feat: Add ZSTD compression as an option for Arrow.

Google APIs · copybara-github · commit 23efea9fc7be · 2021-05-17T10:03:47.000-07:00
Committer: @emkornfield PiperOrigin-RevId: 374220891
diff --git a/google/cloud/bigquery/storage/v1/arrow.proto b/google/cloud/bigquery/storage/v1/arrow.proto
@@ -52,6 +52,9 @@ message ArrowSerializationOptions {
 
     // LZ4 Frame (https://github.com/lz4/lz4/blob/dev/doc/lz4_Frame_format.md)
     LZ4_FRAME = 1;
+
+    // Zstandard compression.
+    ZSTD = 2;
   }
 
   // The compression codec to use for Arrow buffers in serialized record
diff --git a/google/cloud/bigquery/storage/v1/storage.proto b/google/cloud/bigquery/storage/v1/storage.proto
@@ -69,8 +69,7 @@ service BigQueryRead {
       post: "/v1/{read_session.table=projects/*/datasets/*/tables/*}"
       body: "*"
     };
-    option (google.api.method_signature) =
-        "parent,read_session,max_stream_count";
+    option (google.api.method_signature) = "parent,read_session,max_stream_count";
   }
 
   // Reads rows from the stream in the format prescribed by the ReadSession.
@@ -99,8 +98,7 @@ service BigQueryRead {
   // original, primary, and residual, that original[0-j] = primary[0-j] and
   // original[j-n] = residual[0-m] once the streams have been read to
   // completion.
-  rpc SplitReadStream(SplitReadStreamRequest)
-      returns (SplitReadStreamResponse) {
+  rpc SplitReadStream(SplitReadStreamRequest) returns (SplitReadStreamResponse) {
     option (google.api.http) = {
       get: "/v1/{name=projects/*/locations/*/sessions/*/streams/*}"
     };
diff --git a/google/cloud/bigquery/storage/v1/stream.proto b/google/cloud/bigquery/storage/v1/stream.proto
@@ -75,23 +75,20 @@ message ReadSession {
     // Restricted to a maximum length for 1 MB.
     string row_restriction = 2;
 
-    // Optional. Options specific to the Apache Arrow output format.
     oneof output_format_serialization_options {
-      ArrowSerializationOptions arrow_serialization_options = 3
-          [(google.api.field_behavior) = OPTIONAL];
+      // Optional. Options specific to the Apache Arrow output format.
+      ArrowSerializationOptions arrow_serialization_options = 3 [(google.api.field_behavior) = OPTIONAL];
     }
   }
 
   // Output only. Unique identifier for the session, in the form
   // `projects/{project_id}/locations/{location}/sessions/{session_id}`.
   string name = 1 [(google.api.field_behavior) = OUTPUT_ONLY];
 
-  // Output only. Time at which the session becomes invalid. After this time,
-  // subsequent requests to read this Session will return errors. The
-  // expire_time is automatically assigned and currently cannot be specified or
-  // updated.
-  google.protobuf.Timestamp expire_time = 2
-      [(google.api.field_behavior) = OUTPUT_ONLY];
+  // Output only. Time at which the session becomes invalid. After this time, subsequent
+  // requests to read this Session will return errors. The expire_time is
+  // automatically assigned and currently cannot be specified or updated.
+  google.protobuf.Timestamp expire_time = 2 [(google.api.field_behavior) = OUTPUT_ONLY];
 
   // Immutable. Data format of the output data.
   DataFormat data_format = 3 [(google.api.field_behavior) = IMMUTABLE];
@@ -111,11 +108,12 @@ message ReadSession {
   // `projects/{project_id}/datasets/{dataset_id}/tables/{table_id}`
   string table = 6 [
     (google.api.field_behavior) = IMMUTABLE,
-    (google.api.resource_reference) = { type: "bigquery.googleapis.com/Table" }
+    (google.api.resource_reference) = {
+      type: "bigquery.googleapis.com/Table"
+    }
   ];
 
-  // Optional. Any modifiers which are applied when reading from the specified
-  // table.
+  // Optional. Any modifiers which are applied when reading from the specified table.
   TableModifiers table_modifiers = 7 [(google.api.field_behavior) = OPTIONAL];
 
   // Optional. Read options for this session (e.g. column selection, filters).

Original file line number	Diff line number	Diff line change
`@@ -52,6 +52,9 @@ message ArrowSerializationOptions {`
`52`	`52`
`53`	`53`	`// LZ4 Frame (https://github.com/lz4/lz4/blob/dev/doc/lz4_Frame_format.md)`
`54`	`54`	`LZ4_FRAME = 1;`
	`55`	`+`
	`56`	`+ // Zstandard compression.`
	`57`	`+ ZSTD = 2;`
`55`	`58`	`}`
`56`	`59`
`57`	`60`	`// The compression codec to use for Arrow buffers in serialized record`