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

LUCENE-1755: Fix WriteLineDocTask to output a document if it contains either a title or body (or both). 

git-svn-id: https://svn.apache.org/repos/asf/lucene/java/trunk@795792 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Mark Robert Miller 2009-07-20 12:19:06 +00:00
parent 63402f49c7
commit 3adc61c3ac
3 changed files with 85 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
contrib/benchmark/CHANGES.txt
View File

@ -4,6 +4,10 @@ The Benchmark contrib package contains code for benchmarking Lucene in a variety
$Id:$
7/20/2009
LUCENE-1755: Fix WriteLineDocTask to output a document if it contains either
a title or body (or both). (Shai Erera via Mark Miller)
7/14/2009
LUCENE-1725: Fix the example Sort algorithm - auto is now deprecated and no longer works
with Benchmark. Benchmark will now throw an exception if you specify sort fields without

9
contrib/benchmark/src/java/org/apache/lucene/benchmark/byTask/tasks/WriteLineDocTask.java
View File

@ -97,11 +97,12 @@ public class WriteLineDocTask extends PerfTask {
Document doc = docSize > 0 ? docMaker.makeDocument(docSize) : docMaker.makeDocument();
Field f = doc.getField(DocMaker.BODY_FIELD);
String body = f != null ? NORMALIZER.reset(f.stringValue()).replaceAll(" ") : null;
String body = f != null ? NORMALIZER.reset(f.stringValue()).replaceAll(" ") : "";
if (body != null) {
f = doc.getField(DocMaker.TITLE_FIELD);
String title = f != null ? NORMALIZER.reset(f.stringValue()).replaceAll(" ") : "";
f = doc.getField(DocMaker.TITLE_FIELD);
String title = f != null ? NORMALIZER.reset(f.stringValue()).replaceAll(" ") : "";
if (body.length() > 0 || title.length() > 0) {
f = doc.getField(DocMaker.DATE_FIELD);
String date = f != null ? NORMALIZER.reset(f.stringValue()).replaceAll(" ") : "";

80
contrib/benchmark/src/test/org/apache/lucene/benchmark/byTask/tasks/WriteLineDocTaskTest.java
View File

@ -39,7 +39,7 @@ public class WriteLineDocTaskTest extends BenchmarkTestCase {
// class has to be public so that Class.forName.newInstance() will work
public static final class WriteLineDocMaker extends DocMaker {
public Document makeDocument() throws Exception {
Document doc = new Document();
doc.add(new Field(BODY_FIELD, "body", Store.NO, Index.NOT_ANALYZED_NO_NORMS));
@ -52,7 +52,7 @@ public class WriteLineDocTaskTest extends BenchmarkTestCase {
// class has to be public so that Class.forName.newInstance() will work
public static final class NewLinesDocMaker extends DocMaker {
public Document makeDocument() throws Exception {
Document doc = new Document();
doc.add(new Field(BODY_FIELD, "body\r\ntext\ttwo", Store.NO, Index.NOT_ANALYZED_NO_NORMS));
@ -63,6 +63,35 @@ public class WriteLineDocTaskTest extends BenchmarkTestCase {
}
// class has to be public so that Class.forName.newInstance() will work
public static final class NoBodyDocMaker extends DocMaker {
public Document makeDocument() throws Exception {
Document doc = new Document();
doc.add(new Field(TITLE_FIELD, "title", Store.NO, Index.NOT_ANALYZED_NO_NORMS));
doc.add(new Field(DATE_FIELD, "date", Store.NO, Index.NOT_ANALYZED_NO_NORMS));
return doc;
}
}
// class has to be public so that Class.forName.newInstance() will work
public static final class NoTitleDocMaker extends DocMaker {
public Document makeDocument() throws Exception {
Document doc = new Document();
doc.add(new Field(BODY_FIELD, "body", Store.NO, Index.NOT_ANALYZED_NO_NORMS));
doc.add(new Field(DATE_FIELD, "date", Store.NO, Index.NOT_ANALYZED_NO_NORMS));
return doc;
}
}
// class has to be public so that Class.forName.newInstance() will work
public static final class JustDateDocMaker extends DocMaker {
public Document makeDocument() throws Exception {
Document doc = new Document();
doc.add(new Field(DATE_FIELD, "date", Store.NO, Index.NOT_ANALYZED_NO_NORMS));
return doc;
}
}
private static final CompressorStreamFactory csFactory = new CompressorStreamFactory();
private PerfRunData createPerfRunData(File file, boolean setBZCompress,
@ -90,10 +119,13 @@ public class WriteLineDocTaskTest extends BenchmarkTestCase {
String line = br.readLine();
assertNotNull(line);
String[] parts = line.split(Character.toString(WriteLineDocTask.SEP));
assertEquals(3, parts.length);
int numExpParts = expBody == null ? 2 : 3;
assertEquals(numExpParts, parts.length);
assertEquals(expTitle, parts[0]);
assertEquals(expDate, parts[1]);
assertEquals(expBody, parts[2]);
if (expBody != null) {
assertEquals(expBody, parts[2]);
}
assertNull(br.readLine());
} finally {
br.close();
@ -149,4 +181,44 @@ public class WriteLineDocTaskTest extends BenchmarkTestCase {
doReadTest(file, false, "title text", "date text", "body text two");
}
public void testEmptyBody() throws Exception {
// WriteLineDocTask threw away documents w/ no BODY element, even if they
// had a TITLE element (LUCENE-1755). It should throw away documents if they
// don't have BODY nor TITLE
File file = new File(getWorkDir(), "one-line");
PerfRunData runData = createPerfRunData(file, false, null, NoBodyDocMaker.class.getName());
WriteLineDocTask wldt = new WriteLineDocTask(runData);
wldt.doLogic();
wldt.close();
doReadTest(file, false, "title", "date", null);
}
public void testEmptyTitle() throws Exception {
File file = new File(getWorkDir(), "one-line");
PerfRunData runData = createPerfRunData(file, false, null, NoTitleDocMaker.class.getName());
WriteLineDocTask wldt = new WriteLineDocTask(runData);
wldt.doLogic();
wldt.close();
doReadTest(file, false, "", "date", "body");
}
public void testJustDate() throws Exception {
File file = new File(getWorkDir(), "one-line");
PerfRunData runData = createPerfRunData(file, false, null, JustDateDocMaker.class.getName());
WriteLineDocTask wldt = new WriteLineDocTask(runData);
wldt.doLogic();
wldt.close();
BufferedReader br = new BufferedReader(new InputStreamReader(new FileInputStream(file), "utf-8"));
try {
String line = br.readLine();
assertNull(line);
} finally {
br.close();
}
}
}