talia_import.xsd

<?xml version="1.0" encoding="UTF-8"?>
<!-- 
    Schema file for the Hyper Import Format. This format can be used to check the Hyper XML
    that will be imported in Talia. The format is quite particular for the Discovery
    project; it is based on the original data format of the Hyper software.
-->
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">
    
    <!-- 
        The 'relations' element can be used in almost all sources. Each relation
        needs a predicate and an object. The predicate should refer to a predicate
        defined in the ontology, and the object should refer to an actual Source
    -->
    <xs:complexType name="relationsType">
        <xs:sequence>
            <xs:element name="relation" maxOccurs="unbounded">
                <xs:complexType>
                    <xs:sequence>
                        <xs:element name="predicate" type="xs:string"/>
                        <xs:element name="object" type="xs:string"/>
                    </xs:sequence>
                </xs:complexType>
                <xs:unique name="predicateAndObject">
                    <xs:selector xpath="."/>
                    <xs:field xpath="object"/>
                    <xs:field xpath="predicate"/>
                </xs:unique>
            </xs:element>
        </xs:sequence>
    </xs:complexType>

    <!-- Special date type, accepts a subset of ISO 8601 without much checking -->
    <xs:simpleType name="taliaDate">
        <xs:restriction base="xs:string">
            <xs:pattern value="\d{4}(-\d{2})?(-\d{2})?(.\d{2}:\d{2}.?)?"/>
        </xs:restriction>
    </xs:simpleType>
    
    <!--
        A 'catalog' element descries a catalog of elements. At the momet it
        only accepts one parameter (apart from the siglm:
        * position - Gives a position for ordering the catalogs on the site
    -->
    <xs:complexType name="catalogType">
        <xs:sequence>
            <xs:element name="siglum" type="xs:string"></xs:element>
            <xs:element name="position" type="xs:integer"></xs:element>
        </xs:sequence>
    </xs:complexType>
    
    <!--
        These are the base elements that are common to all sources:
        * catalog - The catalog of which this source is part
        * siglum - The unique (within this installation) siglum of the source
        * title - The title of the surce
        * type - Describes the internal type of the source (if any)
        * subtype - Describes the internal subtype of the source's type (if any)
    -->
    <xs:complexType name="sourceType">
        <xs:sequence>
            <xs:element name="catalog" type="xs:string" minOccurs="0"/>
            <xs:element name="siglum" type="xs:string"/>
            <xs:element name="title" type="xs:string"/>
            <xs:element name="type" type="xs:string" minOccurs="0"/>
            <xs:element name="subtype" type="xs:string" minOccurs="0"/>
        </xs:sequence>
    </xs:complexType>

    <!--
        A book element usually represents a (physical) book and consist of 
        sub-elements like chapters and pages.
        * coprightNote - String with a text describing the conditions for using this book
        * description - A description of the book
        * date - Date when the original book was published
        * publisher - The person or organisation who published this book
        * publishingPlace - The place where this book was published
        * ordering - Position of the book within the current collection
        * relations - Semantic relations for this Source
    -->
    <xs:complexType name="bookType">
        <xs:complexContent>
            <xs:extension base="sourceType">
                <xs:sequence>
                    <xs:element name="copyrightNote" type="xs:string"/>
                    <xs:element name="description" type="xs:string"/>
                    <xs:element name="date" type="taliaDate" default="0001-01-01"/>
                    <xs:element name="collocation" type="xs:string"/>
                    <xs:element name="publisher" type="xs:string"/>
                    <xs:element name="publishingPlace" type="xs:string"/>
                    <xs:element name="ordering" type="xs:int"/>
                    <xs:element name="relations" type="relationsType" minOccurs="0"/>
                </xs:sequence>
            </xs:extension>
        </xs:complexContent>
    </xs:complexType>
    
