From dab89fb9ef39d76d31fe549bf18297fbe878436f Mon Sep 17 00:00:00 2001
From: Wendell Piez <wendell.piez@nist.gov>
Date: Mon, 8 Jul 2019 14:06:47 -0400
Subject: [PATCH] More work on datatyping #436 including now in JSON Schema

---
 build/metaschema/json/json-schema-metamap.xsl | 85 +++++++++++++++++--
 build/metaschema/lib/metaschema.xsd           | 11 ++-
 build/metaschema/xml/oscal-datatypes.xsd      | 15 ++--
 build/metaschema/xml/produce-xsd.xsl          | 71 +++++++++++-----
 4 files changed, 143 insertions(+), 39 deletions(-)

diff --git a/build/metaschema/json/json-schema-metamap.xsl b/build/metaschema/json/json-schema-metamap.xsl
index 12056b55ff..a42988bdc9 100644
--- a/build/metaschema/json/json-schema-metamap.xsl
+++ b/build/metaschema/json/json-schema-metamap.xsl
@@ -329,7 +329,7 @@
         <string key="type">object</string>
     </xsl:template>
 
-    <xsl:template match="define-field | define-flag" mode="object-type">
+    <xsl:template match="define-field[empty(flag)] | define-flag" mode="object-type">
         <string key="type">string</string>
     </xsl:template>
     
@@ -341,7 +341,9 @@
             <xsl:when test="exists(key('definition-by-name',@ref)/@as-type)">
                 <xsl:apply-templates mode="#current" select="key('definition-by-name',@ref)"/>
             </xsl:when>
-            <xsl:otherwise>string</xsl:otherwise>
+            <xsl:otherwise>
+                <string key="type">string</string> 
+            </xsl:otherwise>
         </xsl:choose>
     </xsl:template>
 
@@ -351,20 +353,89 @@
     
     <xsl:template priority="2" match="*[@as-type='integer']" mode="object-type">
         <string key="type">integer</string>
+        <number key="multipleOf">1.0</number>
+    </xsl:template>
+
+    <xsl:template priority="2" match="*[@as-type='positiveInteger']" mode="object-type">
+        <string key="type">integer</string>
+        <number key="multipleOf">1.0</number>
+        <number key="minimum">1</number>
+    </xsl:template>    
+    
+    <xsl:template priority="2" match="*[@as-type='nonNegativeInteger']" mode="object-type">
+        <string key="type">integer</string>
+        <number key="multipleOf">1.0</number>
+        <number key="minimum">0</number>
     </xsl:template>
     
-    <!-- Types are listed in ../xml/produce-xsl.xsl and ../xml/oscal-datatypes -->
+    <!-- Types are listed in ../xml/produce-xsd.xsl and ../xml/oscal-datatypes.xsd -->
+    
     <xsl:variable name="numeric-types" as="element()*">
+        <type>decimal</type>
+        <type>float</type>
         <type>double</type>
-        <type>percent</type>
-    </xsl:variable>        
-        
+    </xsl:variable>
+    
     <xsl:template priority="2" match="*[@as-type=$numeric-types]" mode="object-type">
         <string key="type">number</string>
     </xsl:template>
     
+    <xsl:template priority="3" match="*[@as-type = $datatypes/*/@key]" mode="object-type">
+        <xsl:copy-of select="key('datatypes-by-name',@as-type,$datatypes)/*"/>
+    </xsl:template>
     
