[MS-SSCLRT]:
Microsoft SQL Server CLR Types Serialization Formats
Intellectual Property Rights Notice for Open Specifications Documentation


Technical Documentation. Microsoft publishes Open Specifications documentation for
protocols, file formats, languages, standards as well as overviews of the interaction among each
of these technologies.



Copyrights. This documentation is covered by Microsoft copyrights. Regardless of any other
terms that are contained in the terms of use for the Microsoft website that hosts this
documentation, you may make copies of it in order to develop implementations of the
technologies described in the Open Specifications and may distribute portions of it in your
implementations using these technologies or your documentation as necessary to properly
document the implementation. You may also distribute in your implementation, with or without
modification, any schema, IDL’s, or code samples that are included in the documentation. This
permission also applies to any documents that are referenced in the Open Specifications.



No Trade Secrets. Microsoft does not claim any trade secret rights in this documentation.



Patents. Microsoft has patents that may cover your implementations of the technologies
described in the Open Specifications. Neither this notice nor Microsoft's delivery of the
documentation grants any licenses under those or any other Microsoft patents. However, a given
Open Specification may be covered by Microsoft Open Specification Promise or the Community
Promise. If you would prefer a written license, or if the technologies described in the Open
Specifications are not covered by the Open Specifications Promise or Community Promise, as
applicable, patent licenses are available by contacting iplg@microsoft.com.



Trademarks.The names of companies and products contained in this documentation may be
covered by trademarks or similar intellectual property rights. This notice does not grant any
licenses under those rights. For a list of Microsoft trademarks, visit
www.microsoft.com/trademarks.



Fictitious Names. The example companies, organizations, products, domain names, email
addresses, logos, people, places, and events depicted in this documentation are fictitious. No
association with any real company, organization, product, domain name, email address, logo,
person, place, or event is intended or should be inferred.

Reservation of Rights. All other rights are reserved, and this notice does not grant any rights
other than specifically described above, whether by implication, estoppel, or otherwise.
Tools. The Open Specifications do not require the use of Microsoft programming tools or
programming environments in order for you to develop an implementation. If you have access to
Microsoft programming tools and environments you are free to take advantage of them. Certain
Open Specifications are intended for use in conjunction with publicly available standard
specifications and network programming art, and assumes that the reader either is familiar with the
aforementioned material or has immediate access to it.

1 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Revision Summary
Date

Revision
History

Revision
Class

Comments

08/07/2009

0.1

Major

First release.

11/06/2009

0.1.1

Editorial

Revised and edited the technical content.

03/05/2010

0.2

Minor

Updated the technical content.

04/21/2010

1.0

Major

Updated and revised the technical content.

06/04/2010

1.0.1

Editorial

Revised and edited the technical content.

06/22/2010

2.0

Major

Significantly changed the technical content.

09/03/2010

3.0

Major

Significantly changed the technical content.

02/09/2011

3.1

Minor

Clarified the meaning of the technical content.

07/07/2011

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

11/03/2011

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

01/19/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

02/23/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

03/27/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

05/24/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

06/29/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

07/16/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

10/08/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

10/23/2012

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

03/26/2013

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

06/11/2013

3.1

No change

No changes to the meaning, language, or formatting of
the technical content.

2 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

3 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Contents
1

Introduction ............................................................................................................. 5
1.1 Glossary ............................................................................................................... 5
1.2 References ............................................................................................................ 5
1.2.1 Normative References ....................................................................................... 6
1.2.2 Informative References ..................................................................................... 6
1.3 Overview .............................................................................................................. 6
1.4 Relationship to Protocols and Other Structures .......................................................... 7
1.5 Applicability Statement ........................................................................................... 7
1.6 Versioning and Localization ..................................................................................... 7
1.7 Vendor-Extensible Fields ......................................................................................... 8

2

Structures ................................................................................................................ 9
2.1 GEOGRAPHY and GEOMETRY Structures ................................................................... 9
2.1.1 Basic GEOGRAPHY Structure (Version 1) ............................................................. 9
2.1.2 Basic GEOGRAPHY Structure (Version 2) ........................................................... 11
2.1.3 FIGURE Structure ........................................................................................... 14
2.1.4 SHAPE Structure ............................................................................................ 15
2.1.5 GEOGRAPHY POINT Structure .......................................................................... 15
2.1.6 GEOMETRY POINT Structure ............................................................................ 16
2.1.7 SEGMENT Structure ........................................................................................ 17
2.2 HIERARCHYID Structure ....................................................................................... 17
2.2.1 Logical Definition ............................................................................................ 17
2.2.2 Physical Representation .................................................................................. 17
2.3 CLR UDTs ........................................................................................................... 19
2.3.1 Native UDT Serialization .................................................................................. 19
2.3.1.1 Binary Format of Each Byte ........................................................................ 20
2.3.1.2 Binary Format of Primitive Types ................................................................ 20
2.3.1.3 Nested Structures ..................................................................................... 22
2.3.2 User-Defined UDT Serialization ........................................................................ 22

3

Structure Examples ................................................................................................ 23
3.1 GEOGRAPHY and GEOMETRY Structure Examples .................................................... 23
3.1.1 Example of an Empty Point Structure ................................................................ 23
3.1.2 Example of a Geometry Point Structure ............................................................. 23
3.1.3 Example of a Linestring Structure ..................................................................... 24
3.1.4 Example of a Geometry Collection Structure ...................................................... 25
3.1.5 Example of an Object Serialized in Version 2 ..................................................... 28
3.2 HIERARCHYID Examples ....................................................................................... 29
3.3 CLR UDT Serialization Example .............................................................................. 30

4

Security Considerations.......................................................................................... 33

5

Appendix A: Product Behavior ................................................................................ 34

6

Change Tracking..................................................................................................... 35

7

Index ..................................................................................................................... 36

4 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

1

Introduction
This document specifies the binary format of the GEOGRAPHY, GEOMETRY, HIERARCHYID, and
common language runtime (CLR) user-defined type (UDT) structures managed by the
protocol server. The protocol server provides the geography, geometry, and hierarchyid protocol
server data types as well as the CLR UDTs that use these structures.
The first two of these protocol server types implement the OpenGIS Consortium’s (OGC) Simple
Feature Specification (SFS) [OGCSFS] section 8. Thus, the content of these structures closely
mirrors the SFS.
The structures that are used to transfer geography and geometry data types are identical. In this
document, the term "GEOGRAPHY structure" refers to both the GEOGRAPHY and GEOMETRY
structures, except where it is necessary to distinguish between the two structures. Likewise,
"geography data type" refers to both the geography and geometry protocol server data types.
CLR UDTs enable users to extend the protocol server type system by creating new types. These
types can include any fields and methods defined by the user. The exact structure depends on the
user who is implementing CLR UDTs. The protocol client program must contain the knowledge of the
internal structure of each CLR UDT before it can read that type’s binary format.
Sections 1.7 and 2 of this specification are normative and can contain the terms MAY, SHOULD,
MUST, MUST NOT, and SHOULD NOT as defined in RFC 2119. All other sections and examples in this
specification are informative.

1.1

Glossary

The following terms are defined in [MS-GLOS]:
little-endian
The following terms are specific to this document:
common language runtime (CLR): The common language runtime is the infrastructure that
the .NET Framework uses to execute all managed applications. The runtime supplies managed
code with services such as cross-language integration, code access security, object lifetime
management, and debugging and profiling support.
user-defined type (UDT): User-defined types can extend the scalar type system of the protocol
server database, enabling storage of CLR objects in a protocol server database. UDTs can
contain multiple elements and they can have behaviors, which differentiates them from the
traditional alias data types that consist of a single protocol server system data type.
MAY, SHOULD, MUST, SHOULD NOT, MUST NOT: These terms (in all caps) are used as
described in [RFC2119]. All statements of optional behavior use either MAY, SHOULD, or
SHOULD NOT.

