<?xml version='1.0' ?>
<!DOCTYPE rfc SYSTEM 'rfc2629.dtd'>
<rfc category='bcp' ipr="trust200811" docName='draft-ietf-geopriv-civic-address-recommendations-02' >
    <?rfc toc='yes' ?>
    <?rfc tocompact='no' ?>
    <?rfc compact='yes' ?>
    <?rfc subcompact='yes' ?>
    <front> 
      <title abbrev='Civic Address Considerations'>
		Considerations for Civic Addresses in PIDF-LO - Guidelines and IANA Registry Definition
      </title>

      <author initials='K. H.' surname='Wolf' fullname='Karl Heinz Wolf'>
        <organization abbrev='nic.at'>
          nic.at GmbH
        </organization>
        <address>
          <postal>
            <street>Karlsplatz 1/2/9</street>
            <city>Wien</city>
            <code>A-1010</code>
            <country>Austria</country>
          </postal>
          <phone>+43 1 5056416 37</phone>
          <email>karlheinz.wolf@nic.at</email>
          <uri>http://www.nic.at/</uri>
        </address>
      </author>

      
      <author initials='A.' surname='Mayrhofer' fullname='Alexander Mayrhofer'>
        <organization abbrev='nic.at'>
          nic.at GmbH
        </organization>
        <address>
          <postal>
            <street>Karlsplatz 1/2/9</street>
            <city>Wien</city>
            <code>A-1010</code>
            <country>Austria</country>
          </postal>
          <phone>+43 1 5056416 34</phone>
          <email>alexander.mayrhofer@nic.at</email>
          <uri>http://www.nic.at/</uri>
        </address>
      </author>

      <date month='Feb' year='2009' day='18'/>
      <area>RAI</area>
      <workgroup>GEOPRIV</workgroup>
      <keyword>Civic Addresses</keyword>
      <keyword>PIDF-LO</keyword>
      
      <abstract>

	<t>
	This document provides a guideline for creating civic address consideration documents for individual countries, as required by RFC 4776. Furthermore, this document also creates an IANA Registry referring to such address 
	consideration documents.
<!-- Since civic addresses may have a different format in individual countries, such address considerations are necessary in order to map the civic address fields to the PIDF Location Object (PIDF-LO) elements. -->
	</t>

      </abstract>
    </front>


    <middle>

      <section anchor='intro' title='Introduction'>
        <t>
	The <xref target='RFC4119'>"Presence Information Data Format Location Object" (PIDF-LO)</xref> is an object format for carrying geographical
   	information on the Internet. PIDF-LO can be used to convey civic
	address information, and supports a range of "civic address types"
	(CATypes) to hold individual attributes of such addresses
	(see Section 2.2.1 of <xref target='RFC4119'/> and Section 3.1 of 
	<xref target='RFC5139'/>).
	</t>

<t>In many use cases, PIDF-LOs are populated with data from long-established sources, like postal or governmental building registers, line information databases and yellow / white pages of infrastructure providers, or official residents registers. The structure and format of data from such sources is almost always 
different from PIDF-LO's CAtypes definition - additionally, structure and format of those sources differs from country to country.</t>

<t>To make use of such existing data sources, transposing 
that data into PIDF-LO format is required.
With no guidelines
available on how to map source Fields into CAtype Elements, different creators
of PIDF-LO documents might end up with different results,
even when using the same data source - which reduces 
interopability and increases the risk of misinterpretation by 
receivers.
</t>

<t>
Therefore, 
civic address considerations are necessary
to ensure uniform usage of PIDF-LO Elements for such data sources. <xref target='RFC4776'/> explicitly requests such documents to be provided, but does neither define their structure 
nor a way to publish them.
This memo provides documentation on how to create such civic address considerations, and requests the creation of an IANA Registry to store references 
to such documents.
</t>

<t>For some countries Section 3.4 of <xref target='RFC4776'/> already contains considerations on the use of administrative sub-division elements. It's important to note that those examples are outdated, because <xref target='RFC5139'>RFC 5139</xref> disallows the use of the 'A6' elements for street names.
</t>

<t>
The guidelines in this document have been created with a focus on
formal application of PIDF-LO (such as conveying location during an emergency call). It is not intended to forbid other, more informal uses of PIDF-LO 
that may not follow any formal mapping specifications. An example usecase of 
such informal 
usage may be the transmission of PIDF-LO documents during an instant messaging session 
between humans. Such use may, however, imply some drawbacks like prohibiting automatic processing of civic addresses from such a PIDF-LO. 
</t>

<!--
However, such informal PIDF-LOs are typically not created from "official" data sources described in considerations documents, and they also don't need to be processed automatically.

Informal usage might be acceptable where addresses are not automatically processed but rather interpreted by humans, for example during an Instant Messaging session.
-->

<!-- 
<t>
This guideline document is based on the experience of writing such a civic address considerations document for Austria. Since there were some difficulties when trying to define a mapping for Austrian civic address fields to PIDF-LO, this document summarizes important experience and issues to consider. Even though every country has it's own address format and therefore other problems will occur, this guideline should help to identify difficulties. As examples, Austrian addresses are used.
</t>
-->

      </section>

      <section anchor='terminology' title='Terminology'>
	<t>
	The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
	"SHOULD", "SHOULD NOT", "RECOMMENDED",  "MAY", and "OPTIONAL" in 
	this document are to be interpreted as described in
	<xref target='RFC2119'>RFC 2119</xref>.
	</t>
        <t>In addition, this document uses "Field" to refer to a field of 
        a civic address data source, and "Element" to refer to 
	a CAType element of a PIDF-LO.</t>
      </section>

<section anchor='reqs' title='Requirements'>

  <t>The following requirements apply to defining civic address mapping 
  considerations:
  </t>
  <vspace blankLines='1'/>
  <list style='symbols'>
    <t>The considerations document MUST identify the data source to which 
    the definitions apply. A brief description of its structure 
    SHOULD be provided as well.</t>
  <vspace blankLines='1'/>
    <t>For any data source, just one active mapping definition should exist
    in order to reduce the risk of ambigious interpretation.
    </t>
  <vspace blankLines='1'/>
    <t>The document MUST include instructions for any Field that occurs in the data. For any of the Fields, the document MUST describe 
    whether the Field is required, optional, or must not be used in the mapping 
    procedure.</t>
  <vspace blankLines='1'/>
    <t>Instructions MUST be included for any CAtype Element that is registered 
    by the time the document is created. Those instructions MUST include 
    information whether an Element is required, optional, or must not be used 
    in that mapping.
    In case the set of CAtypes is revised by the IETF, the address 
    consideration document SHOULD be updated. Until an update is approved, 
    the existing mapping procedure MUST be used.
    </t>
  <vspace blankLines='1'/>
    <t>Address mapping procedures SHOULD be reversible, so that 
    location receipients can identify the corresponding record in the 
    original data source (given they have access to that source).
    </t>
  <vspace blankLines='1'/>
    <t>For any source data Field that is required or optional, at least one example mapping MUST be provided.
    </t>
  </list>
<!--
  <t>FIXME - requirements auf die adressumsetzung, requirements auf das dokument selbst. 
  </t>
!-->

<t>
Although the mapping is defined in a national way and the actual meaning of several PIDF-LO elements may not be clear to an outsider, at least the country element tells in what context this PIDF-LO was created. In case of emergency calls, a PIDF-LO would just be passed to a PSAP in the same country as the location generator anyway. However, in border region there might be exceptions and the PIDF-LO could be sent to a neighboring country. The PIDF-LO can still be passed on to a PSAP in the right country (based on the country element), or the PSAP might be aware of the mapping scheme used in the neighboring country.
</t>

