XML columns in a geodatabase in SQL Server
XML is an open standard for defining data elements within documents. To store XML data in a Microsoft SQL Server database, you can use native XML or ArcSDE XML columns.
You can store user-defined XML documents in either type of XML. The XML_COLUMN_STORAGE DBTUNE parameter controls the type of XML used. By default, this parameter is set to DB_XML. Therefore, by default, the native SQL Server XML data type is used.
If you intend to use XPath queries on native SQL Server XML columns, you must also create full-text catalogs. See your SQL Server documentation for more information.
ArcSDE XML data types require the SQL Server database be enabled for full-text searching and a full-text catalog be created. See Preparing SQL Server to store ArcSDE XML columns and Creating a full-text catalog in SQL Server for information on how to do this.
ArcSDE XML data types are used to store collections of metadata documents to support ArcIMS Metadata Services. Therefore, if you use ArcIMS Metadata Services, you must set the XML_COLUMN_STORAGE parameter to SDE_XML.
XML columns in ArcGIS Desktop
XML columns are not completely supported in the geodatabase. Therefore, the following are true:
- You cannot preview the contents of an XML column in a table in ArcCatalog. The column will be blank or have unintelligible characters in it.
- On the Fields tab of the table Properties dialog box, no data type is listed for an XML column.
- You cannot copy and paste a table containing an XML column from one geodatabase to another.
- If you import or export the table using geoprocessing tools, you have to exclude the XML column.
- You cannot register as versioned a table with an XML column using ArcCatalog.
- If you open a table that contains an XML column in ArcMap, the XML column appears blank.
- To edit a table that contains an XML column in ArcMap, you can use a nonversioned editing session. However, you cannot edit the XML column itself, only the other fields in the table.
XML columns in a SQL Server DBMS
Three ArcSDE system tables are used to manage XML columns: SDE_xml columns, SDE_xml_index_tags, and SDE_xml_indexes. These tables are owned by the ArcSDE administrator user. ArcSDE also creates two additional tables for each XML column that are used to store and index XML documents: the SDE_xml_document and SDE_xml_xpath_index tables. These tables are owned by the user who owns the business table containing the XML column.
ArcSDE creates the following tables, which are used to store and index XML documents.
Do not alter any of these tables using SQL.
When you add an XML column to a business table, a row is added to the XML columns table. This table occurs once in each ArcSDE database.
Field name |
Field type |
Description |
Null? |
---|---|---|---|
column_id |
int |
The XML column's identifier and the table's primary key This value is assigned by ArcSDE at the time the XML column is created. |
NOT NULL |
registration_id |
int |
The identifier of the business table containing the XML column and foreign key to the SDE_table_registry system table |
NOT NULL |
column_name |
nvarchar(32) |
The name of the column that is the XML column in the business table |
NOT NULL |
index_id |
int |
The identifier of the XPath index associated with the XML column (if one exists) and foreign key to the SDE_xml_indexes table |
|
minimum_id |
int |
The value of the initial number used in the business table's XML column to identify individual XML documents |
|
config_keyword |
nvarchar(32) |
The DBTUNE configuration keyword whose parameters determine how the XML document, the XML XPath index tables, and the text indexes created on those tables are defined in the database See Parameters for ArcSDE XML in SQL Server for more information. |
|
xflags |
int |
A value indicating whether the original documents in the XML document table are stored compressed or decompressed By default, documents are compressed; compressed documents provide better performance. |
NOT NULL |
This table occurs once in each ArcSDE database. It contains one row for each XML column that has an XPath index.
Field name |
Field type |
Description |
Null? |
---|---|---|---|
index_id |
int |
The identifier of the XPath index and the table's primary key |
NOT NULL |
index_name |
nvarchar(32) |
The name of the XPath index For XPath indexes associated with an ArcIMS Metadata Service, the name will be ims_xml#, where # is the identifier of the XML column in the Metadata Service's business table. |
NOT NULL |
owner |
nvarchar(32) |
The database user who owns the XML column For ArcIMS Metadata Services, this is the user specified in the service's ArcXML configuration file. |
NOT NULL |
index_type |
int |
A value indicating the type of XPath index With ArcSDE 9.1 and higher, the value will be 2 for the index type SE_XML_INDEX_DEFINITION and 1 for the index type SE_XML_INDEX_TEMPLATE. For XPath indexes associated with an ArcIMS Metadata Service, only the index type SE_XML_INDEX_DEFINITION is supported. |
NOT NULL |
description |
nvarchar(64) |
Text identifying the XPath index If an index definition file was used to create the index, the index description can be specified at the top of the file. |
An XML column can optionally have an XPath index, which lets you search the content of a specific XML element or attribute in each document. The definition of which elements and attributes are included in or excluded from each XPath index is recorded in this table.
