keith-turner commented on code in PR #2718:
URL: https://github.com/apache/accumulo/pull/2718#discussion_r883181843
##########
server/base/src/main/java/org/apache/accumulo/server/metadata/RootGcCandidates.java:
##########
@@ -31,85 +29,73 @@
import org.apache.accumulo.core.metadata.StoredTabletFile;
import org.apache.hadoop.fs.Path;
-import com.google.common.base.Preconditions;
import com.google.gson.Gson;
-import com.google.gson.GsonBuilder;
public class RootGcCandidates {
- private static final Gson GSON = new GsonBuilder().create();
+ // Version 1. Released with Accumulo version 2.1.0
+ private static final int VERSION = 1;
+
+ private final Gson gson = new Gson();
+ private final Data data;
// This class is used to serialize and deserialize root tablet metadata
using GSon. Any changes to
// this class must consider persisted data.
- private static class GSonData {
- int version = 1;
-
- // SortedMap<dir path, SortedSet<file name>>
- SortedMap<String,SortedSet<String>> candidates;
+ private static class Data {
+ private final int version;
+
+ /*
+ * The root tablet will only have a single dir on each volume. Therefore,
root file paths will
+ * have a small set of unique prefixes. The following map is structured to
avoid storing the
+ * same dir prefix over and over in JSon and java.
+ *
+ * SortedMap<dir path, SortedSet<file name>>
+ */
+ private final SortedMap<String,SortedSet<String>> candidates;
+
+ public Data(int version, SortedMap<String,SortedSet<String>> candidates) {
+ this.version = version;
+ this.candidates = candidates;
+ }
}
- /*
- * The root tablet will only have a single dir on each volume. Therefore
root file paths will have
- * a small set of unique prefixes. The following map is structured to avoid
storing the same dir
- * prefix over and over in JSon and java.
- *
- * SortedMap<dir path, SortedSet<file name>>
- */
- private SortedMap<String,SortedSet<String>> candidates;
-
public RootGcCandidates() {
- this.candidates = new TreeMap<>();
+ this.data = new Data(VERSION, new TreeMap<>());
}
- private RootGcCandidates(SortedMap<String,SortedSet<String>> candidates) {
- this.candidates = candidates;
+ public RootGcCandidates(String jsonString) {
+ this.data = gson.fromJson(jsonString, Data.class);
+ checkArgument(data.version == VERSION, "Invalid Root Table GC Candidates
JSON version %s",
+ data.version);
+ data.candidates.forEach((parent, files) -> {
+ checkArgument(!parent.isBlank(), "Blank parent dir in %s",
data.candidates);
+ checkArgument(!files.isEmpty(), "Empty files for dir %s", parent);
+ });
}
- public void add(Iterator<StoredTabletFile> refs) {
- refs.forEachRemaining(ref -> {
- String parent = ref.getPath().getParent().toString();
- candidates.computeIfAbsent(parent, k -> new
TreeSet<>()).add(ref.getFileName());
- });
+ public void add(Stream<StoredTabletFile> refs) {
+ refs.forEach(ref -> data.candidates
+ .computeIfAbsent(ref.getPath().getParent().toString(), k -> new
TreeSet<>())
+ .add(ref.getFileName()));
}
- public void remove(Collection<String> refs) {
- refs.forEach(ref -> {
- Path path = new Path(ref);
- String parent = path.getParent().toString();
- String name = path.getName();
-
- SortedSet<String> names = candidates.get(parent);
- if (names != null) {
- names.remove(name);
- if (names.isEmpty()) {
- candidates.remove(parent);
- }
- }
- });
+ public void remove(Stream<String> refs) {
+ refs.map(Path::new).forEach(
+ path -> data.candidates.computeIfPresent(path.getParent().toString(),
(key, values) -> {
+ values.remove(path.getName());
+ return values.isEmpty() ? null : values;
+ }));
}
- public Stream<String> stream() {
- return candidates.entrySet().stream().flatMap(entry -> {
+ public Stream<String> sortedStream() {
Review Comment:
Why change this to a sorted stream?
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
