honeymoose
/
OpenSearch
mirror of https://github.com/honeymoose/OpenSearch.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Performance improvement when creating a multi-value field cache

This commit is contained in:
acerb 2011-08-26 23:15:26 +03:00 committed by Shay Banon
parent d880a6c1b0
commit 56a2833de4
1 changed files with 8 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
modules/elasticsearch/src/main/java/org/elasticsearch/index/field/data/support/FieldDataLoader.java
View File

@ -41,6 +41,7 @@ public class FieldDataLoader {
field = StringHelper.intern(field);
ArrayList<int[]> ordinals = new ArrayList<int[]>();
int[] idx = new int[reader.maxDoc()];
ordinals.add(new int[reader.maxDoc()]);
int t = 1; // current term number
@ -55,22 +56,15 @@ public class FieldDataLoader {
termDocs.seek(termEnum);
while (termDocs.next()) {
int doc = termDocs.doc();
boolean found = false;
for (int i = 0; i < ordinals.size(); i++) {
int[] ordinal = ordinals.get(i);
if (ordinal[doc] == 0) {
// we found a spot, use it
ordinal[doc] = t;
found = true;
break;
}
}
if (!found) {
// did not find one, increase by one and redo
int[] ordinal = new int[reader.maxDoc()];
int[] ordinal;
if (idx[doc] >= ordinals.size()) {
ordinal = new int[reader.maxDoc()];
ordinals.add(ordinal);
ordinal[doc] = t;
} else {
ordinal = ordinals.get(idx[doc]);
}
ordinal[doc] = t;
idx[doc]++;
}
t++;
} while (termEnum.next());