<t>
A consistent mapping is also very important for checking if two PIDF-LO documents describe the same location. When civic address Fields are put into different PIDF-LO elements, it may be difficult to identify whether or not two PIDF-LOs 
describe identical addresses.
</t>

</section>



	<section anchor='usage' title='Specifying PIDF-LO Element Usage'>
<t>
The purpose of the civic address considerations for an individual data source is to create interopability by specifying a common list of PIDF-LO Elements to be used, and by defining the mapping between these Elements and the Fields of the respective data source.
</t>
<!-- <t>
The motivation for such a civic address consideration is to ensure interoperability. Location recipients certainly want to rely on finding civic address parts in defined elements of PIDF-LO for further processing. Especially when it comes to emergency calling, location information is a critical data where misinterpretation has to be avoided. Therefore, a consistent mapping scheme is required. Since it is not possible to have global PIDF-LO elements which can be unambiguously used in every country in the world, the mapping must be defined on an national level. It has to be ensured, that the mapping is used for all civic addresses in this country.  </t> -->

<section anchor='general_consideration' title='General Considerations and Workflow'>

<t>The workflow for creating an address considerations document is as follows:
</t>
<vspace blankLines='1'/>
<list style='numbers'>
<t>Describe the data source to which the address considerations document applies.</t>
<t>Identify all Fields from the data source, and decide for each of the Fields 
whether or not they are to be used for the purpose of creating PIDF-LO documents. In the considerations document, all Fields must be listed (or at least state which Fields are considered in the mapping and clearly state that the other Fields MUST NOT be used).
</t>
<t>For each of the Fields that is required or optional, specify a clear mapping instruction according to the guidelines below.</t>
<t>Provide a list of all CAtypes registered, and describe their level of usage in this mapping (or combine it with the list of Fields above and clearly list which Elements are not used for the mapping procedure). For Elements that are not described in detail state whether they MUST NOT be used at all or they may be used without further restriction.
</t>
<t>Provide examples of source data and mapping results</t>
</list>
<t>
Civic address fields are designed to be generic containers. In some cases, Fields clearly correspond to such a container, however, in some other cases, identifying the correct container might require some approximation.
For example, in some countries the RD (road) Element might also be appropriate for other thoroughfares, like waterways or tunnels. 
</t>
<t>Fields that are identified to have the same meaning as one of the CAtypes SHOULD be directly mapped to that CAtype Element.
</t>
<t>
Where CAtype usage diverges from the original specification, the mapping definiton of Fields that are mapped to that Element SHOULD include a discussion of the differences.
</t>
<t>Fields that do not fit into an existing CAtype: Even though the list of CAtypes could be extended, it is not feasible to add new elements for every new Field in every data source in every country. Therefore, unless new generic CAtypes are specified by the IETF, only existing elements can be used, which leaves the following options:
</t>
<vspace blankLines='1'/>
<list style='numbers'>
<t>
Concatenate several civic address fields into a single PIDF-LO element (define delimiters if applicable and make sure the separate civic address parts can be retrieved again)
</t><t>
Use a PIDF-LO element that is unused so far
</t>

</list>
<t>Note: Obviously, the first option is required if the number of Fields that are used in the mapping procedure is greater than the number of existing CAtype Elements.</t>

<t>
Note that the xml:lang attribute should be present in PIDF-LO XML documents according to RFC 5139. 
</t>

<!--
Depending on the circumstance and the particular situation in a specific country, the civic address elements could be used for a different purposes that has to be documented in the civic address considerations. When defining the usage, care must be taken to ensure that conflict do not occur. 
</t>
-->
<!--
<t>
Is is important to identify and describe which Fields are to be mapped directly to the corresponding Elements, and which Fields need special consideration. PIDF-LO Elements that are not needed in a specific country can simply be left out of the mapping. The civic address consideration document has to specify clearly that those elements must not be used for representing civic addresses in the described mapping.
</t> 
-->
<!--
FIXME: Was tun wir damit?
<t>
The defined mapping for a particular country aims to become the canonical format for that country. 
</t>
-->

<!--
<t>
All existing civic address parts must find a place in the PIDF-LO. Even rare types of addresses must be considered. Civic addresses can be very complex in some countries. So it is very important to identify the data source that is representing all the possible civic addresses in a country. Perhaps this database is maintained by a governmental company, by an authority, or the post. Moreover, it is important, that this data format is accepted by Public Safety Answering Point (PSAP) operators and they have access to the data source. Even PSAPs within a country may be organized differently and use different data bases for civic addresses. It is crucial to define the PIDF-LO mapping in a way that all PSAPs can use it. It is disireable to have PSAP operators involved in the process of developing civic address considerations, so that their needs are especially considered.
</t>

-->
</section>

	<section anchor='guide' title='Guidelines for Individual Elements'>
<t>
The following sections discuss individual PIDF-LO Elements and describe what to consider for each Element when defining civic address considerations. It is RECOMMENDED to follow a similar structure for considerations documents.
</t>



	<section anchor='country' title='Country'>

<t>
The country element must hold the alpha-2 codes from ISO 3166-1 <xref target='refs.ISO3166-1'></xref> in upper case characters as clarified in Section 3.3 of <xref target='RFC5139'>RFC 5139</xref>.
</t>

<t>This element cannot be redefined on a national basis since it identifies the country itself. This element is used to identify which national mapping for civic addresses has been used in a specific PIDF-LO.
</t>

<t>
Example for Austria: &lt;country&gt;AT&lt;/country&gt;
</t>

	</section>    

	<section anchor='a1a6' title='Country Subdivisions A1-A6'>

<t>
The elements A1 to A6 are used to hold national subdivision identifiers, 
with A1 holding the top-level subdivision identifier. A1 may either contain the
second part of ISO 3166-2 <xref target='refs.ISO3166-2'></xref> (see section 
3.4 of <xref target='RFC5139'>
RFC 5139</xref>), or values as described in the 
address consideration document.
Elements "A2" to "A6" may contain additional levels of subdivisions (see 
section 2.2.1 of RFC 4119).
</t>

<t>For A1, an address consideration document MUST state 
whether ISO 3166-2 codes are to be used exclusively, alternatively it should define a 
list of values to be used (for example, subdivision names). 
In either case, A1 MUST NOT be
redefined for any other use than describing top level subdivisions.
</t>
<t>
For each of the A2 - A6 Elements that is required or optional, the document SHOULD define the set of allowed values, either by listing them, or by referring to such a list.
</t>
<!--
<t>Note that Section 3.4 of RFC 4776 has some examples for a few countries, they may be followed.
</t>
<t>
Some countries may use codes for the subdivisions. Certainly, these codes can be used if they are unique. Also consider exceptions, possibly some bigger cities might have another structure of subdivisions.
</t> -->

<t>Example for Austria:</t>

<artwork align="left"><![CDATA[
A1 province (Bundesland)
A2 political district name or identifier (politscher Bezirk)
A3 commune name or identifier (Gemeinde)
A4 village name or identifier (Ortschaft)
A5 cadastral municipality name or identifier (Katastralgemeindename
   or Katastralgemeindenummer)
]]></artwork>
<t> 
A6 must not be used. For more details see the example in <xref target='a1a6a' />.
</t>

	</section>    

	<section anchor='road' title='Road and Street Names'>
<t>PIDF-LO contains the following Elements related to road names: 
RD, RDSEC, RDBR, RDSUBADDR, PRM, POM (section 3.1 and 3.2 of  <xref target='RFC5139'>RFC 5139</xref>) and PRD, POD, STS (section 3.4 of <xref target='RFC4776'/>). 
Note: The use of the A6 Element for street names is not valid any more (Section 3.2 of <xref target='RFC5139'>RFC 5139</xref>).</t>

<t>Besides the basic specification which of those Elements are required, optional or not to be used, and address considerations document may also describe more complicated dependencies (for example, "RD is optional, but required if any other road name Element is used").
</t>
<t>For any required or optional element, the relation of 
those Elements to Fields of the data source used MUST be described, as well as 
special considerations (like concatenation of Fields into an Element) if they apply.

