Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

SOLR-2606: Fixed sort parsing of fields containing punctuation 

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1152653 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Chris M. Hostetter 2011-07-31 22:08:43 +00:00
parent 865098ae7e
commit 2b7362de2b
5 changed files with 133 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
lucene/src/test-framework/org/apache/lucene/util/_TestUtil.java
View File

@ -281,14 +281,19 @@ public class _TestUtil {
0x2A6DF, 0x2B73F, 0x2FA1F, 0xE007F, 0xE01EF, 0xFFFFF, 0x10FFFF
};
/** Returns random string, all codepoints within the same unicode block. */
/** Returns random string of length between 0-20 codepoints, all codepoints within the same unicode block. */
public static String randomRealisticUnicodeString(Random r) {
return randomRealisticUnicodeString(r, 20);
}
/** Returns random string, all codepoints within the same unicode block. */
/** Returns random string of length up to maxLength codepoints , all codepoints within the same unicode block. */
public static String randomRealisticUnicodeString(Random r, int maxLength) {
final int end = r.nextInt(maxLength);
return randomRealisticUnicodeString(r, 0, 20);
}
/** Returns random string of length between min and max codepoints, all codepoints within the same unicode block. */
public static String randomRealisticUnicodeString(Random r, int minLength, int maxLength) {
final int end = minLength + r.nextInt(maxLength);
final int block = r.nextInt(blockStarts.length);
StringBuilder sb = new StringBuilder();
for (int i = 0; i < end; i++)

4
solr/CHANGES.txt
View File

@ -406,6 +406,10 @@ Bug Fixes
in servlet containers such as WebSphere that do not use a default list
(Jay R. Jaeger, hossman)
* SOLR-2606: Fixed sort parsing of fields containing punctuation that
failed due to sort by function changes introduced in SOLR-1297
(Mitsu Hadeishi, hossman)
Other Changes
----------------------

9
solr/core/src/java/org/apache/solr/search/QueryParsing.java
View File

@ -247,10 +247,11 @@ public class QueryParsing {
// short circuit test for a really simple field name
String field = sp.getId(null);
ParseException qParserException = null;
Exception qParserException = null;
if (field == null || sp.ch() != ' ') {
if (field == null || !Character.isWhitespace(sp.peekChar())) {
// let's try it as a function instead
field = null;
String funcStr = sp.val.substring(start);
QParser parser = QParser.getParser(funcStr, FunctionQParserPlugin.NAME, req);
@ -297,7 +298,9 @@ public class QueryParsing {
}
continue;
}
} catch (ParseException e) {
} catch (IOException ioe) {
throw ioe;
} catch (Exception e) {
// hang onto this in case the string isn't a full field name either
qParserException = e;
}