     <xsl:template match="prose" name="prose"/>
-    
+
+    <xsl:key name="datatypes-by-name" xpath-default-namespace="http://www.w3.org/2005/xpath-functions"
+        match="map" use="@key"/>
+    
+    <xsl:variable name="datatypes" expand-text="false">
+        <map key="date-with-timezone">
+            <string key="type">string</string>
+            <string key="format">date</string>
+            <!--The xs:date with a required timezone.-->
+            <string key="pattern">.+[:Z].*</string>
+        </map>
+        <map key="dateTime-with-timezone">
+            <string key="type">string</string>
+            <string key="format">date-time</string>
+            <!--The xs:dateTime with a required timezone.-->
+            <string key="pattern">.+T.+(Z|[+-].+)</string>
+        </map>
+        <map key="email">
+            <string key="type">string</string>
+            <string key="format">email</string>
+            <!---->
+            <string key="pattern">.+@.+</string>
+        </map>
+        <map key="ip-v4-address">
+            <string key="type">string</string>
+            <string key="format">ipv4</string>
+            <!--The ip-v4-address type specifies an IPv4 address in dot decimal notation.-->
+            <string key="pattern">((25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9][0-9]|[0-9]).){3}(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9][0-9]|[0-9])</string>
+        </map>
+        <map key="ip-v6-address">
+            <string key="type">string</string>
+            <string key="format">ipv6</string>
+            <!--The ip-v6-address type specifies an IPv6 address represented in 8 hextets separated by colons.This is based on the pattern provided here: https://stackoverflow.com/questions/53497/regular-expression-that-matches-valid-ipv6-addresses with some customizations.-->
+            <string key="pattern">(([0-9a-fA-F]{1,4}:){7,7}[0-9a-fA-F]{1,4}|([0-9a-fA-F]{1,4}:){1,7}:|([0-9a-fA-F]{1,4}:){1,6}:[0-9a-fA-F]{1,4}|([0-9a-fA-F]{1,4}:){1,5}(:[0-9a-fA-F]{1,4}){1,2}|([0-9a-fA-F]{1,4}:){1,4}(:[0-9a-fA-F]{1,4}){1,3}|([0-9a-fA-F]{1,4}:){1,3}(:[0-9a-fA-F]{1,4}){1,4}|([0-9a-fA-F]{1,4}:){1,2}(:[0-9a-fA-F]{1,4}){1,5}|[0-9a-fA-F]{1,4}:((:[0-9a-fA-F]{1,4}){1,6})|:((:[0-9a-fA-F]{1,4}){1,7}|:)|[fF][eE]80:(:[0-9a-fA-F]{0,4}){0,4}%[0-9a-zA-Z]{1,}|::([fF]{4}(:0{1,4}){0,1}:){0,1}((25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9][0-9]|[0-9]).){3,3}(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9][0-9]|[0-9])|([0-9a-fA-F]{1,4}:){1,4}:((25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9][0-9]|[0-9]).){3,3}(25[0-5]|2[0-4][0-9]|1[0-9][0-9]|[1-9][0-9]|[0-9]))</string>
+        </map>
+        <map key="hostname">
+            <string key="type">string</string>
+            <string key="format">idn-hostname</string>
+            <!---->
+            <string key="pattern">.+</string>
+        </map>
+        <map key="uri">
+            <string key="type">string</string>
+            <string key="format">uri</string>
+            <!---->
+        </map>
+        <map key="uri-reference">
+            <string key="type">string</string>
+            <string key="format">uri-reference</string>
+            <!---->
+        </map>
+    </xsl:variable>
 
 </xsl:stylesheet>
diff --git a/build/metaschema/lib/metaschema.xsd b/build/metaschema/lib/metaschema.xsd
index b0dea23435..ad8d8fe082 100644
--- a/build/metaschema/lib/metaschema.xsd
+++ b/build/metaschema/lib/metaschema.xsd
@@ -398,20 +398,24 @@
       <xs:enumeration value="boolean" />
       <xs:enumeration value="string" />
       <xs:enumeration value="NCName" />
+      <xs:enumeration value="decimal" />
+      <xs:enumeration value="float" />
       <xs:enumeration value="double" />
       <xs:enumeration value="integer" />
-      <xs:enumeration value="IDREF"/>
+      <xs:enumeration value="nonNegativeInteger"/>
+      <xs:enumeration value="positiveInteger"/>
       