The usage of the Element STS (street suffix) SHOULD be consistent.
In case no suffixes are known in a data source, or it is common to write the street name and the suffix together, the STS Element SHOULD be left out completely. If suffixes may be abbreviated the common abbreviations SHOULD be defined.
</t>

<!-- There are now several elements in PIDF-LO related to roads: RD, STS, PRD, POD, RDSEC, RDBR, RDSUBBR, PRM and POM. The intended purpose of these elements is available at <xref target='RFC4119'>"Presence Information Data Format Location Object" (PIDF-LO)</xref> and <xref target='I-D.ietf-geopriv-revised-civic-lo'>draft-ietf-geopriv-revised-civic-lo</xref>, they might fit the situation in several countries. Define the elements to be used and forbid the use of other elements. Redefine the usage in case other road data is available. Also focus on the element STS, the street suffix. It must be assured that this suffix is used in a consistent way. In case no suffixes are known in a country or it is common to write the street name and the suffix together, it is allowed to forbid the usage of the STS element completely. Suffixes may also be abbreviated. Define the common abbreviations. -->

<t>
Example for Austria:
</t>
<t>
RD: street name
</t>
<t>
All other road Elements must not be used, street suffixes are already included in the "street name" Field, and must not be abbreviated.
</t>

	</section>

	<section anchor='hno' title='House Numbers'>
<t>
PIDF-LO specifies two Elements related to house numbers: HNO ("house number", numeric part only) and HNS ("house number suffix") (see section 3.4 of RFC 4776).
However, in many countries house numbers have a more complex format. In any
case, a clear definition is REQUIRED to minimize confusion potential.
</t>

<t>An address consideration document should provide the following
information with regards to house numbers:
If the structure of house numbers fits the HNO/HNS structure, 
the document MUST mandate to use those fields as described in RFC 4776. 
If the structure of house numbers does not directly fit into those two Elements, 
the document MUST define strategies on how to map source Fields into Elements. Besides HNO and HNS, LOC and BLD could be considered for carrying house number information.
</t>
<t>The document SHOULD describe whether abbreviations of house number information is valid or not. If abbreviations are used, they MUST be clearly defined. If house number consists of more than one number, or multiple prefixes and suffixes may coexist, a delimiter symbol and a clear rule on how to concatenate all this data into the HNO and HNS element might be necessary. Whenever concatenating data into one Element, keep in mind that the location recipient might want to separate the data again.
</t>
<t>
Example from Austria:
</t>
<t>
HNO: concatenate all the data elements of Austrian house numbers into this single PIDF-LO Element in a defined order with delimiter symbols (see <xref target='pidf_mapping' /> for the complete definition).
</t><t>
HNS: Usage not allowed since there may be multiple suffixes for the different parts of the house number.
</t>
<t>
LOC and BLD are not to be used to reflect house number information.
</t>

</section>
	
<section anchor='locnam' title='Local Names'>

<t>
PIDF-LO contains three elements to reflect local names: LMK, LOC, NAM (section 
3.4 of RFC 4776).
Such local names may be of importance for the identification of a location, and 
may either coexist with a valid civic address or (in some cases) no address 
may be assigned so that the local names itself identify the location. 
In rural regions for example, a farm name may be more common than a street 
address to identify a location. Landmarks typically don't have any civic address information assigned. Therefore, local names may either assist in
finding a "street name" type addess, but they might also be the authoritative
(and only) civic location information.
</t>

<t>
For any required or optional Element out of LMK, LOC, NAM the considerations docuiment should state potential
values (source data) for the element. In case that multiple values for 
an Element may occur, a concatenation / selection strategy should be described.
Concatenation using ";" as seperator is recommended, unless this character also appears in the source Fields.
</t>
<t>If local name information and "common" address information is both 
available and used, the document SHOULD discuss the relation between those two
address information types, and expected behaviour of location receipients.
</t>
<!-- So there may be different types of local names that either are the only way to describe a location or may be an additional help with finding the right place. 
Depending on the type of the local name, the elements LMK, LOC or NAM may be suitable. The case where more than one local name is available has also be considered. Either more PIDF-LO elements can be used (if the number of local names has a low limit) or one element should be used with suitable delimiters. -->

<t>
Example from Austria: 
</t><t>
NAM: contains the "Vulgoname" (local name), multiple local names are separated by a semicolon (if applicable)
</t><t>
LMK: contains the farm name (just one name possible) (if applicable)
</t><t>
LOC: can be used without restriction for additional location information (as per RFC 4119)
</t>
<t>The "Vulgoname" is useful to identify the location within its locality, 
since official addresses especially in rural regions might not be well 
known.</t>
	
</section>

	<section anchor='floor' title='Floors'>
<t>
	PIDF-LO defines the element FLR to hold floor information, but 
does not further specify its content. Section 2.1 of RFC 3825 provides 
guidance about floor numbering, but is not directly related to PIDF-LO.
</t>

<t>
An address consideration document SHOULD clearly specify how 
to express floors using the FLR element. Following the above mentioned guidance
is RECOMMENDED, however, local nomenclature might require a completely 
different system. The document SHOULD specify whether only numbers, text, 
or both are allowed in the FLR element. If there are standard values for 
certain floors, they SHOULD be listed. Abbreviations SHOULD be avoided, unless
they are the primary (well known) way of identifying floors.
</t>
<!--
Floor numbering may be different in different countries. Although the floor numbers just have local meaning, it should be defined how to use the floor element. For example, should the FLR element just hold numbers or is text necessary to identify certain floors.
-->

<t>
Example from Austria: 
</t><t>
If floor numbers are to be mapped, the FLR Element MUST be used. Numbers and text are both allowed. The first floor (&lt;FLR&gt;1&lt;/FLR&gt;) is the first "full" floor above the floor at street level. The floor at street level is &lt;FLR&gt;EG&lt;/FLR&gt; or &lt;FLR&gt;0&lt;/FLR&gt;. There might be intermediate 
floors, especially between the floor at street level and the "first floor". Such intermediate floors have names like "Mezzanine", "Erster Halbstock" ("first half floor"), "Zweiter Halbstock" ("second half floor"), and have local meanings.
</t>
	</section>

	<section anchor='addrcodes' title='Address Codes'>

<t>
Address codes are available in several countries in different forms 
(for estates, buildings or usable units for example). These codes 
identify an address record, and MAY be placed in the ADDCODE element in PIDF-LO. Address codes can help the location recipient to determine the location, and 
to identify the original record in the data source. Depending on the type 
of code, the code alone (without any other Elements) may even be sufficient 
to fully identify an address within a country.
</t>
<!--
<t>
The PIDF-LO country element can be used to identify the name space in which
the address code elements are valid. Countries may have more than one type
of address codes (multiple namespaces), so it might be necessary to choose 
the code that is most widely accepted (by PSAPs) or to have identifiers 
for the different codes.
</t>
-->

