This is an automated email from the ASF dual-hosted git repository.
stefanegli pushed a commit to branch OAK-10347
in repository https://gitbox.apache.org/repos/asf/jackrabbit-oak.git
commit fbf842d6b12853d873634a0a1a1f929695488c82
Author: stefan-egli <stefane...@apache.org>
AuthorDate: Thu Jul 13 15:57:24 2023 +0200
OAK-10347 : incorporated review feedback
---
.../document/flatfile/SimpleFlatFileUtil.java | 63 ++++++++--------------
1 file changed, 23 insertions(+), 40 deletions(-)
diff --git
a/oak-run-commons/src/main/java/org/apache/jackrabbit/oak/index/indexer/document/flatfile/SimpleFlatFileUtil.java
b/oak-run-commons/src/main/java/org/apache/jackrabbit/oak/index/indexer/document/flatfile/SimpleFlatFileUtil.java
index b67cebf1c1..3284b6493d 100644
---
a/oak-run-commons/src/main/java/org/apache/jackrabbit/oak/index/indexer/document/flatfile/SimpleFlatFileUtil.java
+++
b/oak-run-commons/src/main/java/org/apache/jackrabbit/oak/index/indexer/document/flatfile/SimpleFlatFileUtil.java
@@ -18,12 +18,13 @@
*/
package org.apache.jackrabbit.oak.index.indexer.document.flatfile;
+import static org.apache.jackrabbit.guava.common.collect.ImmutableList.copyOf;
+import static org.apache.jackrabbit.oak.commons.PathUtils.elements;
+
import java.io.BufferedWriter;
import java.io.File;
import java.io.FileWriter;
import java.io.IOException;
-import java.util.ArrayList;
-import java.util.function.Predicate;
import org.apache.jackrabbit.oak.index.indexer.document.NodeStateEntry;
import org.apache.jackrabbit.oak.plugins.document.DocumentNodeState;
@@ -43,67 +44,49 @@ public class SimpleFlatFileUtil {
private static final String LINE_SEPARATOR =
System.getProperty("line.separator");
- private final FileWriter fw;
private final BufferedWriter bw;
- private final ArrayList<StateInBytesHolder> entryBatch = new ArrayList<>();
- private final Predicate<String> pathPredicate = path -> true;
private final NodeStateEntryWriter entryWriter;
private long totalLines = 0;
- private SimpleFlatFileUtil(File f) throws IOException {
+ private SimpleFlatFileUtil(BufferedWriter bw) throws IOException {
// blobStore is only used for deserialization - so pass null here:
entryWriter = new NodeStateEntryWriter(null);
- fw = new FileWriter(f);
- bw = new BufferedWriter(fw);
+ this.bw = bw;
}
public static void createFlatFileFor(NodeState ns, File f) throws
IOException {
- final SimpleFlatFileUtil h = new SimpleFlatFileUtil(f);
log.info("createFlatFileFor : writing to {}", f.getCanonicalPath());
- h.addEntryAndTraverseChildren(ns);
- h.close();
- log.info("createFlatFileFor : done. wrote {} lines in total.",
h.totalLines);
- }
-
- private void close() throws IOException {
- flush();
- bw.close();
- fw.close();
- }
-
- private void flush() throws IOException {
- for (StateInBytesHolder nsh : entryBatch) {
- String line = entryWriter.toString(nsh.getPathElements(),
nsh.getLine());
- bw.append(line);
- bw.append(LINE_SEPARATOR);
+ try (FileWriter fw = new FileWriter(f);
+ BufferedWriter bw = new BufferedWriter(fw)) {
+ SimpleFlatFileUtil h = new SimpleFlatFileUtil(bw);
+ h.addEntryAndTraverseChildren(ns);
+ log.info("createFlatFileFor : done. wrote {} lines in total.",
h.totalLines);
}
- log.info("flush : wrote another {} nodes, total so far: {} lines.",
- entryBatch.size(), totalLines);
- totalLines += entryBatch.size();
- entryBatch.clear();
}
private void addEntryAndTraverseChildren(NodeState ns) throws IOException {
addEntry(ns);
- if (entryBatch.size() > 999) {
- flush();
- }
for (ChildNodeEntry e : ns.getChildNodeEntries()) {
addEntryAndTraverseChildren(e.getNodeState());
}
}
- private void addEntry(NodeState ns) {
+ private void addEntry(NodeState ns) throws IOException {
DocumentNodeState dns = (DocumentNodeState) ns;
NodeStateEntry e = new NodeStateEntry.NodeStateEntryBuilder(dns,
dns.getPath().toString()).build();
String path = e.getPath();
- if (!NodeStateUtils.isHiddenPath(path) && pathPredicate.test(path)) {
- String jsonText = entryWriter.asJson(e.getNodeState());
- // Here logic differs from NodeStateEntrySorter in sense that
- // Holder line consist only of json and not 'path|json'
- StateInBytesHolder h = new StateInBytesHolder(path, jsonText);
- entryBatch.add(h);
+ if (NodeStateUtils.isHiddenPath(path)) {
+ // skip
+ return;
+ }
+ String jsonText = entryWriter.asJson(e.getNodeState());
+ String line = entryWriter.toString(copyOf(elements(path)), jsonText);
+ bw.append(line);
+ bw.append(LINE_SEPARATOR);
+ totalLines++;
+ if (totalLines % 10000 == 0) {
+ log.info("addEntry : wrote {} lines so far.", totalLines);
}
}
-}
+}
\ No newline at end of file