1.2

References

References to Microsoft Open Specifications documentation do not include a publishing year because
links are to the latest version of the documents, which are updated frequently. References to other
documents include a publishing year when one is available.

5 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

1.2.1

Normative References

We conduct frequent surveys of the normative references to assure their continued availability. If
you have any issue with finding a normative reference, please contact dochelp@microsoft.com. We
will assist you in finding the relevant information. Please check the archive site,
http://msdn2.microsoft.com/en-us/library/E4BD6494-06AD-4aed-9823-445E921C9624, as an
additional source.
[IEEE754] Institute of Electrical and Electronics Engineers, "Standard for Binary Floating-Point
Arithmetic", IEEE 754-1985, October 1985, http://ieeexplore.ieee.org/servlet/opac?punumber=2355
[MS-NRBF] Microsoft Corporation, ".NET Remoting: Binary Format Data Structure".
[MS-TDS] Microsoft Corporation, "Tabular Data Stream Protocol".
[OGCSFS] Herring, J. R., "OpenGIS Implementation Specification for Geographic information –
Simple feature access – Part 1: Common Architecture", OGC 06-103r3 Version 1.2.0, October 2006,
http://portal.opengeospatial.org/files/?artifact_id=18241
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC
2119, March 1997, http://www.ietf.org/rfc/rfc2119.txt

1.2.2

Informative References

[IRE1098] D.A. Huffman, "A Method for the Construction of Minimum-Redundancy Codes",
Proceedings of the I.R.E., September 1952, pp 1098-1102,
http://compression.ru/download/articles/huff/huffman_1952_minimum-redundancy-codes.pdf
[MS-BINXML] Microsoft Corporation, "SQL Server Binary XML Structure".
[MS-CLRUDT] Microsoft Corporation, "CLR User-Defined Types", June 2009,
http://msdn.microsoft.com/en-us/library/ms131120.aspx.

1.3

Overview

The geography and geometry data types are used by the protocol server to represent twodimensional objects. The geography data type is designed to handle ellipsoidal coordinates, defined
from a variety of standard Earth-shape references, and is used specifically to accommodate
geospatial data. The geometry data type is nonspecific and can be used for geospatial and other
spatial applications that use Cartesian coordinates.
Instances of the geometry and geography data types can be composed of a variety of complex
features whose definitions are stored in various structures. These structures are described in detail
later in this document.
The hierarchyid data type is used by a protocol server application to model tree structures in a
more efficient way than was formerly possible. This data type significantly improves on the
performance of current solutions (for instance, recursive queries).
Values of the hierarchyid<1> data type represent nodes in a hierarchy tree. This data type is a
system common language runtime (CLR) type, so applications interpret it the same way they would
interpret any protocol server CLR user-defined type (UDT). The binary structure of the data type,
described in detail later in this document, uses a variant on Huffman encoding to represent the path
from the root of a tree to a particular node in that tree. For more information about Huffman
encoding, see [IRE1098].

6 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

CLR UDTs can represent any type defined by the user. The user implements a CLR UDT as a
structure using the CLR type system. The binary format of a CLR UDT depends on two factors. The
first factor is the CLR UDT’s internal structure, as defined by the user. The second factor is the
serialization format also chosen by the user. To decode the binary format of a CLR UDT, it is
necessary to know these two properties of the CLR UDT.
The user implementing CLR UDTs can include primitive types and other structures. The structures
can include other CLR UDTs. The set of types available for fields may be limited, depending on the
serialization format chosen by the user.
The user can choose between two available serialization formats: protocol server native UDT
serialization, and user-defined UDT serialization. Protocol server native UDT serialization is designed
for simple CLR UDTs that have a simple structure and use only a specified set of simple primitive
types. User-defined UDT serialization is more flexible and enables users to define complex and more
dynamic CLR UDTs.
To learn more about CLR UDTs, see [MS-CLRUDT].

1.4

Relationship to Protocols and Other Structures

All structures described in this document are designed to be transported over Tabular Data Stream
protocol as described in section 2.2.5.5.2 of [MS-TDS].

1.5

Applicability Statement

The spatial data format presented in this document is designed for the native code programmer (C
and C++, for example) and documents the disk representation for the protocol server geography
and geometry data types. Programmers using managed code (the Microsoft .NET Framework) are
encouraged to use the SQL CLR Types library (SQLSysClrTypes.msi) and the corresponding builder
API. Note that Microsoft reserves the right to make changes to this format at any time.
The HIERARCHYID format presented in this document is designed to be used solely with managed
code (the .NET Framework) by using the SQL CLR Types library (SQLSysClrTypes.msi) and the
corresponding APIs. Again, note that Microsoft reserves the right to make changes to this format at
any time.
The format of common language runtime (CLR) user-defined types (UDTs) is designed to be used
solely with managed code by using the same classes that define CLR UDTs in a protocol client
program. As stated earlier in this document, without knowledge of the internal structure of a CLR
UDT and the serialization format that it is using, it is impossible to read the CLR UDT from the
binary data representing it.

1.6

Versioning and Localization

This document describes versions 1 and 2 of the GEOGRAPHY, GEOMETRY structures and version
1 of the HIERARCHY structure.<2> The version number of the GEOGRAPHY and GEOMETRY
structures is stored in the Version field in the structure. The version number of the HIERARCHY
structure is not stored anywhere in the structure. All fields in the GEOGRAPHY, GEOMETRY, and
HIERARCHYID structures contain either numeric or bit flag data. There are no localization
implications for these structures.
The protocol server does not define any versioning scheme for common language runtime (CLR)
user-defined types (UDTs). Any version data created by the user must be part of a CLR UDT itself.

7 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

1.7

Vendor-Extensible Fields

The GEOMETRY, GEOGRAPHY, and HIERARCHY structures do not contain any extensible fields.
All fields of a common language runtime (CLR) user-defined type (UDT) are defined by the user who
creates the type. The serialization format of these fields can also be selected by the user.

8 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

2

Structures

2.1

GEOGRAPHY and GEOMETRY Structures

The GEOGRAPHY and GEOMETRY structures are serialized using the binary format described later
in this section. The GEOGRAPHY structure contains several fixed fields (or header fields) and either
three or four secondary structures that are repeated, as necessary, to describe the geography
fully.<3>
The GEOGRAPHY POINT structure contains the coordinates for an individual point and is repeated
for as many points as are present in the GEOGRAPHY structure. One shape structure will appear
for each OGC simple feature that is contained in the GEOGRAPHY structure. A shape can consist of
multiple figures, each of which is defined by a single figure structure. The GEOGRAPHY structure
contains flags and counts that indicate how many of these subsidiary structures are contained in the
GEOGRAPHY structure.

2.1.1

Basic GEOGRAPHY Structure (Version 1)

Version 1 of the GEOGRAPHY structure is formatted as shown in the following packet diagram. All
double fields contain double-precision floating-point numbers that are 64 bits (8 bytes) long.
Integers and double-precision floating-point numbers are expressed in little-endian format.

0

1

2

3

4

5

6

7

8

9

1
0

1

2

3

4

5

6

7

8

9

2
0

1

2

3

4

5

6

7

8

9

3
0

1

SRID
Version

Serialization Properties

Number of Points (optional, unsigned)

...

Points (optional, variable) (16 * Number of Points
bytes) (variable)
...