<t>
In such cases, a PIDF-LO containing just the country and ADDCODE elements might provide enough
information to retrieve a civic address, given the location recipient has 
access to the respective source database.
</t>
<!-- Note, that when just using the country element and the ADDCODE element, the location recipient necessarily needs to have access to the address code database in order to determine the civic address. -->

<t>
A civic address considerations document SHOULD specify whether 
and in which applications the use of the ADDCODE Element is allowed. If ADDCODE
is used, its relation to the remaining Elements MUST be clearly stated. If 
several namespaces for address codes exist in a country, a mechanism to 
distinguish the different code spaces MUST be described.
</t>

<t>
Examples from Austria:
</t><t>
Statistik Austria provides 4 codes: Adresscode (AdrCD), Adresssubcode (AdrsubCD), Objektnummer (ObjNr) and Nutzungseinheitenlaufnummer (NtzLnr).
</t><t>
The following format SHOULD be used:
</t>
<artwork align="left"><![CDATA[
     <ADDCODE>AdrCD=1234567;AdrsubCD=123;
     ObjNr=2333211;NtzLnr=0001</ADDCODE>
]]></artwork>

	</section>


	<section anchor='other' title='Other Elements'>

<t>

	This section lists all PIDF-LO Elements that have not been discussed so far.
</t><t>
To specify the location inside a building, the following Elements can be useful:
</t>
<vspace blankLines='1'/>
<list style='symbols'>
<t>
UNIT
</t><t>
ROOM
</t><t>
SEAT
</t>
</list>
<t>

The following Elements are to be used for the representation of postal codes:
</t>
<vspace blankLines='1'/>
<list style='symbols'>
<t>
PC
</t><t>
PCN
</t><t>
POBOX
</t>
</list>
<t>
To describe the place-type or the building, the following Elements are available:
</t>
<vspace blankLines='1'/>
<list style='symbols'>
<t>
PLC - Place-type (see <xref target='RFC4589'/>)
</t><t>
BLD - Building (structure)
</t>
</list>
<t> 
For 
any of those Elements that are required or optional in a mapping, the semantics of its contents 
must be described, if it differs from the definition in the PIDF-LO base documents.
</t>
<t>
It is RECOMMENDED that the elements SEAT, UNIT and ROOM remain to be used for identifying a location inside a building. They MAY be used by the owner of the respective building if a considerations document does not restrict their use. For example, an airport could decide to place the gate number in the UNIT element, and a location recipient could identify that PIDF-LO by the value of the PLC Element. The name of the airport could be placed in NAM. 

<!--
In case of an emergency call, this PIDF-LO would be handled by action forces of the airport who are familiar with these definitions. So there is no need to define this mapping in the civic address document. 
-->

</t>

	</section>

	</section>    
</section>

      <section anchor='security' title='Security Considerations'>
        <t>
	RFC 4119 contains general security considerations for handling
	PIDF-LOs. In addition to that, it has to be considered that data 
	from certain data sources (on which the described mapping process is based) are possibly not public, so restrictions as imposed on the original
	data set MUST also be imposed on the resulting PIDF-LO document. The considerations document SHOULD note such restrictions in its Security Considerations section.
	</t>
    </section>

    <section anchor='iana' title='IANA Considerations'>
        <t>
	This document requests creation of the registry "PIDF-LO Civic Address Consideration Registry", according to the following definitions. Furthermore, this document requests registration of a civic address considerations document for Austrian addresses as provided in the Appendix of this document.
        </t>

	<section anchor='iana_reg' title='PIDF-LO Civic Address Consideration Registry'>
        <section anchor='iana_struc' title='Structure'>
	<t>
The IANA Registry contains the following fields:
</t>
<list style='symbols'>
<vspace blankLines='1'/>
<t>Country-Code: Either the ISO 3166 alpha-two code of the country to which the consideration applies or "other" in case the consideration document is not specific to a particular country. This field is to be defined by the requestor.</t>
<vspace blankLines='1'/>
<t>Serial Number: A number that uniquely identifies a considerations document within a certain "Country-Code" field value. Serial Numbers are sequentially assigned by IANA per "Country-Code" value, start at zero, and are never reused. </t>
<vspace blankLines='1'/>
<t>Reference to specification: This field contains a reference to the considerations document.</t>
<vspace blankLines='1'/>
<t>Requestor: The author of the document.</t>
<vspace blankLines='1'/>
<t>Status: One of "active" or "obsolete". When the document is registered by IANA, the status is first set to "active" by IANA. Experts may later request changing the status to "obsolete", especially if there is an updated version of the considerations document available. Authors of consideration documents must contact the experts if they wish to change the status of the document.</t>

	</list>
<t>
Note: The combination of "Country-Code" and "Serial Number" fields uniquely identify a considerations document in the registry (for example, "AT-0", "US-0", "US-1" or "other-0").
</t>

</section>


	<section anchor='iana_temp' title='Registration Template'>

<t>
For registration of address considerations documents in the registry, requestors SHOULD use the following template. The template SHOULD be contained in the considerations document itself.
</t>

        <artwork align="left"><![CDATA[


           <record>
	     <country> <!-- Country-Code --> </country>
	     <serial> <!-- assigned by IANA --> </serial>

	     <reference> 
               <!-- URI to the considerations document --> 
             </reference>

             <requesters>
               <!-- Change accordingly -->
               <xref type="person" data="John_Doe"/>
               <xref type="person" data="Jane_Dale"/>
             </requesters>
	
             <status> <!-- assigned by IANA --> </status>
           </record>


          <people>
            <person id="John_Doe">
              <name> <!-- Firstname Lastname --> </name>
              <org> <!-- Organisation Name --> </org>
              <uri> <!-- mailto: or http: URI --> </uri>
              <updated> <!-- date format YYYY-MM-DD --> </updated>
            </person>
            <!-- repeat person section for each person -->
          </people>



            ]]></artwork>


</section>

	<section anchor='iana_loc' title='Registry Location'>
<t>

Approved registrations are published in the IANA registry named "PIDF-LO Civic Address Consideration Registry", which is available at the following URI: XXX ((TO BE DEFINED BY IANA)).
</t>
<!--
<t>
For each value of the country-code field there may be multiple documents, each with a different serial number. For example, AT 1, AT 2 and so on.
</t>
-->
<t>
Registration are sorted by ascending order by the country code, and by serial number within country code values. Registrations with country code of "other" are put at the end of the list.
</t>


</section>

<section anchor='iana_procedure' title='Registration Procedure'>

<t>
Following the policies outlined in <xref target='RFC5226'/>, new address considerations are added to the registry after Expert Review (see Section 4.1 in RFC 5226). 

