This is the schema file to be referenced in instance documents for validation. It is the top level schema file that integrates the schema components. This version integrates UBIF and SDD. To add further object types, this file has to be changed. Copyright © 2006 TDWG (Taxonomic Databases Working Group, www.tdwg.org). See the file SDD_(c).xsd for authorship and licensing information. Root element for files or data streams. Multiple Dataset objects are completely independent. Potential relationship may be detected by the consumer, but are not expressed in the UBIF format. The sequence of Dataset objects has no semantics and does not have to be preserved. The version of the UBIF standard used is defined in the namespace declaration and needs no separate data element. Metadata describing provider and process that created the current xml document. The provider is not necessarily the creator or owner of the dataset, compare Owner, IPR of dataset itself. Data from one or several object types, with common technical and content metadata. The attribute uri is optional in any first class object and denotes the preferred globally unique identifier. Not two objects may have the same uri. However, although strictly the constraint is global, to simplify dataset integration (where 2 datasets may, e.g., contain the same mediaobject), the uniqueness constraint is enforced only within a dataset, not within the set of datasets. ===== START of Global Uniqueness constraint. This is under discussion and will be settled before betas end. Applications should not rely on its presence or absence until it is settled. Removing it requires no changes in the schema, and existing instance documents will remain valid under such a new version. However, software consuming instance documents that relies on uniquenes may be in trouble if id values in fact are no longer unique within the document, but only within an object class (like specimen, character). - Note: A related constraint, AnyObjectURI, is defined on individual dataset objects. ===== START of TaxonName-related identity constraints ==== ===== START of TaxonHierarchy-related identity constraints ==== Currently we do not distinguish between inner and terminal nodes in this tree. ===== START of Specimen-related identity constraints ==== ===== START of DescriptiveConcept-related identity constraints ==== All concept state (also known as "global states) id values must be unique in an entire project (i. e. not only within a concept). ===== START of Modifier-related identity constraints ==== ===== START of Character-related identity constraints ==== Character-type-specific key, used in type-specific keyrefs. Character-type-specific key, used in type-specific keyrefs. Character-type-specific key, used in type-specific keyrefs. Character-type-specific key, used in type-specific keyrefs. Contains all locally defined character states (= StateDefinition) plus those concept states actually referenced in a character (StateReference; here the ref attribute is automatically also used as a new within-character id). Note that state ids are required to be unique across all characters, not only within each character. In descriptions this constraint is currently redundant, because states are referenced within character references. However, this constraint simplifies application design, and occasionally (e. g., in dependency definitions) states are referenced without a character reference. Note: the "*" at the end of the selector xpath refers to either StateReference (with @ref) or StateDefinition (with @id). This identifies an entire character tree (i. e. not the nodes in the tree) ===== START of CharacterTree-related identity constraints ==== Inner nodes ("Nodes/Node") have id-attribute and are referable, character leaves ("Nodes/Char") lack id-attribute and are not referable. ID-values must be unique across character trees (possibly in multiple datasets!). Inner nodes ("Nodes/Node") have id-attribute and are referable, character leaves ("Nodes/Char") lack id-attribute and are not referable. ===== START of Description-related identity constraints ==== A joint key for CodedDescription or NaturalLanguageDescription elements. References to the various derived character types require definitions of the same character type. This is achieved by corresponding key/keyref pairs for each character type. Element name "State" is currently restricted to this type, but may later be used elsewhere for different purposes, requiring a more specific xpath. Currently State occurs in summary data, sample data, mapping and dependency definitions (OnlyApplicableIf, InapplicableIf). ===== START of Sample-related identity constraints ==== A key to identify a specific sample. This is used when summary statistics are calculated from sample data that are also present in the dataset (compare the "BasedOnSample" element). This identifies an entire Identificationkey (i. e. not the nodes/steps in the key) ===== START of IdentificationKey-related identity constraints ==== Inner leads ("Leads/Lead") have id-attribute and are referable, result leads ("Leads/Result") may have id-attribute and are not referable. Results must be referable only in case of reticulation, but not in case of parent. ID-values must be unique across character trees (possibly in multiple datasets!). Compare IdentificationKeyLeadKey, this one includes the Result@id nodes and is referenced for reticulation elements. ===== START of Agent-related identity constraints ==== ===== START of MediaObject-related identity constraints ==== ===== START of Audience-related identity constraints ==== ===== START of GeographicArea-related identity constraints ==== ===== START of Publication-related identity constraints ==== Dataset itself is derived from Abstract object, as are the base types for which it contains collections. The first elements (Repres., Links, RevisionData, Owner, IPRStatem.) are content metadata for the entire data set / data collection. If, e.g. in response to queries, data subsets are extractet from a larger base data set, these metadata are not required to be specific to each document, but may may describe the base set itself. Link uri should then refer to the base set using rel=BasedUpon, and the dataset-uri may contain a query URI (or be omitted). If a history of the data collection (revised or expanded in various projects or at different institutions, aggregated from multiple sources) is available, this may be reflected in the IPR statements and possibly in the list of Owners. Note: if a data set is an extract from a larger base set, it is recommended to use 'Link rel = BasedOn' to point to the complete data set. Note that this may enable indexing robots to index the entire base data set. The scope of the dataset NOTE: This is the last metadata element of the dataset itself. The following elements contain object collections ('payload'): Taxon names used in the data set. Each object contains a name - either locally defined or representing an external resource defined in a linking mechanism and defines a local id attribute that may be referred to multiple times from within the data set. Optional hierarchy (= tree, biology: taxonomy) of taxon classes defined above. A hierarchy may be incomplete, i. e. some TaxonName object may not be in the hierarchy. TaxonHierarchies may be locally defined or represent an external source. Specimens (including ABCD.Units) are physical objects that have been collected or observed. (Extension point for future core ontology objects. This should include observations and organism interactions) (Extension point for future core ontology objects) (Extension point for future core ontology objects. An occurrence may have multiple geolocalisations, especially later coordinate or grid estimates, adjustments for geo. area name changes etc. This is analogous to taxon identifications. Documentation of persons/organizations involved in the authoring, compiling, editing, etc. of the data set. Publications used in the data set, defined through objects (= local or external link, see under Agents). Named geographical areas (continents, countries, cities, national parks, locally described places, etc.). An example of an external gazetteer referred to is the TDWG Geography standard. Resource definitions containing links like URLs or actually embedding the resource (e. g. encoded images). An Audience is an extension of language (including dialect), with expertise (pupil, beginner, expert) and other parameters (policy makers, farmers, etc.) The language on dataset is required; if no language is known this should be explicitly stated (lang="und"). This value is inherited in all places where an xml:lang attribute is specified in the UBIF schema (on Label, Detail, etc.) if no other value is given there (i.e. this value must be interpreted as a dataset-wide default).