Module io.ebean.api
Package io.ebean

Interface DocumentStore

@NonNullApi public interface DocumentStore
Document storage operations.

  • Method Summary

    Modifier and Type
    Method
    Description
    long
    copyIndex(Query<?> query, String newIndex, int bulkBatchSize)
    Copy from a source index to a new index taking only the documents matching the given query.
    long
    copyIndex(Class<?> beanType, String newIndex)
    Copy the index to a new index.
    long
    copyIndex(Class<?> beanType, String newIndex, long sinceEpochMillis)
    Copy entries from an index to a new index but limiting to documents that have been modified since the sinceEpochMillis time.
    void
    createIndex(String indexName, String alias)
    Create an index given a mapping file as a resource in the classPath (similar to DDL create table).
    void
    dropIndex(String indexName)
    Drop the index from the document store (similar to DDL drop table).
    <T> T
    find(DocQueryContext<T> request)
    Return the bean by fetching it's content from the document store.
    <T> void
    findEach(DocQueryContext<T> query, Consumer<T> consumer)
    Execute the query against the document store with the expectation of a large set of results that are processed in a scrolling resultSet fashion.
    void
    findEach(String indexNameType, String rawQuery, Consumer<RawDoc> consumer)
    Find each processing raw documents.
    <T> void
    findEachWhile(DocQueryContext<T> query, Predicate<T> consumer)
    Execute the query against the document store with the expectation of a large set of results that are processed in a scrolling resultSet fashion.
    void
    findEachWhile(String indexNameType, String rawQuery, Predicate<RawDoc> consumer)
    Find each processing raw documents stopping when the predicate returns false.
    <T> List<T>
    findList(DocQueryContext<T> request)
    Execute the find list query.
    <T> PagedList<T>
    findPagedList(DocQueryContext<T> request)
    Execute the query against the document store returning the paged list.
    void
    indexAll(Class<?> beanType)
    Update the document store for all beans of this type.
    <T> void
    indexByQuery(Query<T> query)
    Update the associated document store using the result of the query.
    <T> void
    indexByQuery(Query<T> query, int bulkBatchSize)
    Update the associated document store index using the result of the query additionally specifying a bulkBatchSize to use for sending the messages to ElasticSearch.
    void
    indexSettings(String indexName, Map<String,Object> settings)
    Modify the settings on an index.
    long
    process(List<DocStoreQueueEntry> queueEntries)
    Process the queue entries sending updates to the document store or queuing them for later processing.

  • Method Details

    • indexByQuery

      <T> void indexByQuery(Query<T> query)
      Update the associated document store using the result of the query.

      This will execute the query against the database creating a document for each bean graph and sending this to the document store.

      Note that the select and fetch paths of the query is set for you to match the document structure needed based on @DocStore and @DocStoreEmbedded so what this query requires is the predicates only.

      This query will be executed using findEach so it is safe to use a query that will fetch a lot of beans. The default bulkBatchSize is used.

      Parameters:
      query - The query that selects object to send to the document store.

    • indexByQuery

      <T> void indexByQuery(Query<T> query, int bulkBatchSize)
      Update the associated document store index using the result of the query additionally specifying a bulkBatchSize to use for sending the messages to ElasticSearch.
      Parameters:
      query - The query that selects object to send to the document store.
      bulkBatchSize - The batch size to use when bulk sending to the document store.

    • indexAll

      void indexAll(Class<?> beanType)
      Update the document store for all beans of this type.

      This is the same as indexByQuery where the query has no predicates and so fetches all rows.

    • find

      @Nullable <T> T find(DocQueryContext<T> request)
      Return the bean by fetching it's content from the document store. If the document is not found null is returned.

      Typically this is called indirectly by findOne() on the query. 

      

 Customer customer =
   database.find(Customer.class)
     .setUseDocStore(true)
     .setId(42)
     .findOne();

    • findList

      <T> List<T> findList(DocQueryContext<T> request)
      Execute the find list query. This request is prepared to execute secondary queries.

      Typically this is called indirectly by findList() on the query that has setUseDocStore(true). 

      

 List<Customer> newCustomers =
  database.find(Customer.class)
    .setUseDocStore(true)
    .where().eq("status, Customer.Status.NEW)
    .findList();

    • findPagedList

      <T> PagedList<T> findPagedList(DocQueryContext<T> request)
      Execute the query against the document store returning the paged list.

      The query should have firstRow or maxRows set prior to calling this method.

      Typically this is called indirectly by findPagedList() on the query that has setUseDocStore(true). 

      

 PagedList<Customer> newCustomers =
  database.find(Customer.class)
    .setUseDocStore(true)
    .where().eq("status, Customer.Status.NEW)
    .setMaxRows(50)
    .findPagedList();

    • findEach

      <T> void findEach(DocQueryContext<T> query, Consumer<T> consumer)
      Execute the query against the document store with the expectation of a large set of results that are processed in a scrolling resultSet fashion.

