Naming Conventions
From XBRLWiki
Revision as of 07:20, 18 October 2012 (edit) Hommes (Talk | contribs) (→Local name (@name)) ← Previous diff |
Current revision (11:51, 26 October 2012) (edit) Hommes (Talk | contribs) |
Current revision
Contents |
Folders
Folders are irrelevant to XML but important to control versioning of released files and ownership of these files.
- Taxonomy files MUST be released as children of a folder.
- The top level folder of any taxonomy MUST represent the owner of the taxonomy files.
- The first level of sub folders MUST represent the content of the taxonomy files:
- dict for metrics, dimensions, domains, members, families and perspectives;
- fws for frameworks, taxonomies, tables, modules and other concepts that constitute the reporting requirements;
- The first level of sub folders MAY represent the content of other taxonomy files:
- ext for models;
- func for functions for (formula) validations;
- The second level of sub folders for the fws folder MUST represent the reporting framework in which the taxonomy resides;
- The third level of sub folders for the reporting framework MUST represent the status of the files within;
- The fourth level of sub folders for the status of the reporting framework MUST represent the release date of the taxonomy files inside.
- The fifth level of sub folders for the release date of the taxonomy contains the folders representing the tables, modules and validations;
- If dates are used to name folders, its notation MUST be: CCYYMMDD (no dashes or other characters).
- Folder names MUST be in lower case.
- Folder names MUST NOT use spaces (if a seperator is needed, an underscore is advised)
Example:
- root
- dict
- fws
- finrep
- normative
- 20131201
- tab
- mod
- val
- 20131201
- normative
- finrep
- ext
- func
File names
File names are irrelevant to XML but the XBRL adoption of XPointer that addresses @id in named files makes it necessary to have rules on the file names:
Schema file names
- File names MUST be in lower case;
- File names MUST NOT be longer than 15 characters;
- File names MUST NOT use spaces (if a seperator is need an underscore is advised);
- File name extension '.xsd' MUST be used for schema files;
- Schema file names MUST represent their technical content according to the following table:
File name | Content | Note |
---|---|---|
dim.xsd | dimensions | Abstract concepts with @substitutionGroup='dimensionItem', related (2.1) label linkbases and the dimension-domain relationships in a definition linkbase |
exp.xsd | explicit domains | Abstract concepts with @type='model:explicitDomainType' and related (2.1) label linkbases |
fam.xsd | families | Abstract concepts with @type='model:familyType', related (2.1) label linkbases and parent-child relationships for family-dimension in a presentation linkbase |
fws.xsd | frameworks | Abstract concepts with @type='model:frameworkType' and related (2.1) label linkbases |
hier.xsd | member hierarchies | Extended linkroles for domain-member, parent-child and (custom) calculation relationships, related (generic) label linkbases and the related definition, presentation and calculation linkbases |
mem.xsd | explicit domain members | Abstract concepts with @type='nonnum:domainItemType' and related (2.1) label linkbases |
met.xsd | metrics | Non abstract concepts with @substitutionGroup='xbrli:item', @type content pointing to types defined in the 'xbrli:' namespace and its related (2.1) label linkbases |
?.xsd | modules | |
pers.xsd | perspectives | Extended linkroles for parent-child relationships between family and dimension, and related (generic) label linkbases |
tab.xsd | tables | Abstract concepts with @type='model:tableGroupType', related (2.1) label linkbases and related custom generic linkbase for tableGroup table relationships |
tax.xsd | taxonomies | Abstract concepts with @type='model:taxonomyType' and related (2.1) label linkbases |
typ.xsd | typed domains | XML Schema 'raw' elements with 'xs:' derived @type, NO @substitutionGroup content and related (generic) label linkbases |
Linkbase file names
- File names MUST be in lower case;
- File name extension '.xml' MUST be used for linkbase files;
- File names MUST NOT use spaces (if a seperator is needed an underscore is advised);
- Linkbase file names are created according to the following patterns:
File name | Linkbase type | Note |
---|---|---|
{main file}-lab-{lang}.xml | 2.1 and generic label | Label linkbases are broken up per language used |
{main-file}-lab-{lang}-{country}.xml | 2.1 and generic label | Label linkbases are broken up per language and detail used |
{main file}-ref.xml | 2.1 and generic reference | |
dim-def.xml | definition linkbase | Dimension-domain and dimension-defaultMember relationships |
{33}-def.xml | definition linkbase | Domain-member relationships |
?.xml | ? linkbase | Table-dimension relationships |
?.xml | ? linkbase | Metrics-table relationships through all arcroles only |
{33}-pre.xml | presentation linkbase | Parent-child relationships |
pers-pre.xml | presentation linkbase | Family-dimension relationships through parent-child arcs |
{33}-cal.xml | calculation linkbase | Domain-member relationships through custom arcs |
?.xml | formula linkbase | |
?-rend.xml | table linkbase | |
{main file}-lab-codes.xml | generic label linkbase | Some concepts of the taxonomy may refer to codes needed for the codes given to the columns and rows of tables (e.g. hier-def.xsd, hier-pre.xsd, hier-cal.xsd) |
tab-pre.xml | custom generic linkbase | tableGroup as parents MUST contain tables (which are a table:table resource) and MAY contain other tableGroups (which are abstract concepts in the model:tableGroupItem substitutionGroup) as their children. Cycles are not allowed. The custom arc is defined here |
Where:
{main file} = corresponds to the name of the schema or linkbase file where the concept or resource is defined without its extension;
{lang} = corresponds to the ISO 639-1 code of the language (in lower case).
