yezhizi commented on code in PR #3264:
URL: https://github.com/apache/kvrocks/pull/3264#discussion_r2536248640
##########
src/types/redis_timeseries.cc:
##########
@@ -1860,6 +1860,54 @@ rocksdb::Status TimeSeries::Create(engine::Context &ctx,
const Slice &user_key,
return createTimeSeries(ctx, ns_key, &metadata, &option);
}
+rocksdb::Status TimeSeries::Alter(engine::Context &ctx, const Slice &user_key,
const TSCreateOption &option,
+ uint8_t mask) {
+ std::string ns_key = AppendNamespacePrefix(user_key);
+ TimeSeriesMetadata metadata;
+ if (auto s = getTimeSeriesMetadata(ctx, ns_key, &metadata); !s.ok()) {
+ return rocksdb::Status::InvalidArgument("key not exists");
+ }
+ auto batch = storage_->GetWriteBatchBase();
+ WriteBatchLogData log_data(kRedisTimeSeries, {"Alter"});
+ if (auto s = batch->PutLogData(log_data.Encode()); !s.ok()) return s;
+
+ // LSB 0 - retention time, LSB 1 - chunk size, LSB 2 - duplicate policy, LSB
4 - Labels
+ metadata.retention_time = (mask & 1) ? option.retention_time :
metadata.retention_time;
+ metadata.chunk_size = (mask & (1 << 1)) ? option.chunk_size :
metadata.chunk_size;
+ metadata.duplicate_policy = (mask & (1 << 2)) ? option.duplicate_policy :
metadata.duplicate_policy;
+
+ std::string bytes;
+ metadata.Encode(&bytes);
+ if (auto s = batch->Put(metadata_cf_handle_, ns_key, bytes); !s.ok()) return
s;
+
+ if (mask & (1 << 4)) {
+ // Delete reverse indexes and labels first.
+ LabelKVList prev_labels;
+ if (auto s = getLabelKVList(ctx, ns_key, metadata, &prev_labels); !s.ok())
return s;
+
+ auto [ns, user_key] = ExtractNamespaceKey(ns_key,
storage_->IsSlotIdEncoded());
+ for (const auto &label : prev_labels) {
+ auto rev_index_key = TSRevLabelKey(ns, label.k, label.v,
user_key).Encode();
+ if (auto s = batch->Delete(index_cf_handle_, rev_index_key); !s.ok())
return s;
+
+ auto it = std::find_if(option.labels.begin(), option.labels.end(),
+ [&label](const auto &kv) { return label.k ==
kv.k; });
Review Comment:
Maybe we could use an `unordered_map` to avoid linear search
--
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]
