This is an automated email from the ASF dual-hosted git repository.
matthieu pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/james-project.git
The following commit(s) were added to refs/heads/master by this push:
new 46ed946fae More Extractors fixes (#1430)
46ed946fae is described below
commit 46ed946fae16fa33761197899ef9b7a8a5b3d5a4
Author: Matthieu Baechler <[email protected]>
AuthorDate: Fri Feb 10 08:24:15 2023 +0000
More Extractors fixes (#1430)
* Avoid double conversion of Duration to/from millis
* Perform IOs into Mono and not synchronously at method call
---
.../james/mailbox/tika/CachingTextExtractor.java | 21 +++++++++++----------
1 file changed, 11 insertions(+), 10 deletions(-)
diff --git
a/mailbox/tika/src/main/java/org/apache/james/mailbox/tika/CachingTextExtractor.java
b/mailbox/tika/src/main/java/org/apache/james/mailbox/tika/CachingTextExtractor.java
index c553cbef63..f7309d6360 100644
---
a/mailbox/tika/src/main/java/org/apache/james/mailbox/tika/CachingTextExtractor.java
+++
b/mailbox/tika/src/main/java/org/apache/james/mailbox/tika/CachingTextExtractor.java
@@ -20,7 +20,6 @@
package org.apache.james.mailbox.tika;
import java.io.ByteArrayInputStream;
-import java.io.IOException;
import java.io.InputStream;
import java.time.Duration;
import java.util.Optional;
@@ -40,6 +39,7 @@ import com.github.benmanes.caffeine.cache.Weigher;
import com.google.common.hash.Hashing;
import reactor.core.publisher.Mono;
+import reactor.core.scheduler.Schedulers;
public class CachingTextExtractor implements TextExtractor {
private final TextExtractor underlying;
@@ -60,7 +60,7 @@ public class CachingTextExtractor implements TextExtractor {
(key, parsedContent) -> computeWeight(parsedContent);
cache = Caffeine.newBuilder()
-
.expireAfterAccess(Duration.ofMillis(cacheEvictionPeriod.toMillis()))
+ .expireAfterAccess(cacheEvictionPeriod)
.maximumWeight(cacheWeightInBytes)
.weigher(weigher)
.evictionListener(removalListener)
@@ -111,14 +111,15 @@ public class CachingTextExtractor implements
TextExtractor {
@Override
public Mono<ParsedContent> extractContentReactive(InputStream inputStream,
ContentType contentType) {
- try {
- byte[] bytes = IOUtils.toByteArray(inputStream);
- String key = Hashing.sha256().hashBytes(bytes).toString();
-
- return Mono.fromFuture(cache.get(key, (a, b) ->
retrieveAndUpdateWeight(bytes, contentType).toFuture()));
- } catch (IOException e) {
- throw new RuntimeException(e);
- }
+ return Mono
+ .fromCallable(() -> IOUtils.toByteArray(inputStream))
+ .subscribeOn(Schedulers.boundedElastic())
+ .flatMap(bytes ->
+ Mono.fromCallable(() ->
Hashing.sha256().hashBytes(bytes).toString())
+ .subscribeOn(Schedulers.parallel())
+ .publishOn(Schedulers.boundedElastic())
+ .flatMap(key -> Mono.fromFuture(cache.get(key, (a, b)
-> retrieveAndUpdateWeight(bytes, contentType).toFuture())))
+ );
}
@Override
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
