Oracle® Data Provider for .NET Developer's Guide 11g Release 1 (11.1) Part Number B28375-01 |
|
|
View PDF |
ODP.NET allows the extraction of data from relational and object-relational tables and views as XML documents. The use of XML documents for insert, update, and delete operations to the database is also allowed. Oracle Database supports XML natively in the database, through Oracle XML DB, a distinct group of technologies related to high-performance XML storage and retrieval. Oracle XML DB is an evolution of the database that encompasses both SQL and XML data models in a highly interoperable manner, providing native XML support.
For samples related to ODP.NET XML support, see the following directory:
ORACLE_BASE\ORACLE_HOME\ODP.NET\Samples
This section includes these topics:
XML support in ODP.NET provides the ability to do the following:
Store XML data natively in the database as the Oracle database native type, XMLType
.
Access relational and object-relational data as XML data from an Oracle Database instance into the Microsoft .NET environment, and process the XML using the Microsoft .NET Framework.
Save changes to the database using XML data.
See Also:
"XQuery Support"For the .NET application developer, these features include the following:
Enhancements to the OracleCommand
, OracleConnection
, and OracleDataReader
classes.
The following XML-specific classes:
OracleXmlType
OracleXmlType
objects are used to retrieve Oracle native XMLType
data.
OracleXmlStream
OracleXmlStream
objects are used to retrieve XML data from OracleXmlType
objects as a read-only .NET Stream
object.
OracleXmlQueryProperties
OracleXmlQueryProperties
objects represent the XML properties used by the OracleCommand
class when the XmlCommandType
property is Query
.
OracleXmlSaveProperties
OracleXmlSaveProperties
objects represent the XML properties used by the OracleCommand
class when the XmlCommandType
property is Insert
, Update
, or Delete
.
Beginning with Oracle Database 10g release 2(10.2), ODP.NET supports the XQuery language through a native implementation of SQL/XML functions, XMLQuery
and XMLTable
. When executing XQuery statements, Oracle XML DB generally evaluates XQuery expressions by compiling them into the same underlying structures as relational queries. Queries are optimized, leveraging both relational-database and XQuery-specific optimization technologies, so that Oracle XML DB serves as a native XQuery engine.The treatment of all XQuery expressions, whether natively compiled or evaluated functionally, is transparent: programmers do not need to change their code to take advantage of XQuery optimizations.
See Also:
Oracle XML DB Developer's Guide to learn more about Oracle's XQuery supportThe read-only Connection
property of the OracleXmlType
class holds a reference to the OracleConnection
object used to instantiate the OracleXmlType
class.
How the OracleXmlType
object obtains a reference to an OracleConnection
object depends on how the OracleXmlType
class is instantiated:
Instantiated from an OracleDataReader
class using the GetOracleXmlType
, GetOracleValue
, or GetOracleValues
method:
The Connection
property is set with a reference to the same OracleConnection
object used by the OracleDataReader
object.
Instantiated by invoking an OracleXmlType
constructor with one of the parameters of type OracleConnection
:
The Connection
property is set with a reference to the same OracleConnection
object provided in the constructor.
Instantiated by invoking an OracleXmlType(OracleClob)
constructor:
The Connection
property is set with a reference to the OracleConnection
object used by the OracleClob
object.
An OracleXmlType
object that is associated with one connection cannot be used with a different connection. For example, if an OracleXmlType
object is obtained using OracleConnection
A
, that OracleXmlType
object cannot be used as an input parameter of a command that uses OracleConnection
B
. By checking the Connection
property of the OracleXmlType
objects, the application can ensure that OracleXmlType
objects are used only within the context of the OracleConnection
referenced by its connection property. Otherwise, ODP.NET raises an exception.
Updating XMLType
columns does not require a transaction. However, encapsulating the entire database update process within a transaction is highly recommended. This allows the updates to be rolled back if there are any errors.
XMLType
columns in the database can be updated using Oracle Data Provider for .NET in a few ways:
If the XMLType
column is fetched into the DataSet
, the XMLType
data is represented as a .NET String
.