      For example, with the ElasticSearch doc store this uses SCROLL.

      Typically this is called indirectly by findEach() on the query that has setUseDocStore(true). 

      

  database.find(Order.class)
    .setUseDocStore(true)
    .where()... // perhaps add predicates
    .findEach((Order order) -> {
      // process the bean ...
    });

    • findEachWhile

      <T> void findEachWhile(DocQueryContext<T> query, Predicate<T> consumer)
      Execute the query against the document store with the expectation of a large set of results that are processed in a scrolling resultSet fashion.

      Unlike findEach() this provides the opportunity to stop iterating through the large query.

      For example, with the ElasticSearch doc store this uses SCROLL.

      Typically this is called indirectly by findEachWhile() on the query that has setUseDocStore(true). 

      

  database.find(Order.class)
    .setUseDocStore(true)
    .where()... // perhaps add predicates
    .findEachWhile(new Predicate<Order>() {
      @Override
      public void accept(Order bean) {
        // process the bean

        // return true to continue, false to stop
        // boolean shouldContinue = ...
        return shouldContinue;
      }
    });

    • findEach

      void findEach(String indexNameType, String rawQuery, Consumer<RawDoc> consumer)
      Find each processing raw documents.
      Parameters:
      indexNameType - The full index name and type
      rawQuery - The query to execute
      consumer - Consumer to process each document

    • findEachWhile

      void findEachWhile(String indexNameType, String rawQuery, Predicate<RawDoc> consumer)
      Find each processing raw documents stopping when the predicate returns false.
      Parameters:
      indexNameType - The full index name and type
      rawQuery - The query to execute
      consumer - Consumer to process each document until false is returned

    • process

      long process(List<DocStoreQueueEntry> queueEntries) throws IOException
      Process the queue entries sending updates to the document store or queuing them for later processing.
      Throws:
      IOException

    • dropIndex

      void dropIndex(String indexName)
      Drop the index from the document store (similar to DDL drop table). 
      

   DocumentStore documentStore = database.docStore();

   documentStore.dropIndex("product_copy");

    • createIndex

      void createIndex(String indexName, String alias)
      Create an index given a mapping file as a resource in the classPath (similar to DDL create table). 
      

   DocumentStore documentStore = database.docStore();

   // uses product_copy.mapping.json resource
   // ... to define mappings for the index

   documentStore.createIndex("product_copy", null);
      Parameters:
      indexName - the name of the new index
      alias - the alias of the index

    • indexSettings

      void indexSettings(String indexName, Map<String,Object> settings)
      Modify the settings on an index.

      For example, this can be used be used to set elasticSearch refresh_interval on an index before a bulk update. 

      

   // refresh_interval -1 ... disable refresh while bulk loading

   Map<String,Object> settings = new LinkedHashMap<>();
   settings.put("refresh_interval", "-1");

   documentStore.indexSettings("product", settings);

 
      

   // refresh_interval 1s ... restore after bulk loading

   Map<String,Object> settings = new LinkedHashMap<>();
   settings.put("refresh_interval", "1s");

   documentStore.indexSettings("product", settings);
      Parameters:
      indexName - the name of the index to update settings on
      settings - the settings to set on the index

    • copyIndex

      long copyIndex(Class<?> beanType, String newIndex)
      Copy the index to a new index.

      This copy process does not use the database but instead will copy from the source index to a destination index. 

      

  long copyCount = documentStore.copyIndex(Product.class, "product_copy");
      Parameters:
      beanType - The bean type of the source index
      newIndex - The name of the index to copy to
      Returns:
      the number of documents copied to the new index

    • copyIndex

      long copyIndex(Class<?> beanType, String newIndex, long sinceEpochMillis)
      Copy entries from an index to a new index but limiting to documents that have been modified since the sinceEpochMillis time.

      To support this the document needs to have a @WhenModified property. 

      

  long copyCount = documentStore.copyIndex(Product.class, "product_copy", sinceMillis);
      Parameters:
      beanType - The bean type of the source index
      newIndex - The name of the index to copy to
      Returns:
      the number of documents copied to the new index

    • copyIndex

      long copyIndex(Query<?> query, String newIndex, int bulkBatchSize)
      Copy from a source index to a new index taking only the documents matching the given query. 
      

  // predicates to select the source documents to copy
  Query<Product> query = database.find(Product.class)
    .where()
      .ge("whenModified", new Timestamp(since))
      .ge("name", "A")
      .lt("name", "D")
      .query();

  // copy from the source index to "product_copy" index
  long copyCount = documentStore.copyIndex(query, "product_copy", 1000);
      Parameters:
      query - The query to select the source documents to copy
      newIndex - The target index to copy the documents to
      bulkBatchSize - The ElasticSearch bulk batch size, if 0 uses the default.
      Returns:
      The number of documents copied to the new index.