Z Values (optional, 8 * Number of Points bytes) (variable)
...
M Values (optional, 8 * Number of Points bytes) (variable)
...
Number of Figures (optional, unsigned)
Figures (optional, 5 * Number of Figure bytes) (variable)
...
Number of Shapes (optional, unsigned)

9 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Shapes (optional, 9 * Number of Shapes bytes) (variable)
...

SRID (4 bytes): (32 bit integer) The spatial reference identifier (SRID) for the geography.
GEOGRAPHY structures MUST use SRID values in the range of 4120 through 4999, inclusive,
with the exception of null geographies. A value of -1 indicates a null geography. When a null
geography is indicated, all other fields are omitted. Default SRID for GEOGRAPHY instances is
4326. Default SRID for GEOMETRY instances is zero (0). For GEOMETRY instance, SRID can be
any value: SRID is not constrained.
Version (1 byte): The version of the GEOGRAPHY structure.<4>
Serialization Properties (1 byte): A bit field that contains individual bit flags that indicate
which optional content is present in the structure, as well as other attributes of the
geography. The first 3 bits of the serialization properties are reserved for future use.

0

1

2

3

4

5

6

7

0

0

0

L

P

V

M

Z

Where the bits are defined as:
Value

Description

Z

The structure has Z values.

(0x01)
M

The structure has M values.

(0x02)
V

Geography is valid.

(0x04)

For GEOGRAPHY structures, V in version 1 is always set.

P

Geography contains a single point. When P is set, Number of Points, Number of
Figures, and Number of Shapes are implicitly assumed to be equal to 1 and are
omitted from the structure. In addition, Figures is implicitly assumed to contain one
figure representing a Stroke with a Point Offset of 0 (zero). Lastly, Shape is implicitly
assumed to contain one shape of type Point, with a Figure Offset of 0 (zero) and
without any parents (Parent Offset set to -1). This is an optimization for the common
case of a single point.

(0x08)

L
(0x10)

Geography contains a single line segment. When L is set, Number of Points is implicitly
assumed to be equal to 2 and does not explicitly appear in the serialized data. Number
of Figures and Number of Shapes are implicitly assumed to be equal to 1 and do not
explicitly appear in the serialized data. In addition, Figures is implicitly assumed to
contain one stroke figure (0x01) with a Point Offset of 0 (zero). Lastly, Shape is
implicitly assumed to contain one shape of type 0x02 (LineString), with a Figure Offset
of 0 and without any parents (Parent Offset set to -1).
P and L are mutually exclusive properties.

10 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Number of Points (optional, unsigned) (4 bytes): The number of points in the
GEOGRAPHY structure. This MUST be a positive number or 0 (zero). If either the P or L bit is
set in the Serialization Properties bit field, this field is omitted from the structure.
Points (optional, variable) (16 * Number of Points bytes) (variable): A sequence of point
structures. The point coordinates are contained in GEOGRAPHY POINT structures in
GEOGRAPHY structures. Likewise, coordinates are contained in GEOMETRY POINT
structures in GEOMETRY structures. Both structures contain a pair of doubles.
If neither the P nor L bit is set in the Serialization Properties bit field, there will be
Number of Points points in the sequence. If the P bit is set, there will be one point. If the L
bit is set, there will be two points.
Z Values (optional, 8 * Number of Points bytes) (variable): A sequence of double values
for the Z value of each point. If the Z bit is set, there will be Number of Points doubles in
the array. If a Z value for an individual point is NULL, it is represented by QNaN [IEEE754].
M Values (optional, 8 * Number of Points bytes) (variable): A sequence of double values
for the M value of each point. If the M bit is set, there will be Number of Points doubles in
the array. If an M value for an individual point is NULL, it is represented as QNaN.
Number of Figures (optional, unsigned) (4 bytes): The number of figures in the structure.
This MUST be a positive number or 0 (zero).
Figures (optional, 5 * Number of Figure bytes) (variable): A sequence of figure structures.
Number of Shapes (optional, unsigned) (4 bytes): The number of shapes in the structure.
This MUST be a positive number.
Shapes (optional, 9 * Number of Shapes bytes) (variable): A sequence of shape
structures.

2.1.2

Basic GEOGRAPHY Structure (Version 2)

Version 2 of the GEOGRAPHY structure is formatted as shown in the following packet diagram. All
double fields contain double-precision floating-point numbers that are 64 bits (8 bytes) long.
Integers and double-precision floating-point numbers are expressed in little-endian format.

0

1

2

3

4

5

6

7

8

9

1
0

1

2

3

4

5

6

7

8

9

2
0

1

2

3

4

5

6

7

8

9

3
0

1

SRID
Version

Serialization Properties

Number of Points (optional, unsigned)

...

Points (optional, variable) (16 * Number of Points
bytes) (variable)
...

Z Values (optional, 8 * Number of Points bytes) (variable)
...

11 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

M Values (optional, 8 * Number of Points bytes) (variable)
...
Number of Figures (optional, unsigned)
Figures (optional, 5 * Number of Figure bytes) (variable)
...
Number of Shapes (optional, unsigned)
Shapes (optional, 9 * Number of Shapes bytes) (variable)
...
Number of Segments (optional)
Segments (optional) (1 * Number of Segments bytes) (variable)
...

SRID (4 bytes): (32 bit integer) The SRID for the geography. GEOGRAPHY structures MUST
use SRID values in the range of 4120 through 4999, inclusive, with the exception of null
geographies. A value of -1 indicates a null geography. When a null geography is indicated, all
other fields are omitted. Default SRID for GEOGRAPHY instances is 4326. Default SRID for
GEOMETRY instances is zero (0). For GEOMETRY instance, SRID can be any value: SRID is not
constrained.
Version (1 byte): The version of the GEOGRAPHY structure. <5>
Serialization Properties (1 byte): A bit field that contains individual bit flags that indicate
which optional content is present in the structure as well as other attributes of the geography.
The first 3 bits of the serialization properties are reserved for future use.

0

1

2

3

4

5

6

7

0

0

H

L

P

V

M

Z

Where the bits are defined as:
Value

Description

Z

The structure has Z values.

(0x01)
M

The structure has M values.

(0x02)

12 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Value

Description

V

Geography is valid. For GEOGRAPHY structures, V in version 2 is always set.

(0x04)
P
(0x08)

L
(0x10)

Geography contains a single point. When P is set, Number of Points, Number of
Figures, and Number of Shapes are implicitly assumed to be equal to 1 and are
omitted from the structure. In addition, Figures is implicitly assumed to contain one
figure representing a Stroke with a Point Offset of 0 (zero). Lastly, Shape is implicitly
assumed to contain one shape of type Point, with a Figure Offset of 0 (zero) and
without any parents (Parent Offset set to -1). This is an optimization for the common
case of a single point.
Geography contains a single line segment. When L is set, Number of Points is implicitly
assumed to be equal to 2 and does not explicitly appear in the serialized data. Number
of Figures and Number of Shapes are implicitly assumed to be equal to 1 and do not
explicitly appear in the serialized data. In addition, Figures is implicitly assumed to
contain one stroke figure (0x01) with a Point Offset of 0 (zero). Lastly, Shape is
implicitly assumed to contain one shape of type 0x02 (LineString), with a Figure Offset
of 0 and without any parents (Parent Offset set to -1).
P and L are mutually exclusive properties.

