ドキュメントを検索する

Document AI ウェアハウスでドキュメントを検索する

このコードサンプルを含む詳細なドキュメントについては、以下をご覧ください。

ドキュメントを検索する

コードサンプル

Java

詳細については、Document AI Warehouse Java API リファレンスドキュメントをご覧ください。

Document AI Warehouse に対する認証を行うには、アプリケーションのデフォルト認証情報を設定します。詳細については、ローカル開発環境の認証の設定をご覧ください。

import com.google.cloud.contentwarehouse.v1.DocumentQuery;
import com.google.cloud.contentwarehouse.v1.DocumentServiceClient;
import com.google.cloud.contentwarehouse.v1.DocumentServiceClient.SearchDocumentsPagedResponse;
import com.google.cloud.contentwarehouse.v1.DocumentServiceSettings;
import com.google.cloud.contentwarehouse.v1.FileTypeFilter;
import com.google.cloud.contentwarehouse.v1.FileTypeFilter.FileType;
import com.google.cloud.contentwarehouse.v1.LocationName;
import com.google.cloud.contentwarehouse.v1.RequestMetadata;
import com.google.cloud.contentwarehouse.v1.SearchDocumentsRequest;
import com.google.cloud.contentwarehouse.v1.SearchDocumentsResponse.MatchingDocument;
import com.google.cloud.contentwarehouse.v1.UserInfo;
import com.google.cloud.resourcemanager.v3.Project;
import com.google.cloud.resourcemanager.v3.ProjectName;
import com.google.cloud.resourcemanager.v3.ProjectsClient;
import java.io.IOException;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.TimeoutException;

public class SearchDocuments {
  public static void main(String[] args) throws IOException, 
        InterruptedException, ExecutionException, TimeoutException { 
    // TODO(developer): Replace these variables before running the sample.
    String projectId = "your-project-id";
    String location = "your-region"; // Format is "us" or "eu".
    String documentQuery = "your-document-query";
    String userId = "your-user-id"; // Format is user:<user-id>

    searchDocuments(projectId, location, documentQuery, userId);
  }

  // Searches all documents for a given Document Query
  public static void searchDocuments(String projectId, String location,
        String documentQuery, String userId) throws IOException, InterruptedException,
          ExecutionException, TimeoutException { 
    String projectNumber = getProjectNumber(projectId);

    String endpoint = "contentwarehouse.googleapis.com:443";
    if (!"us".equals(location)) {
      endpoint = String.format("%s-%s", location, endpoint);
    }

    DocumentServiceSettings documentServiceSettings = 
             DocumentServiceSettings.newBuilder().setEndpoint(endpoint)
             .build(); 

    /*
     * Create the Document Service Client 
     * Initialize client that will be used to send requests. 
     * This client only needs to be created once, and can be reused for multiple requests. 
     */
    try (DocumentServiceClient documentServiceClient = 
            DocumentServiceClient.create(documentServiceSettings)) {  

      /*
       * The full resource name of the location, e.g.:
       * projects/{project_number}/locations/{location} 
       */
      String parent = LocationName.format(projectNumber, location);

      // Define RequestMetadata object for context of the user making the API call
      RequestMetadata requestMetadata = RequestMetadata.newBuilder()
          .setUserInfo(
          UserInfo.newBuilder()
            .setId(userId)
            .build())
            .build();

      // Set file type for filter to 'DOCUMENT'
      FileType documentFileType = FileType.DOCUMENT;

      // Create a file type filter for documents 
      FileTypeFilter fileTypeFilter = FileTypeFilter.newBuilder()
          .setFileType(documentFileType)
          .build();

      // Create document query to search all documents for text given at input
      DocumentQuery query = DocumentQuery.newBuilder()
          .setQuery(documentQuery)
          .setFileTypeFilter(fileTypeFilter)
          .build();

      /*
       * Create the request to search all documents for specified query. 
       * Please note the offset in this request is to only return the specified number of results 
       * to avoid hitting the API quota. 
       */
      SearchDocumentsRequest searchDocumentsRequest = SearchDocumentsRequest.newBuilder()
          .setParent(parent)
          .setRequestMetadata(requestMetadata)
          .setOffset(5)
          .setDocumentQuery(query)
          .build();

      // Make the call to search documents with document service client and store the response
      SearchDocumentsPagedResponse searchDocumentsPagedResponse = 
          documentServiceClient.searchDocuments(searchDocumentsRequest);

      // Iterate through response and print search results for documents matching the search query
      for (MatchingDocument matchingDocument :
          searchDocumentsPagedResponse.iterateAll()) {
        System.out.println(
            "Display Name: " + matchingDocument.getDocument().getDisplayName()
            + "Document Name: " + matchingDocument.getDocument().getName()
            + "Document Creation Time: " + matchingDocument.getDocument().getCreateTime().toString()
            + "Search Text Snippet: " + matchingDocument.getSearchTextSnippet());
      }
    }
  }

  private static String getProjectNumber(String projectId) throws IOException { 
    /*
     * Initialize client that will be used to send requests. 
     * This client only needs to be created once, and can be reused for multiple requests.
     */
    try (ProjectsClient projectsClient = ProjectsClient.create()) { 
      ProjectName projectName = ProjectName.of(projectId); 
      Project project = projectsClient.getProject(projectName);
      String projectNumber = project.getName(); // Format returned is projects/xxxxxx
      return projectNumber.substring(projectNumber.lastIndexOf("/") + 1);
    } 
  }
}