       <!-- specialized datatypes -->
-      <xs:enumeration value="ip-v4-address"/>
-      <xs:enumeration value="ip-v6-address"/>
       <xs:enumeration value="dateTime-with-timezone"/>
       <xs:enumeration value="date-with-timezone"/>
       <!--<xs:enumeration value="percent"/>-->
       <xs:enumeration value="email"/>
       <xs:enumeration value="hostname"/>
+      <xs:enumeration value="ip-v4-address"/>
+      <xs:enumeration value="ip-v6-address"/>
       <xs:enumeration value="uri"/>
       <xs:enumeration value="uri-reference"/>
+      <xs:enumeration value="base64"/>
     </xs:restriction>   
   </xs:simpleType>
   
@@ -420,6 +424,7 @@
       <xs:simpleType>
         <xs:restriction base="xs:string">
           <xs:enumeration value="ID" />
+          <xs:enumeration value="IDREF"/>
         </xs:restriction>
       </xs:simpleType>
     </xs:union>
diff --git a/build/metaschema/xml/oscal-datatypes.xsd b/build/metaschema/xml/oscal-datatypes.xsd
index cac3615420..40f70f7f16 100644
--- a/build/metaschema/xml/oscal-datatypes.xsd
+++ b/build/metaschema/xml/oscal-datatypes.xsd
@@ -45,16 +45,6 @@
     </xs:restriction>
   </xs:simpleType>
   
-  <xs:simpleType name="percent">
-    <xs:annotation>
-      <xs:documentation>The percent type specifies a percentage from 0 to 100.</xs:documentation>
-    </xs:annotation>
-    <xs:restriction base="xs:decimal">
-      <xs:minInclusive value="0"/>
-      <xs:maxInclusive value="100"/>
-    </xs:restriction>
-  </xs:simpleType>
-
   <xs:simpleType name="email">
     <xs:annotation>
       <xs:documentation></xs:documentation>
@@ -89,6 +79,11 @@
     </xs:annotation>
     <xs:restriction base="xs:anyURI">
       <xs:whiteSpace value="collapse"/>
+      <xs:pattern value="\p{L}[\p{L}\d+\-\.]*:.+">
+        <xs:annotation>
+          <xs:documentation>Requires a scheme with colon per RFC 3986</xs:documentation>
+        </xs:annotation>
+      </xs:pattern>
     </xs:restriction>
   </xs:simpleType>
   
diff --git a/build/metaschema/xml/produce-xsd.xsl b/build/metaschema/xml/produce-xsd.xsl
index 0be763eff5..a4fa497cc4 100644
--- a/build/metaschema/xml/produce-xsd.xsl
+++ b/build/metaschema/xml/produce-xsd.xsl
@@ -100,9 +100,14 @@
         <xs:simpleType name="boolean"/>
         <xs:simpleType name="string"/>
         <xs:simpleType name="NCName"/>
+        <xs:simpleType name="decimal"/>
+        <xs:simpleType name="float"/>
         <xs:simpleType name="double"/>
         <xs:simpleType name="integer"/>
+        <xs:simpleType name="nonNegativeInteger"/>
+        <xs:simpleType name="positiveInteger"/>
         <xs:simpleType name="IDREF"/>
+        <xs:simpleType name="ID"/>
     </xsl:variable>
     
     <xsl:template match="namespace"/>
@@ -122,26 +127,48 @@
     <!--<xsl:template match="define-field[@as-type='markup-multiline'][false() (: xml wrap is not on so we get no element :)]"/>-->
     
     <xsl:template match="define-field">
+        <xs:element name="{@name }">
+            <xsl:apply-templates select="." mode="annotated"/>
+            <xs:complexType mixed="true">
+                <xsl:apply-templates select="key | flag"/>
+            </xs:complexType>
+        </xs:element>
+    </xsl:template>
+    
+    <xsl:template priority="5" match="define-field[@as-type='empty']">
         <xs:element name="{@name }">
             <xsl:apply-templates select="." mode="annotated"/>
             <xs:complexType>
-                <xsl:if test="not(@as-type='empty')">
-                  <xsl:attribute name="mixed">true</xsl:attribute>
-                </xsl:if>
-                <xsl:if test="@as-type='markup-line'">
-                    <xs:group ref="{$declaration-prefix}:everything-inline"/>
-                </xsl:if>
-                <xsl:if test="@as-type='markup-multiline'">
-                    <xs:group ref="{$declaration-prefix}:prose"/>
-                </xsl:if>
                 <xsl:apply-templates select="key | flag"/>
             </xs:complexType>
         </xs:element>
     </xsl:template>
     
