Class SQLDataSource

All Implemented Interfaces:
com.isomorphic.base.IAutoConfigurable, com.isomorphic.datasource.Committable, com.isomorphic.datasource.FreeResourcesHandler, com.isomorphic.datasource.IType, com.isomorphic.interfaces.ISQLDataSource, IToJSON, Serializable

public class SQLDataSource extends BasicDataSource implements com.isomorphic.interfaces.ISQLDataSource
A DataSource implementation for SQL database engines. SmartClient Server supports all major database products and allows you to connect a SmartClient databound widget to an existing database table with a few clicks in a simple wizard. DataSources created in this way require zero code and support all four CRUD operations out of the box.

SmartClient Server's SQL support is mature, feature-rich and very customizable. See the client reference documentation for details of configuring and using SQL DataSources - search for "sqlDataSource". For a wider general discussion of data integration options with SmartClient Server, search the client documentation for "serverDataIntegration"

See Also:
  • Method Details

    • getTableColumns

      public List<String> getTableColumns()
      Returns List of column names from the underlying database table DataSource is bound to.

      This is based on the dbName, schema and tableName DataSource attributes to get the list of table columns. Note that the columns in returned list may differ from the DataSource fields as columns in database may have alternative names, there may be extra columns in underlying table or some fields may be missing in underlying table if they are using customSQL features at a DataSource level.

      If, for any reason, DataSource lacks the tableName attribute, getTableColumns() API falls back to returning the getDirectFields().

      Returns:
      List of column names from the underlying database table DataSource is bound to.
    • getPartialWhere

      public static String getPartialWhere(DSRequest dsRequest, String... fieldNames) throws Exception
      Generates partial SQL condition expression for the fieldNames requested for use in the regular whereClause.

      The result is based on the criteria provided in DSRequest parameter and only criteria based on fields that are listed in fieldNames parameter will be generated, everything else will be skipped. If all fields are happen to be requested, then this method returns same SQL expression as is used in $defaultWhereClause Velocity context variable. If there's no criteria for the requested fields, then ('1'='1') will be returned.

      If you want a partial WHERE clause that excludes certain fields instead of naming all included fields, see getWhereWithout(DSRequest, String...). Also see OperationBinding.excludeCriteriaFields for excluding fields from automatic generation.

      If you are using aggregation, and you want to create a WHERE clause that applies to aggregated value, see getPartialHaving(DSRequest, String...) or getHavingWithout(DSRequest, String...).;

      Parameters:
      dsRequest - DSRequest to pick criteria from
      fieldNames - requested field names
      Returns:
      SQL expression filtered by requested fields
      Throws:
      Exception
      See Also:
    • getWhereWithout

      public static String getWhereWithout(DSRequest dsRequest, String... fieldNames) throws Exception
      Generates SQL condition expression without the fieldNames requested to skip from the regular whereClause.

      The result is based on the criteria provided in DSRequest parameter and only criteria based on fields that are not listed in fieldNames parameter will be generated, everything on that list will skipped. If no fields are requested to skip, then this method returns same SQL expression as is used in $defaultWhereClause Velocity context variable. If all fields are requested to be skipped or there's no criteria after the skipping, then ('1'='1') will be returned.

      If you want a partial WHERE clause that includes certain fields instead of naming all excluded fields, see getPartialWhere(DSRequest, String...).

      If you are using aggregation, and you want to create a WHERE clause that applies to aggregated value, see getPartialHaving(DSRequest, String...) or getHavingWithout(DSRequest, String...).;

      Parameters:
      dsRequest - DSRequest to pick criteria from
      fieldNames - field names to skip
      Returns:
      SQL expression filtered by requested fields
      Throws:
      Exception
      See Also:
    • getPartialHaving

      public static String getPartialHaving(DSRequest dsRequest, String... fieldNames) throws Exception
      Generates partial SQL condition expression for the fieldNames requested for use in the groupWhereClause (aka having).

      This is the same as getPartialWhere(DSRequest, String...) except that depending on operationBinding.useHavingClause setting it refers fields by either aliases or full SQL expressions, see OperationBinding.useHavingClause in the SmartClient Reference for details.

      Note that only fields that are selected in the main query may be used.

      Parameters:
      dsRequest - DSRequest to pick criteria from
      fieldNames - requested field names
      Returns:
      SQL expression filtered by requested fields
      Throws:
      Exception
      See Also:
    • getHavingWithout

      public static String getHavingWithout(DSRequest dsRequest, String... fieldNames) throws Exception
      Generates SQL condition expression without the fieldNames requested to skip from groupWhereClause (aka having).