Modifying XMLType
data in the DataSet
does not require special treatment. XMLType
data can be modified in the same way as any data that is stored in the DataSet
. When a change is made and the OracleDataAdapter.Update
method is invoked, the OracleDataAdapter
object ensures that the XMLType
data is handled properly. The OracleDataAdapter
object uses any custom SQL INSERT
, UPDATE
, or DELETE
statements that are provided. Otherwise, valid SQL statements are generated by the OracleCommandBuilder
object as needed to flush the changes to the database.
The OracleCommand
class provides a powerful way of updating XMLType
data, especially with the use of an OracleParameter
object. To update columns in a database table, the new value for the column can be passed as an input parameter of a command.
To update an XMLType
column in the database, a SQL statement can be executed using static values. In addition, input parameters can be bound to SQL statements, anonymous PL/SQL blocks, or stored procedures to update XMLType
columns. The parameter value can be set as .NET Framework Types, ODP.NET Types, or OracleXmlType
objects.
While XMLType
columns can be updated using an OracleXmlType
object, having an instance of an OracleXmlType
class does not guarantee that the XMLType
column in the database can be updated.
Applications can set an XMLType
column in the database to a NULL
value, with or without input binding, as follows:
Setting NULL
values in an XMLType
column with input binding
To set the XMLType
column to NULL
, the application can bind an input parameter whose value is DBNull
.Value
. This indicates to the OracleCommand
object that a NULL
value is to be inserted.
Passing in a null OracleXmlType
object as an input parameter does not insert a NULL
value into the XMLType
column. In this case, the OracleCommand
object raises an exception.
Setting NULL
Values in an XMLType
Column without input binding
The following example demonstrates setting NULL
values in an XMLType
column without input binding:
// Create a table with an XMLType column in the database CREATE TABLE XML_TABLE(NUM_COL number, XMLTYPE_COL xmltype);
An application can set a NULL
value in the XMLType
column by explicitly inserting a NULL
value or by not inserting anything into that column as in the following examples:
insert into xml_table(xmltype_col) values(NULL);
update xml_table t set t.xmltype_col=NULL;
The XMLType
column can be initialized with empty XML data, using a SQL statement:
// Create a table with an XMLType column in the database CREATE TABLE XML_TABLE(NUM_COL number, XMLTYPE_COL xmltype);
INSERT INTO XML_TABLE (NUM_COL, XMLTYPE_COL) VALUES (4, XMLType.createxml('<DOC/>'));
The following are ways that XML data can be updated in an OracleXmlType
object.
The XML data can be updated by passing an XPATH expression and the new value to the Update
method on the OracleXmlType
object.
The XML data can be retrieved on the client side as the .NET Framework XmlDocument
object using the GetXmlDocument
method on the OracleXmlType
object. This XML data can then be manipulated using suitable .NET Framework classes. A new OracleXmlType
can be created with the updated XML data from the .NET Framework classes. This new OracleXmlType
is bound as an input parameter to an update or insert statement.
The following characters in Table 3-15 have special meaning in XML. For more information, refer to the XML 1.0 specifications
Table 3-15 Characters with Special Meaning in XML
Character | Meaning in XML | Entity Encoding |
---|---|---|
< |
Begins an XML tag |
< |
> |
Ends an XML tag |
> |
" |
Quotation mark |
" |
' |
Apostrophe or single quotation mark |
' |
& |
Ampersand |
& |
When these characters appear as data in an XML element, they are replaced with their equivalent entity encoding.
Also certain characters are not valid in XML element names. When SQL identifiers (such as column names) are mapped to XML element names, these characters are converted to a sequence of hexadecimal digits, derived from the Unicode encoding of the character, bracketed by an introductory underscore, a lowercase x
and a trailing underscore. A blank space is not a valid character in an XML element name. If a SQL identifier contains a space character, then in the corresponding XML element name, the space character is replaced by _x0020_
, which is based on Unicode encoding of the space character.
This section discusses retrieving the result set from a SQL query as XML data.
Table 3-16 lists the date and time format handling when retrieving data, for different database releases.
Table 3-16 Date and Time Format Handling When Retrieving Data
Database Release | Date and Time Format Supported |
---|---|
Oracle9i release 2 (9.2.x) and Oracle Database 10g |
Oracle
If the result XML document is used to save changes back to the database, then all To do this, before the query is executed, the application must explicitly perform an
|
Oracle Database 10g release 2 (10.2) or later |
The generated XML For more information on the XML Schema specification, see
|
If the data in any of the select list columns in the query contains any characters with special meaning in XML (see Table 3-15), these characters are replaced with their corresponding entity encoding in the result XML document.
