Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// Copyright 2025 Google LLC
// Copyright 2026 Google LLC
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
Expand Down
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// Copyright 2025 Google LLC
// Copyright 2026 Google LLC
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
Expand Down Expand Up @@ -69,9 +69,6 @@ message Dataset {
// Configuration specific to an unmanaged dataset.
message UnmanagedDatasetConfig {}

// Configuration specific to spanner-based indexing.
message SpannerIndexingConfig {}

// Different states of a dataset.
enum State {
// Default unspecified enum, should not be used.
Expand Down Expand Up @@ -106,14 +103,6 @@ message Dataset {
[(google.api.field_behavior) = OPTIONAL];
}

oneof indexing_source {
// Optional. A lightweight indexing source with low latency and high
// reliability, but lacking advanced features like CMEK and content-based
// search.
SpannerIndexingConfig spanner_indexing_config = 4
[(google.api.field_behavior) = OPTIONAL];
}

// Dataset resource name.
// Format:
// `projects/{project}/locations/{location}/processors/{processor}/dataset`
Expand Down
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// Copyright 2025 Google LLC
// Copyright 2026 Google LLC
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
Expand All @@ -17,6 +17,7 @@ syntax = "proto3";
package google.cloud.documentai.v1beta3;

import "google/api/field_behavior.proto";
import "google/api/resource.proto";
import "google/cloud/documentai/v1beta3/barcode.proto";
import "google/cloud/documentai/v1beta3/geometry.proto";
import "google/protobuf/timestamp.proto";
Expand Down Expand Up @@ -340,7 +341,7 @@ message Document {
repeated DetectedLanguage detected_languages = 2;
}

// Detected non-text visual elements, for example, checkbox, signature, etc.
// Detected non-text visual elements, for example, checkbox, signature etc.
// on the page.
message VisualElement {
// [Layout][google.cloud.documentai.v1beta3.Document.Page.Layout] for
Expand Down Expand Up @@ -530,7 +531,7 @@ message Document {
// A list of visually detected tokens on the page.
repeated Token tokens = 8;

// A list of detected non-text visual elements for example, checkbox,
// A list of detected non-text visual elements, for example, checkbox,
// signature etc. on the page.
repeated VisualElement visual_elements = 9;

Expand Down Expand Up @@ -628,10 +629,11 @@ message Document {
// [Document.text][google.cloud.documentai.v1beta3.Document.text].
TextAnchor text_anchor = 1 [(google.api.field_behavior) = OPTIONAL];

// Required. Entity type from a schema for example, `Address`.
// Required. Entity type from a schema, for example, `Address`.
string type = 2 [(google.api.field_behavior) = REQUIRED];

// Optional. Text value of the entity for example, `1600 Amphitheatre Pkwy`.
// Optional. Text value of the entity, for example, `1600 Amphitheatre
// Pkwy`.
string mention_text = 3 [(google.api.field_behavior) = OPTIONAL];

// Optional. Deprecated. Use `id` field instead.
Expand All @@ -644,7 +646,7 @@ message Document {
// the page where it was found.
PageAnchor page_anchor = 6 [(google.api.field_behavior) = OPTIONAL];

// Optional. Canonical ID. This will be a unique value in the entity list
// Optional. Canonical id. This will be a unique value in the entity list
// for this document.
string id = 7 [(google.api.field_behavior) = OPTIONAL];

Expand Down Expand Up @@ -673,10 +675,10 @@ message Document {
// Relationship between
// [Entities][google.cloud.documentai.v1beta3.Document.Entity].
message EntityRelation {
// Subject entity ID.
// Subject entity id.
string subject_id = 1;

// Object entity ID.
// Object entity id.
string object_id = 2;

// Relationship description.
Expand Down Expand Up @@ -806,7 +808,7 @@ message Document {
// of entities, properties within entities, etc.) in the parent revision.
int32 index = 3;

// The ID of the parent provenance.
// The id of the parent provenance.
int32 id = 2 [deprecated = true];
}

Expand Down Expand Up @@ -846,7 +848,7 @@ message Document {
// The index of the revision that produced this element.
int32 revision = 1 [deprecated = true];