H
(0x20)

Geography is larger than a hemisphere. This bit is added in version 2 of the serialization
format.<6>

Number of Points (optional, unsigned) (4 bytes): The number of points in the
GEOGRAPHY structure. This MUST be a positive number or 0 (zero). If either the P or L bit is
set in the Serialization Properties bit field, this field is omitted from the structure.
Points (optional, variable) (16 * Number of Points bytes) (variable): A sequence of point
structures. The point coordinates are contained in GEOGRAPHY POINT structures in
GEOGRAPHY structures. Likewise, coordinates are contained in GEOMETRY POINT
structures in GEOMETRY structures. Both structures contain a pair of doubles.
If neither the P nor L bit is set in the Serialization Properties bit field, there will be
Number of Points points in the sequence. If the P bit is set, there will be one point. If the L
bit is set, there will be two points.
Z Values (optional, 8 * Number of Points bytes) (variable): A sequence of double values
for the Z value of each point. If the Z bit is set, there will be Number of Points doubles in
the array. If a Z value for an individual point is NULL, it is represented by QNaN [IEEE754].
M Values (optional, 8 * Number of Points bytes) (variable): A sequence of double values
for the M value of each point. If the M bit is set, there will be Number of Points doubles in
the array. If an M value for an individual point is NULL, it is represented by QNaN.
Number of Figures (optional, unsigned) (4 bytes): The number of figures in the structure.
This MUST be a positive number or 0 (zero).
Figures (optional, 5 * Number of Figure bytes) (variable): A sequence of figure structures.
Number of Shapes (optional, unsigned) (4 bytes): The number of shapes in the structure.
This MUST be a positive number.
Shapes (optional, 9 * Number of Shapes bytes) (variable): A sequence of shape
structures.

13 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Number of Segments (optional) (4 bytes): The number of segments in the structure. This
MUST be a positive number. Segments are added in version 2 of the serialization format.<7>
Segments (optional) (1 * Number of Segments bytes) (variable): A sequence of segment
structures.<8>

2.1.3

FIGURE Structure

The FIGURE structure defines the partitions in the Points, Z Values, and M Values sequences for
each constituent of the simple feature represented by the geography. A simple feature may have
more than one part, whereas the collection of simple feature types may contain more than one
simple feature.

0

1

2

3

4

5

6

7

8

9

1
0

1

2

3

4

Figures Attribute (byte)

5

6

7

8

9

2
0

1

2

3

4

5

6

7

8

9

3
0

1

Point Offset (32-bit integer)

...

Figures Attribute (byte) (1 byte): Determines the role of this figure within the GEOMETRY
structure. Valid values in version 1 of the serialization format are as follows:<9>
0 (0x00): Figure is an interior ring in a polygon. Interior rings represent holes in exterior
rings.
1 (0x01): Figure is a stroke. A stroke is a point or a line.
2 (0x02): Figure is an exterior ring in a polygon. An exterior ring represents the outer
boundary of a polygon.
Valid values in version 2 of the serialization format are as follows:<10>
0 (0x00): Figure is a point.
1 (0x01): Figure is a line.
2 (0x02): Figure is an arc.
3 (0x03): Figure is a composite curve, i.e. it contains both line and arc segments.
The order of the coordinates in each ring of a geography polygon (but not a geometry
polygon) is important. The outer rings for polygons are constructed by using the "left-hand"
rule to determine the interior region of a polygon shape. Thus, outer polygon rings have their
GEOGRAPHY POINT coordinate pairs ordered in a counter-clockwise direction. Polygon holes
are constructed using the "right-hand" rule. Thus, the GEOGRAPHY POINT coordinate pairs
of a polygon holes are ordered in a clockwise direction.
Point Offset (32-bit integer) (4 bytes): The offset to the FIGURE structure’s first point in
the Points, Z Values, and M Values sequences.

14 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

2.1.4

SHAPE Structure

The SHAPE structure identifies each simple feature contained in the GEOGRAPHY structure. It links
together the simple feature type, the figure that represents it, and the parent simple feature that
contains the present simple feature (if there is one).

0

1

2

3

4

5

6

7

8

9

1
0

1

2

3

4

5

6

7

8

9

2
0

1

2

3

4

5

6

7

8

3
0

9

1

Parent Offset (32-bit integer)
Figure Offset (32-bit integer)
OpenGIS Type (1 byte)

Parent Offset (32-bit integer) (4 bytes): The offset to the SHAPE structure’s parent
(containing) shape in the Shapes sequence if the shape has a parent, such as an outer ring if
a hole, or a multipart simple feature.
Figure Offset (32-bit integer) (4 bytes): The offset to the SHAPE structure’s Figure in the
Figures sequence.
OpenGIS Type (1 byte) (1 byte): The type of simple feature represented by the SHAPE
structure.
Valid values in version 1 of the serialization format are as follows:<11>
1 (0x01): Point
2 (0x02): LineString
3 (0x03): Polygon
4 (0x04): MultiPoint
5 (0x05): MultiLineString
6 (0x06): MultiPolygon
7 (0x07): GeometryCollection
Version 2 adds the following valid values:<12>
8 (0x08): CircularString
9 (0x09): CompoundCurve
10 (0x0A): CurvePolygon
11 (0x0B): FullGlobe

2.1.5

GEOGRAPHY POINT Structure

The GEOGRAPHY POINT structure contains latitude and longitude coordinates as double values
representing a point located on a spheroid.

15 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

0

1

2

3

4

5

6

7

8

9

1
0

1

2

3

4

5

6

7

8

9

2
0

1

2

3

4

5

6

7

8

9

3
0

1

Latitude (double)
...
Longitude (double)
...

Latitude (double) (8 bytes): The GEOGRAPHY POINT structure’s latitude.
Longitude (double) (8 bytes): The GEOGRAPHY POINT structure’s longitude.
Notes
The example structure provided in this section uses the Well-Known Text (WKT) protocol of
[OGCSFS] section 7.
Latitude and longitude coordinates are stored as decimal degree values. Negative values
are used to designate south latitude and west longitude values.
Latitude values MUST be between -90 and 90 degrees, inclusive.
Longitude values MUST be between -15069 and 15069 degrees, inclusive.
Latitude and Longitude values MUST NOT contain Infinity or NaN [IEEE754].

2.1.6

GEOMETRY POINT Structure

The GEOMETRY POINT structure contains x-coordinates and y-coordinates as double values
representing a point located on a plane.

0

1

2

3

4

5

6

7

8

9

1
0

1

2

3

4

5

6

7

8

9

2
0

1

2

3

4

5

6

7

8

9

3
0

1

X Coordinate (double)
...
Y Coordinate (double)
...

X Coordinate (double) (8 bytes): The GEOMETRY POINT structure's x-coordinate.
Y Coordinate (double) (8 bytes): The GEOMETRY POINT structure's y-coordinate.
X Coordinate and Y Coordinate values MUST NOT contain Infinity or NaN.
Note

The example structure provided in this section uses the WKT protocol of [OGCSFS].

16 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

2.1.7

SEGMENT Structure

The SEGMENT structure defines the structure of a compound curve figure.<13> It contains only
one byte, which represents type of the segment. Segments are stored only for figures whose Figure
Attribute value is 0x03 in version 2 of the serialization format.

0

1

2

3

4

5

6

7

8

9

1
0

1

2

3

4

5

6

7

8

9

2
0

1

2

3

4

5

6

7

8

9

3
0

1

Segment Type

Segment Type (1 byte): Determines the type of the segment within the figure.
Valid values are as follows:
0 (0x00): Segment is a line.
1 (0x01): Segment is an arc.
2 (0x02): Segment is a first line.
3 (0x03): Segment is a first arc.
The first line and first arc segments mark the start of the sequence of segments of the same
type, which are line and arc respectively. Subsequent segments have types line and arc.