The following examples demonstrate how ODP.NET handles the angle bracket characters in the column data:
/* Database Setup connect scott/tiger@oracle drop table specialchars; create table specialchars ("id" number, name varchar2(255)); insert into specialchars values (1, '<Jones>'); commit; */ // C# using System; using System.Data; using System.Xml; using Oracle.DataAccess.Client; class QueryResultAsXMLSample { static void Main() { OracleConnection con = new OracleConnection(); con.ConnectionString = "User Id=scott;Password=tiger;Data Source=oracle;"; con.Open(); // Create the command OracleCommand cmd = new OracleCommand("", con); // Set the XML command type to query. cmd.XmlCommandType = OracleXmlCommandType.Query; // Set the SQL query cmd.CommandText = "select * from specialchars"; // Set command properties that affect XML query behavior. cmd.BindByName = true; // Set the XML query properties cmd.XmlQueryProperties.MaxRows = -1; // Get the XML document as an XmlReader. XmlReader xmlReader = cmd.ExecuteXmlReader(); XmlDocument xmlDocument = new XmlDocument(); xmlDocument.PreserveWhitespace = true; xmlDocument.Load(xmlReader); Console.WriteLine(xmlDocument.OuterXml); // Close and Dispose OracleConnection object con.Close(); con.Dispose(); } }
The following XML document is generated for that table: The XML entity encoding that represents the angle brackets appears in bold.
<?xml version = '1.0'?> <ROWSET> <ROW> <id>1</id > <NAME><Jones></NAME> </ROW> </ROWSET>
If a table or view name has any non-alphanumeric characters other than an underscore (_), the table or view name must be enclosed in quotation marks.
For example, to select all entries from a table with the name test'ing
, the CommandText
property of the OracleCommand
object must be set to the following string:
"select * from \"test'ing\"";
The mapping of SQL identifiers (column names) to XML element names is case- sensitive, and the element names are in exactly the same case as the column names of the table or view.
However, the root tag and row tag names are case-insensitive. The following example demonstrates case-sensitivity in this situation:
//Create the following table create table casesensitive_table ("Id" number, NAME varchar2(255)); //insert name and id insert into casesensitive_table values(1, 'Smith');
The following XML document is generated:
<?xml version = '1.0'?> <ROWSET> <ROW> <Id>1</Id> <NAME>Smith</NAME> </ROW> </ROWSET>
Note that the element name for the Id
column matches the case of the column name.
For each row generated by the SQL query, the SQL identifier (column name) maps to an XML element in the generated XML document, as shown in the following example:
// Create the following table create table emp_table (EMPLOYEE_ID NUMBER(4), LAST_NAME varchar2(25)); // Insert some data insert into emp_table values(205, 'Higgins');
The SQL query, SELECT
*
FROM
EMP_TABLE
, generates the following XML document:
<?XML version="1.0"?> <ROWSET> <ROW> <EMPLOYEE_ID>205</EMPLOYEE_ID> <LAST_NAME>Higgins</LAST_NAME> </ROW> </ROWSET>
The EMPLOYEE_ID
and LAST_NAME
database columns of the employees
table map to the EMPLOYEE_ID
and LAST_NAME
elements of the generated XML document.
This section demonstrates how Oracle database handles the mapping of SQL identifiers to XML element names, when retrieving query results as XML from the database. The demonstration uses the specialchars
table involving the some
id
column.
// Create the specialchars table create table specialchars ("some id" number, name varchar2(255));
Note that the specialchars
table has a column named some
id
that contains a blank space character. The space character is not allowed in an XML element name.
When retrieving the query results as XML, the SQL identifiers in the query select list can contain characters that are not valid in XML element names. When these SQL identifiers (such as column names) are mapped to XML element names, each of these characters is converted to a sequence of hexadecimal digits, derived from the Unicode encoding of the characters, bracketed by an introductory underscore, a lowercase x, and a trailing underscore.
Thus, the SQL query in the following example can be used to get a result as an XML document from the specialchars
table:
select "some id", name from specialchars;
See Also:
"Characters with Special Meaning in XML"You can improve the default mapping of SQL identifiers to XML element names by using the following techniques:
Modify the source. Create an object-relational view over the source schema, and make that view the new source.