This table occurs once in each ArcSDE database. It contains one row for each XPath associated with an XML column's XPath index.
Field name |
Field type |
Description |
Null? |
---|---|---|---|
index_id |
int |
The identifier of the XPath index associated with an XML column (if one exists) and foreign key to the SDE_xml_indexes table |
NOT NULL |
tag_id |
int |
The identifier of an XPath or tag |
NOT NULL |
tag_name |
nvarchar(1024) |
An absolute XPath identifying an XML element or attribute that may occur in an XML document For example, /metadata/mdDateSt identifies an XML element and /metadata/dataIdInfo/tpCat/TopicCatCd/@value identifies an XML attribute. These XPaths must not contain asterisks (*) to refer to a group of XML elements or attributes—each element or attribute is matched exactly using the XPaths specified in this table. Elements can't be named * in a valid XML document. |
NOT NULL |
data_type |
int |
A value indicating whether the XML element or attribute will be indexed as a number, a varchar, or text A 1 indicates the content of the tag will be indexed as text; a 2 indicates the content of the tag will be indexed as a number; a 3 indicates the content of the tag will be indexed as a varchar. |
NOT NULL |
tag_alias |
int |
A number that may be used to identify an XPath For example, the Z39.50 communication protocol uses numeric codes to refer to content that may be searched. This column is not used by the ArcIMS Z39.50 Connector. |
|
description |
nvarchar(64) |
Text identifying the content that should be contained in the XML element or attribute |
|
is_excluded |
int |
A value indicating whether the XML element is included in or excluded from the XPath index 0 = the XPath is included; 1 = the XPath is excluded. |
NOT NULL |
The SDE_xml_doc<column_id> table stores the XML document and maintains a full-text index on the document's content. The ArcSDE database contains one of these tables for each XML column. The number in the table name is the XML column's identifier. This table contains one row for each XML document stored in the XML column.
Field name |
Field type |
Description |
Null? |
---|---|---|---|
sde_xml_id |
int |
The identifier for an XML document stored in the XML column and primary key for the table |
NOT NULL |
doc_property |
int |
A value indicating whether any conflicts were found when adding the content of an XML document to the XPath index 1 = A conflict was found; for example, when an element is supposed to be indexed numerically but the document contains a string in that element instead. NULL value = There were no problems indexing the document. |
NOT NULL |
xml_doc |
varbinarymax |
The XML document |
NOT NULL |
xml_doc_val |
varbinarymax |
The content of the entire XML document with all XML tags and other markup removed A text index is built on this column by default; this index is used to respond to full text queries. For ArcIMS Metadata Services, this index is used to respond to FULLTEXT requests. |
NOT NULL |
sde_time_stamp |
timestamp |
Used to support incremental updates to the text index |
NOT NULL |
The SDE_xml_idx<column_id> table is created for XML columns that have an XPath text index. This table stores the text or number content for each XPath that is indexed.
The ID number in the table name is the internal registration number for the XML column.
Field name |
Field type |
Description |
Null? |
---|---|---|---|
xml_key_column |
int |
The identifier for the indexed value and primary key for the table |
NOT NULL |
sde_xml_id |
int |
The identifier for the XML document that contains the indexed value |
NOT NULL |
tag_id |
int |
The identifier for the tag associated with the XML column's XPath index, which identifies where in the document the value is stored |
NOT NULL |
double_tag |
float |
The indexed value, when the tag is defined as DOUBLE in the XPath index definition |
|
string_tag |
nvarchar(256) |
The indexed value, when the tag is defined as VARCHAR in the XPath index definition |
|
text_tag |
ntext |
The indexed value, when the tag is defined as STRING in the XPath index definition |
|
sde_time_stamp |
timestamp |
Used to support incremental updates to the text index |
NOT NULL |
The following is a diagram of a table with an XML column and the system tables used to track it. Dashed lines indicate implicit relationships; a solid line denotes explicitly defined relationships between tables.
XML columns in an XML document
You cannot export a table containing an XML column to an XML workspace document. You can export it to an XML recordset document, but there is nothing within the document to distinguish the column as XML.