2.2

HIERARCHYID Structure

2.2.1

Logical Definition

A hierarchy tree is an abstract ordered tree. This means that for each node n, there is a "less-than"
(<) order relation on all children of n. This tree has infinite depth. For each node n in the tree, the
children of n are in 1-to-1 correspondence with finite nonempty sequences of integers, which are
called node labels. Given any two children m1 and m2 of n, m1 < m2 if and only if the label of m1
comes before the label of m2 in the lexicographical order in integer sequences. Thus, for a node n,
each child of n has siblings before and after it, and any two children of n have siblings between
them.
The logical representation of a node label for a child of a given node is a sequence of integers
separated by dots (for example, 1, 1.3, or -7.0.-8.9). The hierarchyid data type logically encodes
information about a single node in the hierarchy tree by encoding the path from the root of the tree
to the node. Such a path is logically represented as a sequence of node labels of all children visited
after the root. Each label is followed by a slash, and a slash begins the representation. Thus, a path
that visits only the root is represented by a single slash. For example, /, /1/, /0.3.-7/, /1/3/, and
/0.1/0.2/ are valid hierarchyid paths of lengths 1, 2, 2, 3, and 3, respectively.
The hierarchy data type represents a node in the hierarchy tree based on a binary encoding of the
following logical representation. This encoding is described in section 3.2.

2.2.2

Physical Representation

The logical representation of a node in the hierarchy tree is encoded into a sequence of bits
according to the following representation:

17 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

L0

O0

F0

Li

Oi

Fi

...

Lk

Ok

Fk

W

In the preceding diagram, each L/O pair encodes one integer in the logical representation of the
node; each Fi is a single bit that is 0 (zero) if the integer is followed by a dot in the logical
representation, and 1 if it is followed by a slash. W is a string of 0 to 7 bits, padding the
representation to the nearest byte; all bits in W have value 0.
In the following text, each Li/Oi/Fi triple is referred to as a level. If Fi is 0 (zero), the level is said to
be fake; otherwise, it is said to be real.
Li/Oi pairs encode an integer according to the following description. If the ith integer in the logical
representation of the node is n, the Li/Oi pair encodes n for real levels and n+1 for fake levels. This
is done so that, in varbinary (variable-length binary data) comparisons, fake levels compare above
real levels.
Each Li prefix of an Li/Oi pair specifies a range of integers and a bit size for the following Oi field, as
shown in the following table. Each Oi field has some antiambiguity bits that are always of a fixed
value for a particular Li value. These bits are used to enable unambiguous backward parsing of the
representation. The third column in the table shows the format of Li/Oi pair with antiambiguity bits
in the Oi field, with all other bits of the Oi field shown as dots.
The actual value of the integer encoded is the value of the Oi field (ignoring antiambiguity bits and
interpreting the rest of the bits as an unsigned integer) added to the lower limit of the range
corresponding to the Li field.

Li

Bit size of Oi
(without/with
antiambiguity
bits)

Full format of the Li/Oi pair

Range

000100

48/53

000100..............0.....................0......0...0.1...

-281479271682120
to -4294971465

000101

32/36

000101...................0......0...0.1...

-4294971464 to 4169

000110

12/15

000110.....0...0.1...

-4168 to -73

0010

6/8

0010..0.1...

-72 to -9

00111

3/3

00111...

-8 to -1

01

2/2

01..

0 to 3

100

2/2

100..

4 to 7

101

3/3

101...

8 to 15

110

6/8

110..0.1...

16 to 79

1110

10/13

1110...0...0.1...

80 to 1103

11110

12/15

11110.....0...0.1...

1104 to 5199

111110

32/36

111110...................0......0...0.1...

5200 to 4294972495

111111

48/53

111111..............0.....................0......0...0.1...

4294972496 to

18 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Bit size of Oi
(without/with
antiambiguity
bits)

Li

Full format of the Li/Oi pair

Range
281479271683151

No integer outside of the range -281479271682120 − 281479271683119 can be represented in this
encoding.
Also, note that the encoding used in the hierarchyid data type is limited to 892 bytes.
Consequently, nodes that have too many levels in their representation to fit into 892 bytes cannot
be represented by the hierarchyid data type.
The encoding for the root node is a binary string of length 1. Thus, there is one level and no W field.
Note The encoding represented in the preceding table has three useful properties:
It is parsable. That is, for any binary string, there is at most one interpretation of it as a
sequence of Li/Oi/Fi triples, and there is an efficient parsing algorithm.
The representation is also parsable backward (that is, starting from the last byte). This enables
an algorithm to determine a node’s parent without having to parse the entire binary string.
Comparing two encodings by lexicographical binary comparison is equivalent to conducting
depth-first comparisons on the corresponding tree nodes.

2.3

CLR UDTs

This section describes the binary format of common language runtime (CLR) user-defined types
(UDTs).
Two serialization formats affect the binary format of a CLR UDT.
Native UDT serialization is designed for simple CLR UDTs that have a simple structure and use only
a certain set of simple primitive types.
User-defined UDT serialization is more flexible and lets the user define complex and dynamic CLR
UDTs. For more information, see [MSDN-UDTR].

2.3.1

Native UDT Serialization

Native user-defined type (UDT) serialization is designed to simplify the serialization of simple
common language runtime (CLR) UDTs. Therefore, CLR UDTs that use native UDT serialization have
to adhere to certain limitations, as specified in this section.
All primitive fields MUST be of one of the following types:
BOOL, BYTE, SBYTE, USHORT, SHORT, UINT, INT, ULONG, LONG, FLOAT, DOUBLE, SqlByte,
SqlInt16, SqlInt32, SqlInt64, SqlBoolean, SqlSingle, SqlDouble, SqlDateTime, SqlMoney
CLR UDTs that use native UDT serialization can contain nested structures. Fields defined by these
nested structures must adhere to the same limits that apply to fields of CLR UDTs that use native
UDT serialization.

19 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

The binary format of a CLR UDT that uses native UDT serialization is defined by all of the UDT’s
fields, in the order in which they are defined by the user. The format of each field depends on its
type.

2.3.1.1

Binary Format of Each Byte

Each data type that is formatted by using native user-defined type (UDT) serialization consists of a
series of bytes. Each byte is formatted as 8 bits representing the byte value in binary representation
in a little-endian bit ordering (formatting all bits in order of their significance, starting with the least
significant bit and ending with the most significant bit).

2.3.1.2

Binary Format of Primitive Types

BOOL values are represented as a single byte. Depending on the BOOL value, the byte takes
one of the following two values: 0x01 for True or 0x00 for False.
BYTE values are represented as a single byte.
SBYTE values are represented as a single byte, but the most significant bit is reversed from 0
(zero) to 1 or from 1 to 0.
USHORT values are represented as 2 bytes. The most significant byte is first, followed by the
least significant byte.
SHORT values are represented as 2 bytes. The most significant byte is first, and it has the most
significant bit reversed from 0 (zero) to 1 or from 1 to 0. It is followed by the least significant
byte.
UINT values are represented as 4 bytes, in the order of their significance, starting with the most
significant byte and ending with the least significant byte.
INT values are represented as 4 bytes, in the order of their significance, starting with the most
significant byte and ending with the least significant byte. The most significant byte has the most
significant bit reversed from 0 (zero) to 1 or from 1 to 0.
ULONG values are represented as 8 bytes, in the order of their significance, starting with the
most significant byte and ending with the least significant byte.
LONG values are represented as 8 bytes, in the order of their significance, starting with the most
significant byte and ending with the least significant byte. The most significant byte has the most
significant bit reversed, from 0 (zero) to 1 or from 1 to 0.
FLOAT values are represented as defined by 4-byte, [IEEE754] single-precision, floating-point
format, but the order of the bytes is reversed.
For positive values (including positive 0 (zero)), the most significant bit of the first byte is
reversed from 0 (zero) to 1.
For negative values, all bits of all bytes are reversed, from 0 (zero) to 1 or from 1 to 0.
For negative 0 (zero), all bits remain unchanged.
DOUBLE values are represented as defined by 8-byte, double-precision, floating-point format,
but the order of the bytes is reversed.
For positive values (including positive 0 (zero), the most significant bit of the first byte is
reversed, from 0 (zero) to 1.