Use cursor subqueries and cast-multiset constructs in the SQL query.
Create an alias for the column or attribute names in the SQL query. Prefix the aliases with an at sign (@) to map them to XML attributes instead of XML elements.
Modify the XML document. Use Extensible Stylesheet Language Transformation (XSLT) to transform the XML document. Specify the XSL document and parameters. The transformation is done automatically after the XML document is generated from the relational data. Note that this may have an impact on performance.
Specify the name of the root tag and row tag used in the XML document.
ODP.NET can generate an XML document for data stored in object-relational columns, tables, and views, as shown in the following example:
// Create the following tables and types CREATE TYPE "EmployeeType" AS OBJECT (EMPNO NUMBER, ENAME VARCHAR2(20)); / CREATE TYPE EmployeeListType AS TABLE OF "EmployeeType"; / CREATE TABLE mydept (DEPTNO NUMBER, DEPTNAME VARCHAR2(20), EMPLIST EmployeeListType) NESTED TABLE EMPLIST STORE AS EMPLIST_TABLE; INSERT INTO mydept VALUES (1, 'depta', EmployeeListType("EmployeeType"(1, 'empa')));
The following XML document is generated for the table:
<?xml version = "1.0"?> <ROWSET> <ROW> <DEPTNO>1</DEPTNO> <DEPTNAME>depta</DEPTNAME> <EMPLIST> <EmployeeType> <EMPNO>1</EMPNO> <ENAME>empa</ENAME> </EmployeeType> </EMPLIST> </ROW> </ROWSET>
ODP.NET encloses each item in a collection element, with the database type name of the element in the collection. The mydept
table has a collection in the EMPLIST
database column and each item in the collection is of type EmployeeType
. Therefore, in the XML document, each item in the collection is enclosed in the type name EmployeeType
, which appears in bold in the example.
This section discusses making changes to the database data using XML.
Table 3-17 lists the date and time format handling when saving data, for different database releases.
Table 3-17 Date and Time Format Handling When Saving Data
Database Release | Date and Time Format Supported |
---|---|
Oracle9i release 2 (9.2.x) and Oracle Database 10g |
All The following string is the ISO Date and Time Format notation represented in the Oracle Date and Time Format notation: In addition to using the ISO Format notation in the XML document, before the save is executed, the application must explicitly perform an
|
Oracle Database 10g release 2 (10.2) or later |
The generated XML For more information on the XML Schema specification, see
|
Changes can be saved to database tables and views using XML data. However, insert, update, and delete operations cannot be combined in a single XML document. ODP.NET cannot accept a single XML document and determine which are insert, update, or delete changes.
The insert change must be in an XML document containing only rows to be inserted, the update changes only with rows to be updated, and the delete changes only with rows to be deleted.
For example, using the employees
table that comes with the HR sample schema, you can specify the following query:
select employee_id, last_name from employees where employee_id = 205;
The following XML document is generated:
<?xml version = '1.0'?>
<ROWSET>
<ROW>
<EMPLOYEE_ID>205</EMPLOYEE_ID>
<LAST_NAME>Higgins</LAST_NAME>
</ROW>
</ROWSET>
To change the name of employee 205
from Higgins
to Smith
, specify the employees
table and the XML data containing the changes as follows:
<?xml version = '1.0'?>
<ROWSET>
<ROW>
<EMPLOYEE_ID>205</EMPLOYEE_ID>
<LAST_NAME>Smith</LAST_NAME>
</ROW>
</ROWSET>
If the data in any of the elements in the XML document contains characters that have a special meaning in XML (see Table 3-15), these characters must be replaced with appropriate entity encoding, or be preceded by an escape character in the XML document, so that the data is stored correctly in the database table column. Otherwise, ODP.NET throws an exception.
The following example demonstrates how ODP.NET handles the angle bracket special characters in the column data, using entity encoding:
// Create the following table create table specialchars ("id" number, name varchar2(255));
The following XML document can be used to insert values (1,
'<Jones>')
into the specialchars
table. The XML entity encoding that represents the angle brackets appears in bold.
