Update of /cvsroot/nutch/nutch/src/java/net/nutch/searcher
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv12441
Modified Files:
FetchedSegments.java
Log Message:
Open segment files lazily, so that if they're not used they're not required.
Index: FetchedSegments.java
===================================================================
RCS file: /cvsroot/nutch/nutch/src/java/net/nutch/searcher/FetchedSegments.java,v
retrieving revision 1.9
retrieving revision 1.10
diff -C2 -d -r1.9 -r1.10
*** FetchedSegments.java 7 Oct 2004 04:24:53 -0000 1.9
--- FetchedSegments.java 1 Nov 2004 20:21:59 -0000 1.10
***************
*** 23,26 ****
--- 23,29 ----
private static class Segment {
+ private NutchFileSystem nfs;
+ private File segmentDir;
+
private ArrayFile.Reader fetcher;
private ArrayFile.Reader content;
***************
*** 29,43 ****
public Segment(NutchFileSystem nfs, File segmentDir) throws IOException {
! this.fetcher = new ArrayFile.Reader
! (nfs, new File(segmentDir, FetcherOutput.DIR_NAME).toString());
! this.content = new ArrayFile.Reader
! (nfs, new File(segmentDir, Content.DIR_NAME).toString());
! this.text = new ArrayFile.Reader
! (nfs, new File(segmentDir, ParseText.DIR_NAME).toString());
! this.parsedata = new ArrayFile.Reader
! (nfs, new File(segmentDir, ParseData.DIR_NAME).toString());
}
public FetcherOutput getFetcherOutput(int docNo) throws IOException {
FetcherOutput entry = new FetcherOutput();
fetcher.get(docNo, entry);
--- 32,45 ----
public Segment(NutchFileSystem nfs, File segmentDir) throws IOException {
! this.nfs = nfs;
! this.segmentDir = segmentDir;
}
public FetcherOutput getFetcherOutput(int docNo) throws IOException {
+ if (fetcher == null) {
+ this.fetcher = new ArrayFile.Reader
+ (nfs, new File(segmentDir, FetcherOutput.DIR_NAME).toString());
+ }
+
FetcherOutput entry = new FetcherOutput();
fetcher.get(docNo, entry);
***************
*** 46,49 ****
--- 48,56 ----
public byte[] getContent(int docNo) throws IOException {
+ if (content == null) {
+ this.content = new ArrayFile.Reader
+ (nfs, new File(segmentDir, Content.DIR_NAME).toString());
+ }
+
Content entry = new Content();
content.get(docNo, entry);
***************
*** 52,66 ****
public ParseData getParseData(int docNo) throws IOException {
! ParseData entry = new ParseData();
! parsedata.get(docNo, entry);
! return entry;
}
public ParseText getParseText(int docNo) throws IOException {
! ParseText entry = new ParseText();
! text.get(docNo, entry);
! return entry;
}
}
--- 59,83 ----
public ParseData getParseData(int docNo) throws IOException {
! if (parsedata == null) {
! this.parsedata = new ArrayFile.Reader
! (nfs, new File(segmentDir, ParseData.DIR_NAME).toString());
}
+
+ ParseData entry = new ParseData();
+ parsedata.get(docNo, entry);
+ return entry;
+ }
public ParseText getParseText(int docNo) throws IOException {
! if (text == null) {
! this.text = new ArrayFile.Reader
! (nfs, new File(segmentDir, ParseText.DIR_NAME).toString());
}
+ ParseText entry = new ParseText();
+ text.get(docNo, entry);
+ return entry;
+ }
+
}
***************
*** 99,102 ****
--- 116,124 ----
}
+ public long getFetchDate(HitDetails details) throws IOException {
+ return getSegment(details).getFetcherOutput(getDocNo(details))
+ .getFetchDate();
+ }
+
public ParseText getParseText(HitDetails details) throws IOException {
return getSegment(details).getParseText(getDocNo(details));
-------------------------------------------------------
This SF.Net email is sponsored by:
Sybase ASE Linux Express Edition - download now for FREE
LinuxWorld Reader's Choice Award Winner for best database on Linux.
http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click
_______________________________________________
Nutch-cvs mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/nutch-cvs