The Expert will generally check if the submitted address considerations conforms the civic address guidelines in this document (section <xref target='usage'/>). If in doubt, the Experts SHOULD consult the GEOPRIV mailing list or it's dedicated successor. If possible, the Experts SHOULD check the available documentation on which the address consideration is based.
</t>
<!--
<t>
The address consideration has to be written in a language, the expert reviewer is able to understand.
</t>
-->





    </section>     <!-- end of registration procedure section -->
    </section>     <!-- end of registry definition section -->
    

    <section anchor='austria_request' title='Registration Request for Austria'>
    <t>This document requests registration of the Civic Address Considerations 
    for addresses form the official Austrian Building an Habitation registry, according to the registration procedure described
    above. The required information is contained in Appendix A</t>
    </section>

    </section>     <!-- end of IANA considerations section -->

    <section anchor='ack' title='Acknowledgements'>
      <t>
      The authors would like to thank Martin Thomson and Richard Barnes for reviewing the document, and Gregor Jänin for contributing insights 
      into the Austrian civic address data format.
      </t>
    </section>


  </middle>
  
  <back>

<section anchor='aut' title='Civic Address Considerations Registration for the Austrian building and habitation registry'>



 	<t>
        The Austrian "Gebäude- und Wohnungsregistergesetz" (building 
	and habitation registry law) is the legal basis for the obligation to 
	provide a registry of civic addresses, buildings and their
	usable units (subdivisions of buildings). The registry is
	operated by "Statistik Austria GmbH", a fully governmental 
	owned company. Responsibility for keeping records in the registry 
	up to date is an obligation to the local administration of the 
	individual townships.
	</t>
	<t>
	The data format definition for the individual
	records is publicly available (data access itself is however
	restricted). Hence, an uniform address data base for whole Austria 
	is available. A	detailed description of the Statistik Austria civic address data format is 
	contained in section <xref target='CivicAddresses'/>.
	</t>
<!--
<t>
A guideline of how to use PIDF-LO for Austrian addresses is necessary in order to avoid misinterpretations. This is especially important if the PIDF-LO is conveyed during an emergency call to a Public Safety Answering Point (PSAP). A precise location information is needed in case of emergency to send out responders without any delay to the correct location of the caller. If every data-provider uses its own address mapping to PIDF-LO, confusion and misunderstandings are bound to happen. However, ideally any PSAP should have full access to the data by Statistik Austria. PSAPs must be able to rely that location information is always provided the same way by all data-providers. To address the idiosyncrasies in Austria, the civic address elements are discussed subsequently.
</t>
-->



      <section anchor='CivicAddresses' title='Civic Address Format in Austria'>


        <t>

Statistik Austria data describes estates, buildings and usable units <xref target='refs.merkmalskatalog'></xref>. On a single estate there may be any number of buildings. Apartment houses that have more than one staircase, are split up in separate buildings at every staircase. In every building, there may be several usable units. For example, an apartment house may have several apartments, counting as separate usable units. Moreover, one building may have more than one address, but at least one address. Below, the address Fields for estates (<xref target='table_estates'></xref>), buildings (<xref target='table_building'></xref>) and usable units (<xref target='table_habitation'></xref>) are shown.
</t>
<t>
The ADDCODE, A5 and PCN Elements are optional, the other Elements MUST be used if the data source contains their corresponding Fields. The Elements A1 and A2 (not listed in the tables) SHOULD also be used if data is available. Exception: when using the address codes only (access to the codes is necessary for creator and receiver of the location information), just the ADDCODE and country Elements are mandatory, the other Elements can be used optionally of course.
	</t>

        <texttable anchor="table_estates" title="Civic Address Fields for Estates">

          <preamble></preamble>

          <ttcol align="center">Statistik Austria name</ttcol>

          <ttcol align="center">Explaination</ttcol>

          <ttcol align="center">PIDF-LO Element</ttcol>

          <c>Adresscode</c><c>address identifier</c><c>ADDCODE</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Gemeindename, Gemeindekennziffer</c><c>commune name and identifier</c><c>A3</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Ortschaftsname, Ortschaftskennziffer</c><c>village name and identifier</c><c>A4</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Straßenname, Straßenkennziffer</c><c>street name and identifier</c><c>RD</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Katastralgemeindename, Katastralgemeindenummer</c><c>cadastral municipality and identifier</c><c>A5</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hausnummerntext</c><c>text in front of the house number</c><c>HNO</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hausnummer - 1. Teil - Nummer</c><c>first part of the house number, numeric</c><c>HNO</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hausnummer - 1. Teil - Buchstabe</c><c>first part of the house number, character</c><c>HNO</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hausnummer – Verbindungszeichen Teil 1 -> Bis</c><c>links first and Bis part of house number</c><c>HNO</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hausnummer – Bis-Nummer</c><c>number of bis part of house number</c><c>HNO</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hausnummer – Bis-Buchstabe</c><c>character of bis part of house number</c><c>HNO</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hausnummernbereich</c><c>indicates if all house numbers specified or just odd or even numbers are stated</c><c>HNO</c>
<!--
<c>&nbsp;</c><c></c><c></c>
          <c>Grundstücksnummer</c><c>estate identifier</c><c>HNO - FIXME if necessary</c>
-->

<c>&nbsp;</c><c></c><c></c>
          <c>Postleitzahl</c><c>postal code</c><c>PC</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Postleitzahlengebiet</c><c>postal community code</c><c>PCN</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Vulgoname</c><c>local name</c><c>NAM</c>
<c>&nbsp;</c><c></c><c></c>
          <c>Hofname</c><c>farm name</c><c>LMK</c>


          <postamble></postamble>
        </texttable>



        <texttable anchor="table_building" title="Additional Civic Address Fields for Buildings">

          <preamble></preamble>

          <ttcol align="center">Statistik Austria name</ttcol>

          <ttcol align="center">Explaination</ttcol>

          <ttcol align="center">PIDF-LO Element</ttcol>

          <c>Adressubcode</c><c>address subcode</c><c>ADDCODE</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Objektnummer</c><c>object code</c><c>ADDCODE</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Hausnummer – Verbindungszeichen Teil Bis -> Teil 2</c><c>links Bis and second part of house number</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Hausnummer – 2. Teil – Nummer</c><c>second part of the house number, numeric</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Hausnummer – 2. Teil – Buchstabe</c><c>second part of the house number, character</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Hausnummer – Verbindungszeichen Teil 2-> Teil 3</c><c>links second and third part of house number</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Hausnummer – 3. Teil – Nummer</c><c>third part of the house number, numeric</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Hausnummer – 3. Teil – Buchstabe</c><c>third part of the house number, character</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Gebäudeunterscheidung</c><c> for differentiation of buildings, e.g. Maierweg 27 Hotel vers. Maierweg 27 Appartmenthaus</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <postamble></postamble>
        </texttable>



        <texttable anchor="table_habitation" title="Additional Civic Address Fields for usable units">

          <preamble></preamble>

          <ttcol align="center">Statistik Austria name</ttcol>

          <ttcol align="center">Explaination</ttcol>

          <ttcol align="center">PIDF-LO Element</ttcol>


          <c>Nutzungseinheitenlaufnummer</c><c>usable unit code</c><c>ADDCODE</c>
	  <c>&nbsp;</c><c></c><c></c>


          <c>Türnummer</c><c>door number</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Topnummer</c><c>unit number</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Lagebeschreibung</c><c>for verbal description</c><c>HNO</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Lage</c><c>describes if the usable unit is in the basement, mezzanine, attic floor, ... (but not the floor number)</c><c>FLR</c>
	  <c>&nbsp;</c><c></c><c></c>

          <c>Stockwerk</c><c>floor</c><c>FLR</c>
	  <c>&nbsp;</c><c></c><c></c>

          <postamble></postamble>
        </texttable>


	<t>
	Note: "Floors" in Austria (as in most parts of Europe) are counted 
	differently compared to the US. The "1st floor" in Austria is actually 
	the floor above the floor at street level (2nd floor in US), not 
	considering the fact that in old buildings there might be even more 
	floors between street level and 1st floor, like "mezzanine", 
	"2nd mezzanine". So, an Austrian "1st floor" could well be the 
	"4th floor" according to US nomenclature.
	</t>
<t>
According to <xref target='refs.adrwarten'>Statistik Austria</xref>, 
81.5% of Austrian addresses are of the simple type Musterstraße 1 
(Musterstraße is an example street name). 5% of all addresses have an 
additional character, like Musterstraße 1b. 1% of Austrian addresses 
look like Musterstraße 21A - 23A. For 8% of addresses, an additional 
separator is necessary, like Musterstraße 10 Haus 1 Stiege 2 or 
Musterstraße 20 Gruppe A Reihe 1 Parzelle 13 or Musterstraße 30 Weg 1 
Parzelle 10.  Very seldom, there are so called special addresses 
(0.03%), for example Musterstraße gegenüber 3a, meaning this address is actually opposite of house number 3A. Rather surprisingly, 4.47% of Austrian addresses contain the identifier of the estate since no house number is assigned at all, for example: Musterstraße GNR 1234, or Musterstraße GNR .12/4 Kirche (this type of addresses is common for churches) or a real example in Stockerau: Kolomaniwörth GNR 1583. This identifier is stored by Statistik Austria as Hausnummerntext. Otherwise one could misinterpret this number as a house number, what would be definitely wrong.

