You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@lucene.apache.org by "uschindler (via GitHub)" <gi...@apache.org> on 2023/02/10 13:44:51 UTC
[GitHub] [lucene] uschindler opened a new pull request, #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
uschindler opened a new pull request, #12141:
URL: https://github.com/apache/lucene/pull/12141
This is an alternative approach for #12135
Like in @gsmiller's code, the removal of "clone()" is ok here, as the query only consumes the terms array, but does not modify it or saves it anywhere.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] rmuir commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "rmuir (via GitHub)" <gi...@apache.org>.
rmuir commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1425910302
PointInSetQuery has a lovely homemade "stream" in ctor already: https://github.com/apache/lucene/blob/main/lucene/core/src/java/org/apache/lucene/search/PointInSetQuery.java#L66-L75
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
Re: [PR] Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach) [lucene]
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1880958137
We have another solution for this already.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] rmuir commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "rmuir (via GitHub)" <gi...@apache.org>.
rmuir commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1427025395
and please, please, no `PrefixCodedTerms extends X` where X is some new abstraction used. that's wrong and it puts the onus on PrefixCodedTerms to now deliver on some interface for these crappy join queries.
Instead, if you add a new X, it can use as its code implementation, PrefixCodedTerms. that way the onus is backwards, on the join stuff to implement its own interface.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] gsmiller commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "gsmiller (via GitHub)" <gi...@apache.org>.
gsmiller commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1426790254
+1 that it would be nice to avoid the duplicate prefix-encoding (and storing that duplicate data for two queries in memory).
Ideally, it would be nice if `TermInSetQuery` where pkg-private, just acting as an implementation detail behind factory methods like `KeywordField#newSetQuery`. It might be a bit too extreme, but I wonder if we should consider making this change and asking our users to get their "term in set" query functionality behind exposed factory methods.
Less drastic than essentially deprecating the public definition of `TermInSetQuery` could be to create a new class definition for a "set of terms" that wraps `PrefixCodedTerms` and is used in the ctor. This could provide an abstraction on top of the `PrefixCodedTerms` implementation detail, allowing us to change that in the future without changing our interface.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] rmuir commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "rmuir (via GitHub)" <gi...@apache.org>.
rmuir commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1427024266
more abstractions, copying queries, hope you take PointInSetQuery into account here.
I dont want anyone to be surprised, if i see a goddamn mess that doesnt consistently fix the issue across point+dv, too, i'm gonna veto. sorry.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
Re: [PR] Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach) [lucene]
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler closed pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
URL: https://github.com/apache/lucene/pull/12141
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] rmuir commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "rmuir (via GitHub)" <gi...@apache.org>.
rmuir commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1425853047
If a user has points+dv, they may encounter the same issue. So we may want to consider looking at `PointInSetQuery` too. Maybe it should have a similar streams fix that `TermInSetQuery` has?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] rmuir commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "rmuir (via GitHub)" <gi...@apache.org>.
rmuir commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1426703695
i want to see any proposed solution work for Points, too.
Will hold these database features to the fire!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1427034789
If you do not agree you can veto for sure, but arguing with that before you have seen anything is not acceptable.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1427034648
Sorry I was not planning anything like that. Just move the query and leave the "public" backwards layer available.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on a diff in pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on code in PR #12141:
URL: https://github.com/apache/lucene/pull/12141#discussion_r1102816225
##########
lucene/core/src/java/org/apache/lucene/document/SortedSetDocValuesSetQuery.java:
##########
@@ -51,20 +54,23 @@ final class SortedSetDocValuesSetQuery extends Query implements Accountable {
private final PrefixCodedTerms termData;
private final int termDataHashCode; // cached hashcode of termData
- SortedSetDocValuesSetQuery(String field, BytesRef terms[]) {
+ /**
+ * Creates a new {@link TermInSetQuery} from the given collection of terms. This is most efficient
+ * if it is a sorted set or any other collection which has a presorted and distinct spliterator.
+ */
+ SortedSetDocValuesSetQuery(String field, Collection<BytesRef> terms) {
+ this(field, terms.stream());
+ }
+
+ /** Creates a new {@link TermInSetQuery} from the given array of terms. */
+ SortedSetDocValuesSetQuery(String field, BytesRef... terms) {
+ this(field, Arrays.stream(terms));
+ }
+
+ private SortedSetDocValuesSetQuery(String field, Stream<BytesRef> stream) {
Review Comment:
I checked the code, we have several callers now, some using the array, some the stream one. I'd like to keep the stream abstraction here.
If we change to use streams in general (when we have the possibility to improve IndexOrDocavaluesQuery by lazy initializing the child queries I am happy to refactor this.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
Re: [PR] Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach) [lucene]
Posted by "github-actions[bot] (via GitHub)" <gi...@apache.org>.
github-actions[bot] commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1880903779
This PR has not had activity in the past 2 weeks, labeling it as stale. If the PR is waiting for review, notify the dev@lucene.apache.org list. Thank you for your contribution!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1425860273
Before proceeding from here we should really do a benchmark here. Maybe Greg can test this with his code at Amzn. I don't know how to benchmark this without having a huge database and join queries.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] rmuir commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "rmuir (via GitHub)" <gi...@apache.org>.
rmuir commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1425878966
Right, leaving the PointInSetQuery unoptimized but optimizing the slower one (TermInSetQuery) illustrates the issues with "database features". Anywhere else in the luce search engine, you'd use the faster, more appropriate datastructure (Points in this case)
But with database features user has their `varchar` ID in some database somewhere they are using for joins, which will be slower, and no intention of fixing that :) Hopefully it helps to explain my frustration with such features.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1427010008
Let me work on making a clone of TermInSetQuery in the store package, also pkg-private. We can make the public one still available in the search package but keep the weight hidden.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] rmuir commented on a diff in pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "rmuir (via GitHub)" <gi...@apache.org>.