107
solr/core/src/test/org/apache/solr/search/TestSort.java
View File

@ -24,21 +24,28 @@ import org.apache.lucene.index.IndexReader.AtomicReaderContext;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.search.*;
import org.apache.lucene.search.SortField.Type;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
import org.apache.lucene.util.OpenBitSet;
import org.apache.solr.util.AbstractSolrTestCase;
import org.apache.lucene.util._TestUtil;
import org.apache.solr.request.SolrQueryRequest;
import org.apache.solr.SolrTestCaseJ4;
import org.junit.BeforeClass;
import java.io.IOException;
import java.util.*;
public class TestSort extends AbstractSolrTestCase {
@Override
public String getSchemaFile() { return null; }
@Override
public String getSolrConfigFile() { return null; }
public class TestSort extends SolrTestCaseJ4 {
@BeforeClass
public static void beforeClass() throws Exception {
initCore("solrconfig.xml","schema-minimal.xml");
}
Random r = random;
final Random r = random;
int ndocs = 77;
int iter = 50;
@ -57,6 +64,92 @@ public class TestSort extends AbstractSolrTestCase {
}
}
public void testRandomFieldNameSorts() throws Exception {
SolrQueryRequest req = lrf.makeRequest("q", "*:*");
final int iters = atLeast(5000);
int numberOfOddities = 0;
for (int i = 0; i < iters; i++) {
final StringBuilder input = new StringBuilder();
final String[] names = new String[_TestUtil.nextInt(r,1,10)];
final boolean[] reverse = new boolean[names.length];
for (int j = 0; j < names.length; j++) {
names[j] = _TestUtil.randomRealisticUnicodeString(r, 1, 20);
// reduce the likelyhood that the random str is a valid query or func
names[j] = names[j].replaceFirst("\\{","\\{\\{");
names[j] = names[j].replaceFirst("\\(","\\(\\(");
names[j] = names[j].replaceFirst("(\\\"|\\')","$1$1");
names[j] = names[j].replaceFirst("(\\d)","$1x");
// eliminate pesky problem chars
names[j] = names[j].replaceAll("\\p{Cntrl}|\\p{javaWhitespace}","");
if (0 == names[j].length()) {
numberOfOddities++;
// screw it, i'm taking my toys and going home
names[j] = "last_ditch_i_give_up";
}
reverse[j] = r.nextBoolean();
input.append(r.nextBoolean() ? " " : "");
input.append(names[j]);
input.append(" ");
input.append(reverse[j] ? "desc," : "asc,");
}
input.deleteCharAt(input.length()-1);
SortField[] sorts = null;
try {
sorts = QueryParsing.parseSort(input.toString(), req).getSort();
} catch (RuntimeException e) {
throw new RuntimeException("Failed to parse sort: " + input, e);
}
assertEquals("parsed sorts had unexpected size",
names.length, sorts.length);
for (int j = 0; j < names.length; j++) {
assertEquals("sorts["+j+"] had unexpected reverse: " + input,
reverse[j], sorts[j].getReverse());
final Type type = sorts[j].getType();
if (Type.SCORE.equals(type)) {
numberOfOddities++;
assertEquals("sorts["+j+"] is (unexpectedly) type score : " + input,
"score", names[j]);
} else if (Type.DOC.equals(type)) {
numberOfOddities++;
assertEquals("sorts["+j+"] is (unexpectedly) type doc : " + input,
"_docid_", names[j]);
} else if (Type.CUSTOM.equals(type) || Type.REWRITEABLE.equals(type)) {
numberOfOddities++;
// our orig string better be parsable as a func/query
QParser qp =
QParser.getParser(names[j], FunctionQParserPlugin.NAME, req);
try {
Query q = qp.getQuery();
assertNotNull("sorts["+j+"] had type " + type +
" but parsed to null func/query: " + input, q);
} catch (Exception e) {
assertNull("sorts["+j+"] had type " + type +
" but errored parsing as func/query: " + input, e);
}
} else {
assertEquals("sorts["+j+"] had unexpected field: " + input,
names[j], sorts[j].getField());
}
}
}
assertTrue("Over 0.2% oddities in test: " +
numberOfOddities + "/" + iters +
" have func/query parsing semenatics gotten broader?",
numberOfOddities < 0.002 * iters);
}
public void testSort() throws Exception {
Directory dir = new RAMDirectory();
Field f = new Field("f","0", Field.Store.NO, Field.Index.NOT_ANALYZED_NO_NORMS);

18
solr/core/src/test/org/apache/solr/search/function/TestFunctionQuery.java
View File

@ -434,9 +434,15 @@ public class TestFunctionQuery extends SolrTestCaseJ4 {
@Test
public void testSortByFunc() throws Exception {
assertU(adoc("id", "1", "const_s", "xx", "x_i", "100", "1_s", "a"));
assertU(adoc("id", "2", "const_s", "xx", "x_i", "300", "1_s", "c"));
assertU(adoc("id", "3", "const_s", "xx", "x_i", "200", "1_s", "b"));
assertU(adoc("id", "1", "const_s", "xx",
"x_i", "100", "1_s", "a",
"x:x_i", "100", "1-1_s", "a"));
assertU(adoc("id", "2", "const_s", "xx",
"x_i", "300", "1_s", "c",
"x:x_i", "300", "1-1_s", "c"));
assertU(adoc("id", "3", "const_s", "xx",
"x_i", "200", "1_s", "b",
"x:x_i", "200", "1-1_s", "b"));
assertU(commit());
String desc = "/response/docs==[{'x_i':300},{'x_i':200},{'x_i':100}]";
@ -494,6 +500,12 @@ public class TestFunctionQuery extends SolrTestCaseJ4 {
assertJQ(req("q",q, "fl","x_i", "sort", "1_s asc")
,asc
);
assertJQ(req("q",q, "fl","x_i", "sort", "x:x_i desc")
,desc
);
assertJQ(req("q",q, "fl","x_i", "sort", "1-1_s asc")
,asc
);
// really ugly field name that isn't a java Id, and can't be
// parsed as a func, but sorted fine in Solr 1.4