	</t>


      </section>

      <section anchor='samples' title='Sample Addresses'>
        <t>
	In order to clarify the Austrian civic address format, this section provides some exemplary addresses:
	</t>

        <artwork align="left"><![CDATA[
1234 Musterstadt, Hauptstraße 1a – 5a Block 1b Haus 2c Stiege 1
Postleitzahl: 1234
Stadt: Musterstadt
Straße: Hauptstraße
Hausnummer - 1.  Teil - Nummer: 1
Hausnummer - 1.  Teil - Buchstabe: a
Hausnummer - Verbindungszeichen Teil 1 -> Bis: -
Hausnummer - 2.  Teil - Nummer: 5
Hausnummer - 2.  Teil - Buchstabe: a
Hausnummer - Verbindungszeichen Teil Bis -> Teil 2: Block
Hausnummer - 2.  Teil - Nummer: 1
Hausnummer - 2.  Teil - Buchstabe: b
Hausnummer - Verbindungszeichen Teil 2-> Teil 3: Haus
Hausnummer - 3.  Teil - Nummer: 2
Hausnummer - 3.  Teil - Buchstabe: c
Gebäudeunterscheidung: Stiege 1

1234 Musterstadt, Musterstraße 13 Hotel
Postleitzahl: 1234
Stadt: Musterstadt
Straße: Musterstraße
Hausnummer - 1. Teil - Nummer: 13
Gebäudeunterscheidung: Hotel

6020 Innsbruck, Anichstraße vor 35
Postleitzahl: 6020
Stadt: Innsbruck
Straße: Anichstraße
Hausnummerntext: vor ("in front of")
Hausnummer: 35

6173 Oberperfuss, Riedl 3097 (Pfarrkirche)
Postleitzahl: 6173
Stadt: Oberperfuss
Straße: Riedl
Hausnummerntext: 3097
(since the estate identifier is 81305 3097 where 81305 is the 
Katastralgemeindenummer (cadastral municipality) and no house
number is assigned)
Vulgoname: Pfarrkirche

            ]]></artwork>

    </section>

      <section anchor='codes' title='Address Codes in Austria'>
<t>
Statistik Austria registers 4 codes: Adresscode, Adresssubcode, Objektnummer and the Nutzungseinheitenlaufnummer. The Adresscode (7 digits) is a unique code for an address in Austria. The Adressregister maps the Adresscode to the civic address. If there is a building located at an address, there is also an Adresssubcode (3 digits) assigned. Every building at an address has its own Adresssubcode (assigned sequentially starting with 001, 002, 003 and so on) in order to distinguish between buildings at the same address. Furthermore, every building located in Austria has its own unique code, the Objektnummer (7 digits). This code identifies the building independent of the Adresscode. That's because addresses are subject to change while the building may persist. To differ multiple usable units inside a building, the Nutzungseinheitenlaufnummer (4 digits) is used. This code is also assigned in sequential order for each building.
</t><t>
Besides, every address and building is geocoded by Statistik Austria. Hence, if every PIDF-LO location object would carry data in the format of Statistik Austria and every PSAP would use the database of Statistik Austria for mapping, a time saving, definite mapping without irregularities could be achieved.
</t><t>
Besides these codes, Statistik Austria maintains reference numbers for communes, localities or streets, to mention just a few.
</t>

    </section>


      <section anchor='pidf' title='Austrian Addresses in PIDF-LO'>
<t>
The following subsections define the mapping procedure.
</t>


      <section anchor='country-A' title='Country'>
        <t>
The country element for Austria must be set to AT, since this is the ISO 3166-1 <xref target='refs.ISO3166-1'></xref> alpha-2 code for Austria. 
<artwork align="left"><![CDATA[
<country>AT</country>
]]></artwork>
The usage of the ISO 3166 code is demanded by <xref target='RFC4119'>RFC 4119</xref> and <xref target='RFC5139'>RFC 5139</xref> proposes to use upper case characters only.
	</t>
    </section>

      <section anchor='a1a6a' title='Country Subdivisions A1-A6'>

The Elements A1-A6 are used in Austria for the following data:

<artwork align="left"><![CDATA[
A1 province (Bundesland), Section A.4.2.1
A2 political district name or identifier (politscher Bezirk), 
   Section A.4.2.2
A3 commune name or identifier (Gemeinde), Section A.4.2.3
A4 village name or identifier (Ortschaft), Section A.4.2.4
A5 cadastral municipality name or identifier (Katastralgemeindename
   or Katastralgemeindenummer), Section A.4.2.5
]]></artwork>

<t>
Element A6 must not be used.
</t>
<t>
Last, there is an exception to mention concerning the Austrian capital Vienna (Wien). The city of Vienna is equal to its political district and even the province is called Vienna. Nevertheless, Vienna is separated in 23 districts within the same political district. Consequently, an address in Vienna would look like:
</t>
<artwork align="left"><![CDATA[
<country>AT</country>
<A1>Wien</A1>
<A2>Wien</A2>
<A3>Wien</A3>
<A4>Favoriten</A4> or <A4>10<A4>
<A5>Inzersdorf Stadt<A5>
]]></artwork>
<t>
The element A4, holding the city division, can hold the name or the number of the district.
</t>


	<section anchor='a1' title='A1 Element'>
        <t>
As proposed in <xref target='RFC5139'>RFC 5139</xref>, for the PIDF-LO element A1, the second part of ISO 3166-2 <xref target='refs.ISO3166-2'></xref> can be used. However, in Austria it is also common to write out the names of the states. <xref target='table_a1'></xref> shows the possible values of the A1 element for Austrian states.
	</t>


        <texttable anchor="table_a1" title="A1 element format for Austria (Note: values are shown in UTF-8, which is recommended to be used for PIDF-LO)">

          <preamble></preamble>

          <ttcol align="center">Bundesland</ttcol>

          <ttcol align="center">second part of ISO 3166-2 code</ttcol>

          <c>Burgenland</c><c>1</c>
	  <c>&nbsp;</c><c></c>
          <c>K=U+00E4rnten</c><c>2</c>
	  <c>&nbsp;</c><c></c>
          <c>Nieder=U+00F6sterreich</c><c>3</c>
	  <c>&nbsp;</c><c></c>
          <c>Ober=U+00F6sterreich</c><c>4</c>
	  <c>&nbsp;</c><c></c>
          <c>Salzburg</c><c>5</c>
	  <c>&nbsp;</c><c></c>
          <c>Steiermark</c><c>6</c>
	  <c>&nbsp;</c><c></c>
          <c>Tirol</c><c>7</c>
	  <c>&nbsp;</c><c></c>
          <c>Vorarlberg</c><c>8</c>
	  <c>&nbsp;</c><c></c>
          <c>Wien</c><c>9</c>

