program/everything/xml/scheme.xsd

   1 <?xml version="1.0" encoding="UTF-8"?>
   2 <?xml-stylesheet type="text/xsl" href="xs3p.xsl"?>
   3 <xs:schema attributeFormDefault="unqualified" elementFormDefault="qualified" xmlns:xs="http://www.w3.org/2001/XMLSchema">
   4 <!--This is the main element, required. It contains crawler and/data entries-->
   5   <xs:element name="crawler_output">
   6     <xs:complexType>
   7       <xs:sequence>
   8 <!--Crawler entries contain the information of the crawler, there can be multiple-->
   9         <xs:element name="crawler" maxOccurs="unbounded" minOccurs="0">
  10           <xs:complexType>
  11             <xs:simpleContent>
  12               <xs:extension base="xs:string">
  13                 <xs:attribute type="xs:string" name="name" use="optional"/>
  14                 <xs:attribute type="xs:string" name="venue" use="optional"/>
  15                 <xs:attribute type="xs:string" name="freq" use="optional"/>
  16                 <xs:attribute type="xs:string" name="def_loc" use="optional"/>
  17                 <xs:attribute type="xs:string" name="adress" use="optional"/>
  18                 <xs:attribute type="xs:anyURI" name="website" use="optional"/>
  19                 <xs:attribute type="xs:anyURI" name="url" use="optional"/>
  20               </xs:extension>
  21             </xs:simpleContent>
  22           </xs:complexType>
  23         </xs:element>
  24 <!--Data entries contain the information of a single crawled entry, there can be multiple-->
  25         <xs:element name="data" maxOccurs="unbounded" minOccurs="0">
  26           <xs:complexType>
  27             <xs:sequence>
  28               <xs:element name="entry">
  29                 <xs:complexType>
  30                   <xs:sequence>
  31 <!--These four fields contain the user data-->
  32                     <xs:element type="xs:string" name="where"/>
  33                     <xs:element type="xs:string" name="what"/>
  34                     <xs:element type="xs:string" name="date"/>
  35                     <xs:element type="xs:string" name="time"/>
  36 <!--These fields contain the raw original title and summary-->
  37                     <xs:element type="xs:string" name="full_title"/>
  38                     <xs:element type="xs:string" name="full_summary"/>
  39 <!--These fields contain some other information from the rss-->
  40                     <xs:element type="xs:anyURI" name="link" maxOccurs="unbounded" minOccurs="0"/>
  41                     <xs:element type="xs:string" name="pub_date" maxOccurs="unbounded" minOccurs="0"/>
  42 <!--Extracted URIs is a list of urls, this can be empty-->
  43                     <xs:element name="extracted_uris" maxOccurs="unbounded" minOccurs="0">
  44                       <xs:complexType>
  45                         <xs:sequence>
  46                           <xs:element type="xs:anyURI" name="url" maxOccurs="unbounded" minOccurs="0"/>
  47                         </xs:sequence>
  48                       </xs:complexType>
  49                     </xs:element>
  50                   </xs:sequence>
  51                 </xs:complexType>
  52               </xs:element>
  53             </xs:sequence>
  54 <!--These fields specify the crawler name and the date crawled-->
  55             <xs:attribute type="xs:string" name="from"/>
  56             <xs:attribute type="xs:dateTime" name="date"/>
  57           </xs:complexType>
  58         </xs:element>
  59       </xs:sequence>
  60     </xs:complexType>
  61   </xs:element>
  62 </xs:schema>