<?xml version = '1.0'?> <ROWSET> <ROW> <id>1</id > <NAME><Jones></NAME> </ROW> </ROWSET>
If a table or view name has any non-alphanumeric characters other than an underscore (_), the table or view name must be enclosed in quotation marks.
For example, to save changes to a table with the name test'ing
, the OracleCommand.XmlSaveProperties.TableName
property must be set to "\"test'ing\""
.
For each XML element that represents a row of data in the XML document, the child XML elements map to database column names. The mapping of the child element name to the column name is always case-sensitive, but the root tag and row tag names are case-insensitive. The following example demonstrates this case-sensitivity:
//Create the following table create table casesensitive_table ("Id" number, NAME varchar2(255));
The following XML document can be used to insert values (1
, Smith
) into the casesensitive_table
:
<?xml version = '1.0'?> <ROWSET> <ROW> <Id>1</Id> <NAME>Smith</NAME> </ROW> </ROWSET>
Note that the element name for the Id
column matches the case of the column name.
This section describes how Oracle database handles the mapping of XML element names to column names when using XML for data manipulation in the database. The following specialchars
table involving the some
id
column demonstrates this handling.
// Create the specialchars table create table specialchars ("some id" number, name varchar2(255));
Note that the specialchars
table has a column named some
id
that contains a blank space character. The space character is not allowed in an XML element name.
When an XML document is used to save changes to a table or view, the OracleCommand.XmlSaveProperties.UpdateColumnsList
property is used to specify the list of columns to update or insert.
When an XML document is used to save changes to a column in a table or view, and the corresponding column name contains any of the characters that are not valid in an XML element name, the escaped column name must be specified in the UpdateColumnsList
property as in the following example.
The following XML document can be used to insert values (2
, <Jones>
) into the specialchars
table:
<?xml version = '1.0'?> <ROWSET> <ROW> <some_x0020_id>2</some_x0020_id> <NAME><Jones></NAME> </ROW> </ROWSET>
The following example specifies the list of columns to update or insert:
/* Database Setup connect scott/tiger@oracle drop table specialchars; create table specialchars ("some id" number, name varchar2(255)); insert into specialchars values (1, '<Jones>'); commit; */ // C# using System; using System.Data; using System.Xml; using Oracle.DataAccess.Client; class InsertUsingXmlDocSample { static void Main() { OracleConnection con = new OracleConnection(); con.ConnectionString = "User Id=scott;Password=tiger;Data Source=oracle;"; con.Open(); Console.WriteLine("Connected Successfully"); // Create the command OracleCommand cmd = new OracleCommand("", con); // Set the XML command type to query. cmd.XmlCommandType = OracleXmlCommandType.Insert; // Set the XML document cmd.CommandText = "<?xml version = '1.0'?>\n" + "<ROWSET>\n" + "<ROW>\n" + "<some_x0020_id>2</some_x0020_id>\n" + "<NAME><Jones></NAME>\n" + "</ROW>\n" + "</ROWSET>\n"; cmd.XmlSaveProperties.Table = "specialchars"; string[] ucols = new string[2]; ucols[0] = "some_x0020_id"; ucols[1] = "NAME"; cmd.XmlSaveProperties.UpdateColumnsList = ucols; // Insert rows int rows = cmd.ExecuteNonQuery(); Console.WriteLine("Number of rows inserted successfully : {0} ", rows); // Close and Dispose OracleConnection object con.Close(); con.Dispose(); } }
You can improve the default mapping by using the following techniques:
Modify the target. Create an object-relational view over the target schema, and make the view the new target.
Modify the XML document. Use XSLT to transform the XML document. Specify the XSL document and parameters. The transformation is done before the changes are saved. Note that this is may have an impact on performance.
Specify the name of the row tag used in the XML document.
Changes in an XML document can also be saved to object-relational data. Each item in a collection can be specified in one of the following ways in the XML document:
By enclosing the database type name of the item as the XML element name.
By enclosing the name of the database column holding the collection with _ITEM
appended as the XML element name.
Oracle Database does not save changes to multiple relational tables that have been joined together. Oracle recommends that you create a view on those relational tables, and then update that view. If the view cannot be updated, triggers can be used instead.
See Also:
Oracle Database SQL Language Reference for the description and syntax of theCREATE
VIEW
statement