LUCENE-1479: if date is missing, don't skip the doc; just don't add 'docdate' field to the doc

git-svn-id: https://svn.apache.org/repos/asf/lucene/java/trunk@733697 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Michael McCandless 2009-01-12 11:37:23 +00:00
parent f0300a643e
commit 03d6b2aa24
2 changed files with 371 additions and 27 deletions

View File

@ -33,7 +33,6 @@ import java.util.zip.GZIPInputStream;
import org.apache.lucene.benchmark.byTask.utils.Config; import org.apache.lucene.benchmark.byTask.utils.Config;
/** /**
* A DocMaker using the (compressed) Trec collection for its input. * A DocMaker using the (compressed) Trec collection for its input.
* <p> * <p>
@ -44,7 +43,14 @@ import org.apache.lucene.benchmark.byTask.utils.Config;
*/ */
public class TrecDocMaker extends BasicDocMaker { public class TrecDocMaker extends BasicDocMaker {
private static final String newline = System.getProperty("line.separator"); private static final String DATE = "Date: ";
private static final String DOCHDR = "<DOCHDR>";
private static final String TERM_DOCHDR = "</DOCHDR>";
private static final String TERM_DOCNO = "</DOCNO>";
private static final String DOCNO = "<DOCNO>";
private static final String TERM_DOC = "</DOC>";
private static final String DOC = "<DOC>";
private static final String NEW_LINE = System.getProperty("line.separator");
protected ThreadLocal dateFormat = new ThreadLocal(); protected ThreadLocal dateFormat = new ThreadLocal();
protected File dataDir = null; protected File dataDir = null;
@ -135,25 +141,37 @@ public class TrecDocMaker extends BasicDocMaker {
} }
// read until finding a line that starts with the specified prefix // read until finding a line that starts with the specified prefix
protected StringBuffer read (String prefix, StringBuffer sb, boolean collectMatchLine, boolean collectAll) throws Exception { protected StringBuffer read(String prefix, StringBuffer sb,
boolean collectMatchLine, boolean collectAll,
String terminatingTag) throws Exception {
sb = (sb==null ? new StringBuffer() : sb); sb = (sb==null ? new StringBuffer() : sb);
String sep = ""; String sep = "";
while (true) { while (true) {
String line = reader.readLine(); String line = reader.readLine();
if (line==null) { if (line == null) {
openNextFile(); openNextFile();
continue; continue;
} }
if (line.startsWith(prefix)) { if (line.startsWith(prefix)) {
if (collectMatchLine) { if (collectMatchLine) {
sb.append(sep+line); sb.append(sep).append(line);
sep = newline; sep = NEW_LINE;
} }
break; break;
} }
if (terminatingTag != null && line.startsWith(terminatingTag)) {
// didn't find the prefix that was asked, but the terminating
// tag was found. set the length to 0 to signal no match was
// found.
sb.setLength(0);
break;
}
if (collectAll) { if (collectAll) {
sb.append(sep+line); sb.append(sep).append(line);
sep = newline; sep = NEW_LINE;
} }
} }
//System.out.println("read: "+sb); //System.out.println("read: "+sb);
@ -165,22 +183,31 @@ public class TrecDocMaker extends BasicDocMaker {
openNextFile(); openNextFile();
} }
// 1. skip until doc start // 1. skip until doc start
read("<DOC>",null,false,false); read(DOC,null,false,false,null);
// 2. name // 2. name
StringBuffer sb = read("<DOCNO>",null,true,false); StringBuffer sb = read(DOCNO,null,true,false,null);
String name = sb.substring("<DOCNO>".length()); String name = sb.substring(DOCNO.length(), sb.indexOf(TERM_DOCNO, DOCNO.length()));
name = name.substring(0,name.indexOf("</DOCNO>"))+"_"+iteration; name = name + "_" + iteration;
// 3. skip until doc header // 3. skip until doc header
read("<DOCHDR>",null,false,false); read(DOCHDR,null,false,false,null);
boolean findTerminatingDocHdr = false;
// 4. date // 4. date
sb = read("Date: ",null,true,false); sb = read(DATE,null,true,false,TERM_DOCHDR);
String dateStr = sb.substring("Date: ".length()); String dateStr = null;
if (sb.length() != 0) {
// Date found.