20 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

For negative values, all bits of all bytes are reversed, from 0 (zero) to 1 or from 1 to 0.
For negative 0 (zero), all bits remain unchanged.
SqlByte values are represented as 2 bytes. The first byte is a BOOL value that indicates whether
or not the SqlByte value is NULL (True indicates that this value is not NULL; False indicates
that it is NULL). The second byte is the actual BYTE value representing the SqlByte value.
SqlInt16 values are represented as 3 bytes. The first byte is a BOOL value that indicates
whether or not the SqlInt16 value is NULL (True indicates that this value is not NULL; False
indicates that it is NULL). The other 2 bytes are the actual SHORT value representing the
SqlInt16 value.
SqlInt32 values are represented as 5 bytes. The first byte is a BOOL value that indicates
whether or not the SqlInt32 value is NULL (True indicates that this value is not NULL; False
indicates that it is NULL). The other 4 bytes are the actual INT value representing the SqlInt32
value.
SqlInt64 values are represented as 9 bytes. The first byte is a BOOL value that indicates
whether or not the SqlInt64 value is NULL (True indicates that this value is not NULL; False
indicates that it is NULL). The other 8 bytes are the actual LONG value representing the
SqlInt64 value.
SqlBoolean values are represented as a single byte. Depending on the value of SqlBoolean,
this byte can have any of the following three values: 0x00 for NULL, 0x01 for False, or 0x02 for
True.
SqlSingle values are represented as 5 bytes. The first byte is a BOOL value that indicates
whether or not the SqlSingle value is NULL (True indicates that this value is not NULL; False
indicates that it is NULL). The other 4 bytes are the actual FLOAT value representing the
SqlSingle value.
SqlDouble values are represented as 9 bytes. The first byte is a BOOL value that indicates
whether or not the SqlDouble value is NULL (True indicates that this value is not NULL; False
indicates that it is NULL). The other 8 bytes are the actual DOUBLE value representing the
SqlDouble value.
SqlDateTime values are represented as 9 bytes. The first byte is a BOOL value that indicates
whether or not the SqlDateTime value is NULL (True indicates that this value is not NULL;
False indicates that it is NULL). The next 4 bytes are an INT value representing the date as the
number of days elapsed since 1/1/1900 (for dates before 1/1/1900, this will be a negative
value). The final 4 bytes are an INT value representing the number of ticks elapsed since
midnight of the day represented by the date part. The following rules can be used to calculate the
number of elapsed ticks from the number of elapsed milliseconds:
Each second consists of 300 ticks. All ticks represent values with the number of milliseconds
ending in 0, 3, or 7. For example: 000, 003, 007, 010, 013, 017, 020, …, 990, 993, 997.
The valid range for SqlDateTime values is from 1753-1-1 00:00:00.000 through 9999-12-31
23:59:59.997.
SqlMoney values are represented as 9 bytes. The first byte is a BOOL value that indicates
whether or not the SqlMoney value is NULL (True indicates that this value is not NULL; False
indicates that it is NULL). The other 8 bytes are a LONG value representing the SqlMoney value
multiplied by 10000.

21 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

2.3.1.3

Nested Structures

Common language runtime (CLR) user-defined types (UDTs) that use native UDT serialization can
include nested structures. Nested structures are also created by the user and are represented by
formatting all their fields in the order in which they are defined by the user who created the nested
structure. The knowledge of the definition of the nested structure is necessary to decode it—the
serialization format only contains the serialized data of the nested structure, but not the definition of
the structure itself.

2.3.2

User-Defined UDT Serialization

User-defined User-defined type (UDT) serialization is used when native UDT serialization does not
provide enough flexibility to express more complex and dynamic structures. The user-defined
approach lets users implement their own serialization formats using types defined in .NET Remoting
Binary Format.
For more details about .NET Remoting Binary Format, see [MS-NRBF].

22 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

3

Structure Examples

3.1

GEOGRAPHY and GEOMETRY Structure Examples

The following examples illustrate how a selection of simple features is represented in the structures
defined in this document.

3.1.1

Example of an Empty Point Structure

POINT EMPTY is designed to handle a non-null condition when a function returns an empty set. This
may occur, for instance, when two disjoint spatial features are intersected.
POINT EMPTY is represented by the following binary string:

0x00000000 01 04 00000000 00000000 01000000 FFFFFFFF FFFFFFFF 01

This string is interpreted as shown in the following table.
Binary value

Description

00000000

SRID = 0

01

Version = 1

04

Serialization Properties = V (is valid)

00000000

Number of Points = 0 (no points)

00000000

Number of Figures = 0 (no figures)

01000000

Number of Shapes = 1

FFFFFFFF

1st Shape Parent Offset = -1 (no parent)

FFFFFFFF

1st Shape Figure Offset = -1 (no figure)

01

1st Shape OpenGIS Type = 1 (point)

3.1.2

Example of a Geometry Point Structure

POINT(5 10) holds a 0-dimension feature that represents a point location. The following figure
shows a geometry point feature located at the intersection of 5 on the x-axis and 10 on the y-axis
(the actual point is surrounded by a circular symbol to make it easier to see).

Figure 1: A geometry point
POINT (5 10) in SRID 4326 is represented by the following binary string:

23 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

0xE6100000 01 0C 0000000000001440 0000000000002440

This string is interpreted as shown in the following table.
Binary value

Description

E6100000

SRID = 4326

01

Version = 1

0C

Serialization Properties = V + P (geometry is valid, single point)

0000000000001440

X=5

0000000000002440

Y = 10

3.1.3

Example of a Linestring Structure

A LINESTRING is an ordered series of connected points. The LINESTRING (0 1 1, 3 2 2, 4 5 NULL)
contains a Z value for each point location, with the last Z value being NULL. The following figure
represents the x and y coordinates only for a geometry type.

Figure 2: A geometry linestring
LINESTRING (0 1 1, 3 2 2, 4 5 NULL) is represented by the following binary string:
0xE6100000 01 05 03000000 0000000000000000 000000000000F03F 0000000000000840
0000000000000040 0000000000001040 0000000000001440 000000000000F03F
0000000000000040 000000000000F8FF 01000000 01 00000000 01000000 FFFFFFFF 00000000 02
This string is interpreted as shown in the following table.
Binary value

Description

E6100000

SRID = 4326

01

Version = 1

05

Serialization Properties = V + Z (geometry is valid, has Z values)

03000000

Number of Points = 3

24 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Binary value

Description

0000000000000000

1st point X = 0

000000000000F03F

1st point Y = 1

0000000000000840

2nd point X = 3

0000000000000040

2nd point Y = 2

0000000000001040

3rd point X = 4

0000000000001440

3rd point Y = 5

000000000000F03F

1st point Z = 1

0000000000000040

2nd point Z= 2

000000000000F8FF

3rd point Z = QNaN

01000000

Number of Figures = 1

01

1st Figure Attribute = 1 (stroke)

00000000

1st Figure Point Offset = 0 (figure starts with 1st point)

01000000

Number of Shapes = 1

FFFFFFFF

1st Shape Parent Offset = -1 (no parent)

00000000

1st Shape Figure Offset = 0 (shape starts with 1st figure)

02

1st Shape OpenGIS Type = 2 (linestring)

3.1.4

Example of a Geometry Collection Structure

A GEOMETRYCOLLECTION is a heterogeneous collection of simple features. The following figure
shows a geography containing a single point, a single linestring, and a polygon with an interior ring
(hole).