// The ID of this operation. Needs to be unique within the scope of the
// The Id of this operation. Needs to be unique within the scope of the
// revision.
int32 id = 2 [deprecated = true];

Expand All @@ -871,7 +873,7 @@ message Document {

// Who/what made the change
oneof source {
// If the change was made by a person specify the name or ID of that
// If the change was made by a person specify the name or id of that
// person.
string agent = 4;

Expand All @@ -880,7 +882,7 @@ message Document {
string processor = 5;
}

// ID of the revision, internally generated by doc proto storage.
// Id of the revision, internally generated by doc proto storage.
// Unique within the context of the document.
string id = 1;

Expand All @@ -889,7 +891,7 @@ message Document {
// index into the `revisions` field.
repeated int32 parent = 2 [deprecated = true];

// The revisions that this revision is based on. Must include all the IDs
// The revisions that this revision is based on. Must include all the ids
// that have anything to do with this revision - eg. there are
// `provenance.parent.revision` fields that index into this field.
repeated string parent_ids = 7;
Expand Down Expand Up @@ -1013,18 +1015,18 @@ message Document {
message LayoutImageBlock {
// Source of the image.
oneof image_source {
// Optional. Asset ID of the inline image. If set, find the image
// Optional. Asset id of the inline image. If set, find the image
// content in the blob_assets field.
string blob_asset_id = 4 [(google.api.field_behavior) = OPTIONAL];

// Optional. Google Cloud Storage URI of the image.
// Optional. Google Cloud Storage uri of the image.
string gcs_uri = 5 [(google.api.field_behavior) = OPTIONAL];

// Optional. Data URI of the image.
// Optional. Data uri of the image.
// It is composed of four parts: a prefix (data:), a MIME type
// indicating the type of data, an optional base64 token if
// non-textual, and the data itself:
// data:[<mediatype>][;base64],<data>.
// data:[<mediatype>][;base64],<data>
string data_uri = 6 [(google.api.field_behavior) = OPTIONAL];
}

Expand Down Expand Up @@ -1103,18 +1105,18 @@ message Document {
message ImageChunkField {
// Source of the image.
oneof image_source {
// Optional. Asset ID of the inline image. If set, find the image
// Optional. Asset id of the inline image. If set, find the image
// content in the blob_assets field.
string blob_asset_id = 1 [(google.api.field_behavior) = OPTIONAL];

// Optional. Google Cloud Storage URI of the image.
// Optional. Google Cloud Storage uri of the image.
string gcs_uri = 2 [(google.api.field_behavior) = OPTIONAL];

// Optional. Data URI of the image.
// Optional. Data uri of the image.
// It is composed of four parts: a prefix (data:), a MIME type
// indicating the type of data, an optional base64 token if
// non-textual, and the data itself:
// data:[<mediatype>][;base64],<data>.
// data:[<mediatype>][;base64],<data>
string data_uri = 3 [(google.api.field_behavior) = OPTIONAL];
}

Expand Down Expand Up @@ -1171,7 +1173,7 @@ message Document {
// in this document, for example, image bytes, such that it can be referenced
// by other fields in the document via asset ID.
message BlobAsset {
// Optional. The ID of the blob asset.
// Optional. The id of the blob asset.
string asset_id = 1 [(google.api.field_behavior) = OPTIONAL];

// Optional. The content of the blob asset, for example, image bytes.
Expand Down Expand Up @@ -1205,7 +1207,7 @@ message Document {
VALIDATION_RESULT_TYPE_NOT_APPLICABLE = 4;
}

// The name of the validation rule.
// The display name of the validation rule.
string rule_name = 1;

// The description of the validation rule.
Expand All @@ -1229,14 +1231,17 @@ message Document {

// Entity revision.
message EntitiesRevision {
// The revision ID.
// The revision id.
string revision_id = 1;

// The entities in this revision.
repeated Entity entities = 2;

// The entity validation output for this revision.
EntityValidationOutput entity_validation_output = 3;

// Optional. The history of this revision.
Provenance provenance = 4 [(google.api.field_behavior) = OPTIONAL];
}