{country} = corresponds to the ISO 639-2 code of the region or country (in lower case).
{33} = the name of schema file where the children of the hosted relationships are created.
Comment-08 Comment-09 Comment-10 Comment-13 Comment-19
Namespaces
Namespaces are the unique identifier of a schema file and part of the key on all the content that is created in that schema file. A namespace can be written as an URI or URN. With an URI there is an expectancy that it really identifies the schema. An URN is 'just' a name. A much used practice is to express URI's as URL's without the extension of the actual file it addresses. These URI's are being used as URL's to store the schema file on a server that can be called from software.
A (target)namespace in a schema is often abbreviated with a namespace prefix. This allows for shorthand qualified names to be used inside schema's. Not all XML software can handle schema's that have no namespace prefix assigned to them, and will generate a warning or error. As a consequence two strings are being created and have naming conventions assigned to them.
Target namespace
- Namespaces MUST be in lower case;
- Namespaces MUST reflect URI's;
- Namespaces MUST reflect the actual location (URL) that the schemas are accessable by software (no GUI);
DTS Authors are free to assign any webserver address, however since the URI represents the physical location, the names of folders will automatically form the 'end' of the URI assigned.
Example for the finrep metrics schema at EBA:
xbrl.eba.europa.eu/.../dict/met.xsd
Where the dots represent any folder structure EBA finds appropriate.
Different concepts are stored in different schemas, each carrying their own namespace, location and prefix. The table below shows how their names are constructed.
Namespace prefix
- Namespace prefix' MUST be in lower case;
- Namespace prefix' MUST only use characters a-z0-9, -, _;
- Namespace prefix' MUST be a concatenation of the owner of the schema file, underscore, name of the schema file without the extension which MAY be followed by another underscore and a two letter subdivision;
Examples:
eba_met
eba_mod_xx
See also: Reserved namespaces
Type of concept | Location | Namespace | Prefix |
---|---|---|---|
Dimensions | {oloc}/dict/dim/dim.xsd | {ons}/dict/dim | {opre}_dim |
Explicit domains | {oloc}/dict/dom/exp.xst | {ons}/dict/exp | {opre}_exp |
Families | {oloc}/dict/dim/fam.xsd | {ons}/dict/fam | {opre}_fam |
Frameworks | {oloc}/fws/fws.xsd | {ons}/fws | {opre}_fws |
Hierarchies | {oloc}/dict/dom/{dc}/hier.xsd | {ons}/dict/dom/{dc} | {opre}_?? |
Members | {oloc}/dict/dom/{dc}/mem.xsd | {ons}/dict/dom/{dc} | {opre}_{dc} |
Metrics | {oloc}/dict/met/met.xsd | {ons}/dict/met | {opre}_met |
Perspectives | {oloc}/dict/dim/pers.xsd | {ons}/dict/pers | {opre}_pers |
TableGroups | {oloc}/fws/{framework}/{normative}/{pub-date}/tab/tab.xsd | {ons}/fws/{framework}/{normative}/{pub-date}/tab | {opre}_tab |
Taxonomies | {oloc}/fws/{framework}/{normative}/{pub-date}/tax.xsd | {ons}/fws/{framework}/{normative}/{pub-date} | {opre}_tax |
Typed domains | {oloc}/dict/dom/typ.xsd | {ons}/dict/typ | {opre}_typ |
Where:
{oloc} = official location of taxonomy files of the owner of the concepts
{ons} = its base namespace
{opre} = the prefix of its base namespace
{dc} = the code of a domain as used in the DPM
{framework} = the code of the framework in the DPM
{normative} = the term 'optional' or 'normative'
{pub-date} = the date of publication of the files in the subfolders
Example from EBA (metrics):
Location: http://www.eba.europa.eu/eu/fr/xbrl/dict/met/met.xsd
Namespace: http://www.eba.europa.eu/xbrl/dict/met
Prefix: eba_met
Concepts
Local name (@name)
- The local name of a concept that is a Metric is formed by multiple parts: {Assigned letter}{Periodtype letter}
- The local name of a concept that is a Dimension is formed by the code used in the data point model; a short sequence of capital case letters (usually two, but it is not limited to two letters).
