syntax = "proto3";

package yandex.cloud.ai.vision.v1;

import "yandex/cloud/ai/vision/v1/primitives.proto";

option go_package = "github.com/yandex-cloud/go-genproto/yandex/cloud/ai/vision/v1;vision";
option java_package = "yandex.cloud.api.ai.vision.v1";

message TextAnnotation {
  // Pages of the recognized file.
  //
  // For JPEG and PNG files contains only 1 page.
  repeated Page pages = 1;
}

message Page {
  // Page width in pixels.
  int64 width = 1;

  // Page height in pixels.
  int64 height = 2;

  // Recognized text blocks in this page.
  repeated Block blocks = 3;

  // Recognized entities
  repeated Entity entities = 4;
}

message Entity {
  // Entity name
  string name = 1;

  // Recognized entity text
  string text = 2;
}

message Block {
  // Area on the page where the text block is located.
  Polygon bounding_box = 1;

  // Recognized lines in this block.
  repeated Line lines = 2;
}

message Line {
  // Area on the page where the line is located.
  Polygon bounding_box = 1;

  // Recognized words in this line.
  repeated Word words = 2;

  // Confidence of the OCR results for the line. Range [0, 1].
  double confidence = 3;
}

message Word {
  // Area on the page where the word is located.
  Polygon bounding_box = 1;

  // Recognized word value.
  string text = 2;

  // Confidence of the OCR results for the word. Range [0, 1].
  double confidence = 3;

  message DetectedLanguage {
    // Detected language code.
    string language_code = 1;

    // Confidence of detected language. Range [0, 1].
    double confidence = 2;
  }

  // A list of detected languages together with confidence.
  repeated DetectedLanguage languages = 4;

  // Id of recognized word in entities array
  int64 entity_index = 5;
}