    <!-- 
        The author of an element in the database. The author will have a siglum like all
        sources, and in addition some business card data. The 'from_date' and 'to_date'
        contain the dates between which the author was a member of the current site
        (or institution)
    -->
    <xs:complexType name="authorType">
        <xs:sequence>
            <xs:element name="siglum" type="xs:string"></xs:element>
            <xs:element name="name" type="xs:string"></xs:element>
            <xs:element name="surname" type="xs:string"></xs:element>
            <xs:element name="status" type="xs:string"></xs:element>
            <xs:element name="institution" type="xs:string"></xs:element>
            <xs:element name="position" type="xs:string"></xs:element>
            <xs:element name="street" type="xs:string"></xs:element>
            <xs:element name="zip" type="xs:string"></xs:element>
            <xs:element name="city" type="xs:string"></xs:element>
            <xs:element name="country" type="xs:string"></xs:element>
            <xs:element name="telephone"></xs:element>
            <xs:element name="fax" type="xs:string"></xs:element>
            <xs:element name="email" type="xs:string"></xs:element>
            <xs:element name="webpage" type="xs:string"></xs:element>
            <xs:element name="from_date" type="taliaDate" default="0001-01-01"></xs:element>
            <xs:element name="to_date" type="taliaDate" default="0001-01-01"></xs:element>
        </xs:sequence>
    </xs:complexType>
    
    <!--
        Type for assigning an author
    -->
    <xs:complexType name="authorsType">
        <xs:sequence>
            <xs:element name="author" type="xs:string" maxOccurs="unbounded"/>
        </xs:sequence>
    </xs:complexType>

    <!--
        A secondary source that was contributed to the current site.
        'publishing_date' - The date on which the resource was published online
        'publisher' - The person or institution responsible for the publication
        'language' - language code for the language of this publication
        'alreadyPublished' - Indicates if this element has already been published somewhere eles
        'copyrightNote' - The conditions for using this Source
        'dimensionsX, dimensionY' - Physical dimensions of the contribution (if applicabele)
        'file_size, file_name, ...' - Points to the file contaiining the data
        'authors' - List of authors of this resource
    -->
    <xs:complexType name="contributionType">
        <xs:complexContent>
            <xs:extension base="sourceType">
                <xs:sequence>
                    <xs:element name="publishing_date" type="taliaDate" default="0001"/>
                    <xs:element name="publisher" type="xs:string"/>
                    <xs:element name="language" type="xs:language" />
                    <xs:element name="alreadyPublished" type="xs:string"/>
                    <xs:element name="relations" type="relationsType"/>
                    <xs:element name="copyrightNote" type="xs:string"/>
                    <xs:element name="subtype" type="xs:string" minOccurs="0"></xs:element>
                    <xs:element name="dimensionX" type="xs:int" default="0" minOccurs="0"/>
                    <xs:element name="dimensionY" type="xs:int" default="0" minOccurs="0"/>
                    <xs:element name="file_size" type="xs:int" default="0"/>
                    <xs:element name="file_name" type="xs:string"/>
                    <xs:element name="file_url" type="xs:string"/>
                    <xs:element name="file_content_type" type="xs:string"/>
                    <xs:element name="authors" type="authorsType"/>
                </xs:sequence>
            </xs:extension>
        </xs:complexContent>
    </xs:complexType>

    <!--
        A note is a single, coherent, piece of text that appears on a page.
        - 'page' - Page on which the note appears
        - 'position' - Gives the order in which the positions are sorted
        - 'coordinates' - A list of coordinates that describe a polygon on the
                          site that is the outline of the current not
    -->
    <xs:complexType name="notesType">
        <xs:sequence>
            <xs:element name="note" maxOccurs="unbounded">
                <xs:complexType>
                    <xs:sequence>
                        <xs:element name="page" type="xs:string"/>
                        <xs:element name="position" type="xs:int"/>
                        <xs:element name="coordinates" type="xs:string"/>
                    </xs:sequence>
                </xs:complexType>
                <xs:unique name="notePosition">
                    <xs:selector xpath="note/position"/>
                    <xs:field xpath="."/>
                </xs:unique>

            </xs:element>
        </xs:sequence>
    </xs:complexType>

    <!--
        A paragraph is a paragraph of text, that may stretch over several notes
        (especially if the paragraph strethes over more than one page. The 
        pargraph referes all the notes that are part of it.
    -->
    <xs:complexType name="paragraphType">
        <xs:complexContent>
            <xs:extension base="sourceType">
                <xs:sequence>
                    <xs:element name="notes" type="notesType"/>
                </xs:sequence>
            </xs:extension>
        </xs:complexContent>
    </xs:complexType>

    <!--
        A page is a direct sub-part of a book. The page must declare the
        'isSubPartOf' relation to associate it to a book.
        