- The local name of a concept that is a Family is formed by the numeric code in the data point model, preceded by a lower case 'f'.
- The local name of a concept that is a Domain is formed by the code used in the data point model; a short sequence of capital case letters (usually two, but it is not limited to two letters).
- The local name of a concept that is a Explicit domain member is formed by the numeric code in the data point model, preceded by a lower case 'x' EXCEPT when the members are part of a widely accepted standard (like ISO), in this case the @name will represent the international accepted code.
- The local name of a concept that is TableGroup is formed by the code in the data point model, preceded by a lower case 'tg'.
Concept type | Model data type | XBRL data type | Assigned letter | Measure |
---|---|---|---|---|
Metrics | Boolean | xbrli:booleanItemType | b | no unit |
Metrics | Date | xbrli:dateItemType | d | no unit |
Metrics | Decimal | xbrli:decimalItemType | p | xbrli:pure |
Metrics | Integer | xbrli:integerItemType | i | xbrli:pure |
Metrics | Monetary | xbrli:monetaryItemType | m | iso4217: |
Metrics | Percent | num:percentItemType | p | xbrli:pure |
Metrics | Text | xbrli:stringItemType | s | no unit |
Explicit domain | xbrli:qnameItemType | e | no unit | |
Typed domain | Domain corresponding data type, codification letter and measure. | |||
Dimension | xbrli:stringItemType | |||
Families | model:familyType | |||
Frameworks | model:frameworkType | |||
TableGroups | model:tableGroupType | |||
Taxonomies | model:taxonomyType | |||
Explicit domain member | nonnum:domainItemType |
Periodtype letters are:
i = instant
d = duration
Examples:
Monetary instant concept: @name="mi" @id="eba_mi"
Dimension concept @name="CR" @id="eba_CR"
Family concept @name="f3" @id="eba_f3"
Explicit member @name="x1" @id="eba_x1"
Identifier (@id)
The identifier of a concept is constructed as: {opre}_{name}
Where:
{opre} = the prefix of the base namespace of the owner of the concept
{name} = as put in the @name
Types
Codes and enumeration values
Linkroles
URI
- The URI of perspective linkroles is constructed as: {ons}/role/dict/pers/{code}
- The URI of hierarchy linkroles for explicit domain members is constructed as: {ons}/role/dict/dom/{dom-code}/{hierarchy-code}
- The URI of table linkroles is constructed as: {ons}/role/fws/{framework}/{normative}/{pub-date}/tab/{table}
Where:
{ons} = the base namespace of the owner
{code} = the numeric code used in the data point model
{dom-code} = the sequence of capitals representing the domain used in the data point model
{hierarchy-code} = the numeric hierarchy code
{framework} = the code of the framework in the DPM
{normative} = the term 'optional' or 'normative'
{pub-date} = the date of publication of the files in the subfolders
{table} = the code of the table in the DPM
Identifier (@id)
- The @id on perspective linkroles is constructed as: {opre}_p{code}
- The @id on hierarchy linkroles for explicit domain member is constructed as: {opre}_r{code}
- The @id on table linkroles is constructed as: {opre}_t{code}
Where:
{opre} = the prefix if the base namespace of the owner
{code} = the numeric code used in the data point model
Resources
- The @id on predefined axes resources is constructed as: {opre}_a{code}
- The @id on variable axes resources is constructed as: {opre}_a{code}
- The @id on coordinate resources is constructed as: {opre}_c{code}
- The @id on base item hierarchy reference resources is constructed as: {opre}_h{code}
- The @id on dimension hierarchy reference resources is constructed as: {opre}_h{code}
- The @id on 2.1 label resources is constructed as: ??
- The @id on generic label resources is constructed as: ??
- The @id on 2.1 reference resources is constructed as: ??
- The @id on generic reference resources is constructed as: ??
- The @id on parameter resources is constructed as: ??
- The @id on fact variable resources is constructed as: ??
- The @id on general variable resources is constructed as: ??
- The @id on assertion resources is constructed as: ??