dateStr = sb.substring(DATE.length());
findTerminatingDocHdr = true;
}
// 5. skip until end of doc header // 5. skip until end of doc header
read("</DOCHDR>",null,false,false); if (findTerminatingDocHdr) {
read(TERM_DOCHDR,null,false,false,null);
}
// 6. collect until end of doc // 6. collect until end of doc
sb = read("</DOC>",null,false,true); sb = read(TERM_DOC,null,false,true,null);
// this is the next document, so parse it // this is the next document, so parse it
Date date = parseDate(dateStr); Date date = dateStr != null ? parseDate(dateStr) : null;
HTMLParser p = getHtmlParser(); HTMLParser p = getHtmlParser();
DocData docData = p.parse(name, date, sb, getDateFormat(0)); DocData docData = p.parse(name, date, sb, getDateFormat(0));
addBytes(sb.length()); // count char length of parsed html text (larger than the plain doc body text). addBytes(sb.length()); // count char length of parsed html text (larger than the plain doc body text).
@ -202,18 +229,14 @@ public class TrecDocMaker extends BasicDocMaker {
} }
protected Date parseDate(String dateStr) { protected Date parseDate(String dateStr) {
Date date = null; for (int i = 0; i < DATE_FORMATS.length; i++) {
for (int i=0; i<DATE_FORMATS.length; i++) {
try { try {
date = getDateFormat(i).parse(dateStr.trim()); return getDateFormat(i).parse(dateStr.trim());
return date; } catch (ParseException e) {}
} catch (ParseException e) {
}
} }
// do not fail test just because a date could not be parsed // do not fail test just because a date could not be parsed
System.out.println("ignoring date parse exception (assigning 'now') for: "+dateStr); System.out.println("ignoring date parse exception (assigning 'null') for: "+dateStr);
date = new Date(); // now return null;
return date;
} }

View File

@ -0,0 +1,321 @@
package org.apache.lucene.benchmark.byTask.feeds;
/**
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
import java.io.BufferedReader;
import java.io.StringReader;
import java.util.Date;
import junit.framework.TestCase;
public class TrecDocMakerTest extends TestCase {
/** A TrecDocMaker which works on a String and not files. */
private static class StringableTrecDocMaker extends TrecDocMaker {
private String docs = null;
public StringableTrecDocMaker(String docs, boolean forever) {
this.docs = docs;
this.forever = forever;
}
protected void openNextFile() throws NoMoreDataException, Exception {
if (reader != null) {
if (!forever) {
throw new NoMoreDataException();
}
++iteration;
}
reader = new BufferedReader(new StringReader(docs));
}
}
private void assertDocData(DocData dd, String expName, String expTitle, String expBody, Date expDate) {
assertNotNull(dd);
assertEquals(expName, dd.getName());
assertEquals(expTitle, dd.getTitle());
assertTrue(dd.getBody().indexOf(expBody) != -1);
assertEquals(expDate, dd.getDate());
}
private void assertNoMoreDataException(StringableTrecDocMaker stdm) throws Exception {
boolean thrown = false;
try {
stdm.getNextDocData();
} catch (NoMoreDataException e) {
thrown = true;
}
assertTrue("Expecting NoMoreDataException", thrown);
}
public void testOneDocument() throws Exception {
String docs = "<DOC>\r\n" +
"<DOCNO>TEST-000</DOCNO>\r\n" +
"<DOCHDR>\r\n" +
"http://lucene.apache.org.trecdocmaker.test\r\n" +
"HTTP/1.1 200 OK\r\n" +
"Date: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Server: Apache/1.3.27 (Unix)\r\n" +
"Last-Modified: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Content-Length: 614\r\n" +
"Connection: close\r\n" +
"Content-Type: text/html\r\n" +
"</DOCHDR>\r\n" +
"<html>\r\n" +
"\r\n" +
"<head>\r\n" +
"<title>\r\n" +
"TEST-000 title\r\n" +
"</title>\r\n" +
"</head>\r\n" +
"\r\n" +
"<body>\r\n" +
"TEST-000 text\r\n" +
"\r\n" +
"</body>\r\n" +
"\r\n" +
"</DOC>";
StringableTrecDocMaker stdm = new StringableTrecDocMaker(docs, false);