rmuir commented on code in PR #12141:
URL: https://github.com/apache/lucene/pull/12141#discussion_r1102797942
##########
lucene/core/src/java/org/apache/lucene/document/SortedSetDocValuesSetQuery.java:
##########
@@ -51,20 +54,23 @@ final class SortedSetDocValuesSetQuery extends Query implements Accountable {
private final PrefixCodedTerms termData;
private final int termDataHashCode; // cached hashcode of termData
- SortedSetDocValuesSetQuery(String field, BytesRef terms[]) {
+ /**
+ * Creates a new {@link TermInSetQuery} from the given collection of terms. This is most efficient
+ * if it is a sorted set or any other collection which has a presorted and distinct spliterator.
+ */
+ SortedSetDocValuesSetQuery(String field, Collection<BytesRef> terms) {
+ this(field, terms.stream());
+ }
+
+ /** Creates a new {@link TermInSetQuery} from the given array of terms. */
+ SortedSetDocValuesSetQuery(String field, BytesRef... terms) {
+ this(field, Arrays.stream(terms));
+ }
+
+ private SortedSetDocValuesSetQuery(String field, Stream<BytesRef> stream) {
Review Comment:
do we really need 3 ctors or could we just have a single package private ctor taking stream. The XXXField classes could just call .stream method? I just don't know if we need so much "sugar" for classes that are entirely package private.
Unfortunately, the TermInSetQuery is a public class, so maybe there are different concerns there.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on a diff in pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on code in PR #12141:
URL: https://github.com/apache/lucene/pull/12141#discussion_r1102800225
##########
lucene/core/src/java/org/apache/lucene/document/SortedSetDocValuesSetQuery.java:
##########
@@ -51,20 +54,23 @@ final class SortedSetDocValuesSetQuery extends Query implements Accountable {
private final PrefixCodedTerms termData;
private final int termDataHashCode; // cached hashcode of termData
- SortedSetDocValuesSetQuery(String field, BytesRef terms[]) {
+ /**
+ * Creates a new {@link TermInSetQuery} from the given collection of terms. This is most efficient
+ * if it is a sorted set or any other collection which has a presorted and distinct spliterator.
+ */
+ SortedSetDocValuesSetQuery(String field, Collection<BytesRef> terms) {
+ this(field, terms.stream());
+ }
+
+ /** Creates a new {@link TermInSetQuery} from the given array of terms. */
+ SortedSetDocValuesSetQuery(String field, BytesRef... terms) {
+ this(field, Arrays.stream(terms));
+ }
+
+ private SortedSetDocValuesSetQuery(String field, Stream<BytesRef> stream) {
Review Comment:
thats fine. I just made this query with same ctors like TermInSetQuery. But yes as it is pkg private we can make it stupid simple.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on a diff in pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on code in PR #12141:
URL: https://github.com/apache/lucene/pull/12141#discussion_r1102785415
##########
lucene/core/src/java/org/apache/lucene/index/PrefixCodedTerms.java:
##########
@@ -104,6 +105,16 @@ public PrefixCodedTerms finish() {
}
}
+ public static Collector<BytesRef, Builder, PrefixCodedTerms> collector(String field) {
Review Comment:
sorry I placed this here, because I found no other common place.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] uschindler commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "uschindler (via GitHub)" <gi...@apache.org>.
uschindler commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1426666265
I am still not happy with this idea at all, because next to sorting we still build the PrefixCodedTerms instance 2 times on construction.
I think the original idea of @gsmiller is good and in combination with
It would be good to find a way to share the PrefixCodedTerms instance between both queries (also for memory usage). The problem is that both queries (TermInSet and SortedDocValuesSetQuery) live in different packages. If both would be in same package we could make a "pkg private" ctor vor both so they can share their PrefixCodedTerms without making it public.
My idea would be to have TermInSetQuery in search package be deprecated and like for the other queries now create a factory only. Maybe just have a public subclass in the serach package. I will think a bit about it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] gsmiller commented on pull request #12141: Avoid duplicate sorting in KeywordField#newSetQuery (alternative approach)
Posted by "gsmiller (via GitHub)" <gi...@apache.org>.
gsmiller commented on PR #12141:
URL: https://github.com/apache/lucene/pull/12141#issuecomment-1426788802
Thanks @uschindler for the alternate approach. It helped me understand your earlier suggestion to use streams, which I wasn't totally clear on (I thought you were originally suggesting to do away with prefix-encoding altogether and reference the streams directly inside the query implementations to iterate the terms, which was confusing).
I'm not setup to re-run our internal benchmarks at the moment (where we see a large amount of time spent sorting terms), but I at least ran my simple "benchmark" test case that does some simple timing over the query initialization (see below). The results for this PR were as good as my initial proposal to share prefix-encoded terms. So, from a pure performance point-of-view, this appears to be just as efficient as what I'd come up with initially.
Simple test case "benchmark":
```
public void testSortPerformance() {
int len = 50000;
BytesRef[] terms = new BytesRef[len];
for (int i = 0; i < len; i++) {
String s = TestUtil.randomSimpleString(random(), 10, 20);
terms[i] = new BytesRef(s);
}
int iters = 300;
for (int i = 0; i < iters; i++) {
KeywordField.newSetQuery("foo", terms);
}
long minTime = Long.MAX_VALUE;
for (int i = 0; i < iters; i++) {
long t0 = System.nanoTime();
KeywordField.newSetQuery("foo", terms);
minTime = Math.min(minTime, System.nanoTime() - t0);
}
System.err.println("Time: " + minTime / 1_000_000);
}
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org