Figure 3: A geometry collection containing a point, a linestring, and a polygon with a hole
A GEOMETRYCOLLECTION (POINT (4 0), LINESTRING (4 2, 5 3), POLYGON ((0 0, 3 0, 3 3, 0 3, 0 0),
(1 1, 1 2, 2 2, 2 1, 1 1))) is represented by the following binary string:

25 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

0xE6100000 01 04 0D000000
0000000000000000 0000000000001040 0000000000000040 0000000000001040 0000000000000840
0000000000001440 0000000000000000 0000000000000000 0000000000000000 0000000000000840
0000000000000840 0000000000000840 0000000000000840 0000000000000000 0000000000000000
0000000000000000 000000000000F03F 000000000000F03F 0000000000000040 000000000000F03F
0000000000000040 0000000000000040 000000000000F03F 0000000000000040 000000000000F03F
000000000000F03F
04000000 01 00000000 01 01000000 02 03000000 00 08000000
04000000 FFFFFFFF 00000000 07 00000000 00000000 01 00000000 01000000 02
00000000 02000000 03

This string is interpreted as shown in the following table.
Binary value

Description

E6100000

SRID = 4326

01

Version = 1

04

Serialization Properties = V (geography is valid)

0D000000

Number of Points = 13

0000000000000000

1st point latitude = 0

0000000000001040

1st point longitude = 4

0000000000000040

2nd point latitude = 2

0000000000001040

2nd point longitude = 4

0000000000000840

3rd point latitude = 3

0000000000001440

3rd point longitude = 5

0000000000000000

4th point latitude = 0

0000000000000000

4th point longitude = 0

0000000000000000

5th point latitude = 0

0000000000000840

5th point longitude = 3

0000000000000840

6th point latitude = 3

0000000000000840

6th point longitude = 3

0000000000000840

7th point latitude = 3

0000000000000000

7th point longitude = 0

0000000000000000

8th point latitude = 0

0000000000000000

8th point longitude = 0

000000000000F03F

9th point latitude = 1

000000000000F03F

9th point longitude = 1

26 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

Binary value

Description

0000000000000040

10th point latitude = 2

000000000000F03F

10th point longitude = 1

0000000000000040

11th point latitude = 2

0000000000000040

11th point longitude = 2

000000000000F03F

12th point latitude = 1

0000000000000040

12th point longitude = 2

000000000000F03F

13th point latitude = 1

000000000000F03F

13th point longitude = 1

04000000

Number of Figures = 4

01

1st Figure Attribute = 1 (stroke)

00000000

1st Figure Point Offset = 0 (figure starts with 1st point)

01

2nd Figure Attribute = 1 (stroke)

01000000

2nd Figure Point Offset = 1 (figure starts with 2nd point)

02

3rd Figure Attribute = 2 (exterior polygon ring)

03000000

3rd Figure Point Offset = 3 (figure starts with 4th point)

00

4th Figure Attribute = 0 (interior polygon ring)

08000000

4th Figure Point Offset = 8 (figure starts with 9th point)

04000000

Number of Shapes = 4

FFFFFFFF

1st Shape Parent Offset = -1 (no parent)

00000000

1st Shape Figure Offset = 0 (shape starts with 1st figure)

07

1st Shape OpenGIS Type = 7 (GeometryCollection)

00000000

2nd Shape Parent Offset = 0 (parent shape is 1st shape)

00000000

2nd Shape Figure Offset = 0 (shape starts with 1st figure)

01

2nd Shape OpenGIS Type = 1 (Point)

00000000

3rd Shape Parent Offset = 0 (parent shape is 1st shape)

01000000

3rd Shape Figure Offset = 1 (shape starts with 2nd figure)

02

3rd Shape OpenGIS Type = 2 (LineString)

00000000

4th Shape Parent Offset = 0 (parent shape is 1st shape)

02000000

4th Shape Figure Offset = 2 (shape starts with 3rd figure)

03

4th Shape OpenGIS Type = 3 (Polygon)

27 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

3.1.5

Example of an Object Serialized in Version 2

This CURVEPOLYGON instance is a surface whose boundary is a curve; in this case, the curve is a
COMPOUNDCURVE. It is an instance of geography type that is a hole, so it is larger than a
hemisphere.<14>

Figure 4: A curve polygon hole
A CURVEPOLYGON(COMPOUNDCURVE((0 0, 0 2, 2 2), CIRCULARSTRING (2 2, 1 0, 0 0))) is
represented by the following binary string:
E6100000 02 24 05000000
0000000000000000 0000000000000000 0000000000000040 0000000000000000 0000000000000040
0000000000000040 0000000000000000 000000000000F03F 0000000000000000 0000000000000000
01000000 03 00000000
01000000 FFFFFFFF 00000000 0A
03000000 02 00 03

This string is interpreted as shown in the following table.

28 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

3.2

Binary value

Description

E6100000

SRID = 4326

02

Version = 2

24

Serialization Properties = VH (geography which is valid and larger than a
hemisphere)

05000000

Number of Points = 5

0000000000000000

1st point latitude = 0

0000000000000000

1st point longitude = 0

0000000000000040

2nd point latitude = 2

0000000000000000

2nd point longitude = 0

0000000000000040

3rd point latitude = 2

0000000000000040

3rd point longitude = 2

0000000000000000

4th point latitude = 0

000000000000F03F

4th point longitude = 1

0000000000000000

5th point latitude = 0

0000000000000000

5th point longitude = 0

01000000

Number of Figures = 1

03

1st Figure Attribute = 3 (compound curve)

00000000

1st Figure Point Offset = 0 (figure starts with 1st point)

01000000

Number of Shapes = 1

FFFFFFFF

1st Shape Parent Offset = -1 (no parent)

00000000

1st Shape Figure Offset = 0 (shape starts with 1st figure)

0A

1st Shape OpenGIS Type = 10 (CurvePolygon)

03000000

Number of Segments = 3

02

1st Segment Segment Type = 2 (First Line)

00

2nd Segment Segment Type = 0 (Line)

03

3rd Segment Segment Type = 3 (First Arc)

HIERARCHYID Examples

The root node is represented by a HIERARCHYID structure.
Example 1

29 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

The first child of the root node, with a logical representation of /1/, is represented as the following
bit sequence:
01011000
The first two bits, 01, are the L1 field, meaning that the first node has a label between 0 (zero) and
3. The next two bits, 01, are the O1 field and are interpreted as the integer 1. Adding this to the
beginning of the range specified by the L1 yields 1. The next bit, with the value 1, is the F1 field,
which means that this is a "real" level, with 1 followed by a slash in the logical representation. The
final three bits, 000, are the W field, padding the representation to the nearest byte.
Example 2
As a more complicated example, the node with logical representation /1/-2.18/ (the child with label
-2.18 of the child with label 1 of the root node) is represented as the following sequence of bits (a
space has been inserted after every grouping of 8 bits to make the sequence easier to follow):
01011001 11111011 00000101 01000000
The first three fields are the same as in the first example. That is, the first two bits (01) are the L1
field, the second two bits (01) are the O1 field, and the fifth bit (1) is the F1 field. This encodes the
/1/ portion of the logical representation.
The next 5 bits (00111) are the L2 field, so the next integer is between -8 and -1. The following 3
bits (111) are the O2 field, representing the offset 7 from the beginning of this range. Thus, the L2
and O2 fields together encode the integer -1. The next bit (0) is the F2 field. Because it is 0 (zero),
this level is fake, and 1 has to be subtracted from the integer yielded by the L2 and O2 fields.
Therefore, the L2, O2, and F2 fields together represent -2 in the logical representation of this node.
The next 3 bits (110) are the L3 field, so the next integer is between 16 and 79. The subsequent 8
bits (00001010) are the L4 field. Removing the anti-ambiguity bits from there (the third bit (0) and
the fifth bit (1)) leaves 000010, which is the binary representation of 2. Thus, the integer encoded
by the L3 and O3 fields is 16+2, which is 18. The next bit (1) is the F3 field, representing the slash
(/) after the 18 in the logical representation. The final 6 bits (000000) are the W field, padding the
physical representation to the nearest byte.

