feat: Introduce RagFileMetadataConfig for importing metadata to Rag

Google APIs · copybara-github · commit 4cdc2aa1af11 · 2025-06-11T12:41:51.000-07:00
PiperOrigin-RevId: 770274285
diff --git a/google/cloud/aiplatform/v1beta1/vertex_rag_data.proto b/google/cloud/aiplatform/v1beta1/vertex_rag_data.proto
@@ -462,6 +462,10 @@ message RagFile {
 
   // Output only. State of the RagFile.
   FileStatus file_status = 13 [(google.api.field_behavior) = OUTPUT_ONLY];
+
+  // Output only. The metadata for metadata search. The contents will be
+  // be in JSON format.
+  string user_metadata = 15 [(google.api.field_behavior) = OUTPUT_ONLY];
 }
 
 // A RagChunk includes the content of a chunk of a RagFile, and associated
@@ -592,13 +596,66 @@ message RagFileParsingConfig {
   bool use_advanced_pdf_parsing = 2 [deprecated = true];
 }
 
+// Metadata config for RagFile.
+message RagFileMetadataConfig {
+  // Specifies the metadata schema source.
+  oneof metadata_schema_source {
+    // Google Cloud Storage location. Supports importing individual files as
+    // well as entire Google Cloud Storage directories. Sample formats:
+    // - `gs://bucket_name/my_directory/object_name/metadata_schema.json`
+    // - `gs://bucket_name/my_directory`
+    // If providing a directory, the metadata schema will be read from
+    // the files that ends with "metadata_schema.json" in the directory.
+    GcsSource gcs_metadata_schema_source = 1;
+
+    // Google Drive location. Supports importing individual files as
+    // well as Google Drive folders.
+    // If providing a folder, the metadata schema will be read from
+    // the files that ends with "metadata_schema.json" in the directory.
+    GoogleDriveSource google_drive_metadata_schema_source = 2;
+
+    // Inline metadata schema source. Must be a JSON string.
+    string inline_metadata_schema_source = 3;
+  }
+
+  // Specifies the metadata source.
+  oneof metadata_source {
+    // Google Cloud Storage location. Supports importing individual files as
+    // well as entire Google Cloud Storage directories. Sample formats:
+    // - `gs://bucket_name/my_directory/object_name/metadata.json`
+    // - `gs://bucket_name/my_directory`
+    // If providing a directory, the metadata will be read from
+    // the files that ends with "metadata.json" in the directory.
+    GcsSource gcs_metadata_source = 4;
+
+    // Google Drive location. Supports importing individual files as
+    // well as Google Drive folders.
+    // If providing a directory, the metadata will be read from
+    // the files that ends with "metadata.json" in the directory.
+    GoogleDriveSource google_drive_metadata_source = 5;
+
+    // Inline metadata source. Must be a JSON string.
+    string inline_metadata_source = 6;
+  }
+}
+
 // Config for uploading RagFile.
 message UploadRagFileConfig {
   // Specifies the size and overlap of chunks after uploading RagFile.
   RagFileChunkingConfig rag_file_chunking_config = 1 [deprecated = true];
 
   // Specifies the transformation config for RagFiles.
   RagFileTransformationConfig rag_file_transformation_config = 3;
+
+  // Specifies the metadata config for RagFiles.
+  // Including paths for metadata schema and metadata.
+  // Alteratively, inline metadata schema and metadata can be provided.
+  RagFileMetadataConfig rag_file_metadata_config = 4;
+
+  // Optional. Specifies the parsing config for RagFiles.
+  // RAG will use the default parser if this field is not set.
+  RagFileParsingConfig rag_file_parsing_config = 5
+      [(google.api.field_behavior) = OPTIONAL];
 }
 
 // Config for importing RagFiles.
@@ -668,6 +725,10 @@ message ImportRagFilesConfig {
   RagFileParsingConfig rag_file_parsing_config = 8
       [(google.api.field_behavior) = OPTIONAL];
 
+  // Specifies the metadata config for RagFiles.
+  // Including paths for metadata schema and metadata.
+  RagFileMetadataConfig rag_file_metadata_config = 17;
+
   // Optional. The max number of queries per minute that this job is allowed to
   // make to the embedding model specified on the corpus. This value is specific
   // to this job and not shared across other import jobs. Consult the Quotas