          <postamble></postamble>
        </texttable>
</section>

      <section anchor='a2' title='A2 Element'>
        <t>
Names of the Austrian political districts are available at Statistik Austria  <xref target='refs.bezirke'></xref>. These names, the unique code for the politcal district or both can be used for the A2 element. If the content of the A2 elment is numeric, obviously the code is provieded (there is no political district in Austria with a number in its name). In case both, the name and the code are provided, they are seperated by a semicolon, and the name must be listed first.
<t></t>
The district of "Bruck an der Leitha" could be represented by:

<artwork align="left"><![CDATA[
<A2>Bruck an der Leitha<A2> or <A2>307</A2> or
<A2>Bruck an der Leitha;307</A2>
]]></artwork>
	</t>


    </section>    

      <section anchor='a3' title='A3 Element'>
	<t>The element A3 holds the Gemeindename (commune name) or the identifier of the Gemeinde, or both separated by a semicolon (the name must be listed first). If the content of the A3 element consists of a number only, it is obvious that just the identifier is provided. Statistik Austria maintains a table with the Gemeindenamen and identifiers <xref target='refs.gemeinden'></xref>, which must be used as the content for the A3 element, no other spelling is allowed.</t>
<t>
Sample:
</t>
<artwork align="left"><![CDATA[
<A3>Neusiedl am See</A3>
or
<A3>10713</A3>
or
<A3>Neusiedl am See;10713</A3>
]]></artwork>

</section>

<section anchor='a4' title='A4 Element'>

<t>
The element A4 holds the Ortschaftsname (village name), the Ortschaftskennziffer (the identifier), or both separated by a semicolon (the name must be listed first). If the content of the A4 element consists of a number only, it is obvious that just the identifier is provided since there are no Ortschaftsnamen in Austria which contain a number. Statistik Austria maintains a table with the Ortschaftsnamen and identifiers <xref target='refs.ortschaften'></xref>, which must be used as the content for the A4 element, no other spelling is allowed.
</t><t>
Sample:
</t>
<artwork align="left"><![CDATA[
<A4>Wilfleinsdorf</A4> or <A4>03448</A4> or
<A4>Wilfleinsdorf;03448</A4>
]]></artwork>

</section>

<section anchor='a5' title='A5 Element'>

<t>
The element A5 holds the Katastralgemeindename (cadastral municipality), the Katastralgemeindekennziffer (the identifier), or both separated by a semicolon (the name must be listed first). If the content of the A5 element consists of a number only, it is obvious that just the identifier is provided since there are no Katastragemeindenamen in Austria which contain a number.
<!--
FIXME: there is no Katastralgemeindeliste bei der statistik austria! BEV, dataweb - keinz zugang. beispiel aus der wikipedia.
!-->
</t><t>
Sample (Vienna, Fünfhaus):
</t>
<artwork align="left"><![CDATA[
<A5>Oberbaumgarten</A5> or <A5>1208</A5> or 
<A5>Oberbaumgarten;1208</A5>
]]></artwork>


</section>

</section>

<section anchor='street_a' title='Road and Street Names'>

<t>
The PIDF-LO element RD holds the complete street name, including the street suffix. No abbreviations are allowed. No other elements are needed for streets and must not be used.
</t>

</section>


		<section anchor='pidf_mapping' title='House Numbers'>

<t>
Statistik Austria lists
 14 data fields related to the house number of a building plus another 5 fields for distinction of different usable units inside a building (including the floor, which has a separate element in PIDF-LO). 

Unfortunately, PIDF-LO only defines a single house number element (HNO, numeric part only) and a house number suffix element (HNS). Therefore, this section defines a mapping in order to accomodate all data: All house number data is concatenate into a single HNO element, even though it is expected to hold numeric part only.

<!--
wegdamit XXX FIXME: As delimiter symbols the connecting-symbols as defined by Statistik Austria have to be used (no abbreviations allowed).

If the location recipient does not need to separate the data elements again, the house number parts may be simply concatenated with spaces in between (no spaces between the numeric part of a house number and its related character). 
!-->


</t><t>
In order to allow automatic procession of the HNO Element, it is necessary to use a semicolon as delimiter symbol (Austrian house numbers do not contain semicolons). The house number parts MUST be provided in the order as they are listed by the Statistik Austria document <xref target='refs.merkmalskatalog'></xref>. For user interface representation, the semicolon separated format can be transformed by replacing semicolons by spaces (multiple spaces should be combined) and no space should be present between a numeric part of a house number part and its related character.

</t><t>
It is not allowed to use the HNS element for Austrian addresses, since there are addresses that do not have just a single suffix.
</t>
<!--
 For example, the address Lazarettgasse 13A could be mapped by:
<artwork align="left"><![CDATA[
<HNO>13</HNO> <HNS>A</HNS>  <!-- bad example -->
<!--
]]></artwork>
However, the building at Lazarettgasse has the house number 13A - 13C. Consequently, just the HNO element should be used:
<artwork align="left"><![CDATA[
<HNO>13A - 13C</HNO>
]]></artwork>
And even for addresses with a house number consisting of a single number and a single prefix, just HNO should be used because of uniformity:
<artwork align="left"><![CDATA[
<HNO>13A</HNO>
]]></artwork>
Addresses with a house number text would look like:
<artwork align="left"><![CDATA[
<HNO>vor 1 - 1A</HNO>
]]></artwork>
with no HNS element.
</t>
-->
<t>
The house number "vor 1 - 1A" (consisting of a house number text "vor", first part of the house number numeric "1", "-" as the link of the first and Bis part, "1" as house number bis part numeric, "A" as character of the bis part) would be mapped to:
<artwork align="left"><![CDATA[
<HNO>vor;1;;-;1;A;;;;;;;;;;;</HNO>
]]></artwork>
</t>


	        </section>

<section anchor='locals_a' title='Local Names'>
<t>
NAM: contains the Vulgoname (local name), multiple local names are separated by a semicolon (if applicable)
</t><t>
LMK: contains the farm name (just one name possible) (if applicable)
</t><t>
LOC: can be used without restriction for additional location information (as per RFC 4119) 
</t>
</section>


<section anchor='floor_a' title='Floors'>

<t>
The floor element may contain numbers or text describing the floor. The first floor (&lt;FLR&gt;1&lt;/FLR&gt;) is the floor above the floor at street level. The floor at street level is &lt;FLR&gt;EG&lt;/FLR&gt; or &lt;FLR&gt;0&lt;/FLR&gt;. Other floors may have names like mezzanine, for example. The Statistik Austria data elements Lage and Stockwerk are concatenated if necessary.
</t>

</section>

      <section anchor='pidf_ADDCODE' title='Additional Code Element'>
        <t>
The element additional code may be used to hold the codes provided by Statistik Austria. There is an Adresscode, Adressubcode, Objektnummer and a Nutzungseinheitenlaufnummer. These unique codes identify the location. Actually, these codes alone would be enough, but requires that the location recipient has access to the database of Statistik Austria.
</t><t>
If the additional code in a PIDF-LO document is going to hold the codes from Statistik Austria, the following format should be used:
<artwork align="left"><![CDATA[
<ADDCODE>AdrCD=1234567;AdrsubCD=123;
ObjNr=2333211;NtzLnr=0001</ADDCODE>
]]></artwork>
It is not necessary to provide all codes, but there are some restrictions: The Adresssubcode cannot be used without an Adresscode. More restrictions are definded by Statistik Austria.

