {
|
255 |
|
|
inx = path.lastIndexOf("/");
|
256 |
1831
|
tao
|
|
257 |
2068
|
jones
|
nodes.addElement(path.substring(inx + 1));
|
258 |
|
|
path = path.substring(0, Math.abs(inx));
|
259 |
|
|
} while (inx > 0);
|
260 |
1831
|
tao
|
|
261 |
2068
|
jones
|
// nested statements
|
262 |
|
|
int i = 0;
|
263 |
|
|
for (i = 0; i < nodes.size() - 1; i++) {
|
264 |
|
|
nestedStmts.append("(SELECT nodeid FROM xml_nodes"
|
265 |
|
|
+ " WHERE nodename LIKE '" + (String) nodes.elementAt(i)
|
266 |
|
|
+ "'" + " AND parentnodeid IN ");
|
267 |
|
|
}
|
268 |
|
|
// for the last statement: it is without " AND parentnodeid IN "
|
269 |
|
|
nestedStmts.append("(SELECT nodeid FROM xml_nodes"
|
270 |
|
|
+ " WHERE nodename LIKE '" + (String) nodes.elementAt(i) + "'");
|
271 |
|
|
// node.size() number of closing brackets
|
272 |
|
|
for (i = 0; i < nodes.size(); i++) {
|
273 |
|
|
nestedStmts.append(")");
|
274 |
|
|
}
|
275 |
1831
|
tao
|
|
276 |
2068
|
jones
|
return nestedStmts.toString();
|
277 |
1831
|
tao
|
}
|
278 |
|
|
|
279 |
|
|
/**
|
280 |
|
|
* create a String description of the query that this instance represents.
|
281 |
|
|
* This should become a way to get the XML serialization of the query.
|
282 |
|
|
*/
|
283 |
2068
|
jones
|
public String toString()
|
284 |
|
|
{
|
285 |
1831
|
tao
|
|
286 |
2068
|
jones
|
return this.printSQL(true);
|
287 |
1831
|
tao
|
}
|
288 |
2068
|
jones
|
}
|