// Original source document from the user.
Expand Down Expand Up @@ -1320,7 +1325,7 @@ message Document {
// The entity revision ID that `document.entities` field is based on.
// If this field is set and `entities_revisions` is not empty, the entities in
// `document.entities` field are the entities in the entity revision with this
// ID and `document.entity_validation_output` field is the
// id and `document.entity_validation_output` field is the
// `entity_validation_output` field in this entity revision.
string entities_revision_id = 23;
}
Expand All @@ -1347,7 +1352,7 @@ message RevisionRef {
// Reads the revision by the predefined case.
RevisionCase revision_case = 1;

// Reads the revision given by the ID.
// Reads the revision given by the id.
string revision_id = 2;

// Reads the revision generated by the processor version.
Expand Down
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// Copyright 2025 Google LLC
// Copyright 2026 Google LLC
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
Expand All @@ -16,6 +16,7 @@ syntax = "proto3";

package google.cloud.documentai.v1beta3;

import "google/cloud/documentai/v1beta3/document.proto";
import "google/protobuf/field_mask.proto";

option csharp_namespace = "Google.Cloud.DocumentAI.V1Beta3";
Expand All @@ -26,6 +27,12 @@ option java_package = "com.google.cloud.documentai.v1beta3";
option php_namespace = "Google\\Cloud\\DocumentAI\\V1beta3";
option ruby_package = "Google::Cloud::DocumentAI::V1beta3";

// A set of inline documents.
message Documents {
// The list of documents.
repeated Document documents = 1;
}

// Payload message of raw document content (bytes).
message RawDocument {
// Inline document content.
Expand Down Expand Up @@ -64,6 +71,12 @@ message GcsPrefix {
string gcs_uri_prefix = 1;
}

// Specifies a set of raw documents.
message RawDocuments {
// Specifies raw document content and mime type.
repeated RawDocument documents = 1;
}

// The common config to specify a set of documents used as input.
message BatchDocumentsInputConfig {
// The source. Make sure that the caller of the API has storage.objects.get
Expand Down
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// Copyright 2025 Google LLC
// Copyright 2026 Google LLC
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
Expand Down Expand Up @@ -286,6 +286,7 @@ service DocumentProcessorService {
// the specified processor.
rpc ReviewDocument(ReviewDocumentRequest)
returns (google.longrunning.Operation) {
option deprecated = true;
option (google.api.http) = {
post: "/v1beta3/{human_review_config=projects/*/locations/*/processors/*/humanReviewConfig}:reviewDocument"
body: "*"
Expand Down Expand Up @@ -393,6 +394,9 @@ message ProcessOptions {

// Optional. Whether to include table annotations in layout parser response.
bool enable_table_annotation = 6 [(google.api.field_behavior) = OPTIONAL];

// Optional. Whether to split table.
bool enable_table_split = 8 [(google.api.field_behavior) = OPTIONAL];
}

// A list of individual page numbers.
Expand Down Expand Up @@ -475,7 +479,7 @@ message ProcessRequest {

// Whether human review should be skipped for this request. Default to
// `false`.
bool skip_human_review = 3;
bool skip_human_review = 3 [deprecated = true];

// Specifies which fields to include in the
// [ProcessResponse.document][google.cloud.documentai.v1beta3.ProcessResponse.document]
Expand Down Expand Up @@ -614,7 +618,7 @@ message BatchProcessRequest {

// Whether human review should be skipped for this request. Default to
// `false`.
bool skip_human_review = 4;
bool skip_human_review = 4 [deprecated = true];

// Inference-time options for the process API
ProcessOptions process_options = 7;
Expand Down Expand Up @@ -1458,3 +1462,11 @@ message ImportProcessorVersionMetadata {
// The basic metadata for the long-running operation.
CommonOperationMetadata common_metadata = 1;
}

// The long-running operation metadata for the
// [UpdateProcessorVersion][google.cloud.documentai.v1beta3.DocumentProcessorService.UpdateProcessorVersion]
// method.
message UpdateProcessorVersionMetadata {
// The basic metadata for the long-running operation.
CommonOperationMetadata common_metadata = 1;
}
Loading