By setting the country element to AT (see <xref target='country'></xref>), indicating an Austrian address, the Additional Code element is expected to hold codes from Statistik Austria only. When creating PIDF-LO documents using address codes by Statistik Austria, the country and ADDCODE elements are mandatory.
	</t>
    </section>

<section anchor='other-a' title='Other Elements'>

<t>
The elements PC and PCN can hold the data form Statistik Austria, the POBOX can be used if the post assigned a post office box. At least the PC element should be present.
</t><t>
PC: Postleitzahl (postal code)
</t><t>
PCN: Postleitzahlengebiet (postal community name)
</t><t>
POBOX: Postfach
</t><t>
The elements UNIT, ROOM, SEAT, PLC and BLD may be used without further restriction.

</t>

</section>

<section anchor='nouse' title='Elements not to be used'>

The following listing shows all PIDF-LO elements that should not be used for representing Austrian addresses:

<artwork align="left"><![CDATA[
A6
STS
HNS
PRD
POD
RDBR
RDSUBBR
PRM
POM
]]></artwork>

    </section>

</section>


      <section anchor='example' title='Example'>
        <t>

This section shows an example mapping of an Austrian address to PIDF-LO.


<artwork align="left"><![CDATA[
Address:

Bundesland: Wien
Politischer Bezirk: Wien
Gemeindename: Wien
9. Bezirk
Strasse: Lazarettgasse
Hausnummer - 1. Teil - Nummer: 13
Hausnummer - 1. Teil - Buchstabe: A
Hausnummer - Verbindungszeichen Teil 1-Bis: -
Hausnummer - Bis-Nummer: 13
Hausnummer - Bis-Buchstabe: C
Postleitzahl: 1090

]]></artwork>


PIDF-LO:

<artwork align="left"><![CDATA[

<?xml version="1.0" encoding="UTF-8" ?>
  <presence xmlns="urn:ietf:params:xml:ns:pidf"
     xmlns:gp="urn:ietf:params:xml:ns:pidf:geopriv10"
     xmlns:cl="urn:ietf:params:xml:ns:pidf:geopriv10:civicAddr"
     entity="pres:123@examplehost">
   <tuple id="abcd123456">
    <status>
     <gp:geopriv>
       <gp:location-info>
         <cl:civicAddress>
           <cl:country>AT</cl:country>
           <cl:A1>Wien</cl:A1>
           <cl:A2>Wien</cl:A2>
           <cl:A3>Wien</cl:A3>
           <cl:A4>9</cl:A4>
           <cl:RD>Lazarettgasse</cl:RD>
           <cl:HNO>;13;A;-;13;C;;;;;;;;;;;;</cl:HNO>
           <cl:PC>1090</cl:PC>
         </cl:civicAddress>
       </gp:location-info>
       <gp:usage-rules>
         <gp:retransmission-allowed>yes</gp:retransmission-allowed>
         <gp:retention-expiry>2009-11-10T12:00:00Z</gp:retention-expiry>
       </gp:usage-rules>
     </gp:geopriv>
    </status>
    <timestamp>2009-02-09T12:00:00Z</timestamp>
   </tuple>
  </presence>
]]></artwork>

	</t>
    </section>



<section anchor='iana_austria' title='IANA Registration Record'>

<t>

        <artwork align="left"><![CDATA[


           <record>
	     <country>AT</country>
	     <serial> <!-- assigned by IANA --> </serial>

	     <reference> 
               <!-- RFC-Editor: the URI to his RFC / --> 
             </reference>

             <requesters>
               <xref type="person" data="Alexander_Mayrhofer"/>
               <xref type="person" data="Karl_Heinz_Wolf"/>
             </requesters>
	
             <status> <!-- assigned by IANA --> </status>
           </record>


          <people>
            <person id="Alexander_Mayrhofer">
              <name>Alexander Mayrhofer</name>
              <org>nic.at Gmbh</org>
              <uri>mailto:alexander.mayrhofer@nic.at</uri>
              <updated>2009-01-09</updated>
            </person>
            <person id="Karl_Heinz_Wolf">
              <name>Karl Heinz Wolf</name>
              <org>nic.at Gmbh</org>
              <uri>mailto:karlheinz.wolf@nic.at</uri>
              <updated>2009-01-09</updated>
            </person>
          </people>



            ]]></artwork>

</t>

</section>   

</section>   


    <references title='Normative References'>


      <?rfc include="reference.RFC.2119" ?>
      <?rfc include="reference.RFC.4119" ?>
      <?rfc include="reference.RFC.4776" ?>
      <?rfc include="reference.RFC.5139" ?>
      <?rfc include="reference.RFC.5226" ?>
      <?rfc include="reference.RFC.4589" ?>

    </references>

    <references title='Informative References'>



      <reference anchor='refs.adrwarten'>
        <front>
          <title abbrev='Theoretisches Handbuch'>Handbuch Adress-GWR-Online Teil A Theoretisches Handbuch Kapitel 2 Warten von Adressen im Adress-GWR-Online</title>
          <author initials='' surname='' fullname=''>
            <organization abbrev='Statistik AT'>Statistik Austria</organization>
          </author>
          <date month='Jan' year='2005'/>
        </front>
      </reference>

      <reference anchor='refs.merkmalskatalog'>
        <front>
          <title abbrev='Merkmalskatalog'>Handbuch Adress-GWR-Online Teil C Anhang 2 Merkmalskatalog</title>
          <author initials='' surname='' fullname=''>
            <organization abbrev='Statistik AT'>Statistik Austria</organization>
          </author>
          <date month='Sept' year='2004'/>
        </front>
      </reference>


      <reference anchor='refs.ISO3166-1'>
        <front>
          <title abbrev='ISO3166-1'>Codes for the representation of names of countries and their subdivisions - Part 1: Country codes</title>
          <author initials='' surname='' fullname=''>
            <organization abbrev='ISO'>International Organization for Standardization</organization>
          </author>
          <date month='' year='1997'/>
        </front>
<seriesInfo name="ISO Standard" value="3166-1:1997"/>
      </reference>

      <reference anchor='refs.ISO3166-2'>
        <front>
          <title abbrev='ISO3166-2'>Codes for the representation of names of countries and their subdivisions - Part 2: Country subdivision code</title>
          <author initials='' surname='' fullname=''>
            <organization abbrev='ISO'>International Organization for Standardization</organization>
          </author>
          <date month='' year='1998'/>
        </front>
<seriesInfo name="ISO Standard" value="3166-2:1998"/>
      </reference>

      <reference anchor='refs.bezirke'>
        <front>
          <title abbrev='Bezirke'>Politische Bezirke, Gebietsstand 2008</title>
          <author initials='' surname='' fullname=''>
            <organization abbrev='Statistik AT'>Statistik Austria</organization>
          </author>
          <date month='Feb' year='2008'/>
        </front>
      </reference>


      <reference anchor='refs.gemeinden'>
        <front>
          <title abbrev='Gemeinden'>Gemeindeliste sortiert nach Gemeindekennziffer, Gebietsstand 2008</title>
          <author initials='' surname='' fullname=''>
            <organization abbrev='Statistik AT'>Statistik Austria</organization>
          </author>
          <date month='Feb' year='2008'/>
        </front>
      </reference>

      <reference anchor='refs.ortschaften'>
        <front>
          <title abbrev='Ortschaften'>Gemeinden mit Ortschaften und Postleitzahlen, Gebietsstand 2008</title>
          <author initials='' surname='' fullname=''>
            <organization abbrev='Statistik AT'>Statistik Austria</organization>
          </author>
          <date month='Feb' year='2008'/>
        </front>
      </reference>






    </references>

  </back>
  
</rfc>