Node.js

詳細については、Document AI Warehouse Node.js API リファレンスドキュメントをご覧ください。

/**
 * TODO(developer): Uncomment these variables before running the sample.
 * const projectNumber = 'YOUR_PROJECT_NUMBER';
 * const location = 'YOUR_PROJECT_LOCATION'; // Format is 'us' or 'eu'
 * const userId = 'user:xxx@example.com'; // Format is "user:xxx@example.com"
 * const documentQueryText = 'YOUR_DOCUMENT_QUERY'
 */

// Import from google cloud
const {DocumentServiceClient} = require('@google-cloud/contentwarehouse').v1;

const apiEndpoint =
  location === 'us'
    ? 'contentwarehouse.googleapis.com'
    : `${location}-contentwarehouse.googleapis.com`;

// Create service client
const serviceClient = new DocumentServiceClient({apiEndpoint: apiEndpoint});

// Get Document Schema
async function searchDocuments() {
  // Initialize request argument(s)
  const searchRequest = {
    // The full resource name of the location, e.g.:
    // projects/{project_number}/locations/{location}
    parent: `projects/${projectNumber}/locations/${location}`,

    // Document Text Query
    documentQuery: {
      query: documentQueryText,
      // File Type Filter
      fileTypeFilter: {
        fileType: 'DOCUMENT',
      },
    },

    // Histogram Query
    histogramQueries: [
      {
        histogramQuery: 'count("DocumentSchemaId")',
      },
    ],
    requestMetadata: {userInfo: {id: userId}},
  };

  // Make Request
  const response = serviceClient.searchDocuments(searchRequest);

  // Print out response
  response.then(
    result => console.log(`Document Found: ${JSON.stringify(result)}`),
    error => console.log(`${error}`)
  );
}

Python

詳細については、Document AI Warehouse Python API リファレンスドキュメントをご覧ください。


from google.cloud import contentwarehouse

# TODO(developer): Uncomment these variables before running the sample.
# project_number = 'YOUR_PROJECT_NUMBER'
# location = 'YOUR_PROJECT_LOCATION' # Format is 'us' or 'eu'
# document_query_text = 'YOUR_DOCUMENT_QUERY'
# user_id = 'user:YOUR_SERVICE_ACCOUNT_ID' # Format is "user:xxxx@example.com"


def search_documents_sample(
    project_number: str, location: str, document_query_text: str, user_id: str
) -> None:
    # Create a client
    client = contentwarehouse.DocumentServiceClient()

    # The full resource name of the location, e.g.:
    # projects/{project_number}/locations/{location}
    parent = client.common_location_path(project=project_number, location=location)

    # File Type Filter
    # Options: DOCUMENT, FOLDER
    file_type_filter = contentwarehouse.FileTypeFilter(
        file_type=contentwarehouse.FileTypeFilter.FileType.DOCUMENT
    )

    # Document Text Query
    document_query = contentwarehouse.DocumentQuery(
        query=document_query_text,
        file_type_filter=file_type_filter,
    )

    # Histogram Query
    histogram_query = contentwarehouse.HistogramQuery(
        histogram_query='count("DocumentSchemaId")'
    )

    request_metadata = contentwarehouse.RequestMetadata(
        user_info=contentwarehouse.UserInfo(id=user_id)
    )

    # Define request
    request = contentwarehouse.SearchDocumentsRequest(
        parent=parent,
        request_metadata=request_metadata,
        document_query=document_query,
        histogram_queries=[histogram_query],
    )

    # Make the request
    response = client.search_documents(request=request)

    # Print search results
    for matching_document in response.matching_documents:
        document = matching_document.document
        # Display name - schema display name.
        # Name.
        # Create date.
        # Snippet - keywords are highlighted with <b> & </b>.
        print(
            f"{document.display_name} - {document.document_schema_name}\n"
            f"{document.name}\n"
            f"{document.create_time}\n"
            f"{matching_document.search_text_snippet}\n"
        )

    # Print histogram
    for histogram_query_result in response.histogram_query_results:
        print(
            f"Histogram Query: {histogram_query_result.histogram_query}\n"
            f"| {'Schema':<70} | {'Count':<15} |"
        )
        for key, value in histogram_query_result.histogram.items():
            print(f"| {key:<70} | {value:<15} |")

次のステップ

他の Google Cloud プロダクトのコードサンプルを検索およびフィルタするには、Google Cloud サンプルブラウザをご覧ください。

ドキュメントを検索する コレクションでコンテンツを整理 必要に応じて、コンテンツの保存と分類を行います。

もっと見る