3.3

CLR UDT Serialization Example

The following example of a common language runtime (CLR) user-defined type (UDT) contains all of
the primitive types described in this document. The CLR UDT is defined in the C# programming
language as follows.
[SqlUserDefinedType(Format.Native)]
public struct SampleNativeUdt : INullable
{
public bool BoolValue;
public byte ByteValue;
public sbyte SByteValue;
public short ShortValue;
public ushort UShortValue;
public int IntValue;
public uint UIntValue;
public long LongValue;
public ulong ULongValue;
public float FloatValue;
public double DoubleValue;
public SqlByte SqlByteValue;

30 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

public
public
public
public
public
public
public
public

SqlInt16 SqlInt16Value;
SqlInt32 SqlInt32Value;
SqlInt64 SqlInt64Value;
SqlDateTime SqlDateTimeValue;
SqlSingle SqlSingleValue;
SqlDouble SqlDoubleValue;
SqlMoney SqlMoneyValue;
SqlBoolean SqlBooleanValue;

// Implementation methods
}

In the preceding example, the CLR UDT’s fields are initialized with the following values.
BoolValue = true;
ByteValue = 1;
SByteValue = -2;
ShortValue = 3;
UShortValue = 4;
IntValue = -5;
UIntValue = 6;
LongValue = 7;
ULongValue = 8;
FloatValue = 1.234568E+08;
DoubleValue = 123456789.0123456;
SqlByteValue = 9;
SqlInt16Value = -10;
SqlInt32Value = 11;
SqlInt64Value = 12;
SqlDateTimeValue = "1/1/2000 12:00:00";
SqlSingleValue = 1.234568E+08;
SqlDoubleValue = 123456789.0123456;
SqlMoneyValue = "$13";
SqlBooleanValue = true;

Binary formatting of this CLR UDT produces the following stream of bytes in hexadecimal notation.
Anything after “--" is a comment intended to improve the readability of this example and is not part
of the binary format for this CLR UDT.
01 -- bool true
01 -- byte 1
7E -- sbyte -2
8003 -- short 3
0004 -- ushort 4
7FFFFFFB -- int -5
00000006 -- uint 6
8000000000000007 -- long 7
0000000000000008 -- ulong 8
CCEB79A3 -- float 123456789.0123456789
3E6290CBABF35BA7 -- double -123456789.0123456789
0109 -- SqlByte [bool true, byte 9]
017FF6 -- SqlInt16 [bool true, short -10]
018000000B -- SqlInt32 [bool true, int 11
01800000000000000C -- SqlInt64 [bool true, long 12]

31 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

0180008EAC80C5C100 -- SqlDateTime [bool true, int 36524 days since 1/1/1900, int 12960000
ticks since midnight]
013314865C -- SqlSingle [bool true, float 1.234568E+08]
01C19D6F34540CA458 -- SqlDouble [bool true, double 123456789.0123456]
01800000000001FBD0 -- SqlMoney [bool true, long 130000]
02 -- SqlBoolean true

32 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

4

Security Considerations
None.

33 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

5

Appendix A: Product Behavior
The information in this specification is applicable to the following Microsoft products or supplemental
software. References to product versions include released service packs:
Microsoft SQL Server 2008 R2
Microsoft SQL Server 2012
Exceptions, if any, are noted below. If a service pack or Quick Fix Engineering (QFE) number
appears with the product version, behavior changed in that service pack or QFE. The new behavior
also applies to subsequent service packs of the product unless otherwise specified. If a product
edition appears with the product version, behavior is different in that product edition.
Unless otherwise specified, any statement of optional behavior in this specification that is prescribed
using the terms SHOULD or SHOULD NOT implies product behavior in accordance with the SHOULD
or SHOULD NOT prescription. Unless otherwise specified, the term MAY implies that the product
does not follow the prescription.
<1> Section 1.3: The Microsoft implementation does not produce values outside of the range
00:00:00.0000000 through 23:59:59.9999999, but it will accept values outside of the range as
described in section 2.4.2 in [MS-BINXML].
<2> Section 1.6: Version 1 denotes SQL Server 2008 R2. Version 2 denotes SQL Server 2012.
<3> Section 2.1: There are three secondary structures in SQL Server 2008 R2 and four secondary
structures in SQL Server 2012.
<4> Section 2.1.1: Version 1 denotes the SQL Server 2008 R2 version of the structure. Version 2
denotes the SQL Server 2012 version of the structure.
<5> Section 2.1.2: Version 1 denotes the SQL Server 2008 R2 version of the structure. Version 2
denotes the SQL Server 2012 version of the structure.
<6> Section 2.1.2: This bit is introduced in SQL Server 2012
<7> Section 2.1.2: This bit is introduced in SQL Server 2012.
<8> Section 2.1.2: This bit is introduced in SQL Server 2012.
<9> Section 2.1.3: These values apply to SQL Server 2008 R2.
<10> Section 2.1.3: These values apply to SQL Server 2012.
<11> Section 2.1.4: These values apply to SQL Server 2008 R2.
<12> Section 2.1.4: These values apply to SQL Server 2012.
<13> Section 2.1.7: This structure is introduced in SQL Server 2012.
<14> Section 3.1.5: This example applies to SQL Server 2012.

34 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

6

Change Tracking
No table of changes is available. The document is either new or has had no changes since its last
release.

35 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013

7

Index
A
Applicability statement 7

B
Basic GEOGRAPHY structure 9

C
Change tracking 35

E
Examples
Empty point structure 23
GEOGRAPHY structure 23
Geometry collection structure 25
Geometry point structure 23
GEOMETRY structure 23
Linestring structure 24

F
FIGURE packet 14
FIGURE structure 14

G
GEOGRAPHY packet (section 2.1.1 9, section 2.1.2
11)
GEOGRAPHY POINT packet 15
GEOGRAPHY POINT structure 15
GEOGRAPHY structure 9
GEOMETRY POINT packet 16
GEOMETRY POINT structure 16
GEOMETRY structure 9
Glossary 5

H

O
Overview 6

P
Product behavior 34

R
References
Informative 5
Normative 5
Relationship to protocols and other structures 7

S
Security considerations 33
SEGMENT packet 17
SHAPE packet 15
SHAPE structure 15
SQL Server versions 34
Structure examples 23
Structures 9
FIGURE 14
GEOGRAPHY 9
GEOGRAPHY POINT 15
GEOMETRY 9
GEOMETRY POINT 16
SHAPE 15

T
Tracking changes 35

V
Vendor-extensible fields 8
Versioning 7

HIERARCHYID examples 29
HIERARCHYID structure 17
Logical definition 17
Physical representation 17

I
Informative references 6
Introduction 5

L
Localization 7

N
Normative references 6

36 / 36
[MS-SSCLRT] — v20130611
Microsoft SQL Server CLR Types Serialization Formats
Copyright © 2013 Microsoft Corporation.
Release: Tuesday, June 11, 2013