        'width, height' - physical dimensions of the page, if aplicable
        'position' - Number that gives the page's number in the current book
        'position_name' - String that describes the position of  the page in the
                          current book/selection (used for presentation)
    -->
    <xs:complexType name="pageType">
        <xs:complexContent>
            <xs:extension base="sourceType">
                <xs:sequence>
                    <xs:element name="width" type="xs:int"/>
                    <xs:element name="height" type="xs:int"/>
                    <xs:element name="position" type="xs:int"/>
                    <xs:element name="position_name" type="xs:string"/>
                    <xs:element name="relations" type="relationsType"/>
                </xs:sequence>
            </xs:extension>
        </xs:complexContent>
    </xs:complexType>
    
    <!--
        A chapter is a collection of pages in the same book (an auxialliare construct)
        'first_page' - The 'position' of the first page in the book.
        'position' - Number to order the existing chapters within the book
        'name' - Name of the currrent chapter
        'book' - Refers to the siglum of the book which is refereced
    -->
    <xs:complexType name="chapterType">
        <xs:complexContent>
            <xs:extension base="sourceType">
                <xs:sequence>
                    <xs:element name="first_page" type="xs:string"/>
                    <xs:element name="position" type="xs:int"/>
                    <xs:element name="name" type="xs:string"/>
                    <xs:element name="book" type="xs:string"/>
                </xs:sequence>
            </xs:extension>
        </xs:complexContent>
    </xs:complexType>

    <!--
        This defines the main list of sources, with each entry 
        refering to one Source that is imported.
    -->
    <xs:element name="sources">
        <xs:complexType>
            <xs:sequence>
                <xs:element name="source" maxOccurs="unbounded">
                    <xs:complexType>
                        <xs:choice>
                            <xs:element name="book" type="bookType"/>
                            <xs:element name="paragraph" type="paragraphType"/>
                            <xs:element name="page" type="pageType"/>
                            <xs:element name="facsimile" type="contributionType"/>
                            <xs:element name="transcription" type="contributionType"/>
                            <xs:element name="text_reconstruction" type="contributionType"/>
                            <xs:element name="chapter" type="chapterType"/>
                            <xs:element name="author" type="authorType"></xs:element>
                            <xs:element name="catalog" type="catalogType"></xs:element>
                        </xs:choice>
                    </xs:complexType>
                </xs:element>
            </xs:sequence>
        </xs:complexType>
        
        <!-- Guruantees that each source has an unique siglum -->
        <xs:key name="sourceKey">
            <xs:selector xpath="source/*/siglum"/>
            <xs:field xpath="."/>
        </xs:key>
        
        <!--
            WARNING: The keys and constraints defined here will not work in all
            cases. This is due to the structure of the format, and how the
            XSD Schema can be applied.
            
            If you run into problems, you may want to ignore these constraints.
        -->
        
        <!-- 
            Keyref to source siglum. 
            This needs to be disabled if the data is in multiple files,
            since then the references cannot be checked.
            
            (This has been disabled since it doesn't work in all cases, due
            to limitations of XSD on the current layout of the schema)
        -->
        <xs:keyref refer="sourceKey" name="relationTarget">
            <xs:selector xpath="source/*/relations/relation/object"></xs:selector>
            <xs:field xpath="."></xs:field>
        </xs:keyref>
        
        <!-- Ordering of the books -->
        <xs:unique name="bookOrder">
            <xs:selector xpath="source/book/ordering"/>
            <xs:field xpath="."/>
        </xs:unique>
        
        <!-- Ordering of the pages -->
        <xs:unique name="pageOrder">
            <xs:selector xpath="source/page"/>
            <xs:field xpath="position"/>
            <xs:field xpath="relations/relation/object"></xs:field>
        </xs:unique>
        
        <!-- Ordering of the chapters -->
        <xs:unique name="chapterOrder">
            <xs:selector xpath="source/chapter"/>
            <xs:field xpath="position"/>
            <xs:field xpath="book"/>
        </xs:unique>
    </xs:element>
</xs:schema>