stdm.setHTMLParser(new DemoHTMLParser());
DocData dd = stdm.getNextDocData();
assertDocData(dd, "TEST-000_0", "TEST-000 title", "TEST-000 text", stdm
.parseDate("Sun, 11 Jan 2009 08:00:00 GMT"));
assertNoMoreDataException(stdm);
}
public void testTwoDocuments() throws Exception {
String docs = "<DOC>\r\n" +
"<DOCNO>TEST-000</DOCNO>\r\n" +
"<DOCHDR>\r\n" +
"http://lucene.apache.org.trecdocmaker.test\r\n" +
"HTTP/1.1 200 OK\r\n" +
"Date: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Server: Apache/1.3.27 (Unix)\r\n" +
"Last-Modified: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Content-Length: 614\r\n" +
"Connection: close\r\n" +
"Content-Type: text/html\r\n" +
"</DOCHDR>\r\n" +
"<html>\r\n" +
"\r\n" +
"<head>\r\n" +
"<title>\r\n" +
"TEST-000 title\r\n" +
"</title>\r\n" +
"</head>\r\n" +
"\r\n" +
"<body>\r\n" +
"TEST-000 text\r\n" +
"\r\n" +
"</body>\r\n" +
"\r\n" +
"</DOC>\r\n" +
"<DOC>\r\n" +
"<DOCNO>TEST-001</DOCNO>\r\n" +
"<DOCHDR>\r\n" +
"http://lucene.apache.org.trecdocmaker.test\r\n" +
"HTTP/1.1 200 OK\r\n" +
"Date: Sun, 11 Jan 2009 08:01:00 GMT\r\n" +
"Server: Apache/1.3.27 (Unix)\r\n" +
"Last-Modified: Sun, 11 Jan 2008 08:01:00 GMT\r\n" +
"Content-Length: 614\r\n" +
"Connection: close\r\n" +
"Content-Type: text/html\r\n" +
"</DOCHDR>\r\n" +
"<html>\r\n" +
"\r\n" +
"<head>\r\n" +
"<title>\r\n" +
"TEST-001 title\r\n" +
"</title>\r\n" +
"</head>\r\n" +
"\r\n" +
"<body>\r\n" +
"TEST-001 text\r\n" +
"\r\n" +
"</body>\r\n" +
"\r\n" +
"</DOC>";
StringableTrecDocMaker stdm = new StringableTrecDocMaker(docs, false);
stdm.setHTMLParser(new DemoHTMLParser());
DocData dd = stdm.getNextDocData();
assertDocData(dd, "TEST-000_0", "TEST-000 title", "TEST-000 text", stdm
.parseDate("Sun, 11 Jan 2009 08:00:00 GMT"));
dd = stdm.getNextDocData();
assertDocData(dd, "TEST-001_0", "TEST-001 title", "TEST-001 text", stdm
.parseDate("Sun, 11 Jan 2009 08:01:00 GMT"));
assertNoMoreDataException(stdm);
}
// If a Date: attribute is missing, make sure the document is not skipped, but
// rather that null Data is assigned.
public void testMissingDate() throws Exception {
String docs = "<DOC>\r\n" +
"<DOCNO>TEST-000</DOCNO>\r\n" +
"<DOCHDR>\r\n" +
"http://lucene.apache.org.trecdocmaker.test\r\n" +
"HTTP/1.1 200 OK\r\n" +
"Server: Apache/1.3.27 (Unix)\r\n" +
"Last-Modified: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Content-Length: 614\r\n" +
"Connection: close\r\n" +
"Content-Type: text/html\r\n" +
"</DOCHDR>\r\n" +
"<html>\r\n" +
"\r\n" +
"<head>\r\n" +
"<title>\r\n" +
"TEST-000 title\r\n" +
"</title>\r\n" +
"</head>\r\n" +
"\r\n" +
"<body>\r\n" +
"TEST-000 text\r\n" +
"\r\n" +
"</body>\r\n" +
"\r\n" +
"</DOC>\r\n" +
"<DOC>\r\n" +
"<DOCNO>TEST-001</DOCNO>\r\n" +
"<DOCHDR>\r\n" +
"http://lucene.apache.org.trecdocmaker.test\r\n" +
"HTTP/1.1 200 OK\r\n" +
"Date: Sun, 11 Jan 2009 08:01:00 GMT\r\n" +
"Server: Apache/1.3.27 (Unix)\r\n" +
"Last-Modified: Sun, 11 Jan 2009 08:01:00 GMT\r\n" +
"Content-Length: 614\r\n" +
"Connection: close\r\n" +
"Content-Type: text/html\r\n" +
"</DOCHDR>\r\n" +
"<html>\r\n" +
"\r\n" +
"<head>\r\n" +
"<title>\r\n" +
"TEST-001 title\r\n" +
"</title>\r\n" +
"</head>\r\n" +
"\r\n" +
"<body>\r\n" +
"TEST-001 text\r\n" +
"\r\n" +
"</body>\r\n" +
"\r\n" +
"</DOC>";
StringableTrecDocMaker stdm = new StringableTrecDocMaker(docs, false);
stdm.setHTMLParser(new DemoHTMLParser());
DocData dd = stdm.getNextDocData();
assertDocData(dd, "TEST-000_0", "TEST-000 title", "TEST-000 text", null);
dd = stdm.getNextDocData();
assertDocData(dd, "TEST-001_0", "TEST-001 title", "TEST-001 text", stdm
.parseDate("Sun, 11 Jan 2009 08:01:00 GMT"));
assertNoMoreDataException(stdm);
}
// When a 'bad date' is input (unparsable date), make sure the DocData date is
// assigned null.