+    <xsl:template priority="5" match="define-field[@as-type='markup-line']">
+        <xs:element name="{@name }">
+            <xsl:apply-templates select="." mode="annotated"/>
+            <xs:complexType mixed="true">
+                <xs:group ref="{$declaration-prefix}:everything-inline"/>
+                <xsl:apply-templates select="key | flag"/>
+            </xs:complexType>
+        </xs:element>
+    </xsl:template>
+    
+    <xsl:template priority="5" match="define-field[@as-type='markup-multiline']">
+        <xs:element name="{@name }">
+            <xsl:apply-templates select="." mode="annotated"/>
+            <xs:complexType mixed="true">
+                <xs:group ref="{$declaration-prefix}:prose"/>
+                <xsl:apply-templates select="key | flag"/>
+            </xs:complexType>
+        </xs:element>
+    </xsl:template>
     
-    <xsl:template match="define-field[@as-type='boolean']">
-        <xs:element name="{@name }" type="xs:boolean">
+    <xsl:template priority="3" match="define-field[exists(@as-type)]">
+        <xs:element name="{@name }">
+            <xsl:call-template name="assign-type">
+                <xsl:with-param name="datatype" select="@as-type"/>
+            </xsl:call-template>
             <xsl:apply-templates select="." mode="annotated"/>
         </xs:element>
     </xsl:template>
@@ -210,8 +237,6 @@
         <xs:group ref="{$declaration-prefix}:prose" maxOccurs="unbounded" minOccurs="0"/>
     </xsl:template>
     
-    
-    
     <xsl:template match="flag | key">
         <xsl:variable name="datatype" select="(@as-type,key('definition-by-name',@ref)/@as-type)[1]"/>
         <xsl:variable name="value-list" select="(valid-values,key('definition-by-name',@ref)/valid-values)[1]"/>
@@ -223,12 +248,9 @@
             <!-- annotate as datatype or string unless an exclusive value-list is given -->
             <xsl:if test="empty($value-list)">
                 <!-- overriding string datatype on attribute -->
-                <xsl:for-each select="$datatype[.=$available-custom-types/@name]">
-                    <xsl:attribute name="type" expand-text="true" select="concat($declaration-prefix,':',.)"/>
-                </xsl:for-each>
-                <xsl:for-each select="$datatype[.=$built-in-types/@name]">
-                    <xsl:attribute name="type" expand-text="true" select="concat('xs:',.)"/>
-                </xsl:for-each>
+                <xsl:call-template name="assign-type">
+                    <xsl:with-param name="datatype" select="$datatype"/>
+                </xsl:call-template>
             </xsl:if>
             
             <xsl:apply-templates select=".[exists(@name)] | key('definition-by-name',@ref)" mode="annotated"/>
@@ -237,6 +259,17 @@
             </xsl:apply-templates>
         </xs:attribute>
     </xsl:template>
+    
+    <xsl:template name="assign-type">
+        <xsl:param name="datatype"/>
+        <xsl:for-each select="$datatype[. = $available-custom-types/@name]">
+            <xsl:attribute name="type" expand-text="true"
+                select="concat($declaration-prefix, ':', .)"/>
+        </xsl:for-each>
+        <xsl:for-each select="$datatype[. = $built-in-types/@name]">
+            <xsl:attribute name="type" expand-text="true" select="concat('xs:', .)"/>
+        </xsl:for-each>
+    </xsl:template>
 
     <!-- When allow-other=yes, we union the enumeration with the declared datatype -->        
     <xsl:template match="valid-values[@allow-other='yes']">