      This is the same as getWhereWithout(DSRequest, String...) except that depending on operationBinding.useHavingClause setting it refers fields by either aliases or full SQL expressions, see OperationBinding.useHavingClause in the SmartClient Reference for details.

      Note that only fields that are selected in the main query may be used.

      Parameters:
      dsRequest - DSRequest to pick criteria from
      fieldNames - field names to skip
      Returns:
      SQL expression filtered by requested fields
      Throws:
      Exception
      See Also:
    • getSQLPrefix

      protected String getSQLPrefix(DSRequest req) throws Exception
      Override this method if you want to programmatically provide a prefix to SQL generated for this DataSource. Scan the client-side docs for sqlPrefix for details. Note, the default implementation of this method returns any operationBinding-level or dataSource-level declared sqlPrefix
      Parameters:
      req - Context DSRequest
      Returns:
      Query prefix text
      Throws:
      Exception
    • getSQLSuffix

      protected String getSQLSuffix(DSRequest req) throws Exception
      Override this method if you want to programmatically provide a suffix to SQL generated for this DataSource. Scan the client-side docs for sqlSuffix for details. Note, the default implementation of this method returns any operationBinding-level or dataSource-level declared sqlSuffix
      Parameters:
      req - Context DSRequest
      Returns:
      Query suffix text
      Throws:
      Exception
    • getSQLClause

      public static String getSQLClause(SQLClauseType clauseType, DSRequest dsRequest) throws Exception
      Returns the text of the generated SQL we will use to implement a clause of the supplied type for the supplied DSRequest. The clause text will be exactly the same as the value of the corresponding $defaultXxxxxClause Velocity variable that you can use when developing your own customSQL queries. Clauses consist of just the variable part and exclude the SQL keywords. The one exception to this is the special clause type "All"; if you specify this clause type, we return a full-formed SQL query suitable for passing straight to the executeQuery() or executeUpdate() method of a JDBC driver or equivalent.

      getSQLClause is not limited to the current DSRequest being processed: you can create a new DSRequest server-side and call getSQLClause() to retrieve the SQL that such a request would generate, which might be used as a sub-select in a larger overall query. This can be particularly useful with aggregated queries.

      Note that we ordinarily generate SQL that is self-contained and context-free, so you can store the clauses that this API returns and use them to construct a query for later execution without any dependence on the DSRequest. However, we are not always able to do this; binary columns and CLOBs cannot be assigned inline, so we have to generate SQL containing '?' markers, which can only be executed as parameterized queries. This in turn means that you need to store the values of any such fields in order to provide them as parameters when you come to execute the query.

      Also note that the query clauses produced by SQLDataSource are database-specific and will only work if run against the same database product they were generated for.

      Parameters:
      clauseType - The type of clause to generate and return
      dsRequest - The DSRequest embodying the fetch or update operation
      Returns:
      The text of the generated SQL we will use to implement a clause of clauseType for the supplied dsRequest
      Throws:
      Exception
    • executeAsBatch

      public static List<Integer> executeAsBatch(List<DSRequest> dsRequests) throws Exception
      Convenience implementation of executeAsBatch(DSRequest[]), if you prefer to pass and return Lists instead of arrays.
      Parameters:
      dsRequests - List of DSRequests to execute as a batch
      Returns:
      List of update counts, each entry pertains to the corresponding (by ordinal position) DSRequest
      Throws:
      Exception
    • executeAsBatch

      public static int[] executeAsBatch(DSRequest[] dsRequests) throws Exception
      Executes an array of DSRequests in a single JDBC Statement.executeBatch() call. The implementation builds up the batch by calling getSQLClause(SQLClauseType,DSRequest) to obtain the query text to execute for each DSRequest, and then adding each query with calls to Statement.addBatch(java.lang.String). This means that many of the sophistications of DSRequest processing are bypassed. Note the following:

      • No DMI, security checks, or other normal consequences of dsRequest.execute() will occur, because the DSRequests are not executed in the normal way
      • executeAsBatch() allocates and frees its own SQL connection, and does not participate in automatic transctions
      • However, the entire batch is committed or rolled back in a single transaction that is specific to the batch, assuming that the underlying database is capaable of transactions

      This method is intended to be used for edge cases where a large number of records must be updated, and the performance overhead of updating them singly makes normal DSRequest processing impractical. exeuteAsBatch() can give dramatic performance improvements in these circumstances, but at the cost of losing much of the functionality of a DSRequest.

      Parameters:
      dsRequests - Array of DSRequests to execute as a batch
      Returns:
      Array of update counts, each entry pertains to the corresponding (by ordinal position) DSRequest
      Throws:
      Exception