public void testBadDate() throws Exception {
String docs = "<DOC>\r\n" +
"<DOCNO>TEST-000</DOCNO>\r\n" +
"<DOCHDR>\r\n" +
"http://lucene.apache.org.trecdocmaker.test\r\n" +
"HTTP/1.1 200 OK\r\n" +
"Date: Bad Date\r\n" +
"Server: Apache/1.3.27 (Unix)\r\n" +
"Last-Modified: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Content-Length: 614\r\n" +
"Connection: close\r\n" +
"Content-Type: text/html\r\n" +
"</DOCHDR>\r\n" +
"<html>\r\n" +
"\r\n" +
"<head>\r\n" +
"<title>\r\n" +
"TEST-000 title\r\n" +
"</title>\r\n" +
"</head>\r\n" +
"\r\n" +
"<body>\r\n" +
"TEST-000 text\r\n" +
"\r\n" +
"</body>\r\n" +
"\r\n" +
"</DOC>";
StringableTrecDocMaker stdm = new StringableTrecDocMaker(docs, false);
stdm.setHTMLParser(new DemoHTMLParser());
DocData dd = stdm.getNextDocData();
assertDocData(dd, "TEST-000_0", "TEST-000 title", "TEST-000 text", null);
assertNoMoreDataException(stdm);
}
public void testForever() throws Exception {
String docs = "<DOC>\r\n" +
"<DOCNO>TEST-000</DOCNO>\r\n" +
"<DOCHDR>\r\n" +
"http://lucene.apache.org.trecdocmaker.test\r\n" +
"HTTP/1.1 200 OK\r\n" +
"Date: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Server: Apache/1.3.27 (Unix)\r\n" +
"Last-Modified: Sun, 11 Jan 2009 08:00:00 GMT\r\n" +
"Content-Length: 614\r\n" +
"Connection: close\r\n" +
"Content-Type: text/html\r\n" +
"</DOCHDR>\r\n" +
"<html>\r\n" +
"\r\n" +
"<head>\r\n" +
"<title>\r\n" +
"TEST-000 title\r\n" +
"</title>\r\n" +
"</head>\r\n" +
"\r\n" +
"<body>\r\n" +
"TEST-000 text\r\n" +
"\r\n" +
"</body>\r\n" +
"\r\n" +
"</DOC>";
StringableTrecDocMaker stdm = new StringableTrecDocMaker(docs, true);
stdm.setHTMLParser(new DemoHTMLParser());
DocData dd = stdm.getNextDocData();
assertDocData(dd, "TEST-000_0", "TEST-000 title", "TEST-000 text", stdm
.parseDate("Sun, 11 Jan 2009 08:00:00 GMT"));
// same document, but the second iteration changes the name.
dd = stdm.getNextDocData();
assertDocData(dd, "TEST-000_1", "TEST-000 title", "TEST-000 text", stdm
.parseDate("Sun, 11 Jan 2009 08:00:00 GMT"));
// Don't test that NoMoreDataException is thrown, since the forever flag is
// turned on.
}
}